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_no_watchdog [HW,ACPI,WDT]
140 Ignore the ACPI-based watchdog interface (WDAT) and let
141 a native driver control the watchdog device instead.
143 acpi_rsdp= [ACPI,EFI,KEXEC]
144 Pass the RSDP address to the kernel, mostly used
145 on machines running EFI runtime service to boot the
146 second kernel for kdump.
148 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
149 Format: To spoof as Windows 98: ="Microsoft Windows"
151 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead
152 of 2 which is mandated by ACPI 6) as the supported ACPI
153 specification revision (when using this switch, it may
154 be necessary to carry out a cold reboot _twice_ in a
155 row to make it take effect on the platform firmware).
157 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
158 acpi_osi="string1" # add string1
159 acpi_osi="!string2" # remove string2
160 acpi_osi=!* # remove all strings
161 acpi_osi=! # disable all built-in OS vendor
163 acpi_osi=!! # enable all built-in OS vendor
165 acpi_osi= # disable all strings
167 'acpi_osi=!' can be used in combination with single or
168 multiple 'acpi_osi="string1"' to support specific OS
169 vendor string(s). Note that such command can only
170 affect the default state of the OS vendor strings, thus
171 it cannot affect the default state of the feature group
172 strings and the current state of the OS vendor strings,
173 specifying it multiple times through kernel command line
174 is meaningless. This command is useful when one do not
175 care about the state of the feature group strings which
176 should be controlled by the OSPM.
178 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
179 to 'acpi_osi="Windows 2000" acpi_osi=!', they all
180 can make '_OSI("Windows 2000")' TRUE.
182 'acpi_osi=' cannot be used in combination with other
183 'acpi_osi=' command lines, the _OSI method will not
184 exist in the ACPI namespace. NOTE that such command can
185 only affect the _OSI support state, thus specifying it
186 multiple times through kernel command line is also
189 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
192 'acpi_osi=!*' can be used in combination with single or
193 multiple 'acpi_osi="string1"' to support specific
194 string(s). Note that such command can affect the
195 current state of both the OS vendor strings and the
196 feature group strings, thus specifying it multiple times
197 through kernel command line is meaningful. But it may
198 still not able to affect the final state of a string if
199 there are quirks related to this string. This command
200 is useful when one want to control the state of the
201 feature group strings to debug BIOS issues related to
204 1. 'acpi_osi="Module Device" acpi_osi=!*' can make
205 '_OSI("Module Device")' FALSE.
206 2. 'acpi_osi=!* acpi_osi="Module Device"' can make
207 '_OSI("Module Device")' TRUE.
208 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
210 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
212 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
213 they all will make '_OSI("Windows 2000")' TRUE.
216 Override the pmtimer bug detection: force the kernel
217 to assume that this machine's pmtimer latches its value
218 and always returns good values.
220 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
221 Format: { level | edge | high | low }
223 acpi_skip_timer_override [HW,ACPI]
224 Recognize and ignore IRQ0/pin2 Interrupt Override.
225 For broken nForce2 BIOS resulting in XT-PIC timer.
227 acpi_sleep= [HW,ACPI] Sleep options
228 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
229 old_ordering, nonvs, sci_force_enable, nobl }
230 See Documentation/power/video.rst for information on
232 s3_beep is for debugging; it makes the PC's speaker beep
233 as soon as the kernel's real-mode entry point is called.
234 s4_nohwsig prevents ACPI hardware signature from being
235 used during resume from hibernation.
236 old_ordering causes the ACPI 1.0 ordering of the _PTS
237 control method, with respect to putting devices into
238 low power states, to be enforced (the ACPI 2.0 ordering
239 of _PTS is used by default).
240 nonvs prevents the kernel from saving/restoring the
241 ACPI NVS memory during suspend/hibernation and resume.
242 sci_force_enable causes the kernel to set SCI_EN directly
243 on resume from S1/S3 (which is against the ACPI spec,
244 but some broken systems don't work without it).
245 nobl causes the internal blacklist of systems known to
246 behave incorrectly in some ways with respect to system
247 suspend and resume to be ignored (use wisely).
249 acpi_use_timer_override [HW,ACPI]
250 Use timer override. For some broken Nvidia NF5 boards
251 that require a timer override, but don't have HPET
253 add_efi_memmap [EFI; X86] Include EFI memory map in
254 kernel's map of available physical RAM.
257 { off | try_unsupported }
258 off: disable AGP support
259 try_unsupported: try to drive unsupported chipsets
260 (may crash computer or cause data corruption)
263 See Documentation/sound/alsa-configuration.rst
266 Allow the default userspace alignment fault handler
267 behaviour to be specified. Bit 0 enables warnings,
268 bit 1 enables fixups, and bit 2 sends a segfault.
270 align_va_addr= [X86-64]
271 Align virtual addresses by clearing slice [14:12] when
272 allocating a VMA at process creation time. This option
273 gives you up to 3% performance improvement on AMD F15h
274 machines (where it is enabled by default) for a
275 CPU-intensive style benchmark, and it can vary highly in
276 a microbenchmark depending on workload and compiler.
278 32: only for 32-bit processes
279 64: only for 64-bit processes
280 on: enable for both 32- and 64-bit processes
281 off: disable for both 32- and 64-bit processes
283 alloc_snapshot [FTRACE]
284 Allocate the ftrace snapshot buffer on boot up when the
285 main buffer is allocated. This is handy if debugging
286 and you need to use tracing_snapshot() on boot up, and
287 do not want to use tracing_snapshot_alloc() as it needs
288 to be done where GFP_KERNEL allocations are allowed.
290 amd_iommu= [HW,X86-64]
291 Pass parameters to the AMD IOMMU driver in the system.
293 fullflush - enable flushing of IO/TLB entries when
294 they are unmapped. Otherwise they are
295 flushed before they will be reused, which
297 off - do not initialize any AMD IOMMU found in
299 force_isolation - Force device isolation for all
300 devices. The IOMMU driver is not
301 allowed anymore to lift isolation
302 requirements as needed. This option
303 does not override iommu=pt
305 amd_iommu_dump= [HW,X86-64]
306 Enable AMD IOMMU driver option to dump the ACPI table
307 for AMD IOMMU. With this option enabled, AMD IOMMU
308 driver will print ACPI tables for AMD IOMMU during
309 IOMMU initialization.
311 amd_iommu_intr= [HW,X86-64]
312 Specifies one of the following AMD IOMMU interrupt
314 legacy - Use legacy interrupt remapping mode.
315 vapic - Use virtual APIC mode, which allows IOMMU
316 to inject interrupts directly into guest.
317 This mode requires kvm-amd.avic=1.
318 (Default when IOMMU HW support is present.)
320 amijoy.map= [HW,JOY] Amiga joystick support
321 Map of devices attached to JOY0DAT and JOY1DAT
323 See also Documentation/input/joydev/joystick.rst
325 analog.map= [HW,JOY] Analog joystick and gamepad support
326 Specifies type or capabilities of an analog joystick
327 connected to one of 16 gameports
328 Format: <type1>,<type2>,..<type16>
331 Power management functions (SPARCstation-4/5 + deriv.)
333 Disable APC CPU standby support. SPARCstation-Fox does
334 not play well with APC CPU idle - disable it if you have
335 APC and your system crashes randomly.
337 apic= [APIC,X86] Advanced Programmable Interrupt Controller
338 Change the output verbosity while booting
339 Format: { quiet (default) | verbose | debug }
340 Change the amount of debugging information output
341 when initialising the APIC and IO-APIC components.
342 For X86-32, this can also be used to specify an APIC
344 Format: apic=driver_name
345 Examples: apic=bigsmp
347 apic_extnmi= [APIC,X86] External NMI delivery setting
348 Format: { bsp (default) | all | none }
349 bsp: External NMI is delivered only to CPU 0
350 all: External NMIs are broadcast to all CPUs as a
352 none: External NMI is masked for all CPUs. This is
353 useful so that a dump capture kernel won't be
357 See Documentation/networking/ipv6.txt.
359 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
360 Limit apic dumping. The parameter defines the maximal
361 number of local apics being dumped. Also it is possible
362 to set it to "all" by meaning -- no limit here.
363 Format: { 1 (default) | 2 | ... | all }.
364 The parameter valid if only apic=debug or
365 apic=verbose is specified.
366 Example: apic=debug show_lapic=all
368 apm= [APM] Advanced Power Management
369 See header of arch/x86/kernel/apm_32.c.
371 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
372 Format: <io>,<irq>,<nodeID>
376 atarimouse= [HW,MOUSE] Atari Mouse
378 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
379 EzKey and similar keyboards
381 atkbd.reset= [HW] Reset keyboard during initialization
383 atkbd.set= [HW] Select keyboard code set
384 Format: <int> (2 = AT (default), 3 = PS/2)
386 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
389 atkbd.softraw= [HW] Choose between synthetic and real raw mode
390 Format: <bool> (0 = real, 1 = synthetic (default))
392 atkbd.softrepeat= [HW]
393 Use software keyboard repeat
395 audit= [KNL] Enable the audit sub-system
396 Format: { "0" | "1" | "off" | "on" }
397 0 | off - kernel audit is disabled and can not be
398 enabled until the next reboot
399 unset - kernel audit is initialized but disabled and
400 will be fully enabled by the userspace auditd.
401 1 | on - kernel audit is initialized and partially
402 enabled, storing at most audit_backlog_limit
403 messages in RAM until it is fully enabled by the
407 audit_backlog_limit= [KNL] Set the audit queue size limit.
408 Format: <int> (must be >=0)
411 bau= [X86_UV] Enable the BAU on SGI UV. The default
412 behavior is to disable the BAU (i.e. bau=0).
413 Format: { "0" | "1" }
416 unset - Disable the BAU.
418 baycom_epp= [HW,AX25]
421 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
423 See header of drivers/net/hamradio/baycom_par.c.
425 baycom_ser_fdx= [HW,AX25]
426 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
427 Format: <io>,<irq>,<mode>[,<baud>]
428 See header of drivers/net/hamradio/baycom_ser_fdx.c.
430 baycom_ser_hdx= [HW,AX25]
431 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
432 Format: <io>,<irq>,<mode>
433 See header of drivers/net/hamradio/baycom_ser_hdx.c.
435 blkdevparts= Manual partition parsing of block device(s) for
436 embedded devices based on command line input.
437 See Documentation/block/cmdline-partition.rst
439 boot_delay= Milliseconds to delay each printk during boot.
440 Values larger than 10 seconds (10000) are changed to
445 Extended command line options can be added to an initrd
446 and this will cause the kernel to look for it.
448 See Documentation/admin-guide/bootconfig.rst
451 Disable BERT OS support on buggy BIOSes.
453 bgrt_disable [ACPI][X86]
454 Disable BGRT to avoid flickering OEM logo.
456 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
457 bttv.radio= Most important insmod options are available as
459 bttv.pll= See Documentation/media/v4l-drivers/bttv.rst
462 bulk_remove=off [PPC] This parameter disables the use of the pSeries
463 firmware feature for flushing multiple hpte entries
466 c101= [NET] Moxa C101 synchronous serial card
468 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
469 Sometimes CPU hardware bugs make them report the cache
470 size incorrectly. The kernel will attempt work arounds
471 to fix known problems, but for some CPUs it is not
472 possible to determine what the correct size should be.
473 This option provides an override for these situations.
476 [NET] Specifies amount of time (in seconds) that
477 the kernel should wait for a network carrier. By default
478 it waits 120 seconds.
480 ca_keys= [KEYS] This parameter identifies a specific key(s) on
481 the system trusted keyring to be used for certificate
483 format: { id:<keyid> | builtin }
485 cca= [MIPS] Override the kernel pages' cache coherency
486 algorithm. Accepted values range from 0 to 7
487 inclusive. See arch/mips/include/asm/pgtable-bits.h
488 for platform specific values (SB1, Loongson3 and
491 ccw_timeout_log [S390]
492 See Documentation/s390/common_io.rst for details.
494 cgroup_disable= [KNL] Disable a particular controller
495 Format: {name of the controller(s) to disable}
496 The effects of cgroup_disable=foo are:
497 - foo isn't auto-mounted if you mount all cgroups in
499 - foo isn't visible as an individually mountable
501 {Currently only "memory" controller deal with this and
502 cut the overhead, others just disable the usage. So
503 only cgroup_disable=memory is actually worthy}
505 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1
506 Format: { { controller | "all" | "named" }
507 [,{ controller | "all" | "named" }...] }
508 Like cgroup_disable, but only applies to cgroup v1;
509 the blacklisted controllers remain available in cgroup2.
510 "all" blacklists all controllers and "named" disables
511 named mounts. Specifying both "all" and "named" disables
514 cgroup.memory= [KNL] Pass options to the cgroup memory controller.
516 nosocket -- Disable socket memory accounting.
517 nokmem -- Disable kernel memory accounting.
519 checkreqprot [SELINUX] Set initial checkreqprot flag value.
520 Format: { "0" | "1" }
521 See security/selinux/Kconfig help text.
522 0 -- check protection applied by kernel (includes
523 any implied execute protection).
524 1 -- check protection requested by application.
525 Default value is set via a kernel config option.
526 Value can be changed at runtime via
527 /sys/fs/selinux/checkreqprot.
530 See Documentation/s390/common_io.rst for details.
533 Prevents the clock framework from automatically gating
534 clocks that have not been explicitly enabled by a Linux
535 device driver but are enabled in hardware at reset or
536 by the bootloader/firmware. Note that this does not
537 force such clocks to be always-on nor does it reserve
538 those clocks in any way. This parameter is useful for
539 debug and development, but should not be needed on a
540 platform with proper driver support. For more
541 information, see Documentation/driver-api/clk.rst.
543 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
545 Forces specified clocksource (if available) to be used
546 when calculating gettimeofday(). If specified
547 clocksource is not available, it defaults to PIT.
548 Format: { pit | tsc | cyclone | pmtmr }
550 clocksource= Override the default clocksource
552 Override the default clocksource and use the clocksource
553 with the name specified.
554 Some clocksource names to choose from, depending on
556 [all] jiffies (this is the base, fallback clocksource)
558 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
559 pxa_timer,timer3,32k_counter,timer0_1
560 [X86-32] pit,hpet,tsc;
561 scx200_hrt on Geode; cyclone on IBM x440
569 clocksource.arm_arch_timer.evtstrm=
572 Enable/disable the eventstream feature of the ARM
573 architected timer so that code using WFE-based polling
574 loops can be debugged more effectively on production
577 clearcpuid=BITNUM [X86]
578 Disable CPUID feature X for the kernel. See
579 arch/x86/include/asm/cpufeatures.h for the valid bit
580 numbers. Note the Linux specific bits are not necessarily
581 stable over kernel options, but the vendor specific
583 Also note that user programs calling CPUID directly
584 or using the feature without checking anything
585 will still see it. This just prevents it from
586 being used by the kernel or shown in /proc/cpuinfo.
587 Also note the kernel might malfunction if you disable
590 cma=nn[MG]@[start[MG][-end[MG]]]
592 Sets the size of kernel global memory area for
593 contiguous memory allocations and optionally the
594 placement constraint by the physical address range of
595 memory allocations. A value of 0 disables CMA
596 altogether. For more information, see
597 include/linux/dma-contiguous.h
599 cmo_free_hint= [PPC] Format: { yes | no }
600 Specify whether pages are marked as being inactive
601 when they are freed. This is used in CMO environments
602 to determine OS memory pressure for page stealing by
606 coherent_pool=nn[KMG] [ARM,KNL]
607 Sets the size of memory pool for coherent, atomic dma
608 allocations, by default set to 256K.
610 com20020= [HW,NET] ARCnet - COM20020 chipset
612 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
614 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
618 ARCnet - COM90xx chipset (memory-mapped buffers)
619 Format: <io>[,<irq>[,<memstart>]]
621 condev= [HW,S390] console device
624 console= [KNL] Output console device and options.
626 tty<n> Use the virtual console device <n>.
630 Use the specified serial port. The options are of
631 the form "bbbbpnf", where "bbbb" is the baud rate,
632 "p" is parity ("n", "o", or "e"), "n" is number of
633 bits, and "f" is flow control ("r" for RTS or
634 omit it). Default is "9600n8".
636 See Documentation/admin-guide/serial-console.rst for more
638 Documentation/networking/netconsole.txt for an
641 uart[8250],io,<addr>[,options]
642 uart[8250],mmio,<addr>[,options]
643 uart[8250],mmio16,<addr>[,options]
644 uart[8250],mmio32,<addr>[,options]
645 uart[8250],0x<addr>[,options]
646 Start an early, polled-mode console on the 8250/16550
647 UART at the specified I/O port or MMIO address,
648 switching to the matching ttyS device later.
649 MMIO inter-register address stride is either 8-bit
650 (mmio), 16-bit (mmio16), or 32-bit (mmio32).
651 If none of [io|mmio|mmio16|mmio32], <addr> is assumed
652 to be equivalent to 'mmio'. 'options' are specified in
653 the same format described for ttyS above; if unspecified,
654 the h/w is not re-initialized.
656 hvc<n> Use the hypervisor console device <n>. This is for
657 both Xen and PowerPC hypervisors.
659 If the device connected to the port is not a TTY but a braille
660 device, prepend "brl," before the device type, for instance
662 For now, only VisioBraille is supported.
665 [KNL] Change console messages format
667 By default we print messages on consoles in
668 "[time stamp] text\n" format (time stamp may not be
669 printed, depending on CONFIG_PRINTK_TIME or
670 `printk_time' param).
672 Switch to syslog format: "<%u>[time stamp] text\n"
673 IOW, each message will have a facility and loglevel
674 prefix. The format is similar to one used by syslog()
675 syscall, or to executing "dmesg -S --raw" or to reading
678 consoleblank= [KNL] The console blank (screen saver) timeout in
679 seconds. A value of 0 disables the blank timer.
683 [KNL] Change the default value for
684 /proc/<pid>/coredump_filter.
685 See also Documentation/filesystems/proc.txt.
687 coresight_cpu_debug.enable
690 Enable/disable the CPU sampling based debugging.
691 0: default value, disable debugging
692 1: enable debugging at boot time
694 cpuidle.off=1 [CPU_IDLE]
695 disable the cpuidle sub-system
698 [CPU_IDLE] Name of the cpuidle governor to use.
700 cpufreq.off=1 [CPU_FREQ]
701 disable the cpufreq sub-system
704 [X86] Delay for N microsec between assert and de-assert
705 of APIC INIT to start processors. This delay occurs
706 on every CPU online, such as boot, and resume from suspend.
709 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
711 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
713 crashkernel=size[KMG][@offset[KMG]]
714 [KNL] Using kexec, Linux can switch to a 'crash kernel'
715 upon panic. This parameter reserves the physical
716 memory region [offset, offset + size] for that kernel
717 image. If '@offset' is omitted, then a suitable offset
718 is selected automatically.
719 [KNL, x86_64] select a region under 4G first, and
720 fall back to reserve region above 4G when '@offset'
721 hasn't been specified.
722 See Documentation/admin-guide/kdump/kdump.rst for further details.
724 crashkernel=range1:size1[,range2:size2,...][@offset]
725 [KNL] Same as above, but depends on the memory
726 in the running system. The syntax of range is
727 start-[end] where start and end are both
728 a memory unit (amount[KMG]). See also
729 Documentation/admin-guide/kdump/kdump.rst for an example.
731 crashkernel=size[KMG],high
732 [KNL, x86_64] range could be above 4G. Allow kernel
733 to allocate physical memory region from top, so could
734 be above 4G if system have more than 4G ram installed.
735 Otherwise memory region will be allocated below 4G, if
737 It will be ignored if crashkernel=X is specified.
738 crashkernel=size[KMG],low
739 [KNL, x86_64] range under 4G. When crashkernel=X,high
740 is passed, kernel could allocate physical memory region
741 above 4G, that cause second kernel crash on system
742 that require some amount of low memory, e.g. swiotlb
743 requires at least 64M+32K low memory, also enough extra
744 low memory is needed to make sure DMA buffers for 32-bit
745 devices won't run out. Kernel would try to allocate at
746 at least 256M below 4G automatically.
747 This one let user to specify own low range under 4G
748 for second kernel instead.
749 0: to disable low allocation.
750 It will be ignored when crashkernel=X,high is not used
751 or memory reserved is below 4G.
754 [KNL] Disable crypto self-tests
759 cs89x0_media= [HW,NET]
760 Format: { rj45 | aui | bnc }
763 See header of drivers/s390/block/dasd_devmap.c.
765 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
766 (one device per port)
767 Format: <port#>,<type>
768 See also Documentation/input/devices/joystick-parport.rst
770 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
772 Documentation/admin-guide/dynamic-debug-howto.rst for
773 details. Deprecated, see dyndbg.
775 debug [KNL] Enable kernel debugging (events log level).
778 [KNL] Enable printing [hashed] pointers early in the
779 boot sequence. If enabled, we use a weak hash instead
780 of siphash to hash pointers. Use this option if you are
781 seeing instances of '(___ptrval___)') and need to see a
782 value (hashed pointer) instead. Cryptographically
783 insecure, please do not use on production kernels.
786 [KNL] verbose self-tests
788 Print debugging info while doing the locking API
790 We default to 0 (no extra messages), setting it to
791 1 will print _a lot_ more information - normally
792 only useful to kernel developers.
794 debug_objects [KNL] Enable object debugging
797 [KNL] Disable object debugging
799 debug_guardpage_minorder=
800 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
801 parameter allows control of the order of pages that will
802 be intentionally kept free (and hence protected) by the
803 buddy allocator. Bigger value increase the probability
804 of catching random memory corruption, but reduce the
805 amount of memory for normal system use. The maximum
806 possible value is MAX_ORDER/2. Setting this parameter
807 to 1 or 2 should be enough to identify most random
808 memory corruption problems caused by bugs in kernel or
809 driver code when a CPU writes to (or reads from) a
810 random memory location. Note that there exists a class
811 of memory corruptions problems caused by buggy H/W or
812 F/W or by drivers badly programing DMA (basically when
813 memory is written at bus level and the CPU MMU is
814 bypassed) which are not detectable by
815 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
816 tracking down these problems.
819 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter
820 enables the feature at boot time. By default, it is
821 disabled and the system will work mostly the same as a
822 kernel built without CONFIG_DEBUG_PAGEALLOC.
823 Note: to get most of debug_pagealloc error reports, it's
824 useful to also enable the page_owner functionality.
825 on: enable the feature
827 debugpat [X86] Enable PAT debugging
829 decnet.addr= [HW,NET]
830 Format: <area>[,<node>]
831 See also Documentation/networking/decnet.txt.
834 [same as hugepagesz=] The size of the default
835 HugeTLB page size. This is the size represented by
836 the legacy /proc/ hugepages APIs, used for SHM, and
837 default size when mounting hugetlbfs filesystems.
838 Defaults to the default architecture's huge page size
841 deferred_probe_timeout=
842 [KNL] Debugging option to set a timeout in seconds for
843 deferred probe to give up waiting on dependencies to
844 probe. Only specific dependencies (subsystems or
845 drivers) that have opted in will be ignored. A timeout of 0
846 will timeout at the end of initcalls. This option will also
847 dump out devices still on the deferred probe list after
851 Format: { on | off | def_only | inf_only | always }
852 on: s390 zlib hardware support for compression on
853 level 1 and decompression (default)
854 off: No s390 zlib hardware support
855 def_only: s390 zlib hardware support for deflate
856 only (compression on level 1)
857 inf_only: s390 zlib hardware support for inflate
859 always: Same as 'on' but ignores the selected compression
860 level always using hardware support (used for debugging)
863 Set number of hash buckets for dentry cache.
865 disable_1tb_segments [PPC]
866 Disables the use of 1TB hash page table segments. This
867 causes the kernel to fall back to 256MB segments which
868 can be useful when debugging issues that require an SLB
872 See Documentation/networking/ipv6.txt.
875 [KNL] Under CONFIG_HARDENED_USERCOPY, whether
876 hardening is enabled for this boot. Hardened
877 usercopy checking is used to protect the kernel
878 from reading or writing beyond known memory
879 allocation boundaries as a proactive defense
880 against bounds-checking flaws in the kernel's
881 copy_to_user()/copy_from_user() interface.
882 on Perform hardened usercopy checks (default).
883 off Disable hardened usercopy checks.
886 Disable RADIX MMU mode on POWER9
889 Disable TLBIE instruction. Currently does not work
890 with KVM, with HASH MMU, or with coherent accelerators.
892 disable_cpu_apicid= [X86,APIC,SMP]
894 The number of initial APIC ID for the
895 corresponding CPU to be disabled at boot,
896 mostly used for the kdump 2nd kernel to
897 disable BSP to wake up multiple CPUs without
898 causing system reset or hang due to sending
901 perf_v4_pmi= [X86,INTEL]
903 Disable Intel PMU counter freezing feature.
904 The feature only exists starting from
905 Arch Perfmon v4 (Skylake and newer).
907 disable_ddw [PPC/PSERIES]
908 Disable Dynamic DMA Window support. Use this if
909 to workaround buggy firmware.
912 See Documentation/networking/ipv6.txt.
914 disable_mtrr_cleanup [X86]
915 The kernel tries to adjust MTRR layout from continuous
916 to discrete, to make X server driver able to add WB
917 entry later. This parameter disables that.
919 disable_mtrr_trim [X86, Intel and AMD only]
920 By default the kernel will trim any uncacheable
921 memory out of your available memory pool based on
922 MTRR settings. This parameter disables that behavior,
923 possibly causing your machine to run very slowly.
925 disable_timer_pin_1 [X86]
926 Disable PIN 1 of APIC timer
927 Can be useful to work around chipset bugs.
929 dis_ucode_ldr [X86] Disable the microcode loader.
931 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
932 this option disables the debugging code at boot.
934 dma_debug_entries=<number>
935 This option allows to tune the number of preallocated
936 entries for DMA-API debugging code. One entry is
937 required per DMA-API allocation. Use this if the
938 DMA-API debugging code disables itself because the
939 architectural default is too low.
941 dma_debug_driver=<driver_name>
942 With this option the DMA-API debugging driver
943 filter feature can be enabled at boot time. Just
944 pass the driver to filter for as the parameter.
945 The filter can be disabled or changed to another
946 driver later using sysfs.
948 driver_async_probe= [KNL]
949 List of driver names to be probed asynchronously.
950 Format: <driver_name1>,<driver_name2>...
952 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
953 Broken monitors, graphic adapters, KVMs and EDIDless
954 panels may send no or incorrect EDID data sets.
955 This parameter allows to specify an EDID data sets
956 in the /lib/firmware directory that are used instead.
957 Generic built-in EDID data sets are used, if one of
958 edid/1024x768.bin, edid/1280x1024.bin,
959 edid/1680x1050.bin, or edid/1920x1080.bin is given
960 and no file with the same name exists. Details and
961 instructions how to build your own EDID data are
962 available in Documentation/driver-api/edid.rst. An EDID
963 data set will only be used for a particular connector,
964 if its name and a colon are prepended to the EDID
965 name. Each connector may use a unique EDID data
966 set by separating the files with a comma. An EDID
967 data set with no connector name will be used for
968 any connectors not explicitly specified.
973 Format: {"off" | "known"}
974 Control how the dt_cpu_ftrs device-tree binding is
975 used for CPU feature discovery and setup (if it
977 off: Do not use it, fall back to legacy cpu table.
978 known: Do not pass through unknown features to guests
979 or userspace, only those that the kernel is aware of.
981 dump_apple_properties [X86]
982 Dump name and content of EFI device properties on
983 x86 Macs. Useful for driver authors to determine
984 what data is available or for reverse-engineering.
986 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
987 module.dyndbg[="val"]
988 Enable debug messages at boot time. See
989 Documentation/admin-guide/dynamic-debug-howto.rst
992 nompx [X86] Disables Intel Memory Protection Extensions.
993 See Documentation/x86/intel_mpx.rst for more
994 information about the feature.
996 nopku [X86] Disable Memory Protection Keys CPU feature found
999 module.async_probe [KNL]
1000 Enable asynchronous probe on this module.
1002 early_ioremap_debug [KNL]
1003 Enable debug messages in early_ioremap support. This
1004 is useful for tracking down temporary early mappings
1005 which are not unmapped.
1007 earlycon= [KNL] Output early console device and options.
1009 When used with no options, the early console is
1010 determined by stdout-path property in device tree's
1011 chosen node or the ACPI SPCR table if supported by
1014 cdns,<addr>[,options]
1015 Start an early, polled-mode console on a Cadence
1016 (xuartps) serial port at the specified address. Only
1017 supported option is baud rate. If baud rate is not
1018 specified, the serial port must already be setup and
1021 uart[8250],io,<addr>[,options]
1022 uart[8250],mmio,<addr>[,options]
1023 uart[8250],mmio32,<addr>[,options]
1024 uart[8250],mmio32be,<addr>[,options]
1025 uart[8250],0x<addr>[,options]
1026 Start an early, polled-mode console on the 8250/16550
1027 UART at the specified I/O port or MMIO address.
1028 MMIO inter-register address stride is either 8-bit
1029 (mmio) or 32-bit (mmio32 or mmio32be).
1030 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
1031 to be equivalent to 'mmio'. 'options' are specified
1032 in the same format described for "console=ttyS<n>"; if
1033 unspecified, the h/w is not initialized.
1037 Start an early, polled-mode console on a pl011 serial
1038 port at the specified address. The pl011 serial port
1039 must already be setup and configured. Options are not
1040 yet supported. If 'mmio32' is specified, then only
1041 the driver will use only 32-bit accessors to read/write
1042 the device registers.
1045 Start an early, polled-mode console on a meson serial
1046 port at the specified address. The serial port must
1047 already be setup and configured. Options are not yet
1051 Start an early, polled-mode console on an msm serial
1052 port at the specified address. The serial port
1053 must already be setup and configured. Options are not
1056 msm_serial_dm,<addr>
1057 Start an early, polled-mode console on an msm serial
1058 dm port at the specified address. The serial port
1059 must already be setup and configured. Options are not
1063 Start an early, polled-mode console on a serial port
1064 of an Actions Semi SoC, such as S500 or S900, at the
1065 specified address. The serial port must already be
1066 setup and configured. Options are not yet supported.
1069 Start an early, polled-mode console on a serial port
1070 of an RDA Micro SoC, such as RDA8810PL, at the
1071 specified address. The serial port must already be
1072 setup and configured. Options are not yet supported.
1075 Use RISC-V SBI (Supervisor Binary Interface) for early
1078 smh Use ARM semihosting calls for early console.
1086 Use early console provided by serial driver available
1087 on Samsung SoCs, requires selecting proper type and
1088 a correct base address of the selected UART port. The
1089 serial port must already be setup and configured.
1090 Options are not yet supported.
1093 Start an early, polled-mode console on a lantiq serial
1094 (lqasc) port at the specified address. The serial port
1095 must already be setup and configured. Options are not
1100 Use early console provided by Freescale LP UART driver
1101 found on Freescale Vybrid and QorIQ LS1021A processors.
1102 A valid base address must be provided, and the serial
1103 port must already be setup and configured.
1107 Start an early, polled-mode, output-only console on the
1108 Freescale i.MX UART at the specified address. The UART
1109 must already be setup and configured.
1112 Start an early, polled-mode console on the
1113 Armada 3700 serial port at the specified
1114 address. The serial port must already be setup
1115 and configured. Options are not yet supported.
1118 Start an early, polled-mode console on a Qualcomm
1119 Generic Interface (GENI) based serial port at the
1120 specified address. The serial port must already be
1121 setup and configured. Options are not yet supported.
1124 Start an early, unaccelerated console on the EFI
1125 memory mapped framebuffer (if available). On cache
1126 coherent non-x86 systems that use system memory for
1127 the framebuffer, pass the 'ram' option so that it is
1128 mapped with the correct attributes.
1131 Use early console provided by Freescale LINFlexD UART
1132 serial driver for NXP S32V234 SoCs. A valid base
1133 address must be provided, and the serial port must
1134 already be setup and configured.
1136 earlyprintk= [X86,SH,ARM,M68k,S390]
1140 earlyprintk=serial[,ttySn[,baudrate]]
1141 earlyprintk=serial[,0x...[,baudrate]]
1142 earlyprintk=ttySn[,baudrate]
1143 earlyprintk=dbgp[debugController#]
1144 earlyprintk=pciserial[,force],bus:device.function[,baudrate]
1145 earlyprintk=xdbc[xhciController#]
1147 earlyprintk is useful when the kernel crashes before
1148 the normal console is initialized. It is not enabled by
1149 default because it has some cosmetic problems.
1151 Append ",keep" to not disable it when the real console
1154 Only one of vga, efi, serial, or usb debug port can
1157 Currently only ttyS0 and ttyS1 may be specified by
1158 name. Other I/O ports may be explicitly specified
1159 on some architectures (x86 and arm at least) by
1160 replacing ttySn with an I/O port address, like this:
1161 earlyprintk=serial,0x1008,115200
1162 You can find the port for a given device in
1163 /proc/tty/driver/serial:
1164 2: uart:ST16650V2 port:00001008 irq:18 ...
1166 Interaction with the standard serial driver is not
1169 The VGA and EFI output is eventually overwritten by
1172 The xen output can only be used by Xen PV guests.
1174 The sclp output can only be used on s390.
1176 The optional "force" to "pciserial" enables use of a
1177 PCI device even when its classcode is not of the
1180 edac_report= [HW,EDAC] Control how to report EDAC event
1181 Format: {"on" | "off" | "force"}
1182 on: enable EDAC to report H/W event. May be overridden
1183 by other higher priority error reporting module.
1184 off: disable H/W event reporting through EDAC.
1185 force: enforce the use of EDAC to report H/W event.
1188 ekgdboc= [X86,KGDB] Allow early kernel console debugging
1191 This is designed to be used in conjunction with
1192 the boot argument: earlyprintk=vga
1195 Format: {"off" | "on" | "skip[mbr]"}
1198 Format: { "old_map", "nochunk", "noruntime", "debug",
1199 "nosoftreserve", "disable_early_pci_dma",
1200 "no_disable_early_pci_dma" }
1201 old_map [X86-64]: switch to the old ioremap-based EFI
1202 runtime services mapping. [Needs CONFIG_X86_UV=y]
1203 nochunk: disable reading files in "chunks" in the EFI
1204 boot stub, as chunking can cause problems with some
1205 firmware implementations.
1206 noruntime : disable EFI runtime services support
1207 debug: enable misc debug output
1208 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose)
1209 attribute may cause the kernel to reserve the
1210 memory range for a memory mapping driver to
1211 claim. Specify efi=nosoftreserve to disable this
1212 reservation and treat the memory by its base type
1213 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM").
1214 disable_early_pci_dma: Disable the busmaster bit on all
1215 PCI bridges while in the EFI boot stub
1216 no_disable_early_pci_dma: Leave the busmaster bit set
1217 on all PCI bridges while in the EFI boot stub
1219 efi_no_storage_paranoia [EFI; X86]
1220 Using this parameter you can use more than 50% of
1221 your efi variable storage. Use this parameter only if
1222 you are really sure that your UEFI does sane gc and
1223 fulfills the spec otherwise your board may brick.
1225 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1226 Add arbitrary attribute to specific memory range by
1227 updating original EFI memory map.
1228 Region of memory which aa attribute is added to is
1231 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1232 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1233 attribute is added to range 0x100000000-0x180000000 and
1234 0x10a0000000-0x1120000000.
1236 If efi_fake_mem=8G@9G:0x40000 is specified, the
1237 EFI_MEMORY_SP(0x40000) attribute is added to
1238 range 0x240000000-0x43fffffff.
1240 Using this parameter you can do debugging of EFI memmap
1241 related features. For example, you can do debugging of
1242 Address Range Mirroring feature even if your box
1243 doesn't support it, or mark specific memory as
1246 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT
1247 that is to be dynamically loaded by Linux. If there are
1248 multiple variables with the same name but with different
1249 vendor GUIDs, all of them will be loaded. See
1250 Documentation/admin-guide/acpi/ssdt-overlays.rst for details.
1253 eisa_irq_edge= [PARISC,HW]
1254 See header of drivers/parisc/eisa.c.
1257 See comment before function elanfreq_setup() in
1258 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1260 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1261 Specifies physical address of start of kernel core
1262 image elf header and optionally the size. Generally
1263 kexec loader will pass this option to capture kernel.
1264 See Documentation/admin-guide/kdump/kdump.rst for details.
1266 enable_mtrr_cleanup [X86]
1267 The kernel tries to adjust MTRR layout from continuous
1268 to discrete, to make X server driver able to add WB
1269 entry later. This parameter enables that.
1271 enable_timer_pin_1 [X86]
1272 Enable PIN 1 of APIC timer
1273 Can be useful to work around chipset bugs
1274 (in particular on some ATI chipsets).
1275 The kernel tries to set a reasonable default.
1277 enforcing [SELINUX] Set initial enforcing status.
1279 See security/selinux/Kconfig help text.
1280 0 -- permissive (log only, no denials).
1281 1 -- enforcing (deny and log).
1283 Value can be changed at runtime via
1284 /sys/fs/selinux/enforce.
1287 Disable Error Record Serialization Table (ERST)
1290 ether= [HW,NET] Ethernet cards parameters
1291 This option is obsoleted by the "netdev=" option, which
1292 has equivalent usage. See its documentation for details.
1296 Permit 'security.evm' to be updated regardless of
1297 current integrity status.
1301 fail_make_request=[KNL]
1302 General fault injection mechanism.
1303 Format: <interval>,<probability>,<space>,<times>
1304 See also Documentation/fault-injection/.
1307 See Documentation/admin-guide/blockdev/floppy.rst.
1309 force_pal_cache_flush
1310 [IA-64] Avoid check_sal_cache_flush which may hang on
1311 buggy SAL_CACHE_FLUSH implementations. Using this
1312 parameter will force ia64_sal_cache_flush to call
1313 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1316 Forcefully enable Physical Address Extension (PAE).
1317 Many Pentium M systems disable PAE but may have a
1318 functionally usable PAE implementation.
1319 Warning: use of this parameter will taint the kernel
1320 and may cause unknown problems.
1323 [FTRACE] will set and start the specified tracer
1324 as early as possible in order to facilitate early
1327 ftrace_dump_on_oops[=orig_cpu]
1328 [FTRACE] will dump the trace buffers on oops.
1329 If no parameter is passed, ftrace will dump
1330 buffers of all CPUs, but if you pass orig_cpu, it will
1331 dump only the buffer of the CPU that triggered the
1334 ftrace_filter=[function-list]
1335 [FTRACE] Limit the functions traced by the function
1336 tracer at boot up. function-list is a comma separated
1337 list of functions. This list can be changed at run
1338 time by the set_ftrace_filter file in the debugfs
1341 ftrace_notrace=[function-list]
1342 [FTRACE] Do not trace the functions specified in
1343 function-list. This list can be changed at run time
1344 by the set_ftrace_notrace file in the debugfs
1347 ftrace_graph_filter=[function-list]
1348 [FTRACE] Limit the top level callers functions traced
1349 by the function graph tracer at boot up.
1350 function-list is a comma separated list of functions
1351 that can be changed at run time by the
1352 set_graph_function file in the debugfs tracing directory.
1354 ftrace_graph_notrace=[function-list]
1355 [FTRACE] Do not trace from the functions specified in
1356 function-list. This list is a comma separated list of
1357 functions that can be changed at run time by the
1358 set_graph_notrace file in the debugfs tracing directory.
1360 ftrace_graph_max_depth=<uint>
1361 [FTRACE] Used with the function graph tracer. This is
1362 the max depth it will trace into a function. This value
1363 can be changed at run time by the max_graph_depth file
1364 in the tracefs tracing directory. default: 0 (no limit)
1366 fw_devlink= [KNL] Create device links between consumer and supplier
1367 devices by scanning the firmware to infer the
1368 consumer/supplier relationships. This feature is
1369 especially useful when drivers are loaded as modules as
1370 it ensures proper ordering of tasks like device probing
1371 (suppliers first, then consumers), supplier boot state
1372 clean up (only after all consumers have probed),
1373 suspend/resume & runtime PM (consumers first, then
1375 Format: { off | permissive | on | rpm }
1376 off -- Don't create device links from firmware info.
1377 permissive -- Create device links from firmware info
1378 but use it only for ordering boot state clean
1379 up (sync_state() calls).
1380 on -- Create device links from firmware info and use it
1381 to enforce probe and suspend/resume ordering.
1382 rpm -- Like "on", but also use to order runtime PM.
1385 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1386 support via parallel port (up to 5 devices per port)
1387 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1388 See also Documentation/input/devices/joystick-parport.rst
1392 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART
1396 gcov_persist= [GCOV] When non-zero (default), profiling data for
1397 kernel modules is saved and remains accessible via
1398 debugfs, even when the module is unloaded/reloaded.
1399 When zero, profiling data is discarded and associated
1400 debugfs files are removed at module unload time.
1402 goldfish [X86] Enable the goldfish android emulator platform.
1403 Don't use this when you are not running on the
1406 gpt [EFI] Forces disk with valid GPT signature but
1407 invalid Protective MBR to be treated as GPT. If the
1408 primary GPT is corrupted, it enables the backup/alternate
1409 GPT to be used instead.
1411 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1412 the "Enable 0" bit of the configuration register.
1415 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1416 the "Enable 0" bit of the configuration register.
1419 grcan.select= [HW] Select which physical interface to use.
1422 grcan.txsize= [HW] Sets the size of the tx buffer.
1423 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1425 grcan.rxsize= [HW] Sets the size of the rx buffer.
1426 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1429 gpio-mockup.gpio_mockup_ranges
1430 [HW] Sets the ranges of gpiochip of for this device.
1431 Format: <start1>,<end1>,<start2>,<end2>...
1433 hardlockup_all_cpu_backtrace=
1434 [KNL] Should the hard-lockup detector generate
1435 backtraces on all cpus.
1438 hashdist= [KNL,NUMA] Large hashes allocated during boot
1439 are distributed across NUMA nodes. Defaults on
1440 for 64-bit NUMA, off otherwise.
1441 Format: 0 | 1 (for off | on)
1443 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1445 hd= [EIDE] (E)IDE hard drive subsystem geometry
1446 Format: <cyl>,<head>,<sect>
1449 Disable Hardware Error Source Table (HEST) support;
1450 corresponding firmware-first mode error processing
1451 logic will be disabled.
1453 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1454 size of <nn>. This works even on boxes that have no
1455 highmem otherwise. This also works to reduce highmem
1456 size on bigger boxes.
1458 highres= [KNL] Enable/disable high resolution timer mode.
1459 Valid parameters: "on", "off"
1464 hpet= [X86-32,HPET] option to control HPET usage
1465 Format: { enable (default) | disable | force |
1467 disable: disable HPET and use PIT instead
1468 force: allow force enabled of undocumented chips (ICH4,
1470 verbose: show contents of HPET registers during setup
1472 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1473 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1475 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
1476 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
1477 On x86-64 and powerpc, this option can be specified
1478 multiple times interleaved with hugepages= to reserve
1479 huge pages of different sizes. Valid pages sizes on
1480 x86-64 are 2M (when the CPU supports "pse") and 1G
1481 (when the CPU supports the "pdpe1gb" cpuinfo flag).
1484 [KNL] Should the hung task detector generate panics.
1487 A nonzero value instructs the kernel to panic when a
1488 hung task is detected. The default value is controlled
1489 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time
1490 option. The value selected by this boot parameter can
1491 be changed later by the kernel.hung_task_panic sysctl.
1493 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1494 terminal devices. Valid values: 0..8
1495 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1496 If specified, z/VM IUCV HVC accepts connections
1497 from listed z/VM user IDs only.
1499 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations
1500 which allow the hypervisor to 'idle' the
1501 guest on lock contention.
1504 Do not unregister boot console at start. This is only
1505 useful for debugging when something happens in the window
1506 between unregistering the boot console and initializing
1509 i2c_bus= [HW] Override the default board specific I2C bus speed
1510 or register an additional I2C bus that is not
1511 registered from board initialization code.
1515 i8042.debug [HW] Toggle i8042 debug mode
1516 i8042.unmask_kbd_data
1517 [HW] Enable printing of interrupt data from the KBD port
1518 (disabled by default, and as a pre-condition
1519 requires that i8042.debug=1 be enabled)
1520 i8042.direct [HW] Put keyboard port into non-translated mode
1521 i8042.dumbkbd [HW] Pretend that controller can only read data from
1522 keyboard and cannot control its state
1523 (Don't attempt to blink the leds)
1524 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1525 i8042.nokbd [HW] Don't check/create keyboard port
1526 i8042.noloop [HW] Disable the AUX Loopback command while probing
1528 i8042.nomux [HW] Don't check presence of an active multiplexing
1530 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1532 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1533 i8042.reset [HW] Reset the controller during init, cleanup and
1534 suspend-to-ram transitions, only during s2r
1535 transitions, or never reset
1536 Format: { 1 | Y | y | 0 | N | n }
1537 1, Y, y: always reset controller
1538 0, N, n: don't ever reset controller
1539 Default: only on s2r transitions on x86; most other
1540 architectures force reset to be always executed
1541 i8042.unlock [HW] Unlock (ignore) the keylock
1542 i8042.kbdreset [HW] Reset device connected to KBD port
1546 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
1547 indicates that the driver is running on unsupported
1549 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
1550 does not match list of supported models.
1552 [HW] Report power status in /proc/i8k
1553 (disabled by default)
1554 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
1557 i915.invert_brightness=
1558 [DRM] Invert the sense of the variable that is used to
1559 set the brightness of the panel backlight. Normally a
1560 brightness value of 0 indicates backlight switched off,
1561 and the maximum of the brightness value sets the backlight
1562 to maximum brightness. If this parameter is set to 0
1563 (default) and the machine requires it, or this parameter
1564 is set to 1, a brightness value of 0 sets the backlight
1565 to maximum brightness, and the maximum of the brightness
1566 value switches the backlight off.
1567 -1 -- never invert brightness
1568 0 -- machine default
1569 1 -- force brightness inversion
1572 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1574 ide-core.nodma= [HW] (E)IDE subsystem
1575 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1576 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1577 .cdrom .chs .ignore_cable are additional options
1578 See Documentation/ide/ide.rst.
1580 ide-generic.probe-mask= [HW] (E)IDE subsystem
1582 Probe mask for legacy ISA IDE ports. Depending on
1583 platform up to 6 ports are supported, enabled by
1584 setting corresponding bits in the mask to 1. The
1585 default value is 0x0, which has a special meaning.
1586 On systems that have PCI, it triggers scanning the
1587 PCI bus for the first and the second port, which
1588 are then probed. On systems without PCI the value
1589 of 0x0 enables probing the two first ports as if it
1592 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1593 Claim all unknown PCI IDE storage controllers.
1596 Format: idle=poll, idle=halt, idle=nomwait
1597 Poll forces a polling idle loop that can slightly
1598 improve the performance of waking up a idle CPU, but
1599 will use a lot of power and make the system run hot.
1601 idle=halt: Halt is forced to be used for CPU idle.
1602 In such case C2/C3 won't be used again.
1603 idle=nomwait: Disable mwait for CPU C-states
1605 ieee754= [MIPS] Select IEEE Std 754 conformance mode
1606 Format: { strict | legacy | 2008 | relaxed }
1609 Choose which programs will be accepted for execution
1610 based on the IEEE 754 NaN encoding(s) supported by
1611 the FPU and the NaN encoding requested with the value
1612 of an ELF file header flag individually set by each
1613 binary. Hardware implementations are permitted to
1614 support either or both of the legacy and the 2008 NaN
1617 Available settings are as follows:
1618 strict accept binaries that request a NaN encoding
1619 supported by the FPU
1620 legacy only accept legacy-NaN binaries, if supported
1622 2008 only accept 2008-NaN binaries, if supported
1624 relaxed accept any binaries regardless of whether
1625 supported by the FPU
1627 The FPU emulator is always able to support both NaN
1628 encodings, so if no FPU hardware is present or it has
1629 been disabled with 'nofpu', then the settings of
1630 'legacy' and '2008' strap the emulator accordingly,
1631 'relaxed' straps the emulator for both legacy-NaN and
1632 2008-NaN, whereas 'strict' enables legacy-NaN only on
1633 legacy processors and both NaN encodings on MIPS32 or
1636 The setting for ABS.fmt/NEG.fmt instruction execution
1637 mode generally follows that for the NaN encoding,
1638 except where unsupported by hardware.
1640 ignore_loglevel [KNL]
1641 Ignore loglevel setting - this will print /all/
1642 kernel messages to the console. Useful for debugging.
1643 We also add it as printk module parameter, so users
1644 could change it dynamically, usually by
1645 /sys/module/printk/parameters/ignore_loglevel.
1648 Ignore RLIMIT_DATA setting for data mappings,
1649 print warning at first misuse. Can be changed via
1650 /sys/module/kernel/parameters/ignore_rlimit_data.
1652 ihash_entries= [KNL]
1653 Set number of hash buckets for inode cache.
1655 ima_appraise= [IMA] appraise integrity measurements
1656 Format: { "off" | "enforce" | "fix" | "log" }
1659 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead.
1660 The builtin appraise policy appraises all files
1663 ima_canonical_fmt [IMA]
1664 Use the canonical format for the binary runtime
1665 measurements, instead of host native format.
1668 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1672 The list of supported hash algorithms is defined
1673 in crypto/hash_info.h.
1676 The builtin policies to load during IMA setup.
1677 Format: "tcb | appraise_tcb | secure_boot |
1680 The "tcb" policy measures all programs exec'd, files
1681 mmap'd for exec, and all files opened with the read
1682 mode bit set by either the effective uid (euid=0) or
1685 The "appraise_tcb" policy appraises the integrity of
1686 all files owned by root.
1688 The "secure_boot" policy appraises the integrity
1689 of files (eg. kexec kernel image, kernel modules,
1690 firmware, policy, etc) based on file signatures.
1692 The "fail_securely" policy forces file signature
1693 verification failure also on privileged mounted
1694 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE
1697 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
1698 Load a policy which meets the needs of the Trusted
1699 Computing Base. This means IMA will measure all
1700 programs exec'd, files mmap'd for exec, and all files
1701 opened for read by uid=0.
1704 Select one of defined IMA measurements template formats.
1705 Formats: { "ima" | "ima-ng" | "ima-sig" }
1709 [IMA] Define a custom template format.
1710 Format: { "field1|...|fieldN" }
1712 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1713 Format: <min_file_size>
1714 Set the minimal file size for using asynchronous hash.
1715 If left unspecified, ahash usage is disabled.
1717 ahash performance varies for different data sizes on
1718 different crypto accelerators. This option can be used
1719 to achieve the best performance for a particular HW.
1721 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1723 Set hashing buffer size. Default: 4k.
1725 ahash performance varies for different chunk sizes on
1726 different crypto accelerators. This option can be used
1727 to achieve best performance for particular HW.
1731 Run specified binary instead of /sbin/init as init
1734 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1735 for working out where the kernel is dying during
1738 initcall_blacklist= [KNL] Do not execute a comma-separated list of
1739 initcall functions. Useful for debugging built-in
1740 modules and initcalls.
1742 initrd= [BOOT] Specify the location of the initial ramdisk
1744 init_on_alloc= [MM] Fill newly allocated pages and heap objects with
1747 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON.
1749 init_on_free= [MM] Fill freed pages and heap objects with zeroes.
1751 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON.
1753 init_pkru= [x86] Specify the default memory protection keys rights
1754 register contents for all processes. 0x55555554 by
1755 default (disallow access to all but pkey 0). Can
1756 override in debugfs after boot.
1758 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1761 int_pln_enable [x86] Enable power limit notification interrupt
1763 integrity_audit=[IMA]
1764 Format: { "0" | "1" }
1765 0 -- basic integrity auditing messages. (Default)
1766 1 -- additional integrity auditing messages.
1768 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
1770 Enable intel iommu driver.
1772 Disable intel iommu driver.
1773 igfx_off [Default Off]
1774 By default, gfx is mapped as normal device. If a gfx
1775 device has a dedicated DMAR unit, the DMAR unit is
1776 bypassed by not enabling DMAR with this option. In
1777 this case, gfx device will use physical address for
1780 With this option iommu will not optimize to look
1781 for io virtual address below 32-bit forcing dual
1782 address cycle on pci bus for cards supporting greater
1783 than 32-bit addressing. The default is to look
1784 for translation below 32-bit and if not available
1785 then look in the higher range.
1786 strict [Default Off]
1787 With this option on every unmap_single operation will
1788 result in a hardware IOTLB flush operation as opposed
1789 to batching them for performance.
1790 sp_off [Default Off]
1791 By default, super page will be supported if Intel IOMMU
1792 has the capability. With this option, super page will
1795 By default, scalable mode will be disabled even if the
1796 hardware advertises that it has support for the scalable
1797 mode translation. With this option set, scalable mode
1798 will be used on hardware which claims to support it.
1799 tboot_noforce [Default Off]
1800 Do not force the Intel IOMMU enabled under tboot.
1801 By default, tboot will force Intel IOMMU on, which
1802 could harm performance of some high-throughput
1803 devices like 40GBit network cards, even if identity
1805 Note that using this option lowers the security
1806 provided by tboot because it makes the system
1807 vulnerable to DMA attacks.
1808 nobounce [Default off]
1809 Disable bounce buffer for untrusted devices such as
1810 the Thunderbolt devices. This will treat the untrusted
1811 devices as the trusted ones, hence might expose security
1812 risks of DMA attacks.
1814 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
1815 0 disables intel_idle and fall back on acpi_idle.
1816 1 to 9 specify maximum depth of C-state.
1820 Do not enable intel_pstate as the default
1821 scaling driver for the supported processors
1823 Use intel_pstate as a scaling driver, but configure it
1824 to work with generic cpufreq governors (instead of
1825 enabling its internal governor). This mode cannot be
1826 used along with the hardware-managed P-states (HWP)
1829 Enable intel_pstate on systems that prohibit it by default
1830 in favor of acpi-cpufreq. Forcing the intel_pstate driver
1831 instead of acpi-cpufreq may disable platform features, such
1832 as thermal controls and power capping, that rely on ACPI
1833 P-States information being indicated to OSPM and therefore
1834 should be used with caution. This option does not work with
1835 processors that aren't supported by the intel_pstate driver
1836 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
1838 Do not enable hardware P state control (HWP)
1841 Only load intel_pstate on systems which support
1842 hardware P state control (HWP) if available.
1844 Enforce ACPI _PPC performance limits. If the Fixed ACPI
1845 Description Table, specifies preferred power management
1846 profile as "Enterprise Server" or "Performance Server",
1847 then this feature is turned on by default.
1849 Allow per-logical-CPU P-State performance control limits using
1850 cpufreq sysfs interface
1852 intremap= [X86-64, Intel-IOMMU]
1853 on enable Interrupt Remapping (default)
1854 off disable Interrupt Remapping
1855 nosid disable Source ID checking
1857 BIOS x2APIC opt-out request will be ignored
1858 nopost disable Interrupt Posting
1860 iomem= Disable strict checking of access to MMIO memory
1861 strict regions from userspace.
1876 nobypass [PPC/POWERNV]
1877 Disable IOMMU bypass, using IOMMU for PCI devices.
1879 iommu.strict= [ARM64] Configure TLB invalidation behaviour
1880 Format: { "0" | "1" }
1882 Request that DMA unmap operations use deferred
1883 invalidation of hardware TLBs, for increased
1884 throughput at the cost of reduced device isolation.
1885 Will fall back to strict mode if not supported by
1886 the relevant IOMMU driver.
1887 1 - Strict mode (default).
1888 DMA unmap operations invalidate IOMMU hardware TLBs
1892 [ARM64, X86] Configure DMA to bypass the IOMMU by default.
1893 Format: { "0" | "1" }
1894 0 - Use IOMMU translation for DMA.
1895 1 - Bypass the IOMMU for DMA.
1896 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH.
1898 io7= [HW] IO7 for Marvel based alpha systems
1899 See comment before marvel_specify_io7 in
1900 arch/alpha/kernel/core_marvel.c.
1902 io_delay= [X86] I/O delay method
1904 Standard port 0x80 based delay
1906 Alternate port 0xed based delay (needed on some systems)
1908 Simple two microseconds delay
1913 See Documentation/admin-guide/nfs/nfsroot.rst.
1915 ipcmni_extend [KNL] Extend the maximum number of unique System V
1916 IPC identifiers from 32,768 to 16,777,216.
1918 irqaffinity= [SMP] Set the default irq affinity mask
1919 The argument is a cpu list, as described above.
1921 irqchip.gicv2_force_probe=
1924 Force the kernel to look for the second 4kB page
1925 of a GICv2 controller even if the memory range
1926 exposed by the device tree is too small.
1928 irqchip.gicv3_nolpi=
1930 Force the kernel to ignore the availability of
1931 LPIs (and by consequence ITSs). Intended for system
1932 that use the kernel as a bootloader, and thus want
1933 to let secondary kernels in charge of setting up
1936 irqchip.gicv3_pseudo_nmi= [ARM64]
1937 Enables support for pseudo-NMIs in the kernel. This
1938 requires the kernel to be built with
1939 CONFIG_ARM64_PSEUDO_NMI.
1942 When an interrupt is not handled search all handlers
1943 for it. Intended to get systems with badly broken
1947 When an interrupt is not handled search all handlers
1948 for it. Also check all handlers each timer
1949 interrupt. Intended to get systems with badly broken
1953 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
1955 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance.
1956 [Deprecated - use cpusets instead]
1957 Format: [flag-list,]<cpu-list>
1959 Specify one or more CPUs to isolate from disturbances
1960 specified in the flag list (default: domain):
1963 Disable the tick when a single task runs.
1965 A residual 1Hz tick is offloaded to workqueues, which you
1966 need to affine to housekeeping through the global
1967 workqueue's affinity configured via the
1968 /sys/devices/virtual/workqueue/cpumask sysfs file, or
1969 by using the 'domain' flag described below.
1971 NOTE: by default the global workqueue runs on all CPUs,
1972 so to protect individual CPUs the 'cpumask' file has to
1973 be configured manually after bootup.
1976 Isolate from the general SMP balancing and scheduling
1977 algorithms. Note that performing domain isolation this way
1978 is irreversible: it's not possible to bring back a CPU to
1979 the domains once isolated through isolcpus. It's strongly
1980 advised to use cpusets instead to disable scheduler load
1981 balancing through the "cpuset.sched_load_balance" file.
1982 It offers a much more flexible interface where CPUs can
1983 move in and out of an isolated set anytime.
1985 You can move a process onto or off an "isolated" CPU via
1986 the CPU affinity syscalls or cpuset.
1987 <cpu number> begins at 0 and the maximum value is
1988 "number of CPUs in system - 1".
1992 Isolate from being targeted by managed interrupts
1993 which have an interrupt mask containing isolated
1994 CPUs. The affinity of managed interrupts is
1995 handled by the kernel and cannot be changed via
1996 the /proc/irq/* interfaces.
1998 This isolation is best effort and only effective
1999 if the automatically assigned interrupt mask of a
2000 device queue contains isolated and housekeeping
2001 CPUs. If housekeeping CPUs are online then such
2002 interrupts are directed to the housekeeping CPU
2003 so that IO submitted on the housekeeping CPU
2004 cannot disturb the isolated CPU.
2006 If a queue's affinity mask contains only isolated
2007 CPUs then this parameter has no effect on the
2008 interrupt routing decision, though interrupts are
2009 only delivered when tasks running on those
2010 isolated CPUs submit IO. IO submitted on
2011 housekeeping CPUs has no influence on those
2014 The format of <cpu-list> is described above.
2018 ivrs_ioapic [HW,X86_64]
2019 Provide an override to the IOAPIC-ID<->DEVICE-ID
2020 mapping provided in the IVRS ACPI table. For
2021 example, to map IOAPIC-ID decimal 10 to
2022 PCI device 00:14.0 write the parameter as:
2023 ivrs_ioapic[10]=00:14.0
2025 ivrs_hpet [HW,X86_64]
2026 Provide an override to the HPET-ID<->DEVICE-ID
2027 mapping provided in the IVRS ACPI table. For
2028 example, to map HPET-ID decimal 0 to
2029 PCI device 00:14.0 write the parameter as:
2030 ivrs_hpet[0]=00:14.0
2032 ivrs_acpihid [HW,X86_64]
2033 Provide an override to the ACPI-HID:UID<->DEVICE-ID
2034 mapping provided in the IVRS ACPI table. For
2035 example, to map UART-HID:UID AMD0020:0 to
2036 PCI device 00:14.5 write the parameter as:
2037 ivrs_acpihid[00:14.5]=AMD0020:0
2039 js= [HW,JOY] Analog joystick
2040 See Documentation/input/joydev/joystick.rst.
2043 When CONFIG_RANDOMIZE_BASE is set, this disables
2044 kernel and module base offset ASLR (Address Space
2045 Layout Randomization).
2048 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print
2049 report on every invalid memory access. Without this
2050 parameter KASAN will print report only for the first
2055 kernelcore= [KNL,X86,IA-64,PPC]
2056 Format: nn[KMGTPE] | nn% | "mirror"
2057 This parameter specifies the amount of memory usable by
2058 the kernel for non-movable allocations. The requested
2059 amount is spread evenly throughout all nodes in the
2060 system as ZONE_NORMAL. The remaining memory is used for
2061 movable memory in its own zone, ZONE_MOVABLE. In the
2062 event, a node is too small to have both ZONE_NORMAL and
2063 ZONE_MOVABLE, kernelcore memory will take priority and
2064 other nodes will have a larger ZONE_MOVABLE.
2066 ZONE_MOVABLE is used for the allocation of pages that
2067 may be reclaimed or moved by the page migration
2068 subsystem. Note that allocations like PTEs-from-HighMem
2069 still use the HighMem zone if it exists, and the Normal
2070 zone if it does not.
2072 It is possible to specify the exact amount of memory in
2073 the form of "nn[KMGTPE]", a percentage of total system
2074 memory in the form of "nn%", or "mirror". If "mirror"
2075 option is specified, mirrored (reliable) memory is used
2076 for non-movable allocations and remaining memory is used
2077 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror"
2078 are exclusive, so you cannot specify multiple forms.
2080 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
2081 Format: <Controller#>[,poll interval]
2082 The controller # is the number of the ehci usb debug
2083 port as it is probed via PCI. The poll interval is
2084 optional and is the number seconds in between
2085 each poll cycle to the debug port in case you need
2086 the functionality for interrupting the kernel with
2087 gdb or control-c on the dbgp connection. When
2088 not using this parameter you use sysrq-g to break into
2089 the kernel debugger.
2091 kgdboc= [KGDB,HW] kgdb over consoles.
2092 Requires a tty driver that supports console polling,
2093 or a supported polling keyboard driver (non-usb).
2094 Serial only format: <serial_device>[,baud]
2095 keyboard only format: kbd
2096 keyboard and serial format: kbd,<serial_device>[,baud]
2097 Optional Kernel mode setting:
2098 kms, kbd format: kms,kbd
2099 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
2101 kgdbwait [KGDB] Stop kernel execution and enter the
2102 kernel debugger at the earliest opportunity.
2104 kmac= [MIPS] korina ethernet MAC address.
2105 Configure the RouterBoard 532 series on-chip
2106 Ethernet adapter MAC address.
2108 kmemleak= [KNL] Boot-time kmemleak enable/disable
2109 Valid arguments: on, off
2111 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
2114 kprobe_event=[probe-list]
2115 [FTRACE] Add kprobe events and enable at boot time.
2116 The probe-list is a semicolon delimited list of probe
2117 definitions. Each definition is same as kprobe_events
2118 interface, but the parameters are comma delimited.
2119 For example, to add a kprobe event on vfs_read with
2120 arg1 and arg2, add to the command line;
2122 kprobe_event=p,vfs_read,$arg1,$arg2
2124 See also Documentation/trace/kprobetrace.rst "Kernel
2125 Boot Parameter" section.
2127 kpti= [ARM64] Control page table isolation of user
2128 and kernel address spaces.
2129 Default: enabled on cores which need mitigation.
2133 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
2134 Default is 0 (don't ignore, but inject #GP)
2136 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface.
2137 Default is false (don't support).
2139 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit
2144 [KVM] Controls the software workaround for the
2145 X86_BUG_ITLB_MULTIHIT bug.
2146 force : Always deploy workaround.
2147 off : Never deploy workaround.
2148 auto : Deploy workaround based on the presence of
2149 X86_BUG_ITLB_MULTIHIT.
2153 If the software workaround is enabled for the host,
2154 guests do need not to enable it for nested guests.
2156 kvm.nx_huge_pages_recovery_ratio=
2157 [KVM] Controls how many 4KiB pages are periodically zapped
2158 back to huge pages. 0 disables the recovery, otherwise if
2159 the value is N KVM will zap 1/Nth of the 4KiB pages every
2160 minute. The default is 60.
2162 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
2163 Default is 1 (enabled)
2165 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
2167 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
2169 kvm-arm.vgic_v3_group0_trap=
2170 [KVM,ARM] Trap guest accesses to GICv3 group-0
2173 kvm-arm.vgic_v3_group1_trap=
2174 [KVM,ARM] Trap guest accesses to GICv3 group-1
2177 kvm-arm.vgic_v3_common_trap=
2178 [KVM,ARM] Trap guest accesses to GICv3 common
2181 kvm-arm.vgic_v4_enable=
2182 [KVM,ARM] Allow use of GICv4 for direct injection of
2185 kvm-intel.ept= [KVM,Intel] Disable extended page tables
2186 (virtualized MMU) support on capable Intel chips.
2187 Default is 1 (enabled)
2189 kvm-intel.emulate_invalid_guest_state=
2190 [KVM,Intel] Enable emulation of invalid guest states
2191 Default is 0 (disabled)
2193 kvm-intel.flexpriority=
2194 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
2195 Default is 1 (enabled)
2198 [KVM,Intel] Enable VMX nesting (nVMX).
2199 Default is 0 (disabled)
2201 kvm-intel.unrestricted_guest=
2202 [KVM,Intel] Disable unrestricted guest feature
2203 (virtualized real and unpaged mode) on capable
2204 Intel chips. Default is 1 (enabled)
2206 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault
2209 Valid arguments: never, cond, always
2211 always: L1D cache flush on every VMENTER.
2212 cond: Flush L1D on VMENTER only when the code between
2213 VMEXIT and VMENTER can leak host memory.
2214 never: Disables the mitigation
2216 Default is cond (do L1 cache flush in specific instances)
2218 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
2219 feature (tagged TLBs) on capable Intel chips.
2220 Default is 1 (enabled)
2222 l1tf= [X86] Control mitigation of the L1TF vulnerability on
2225 The kernel PTE inversion protection is unconditionally
2226 enabled and cannot be disabled.
2229 Provides all available mitigations for the
2230 L1TF vulnerability. Disables SMT and
2231 enables all mitigations in the
2232 hypervisors, i.e. unconditional L1D flush.
2234 SMT control and L1D flush control via the
2235 sysfs interface is still possible after
2236 boot. Hypervisors will issue a warning
2237 when the first VM is started in a
2238 potentially insecure configuration,
2239 i.e. SMT enabled or L1D flush disabled.
2242 Same as 'full', but disables SMT and L1D
2243 flush runtime control. Implies the
2244 'nosmt=force' command line option.
2245 (i.e. sysfs control of SMT is disabled.)
2248 Leaves SMT enabled and enables the default
2249 hypervisor mitigation, i.e. conditional
2252 SMT control and L1D flush control via the
2253 sysfs interface is still possible after
2254 boot. Hypervisors will issue a warning
2255 when the first VM is started in a
2256 potentially insecure configuration,
2257 i.e. SMT enabled or L1D flush disabled.
2261 Disables SMT and enables the default
2262 hypervisor mitigation.
2264 SMT control and L1D flush control via the
2265 sysfs interface is still possible after
2266 boot. Hypervisors will issue a warning
2267 when the first VM is started in a
2268 potentially insecure configuration,
2269 i.e. SMT enabled or L1D flush disabled.
2272 Same as 'flush', but hypervisors will not
2273 warn when a VM is started in a potentially
2274 insecure configuration.
2277 Disables hypervisor mitigations and doesn't
2279 It also drops the swap size and available
2280 RAM limit restriction on both hypervisor and
2285 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst
2291 lapic [X86-32,APIC] Enable the local APIC even if BIOS
2294 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline
2295 value for LAPIC timer one-shot implementation. Default
2296 back to the programmable timer unit in the LAPIC.
2298 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
2301 libata.dma= [LIBATA] DMA control
2302 libata.dma=0 Disable all PATA and SATA DMA
2303 libata.dma=1 PATA and SATA Disk DMA only
2304 libata.dma=2 ATAPI (CDROM) DMA only
2305 libata.dma=4 Compact Flash DMA only
2306 Combinations also work, so libata.dma=3 enables DMA
2307 for disks and CDROMs, but not CFs.
2309 libata.ignore_hpa= [LIBATA] Ignore HPA limit
2310 libata.ignore_hpa=0 keep BIOS limits (default)
2311 libata.ignore_hpa=1 ignore limits, using full disk
2313 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
2317 libata.force= [LIBATA] Force configurations. The format is comma
2318 separated list of "[ID:]VAL" where ID is
2319 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
2320 matching port, link or device. Basically, it matches
2321 the ATA ID string printed on console by libata. If
2322 the whole ID part is omitted, the last PORT and DEVICE
2323 values are used. If ID hasn't been specified yet, the
2324 configuration applies to all ports, links and devices.
2326 If only DEVICE is omitted, the parameter applies to
2327 the port and all links and devices behind it. DEVICE
2328 number of 0 either selects the first device or the
2329 first fan-out link behind PMP device. It does not
2330 select the host link. DEVICE number of 15 selects the
2331 host link and device attached to it.
2333 The VAL specifies the configuration to force. As long
2334 as there's no ambiguity shortcut notation is allowed.
2335 For example, both 1.5 and 1.5G would work for 1.5Gbps.
2336 The following configurations can be forced.
2338 * Cable type: 40c, 80c, short40c, unk, ign or sata.
2339 Any ID with matching PORT is used.
2341 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
2343 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
2344 udma[/][16,25,33,44,66,100,133] notation is also
2347 * [no]ncq: Turn on or off NCQ.
2349 * [no]ncqtrim: Turn off queued DSM TRIM.
2351 * nohrst, nosrst, norst: suppress hard, soft
2354 * rstonce: only attempt one reset during
2355 hot-unplug link recovery
2357 * dump_id: dump IDENTIFY data.
2359 * atapi_dmadir: Enable ATAPI DMADIR bridge support
2361 * disable: Disable this device.
2363 If there are multiple matching configurations changing
2364 the same attribute, the last one is used.
2366 memblock=debug [KNL] Enable memblock debug messages.
2368 load_ramdisk= [RAM] List of ramdisks to load from floppy
2369 See Documentation/admin-guide/blockdev/ramdisk.rst.
2371 lockd.nlm_grace_period=P [NFS] Assign grace period.
2374 lockd.nlm_tcpport=N [NFS] Assign TCP port.
2377 lockd.nlm_timeout=T [NFS] Assign timeout value.
2380 lockd.nlm_udpport=M [NFS] Assign UDP port.
2383 lockdown= [SECURITY]
2384 { integrity | confidentiality }
2385 Enable the kernel lockdown feature. If set to
2386 integrity, kernel features that allow userland to
2387 modify the running kernel are disabled. If set to
2388 confidentiality, kernel features that allow userland
2389 to extract confidential information from the kernel
2392 locktorture.nreaders_stress= [KNL]
2393 Set the number of locking read-acquisition kthreads.
2394 Defaults to being automatically set based on the
2395 number of online CPUs.
2397 locktorture.nwriters_stress= [KNL]
2398 Set the number of locking write-acquisition kthreads.
2400 locktorture.onoff_holdoff= [KNL]
2401 Set time (s) after boot for CPU-hotplug testing.
2403 locktorture.onoff_interval= [KNL]
2404 Set time (s) between CPU-hotplug operations, or
2405 zero to disable CPU-hotplug testing.
2407 locktorture.shuffle_interval= [KNL]
2408 Set task-shuffle interval (jiffies). Shuffling
2409 tasks allows some CPUs to go into dyntick-idle
2410 mode during the locktorture test.
2412 locktorture.shutdown_secs= [KNL]
2413 Set time (s) after boot system shutdown. This
2414 is useful for hands-off automated testing.
2416 locktorture.stat_interval= [KNL]
2417 Time (s) between statistics printk()s.
2419 locktorture.stutter= [KNL]
2420 Time (s) to stutter testing, for example,
2421 specifying five seconds causes the test to run for
2422 five seconds, wait for five seconds, and so on.
2423 This tests the locking primitive's ability to
2424 transition abruptly to and from idle.
2426 locktorture.torture_type= [KNL]
2427 Specify the locking implementation to test.
2429 locktorture.verbose= [KNL]
2430 Enable additional printk() statements.
2432 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
2435 loglevel= All Kernel Messages with a loglevel smaller than the
2436 console loglevel will be printed to the console. It can
2437 also be changed with klogd or other programs. The
2438 loglevels are defined as follows:
2440 0 (KERN_EMERG) system is unusable
2441 1 (KERN_ALERT) action must be taken immediately
2442 2 (KERN_CRIT) critical conditions
2443 3 (KERN_ERR) error conditions
2444 4 (KERN_WARNING) warning conditions
2445 5 (KERN_NOTICE) normal but significant condition
2446 6 (KERN_INFO) informational
2447 7 (KERN_DEBUG) debug-level messages
2449 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
2450 in bytes. n must be a power of two and greater
2451 than the minimal size. The minimal size is defined
2452 by LOG_BUF_SHIFT kernel config parameter. There is
2453 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
2454 that allows to increase the default size depending on
2455 the number of CPUs. See init/Kconfig for more details.
2457 logo.nologo [FB] Disables display of the built-in Linux logo.
2458 This may be used to provide more screen space for
2459 kernel log messages and is useful when debugging
2460 kernel boot problems.
2462 lp=0 [LP] Specify parallel ports to use, e.g,
2463 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
2464 lp=reset first parallel port). 'lp=0' disables the
2465 lp=auto printer driver. 'lp=reset' (which can be
2466 specified in addition to the ports) causes
2467 attached printers to be reset. Using
2468 lp=port1,port2,... specifies the parallel ports
2469 to associate lp devices with, starting with
2470 lp0. A port specification may be 'none' to skip
2471 that lp device, or a parport name such as
2472 'parport0'. Specifying 'lp=auto' instead of a
2473 port specification list means that device IDs
2474 from each port should be examined, to see if
2475 an IEEE 1284-compliant printer is attached; if
2476 so, the driver will manage that printer.
2477 See also header of drivers/char/lp.c.
2480 Sets loops_per_jiffy to given constant, thus avoiding
2481 time-consuming boot-time autodetection (up to 250 ms per
2482 CPU). 0 enables autodetection (default). To determine
2483 the correct value for your kernel, boot with normal
2484 autodetection and see what value is printed. Note that
2485 on SMP systems the preset will be applied to all CPUs,
2486 which is likely to cause problems if your CPUs need
2487 significantly divergent settings. An incorrect value
2488 will cause delays in the kernel to be wrong, leading to
2489 unpredictable I/O errors and other breakage. Although
2490 unlikely, in the extreme case this might damage your
2494 Format: <io>,<irq>,<dma>
2496 lsm.debug [SECURITY] Enable LSM initialization debugging output.
2499 [SECURITY] Choose order of LSM initialization. This
2500 overrides CONFIG_LSM, and the "security=" parameter.
2502 machvec= [IA-64] Force the use of a particular machine-vector
2503 (machvec) in a generic kernel.
2504 Example: machvec=hpzx1
2506 machtype= [Loongson] Share the same kernel image file between different
2508 Example: machtype=lemote-yeeloong-2f-7inch
2510 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
2511 than or equal to this physical address is ignored.
2513 maxcpus= [SMP] Maximum number of processors that an SMP kernel
2514 will bring up during bootup. maxcpus=n : n >= 0 limits
2515 the kernel to bring up 'n' processors. Surely after
2516 bootup you can bring up the other plugged cpu by executing
2517 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus
2518 only takes effect during system bootup.
2519 While n=0 is a special case, it is equivalent to "nosmp",
2520 which also disables the IO APIC.
2522 max_loop= [LOOP] The number of loop block devices that get
2523 (loop.max_loop) unconditionally pre-created at init time. The default
2524 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
2525 of statically allocating a predefined number, loop
2526 devices can be requested on-demand with the
2527 /dev/loop-control interface.
2529 mce [X86-32] Machine Check Exception
2531 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.rst
2533 md= [HW] RAID subsystems devices and level
2534 See Documentation/admin-guide/md.rst.
2537 Format: <first>,<last>
2538 Specifies range of consoles to be captured by the MDA.
2541 Control mitigation for the Micro-architectural Data
2542 Sampling (MDS) vulnerability.
2544 Certain CPUs are vulnerable to an exploit against CPU
2545 internal buffers which can forward information to a
2546 disclosure gadget under certain conditions.
2548 In vulnerable processors, the speculatively
2549 forwarded data can be used in a cache side channel
2550 attack, to access data to which the attacker does
2551 not have direct access.
2553 This parameter controls the MDS mitigation. The
2556 full - Enable MDS mitigation on vulnerable CPUs
2557 full,nosmt - Enable MDS mitigation and disable
2558 SMT on vulnerable CPUs
2559 off - Unconditionally disable MDS mitigation
2561 On TAA-affected machines, mds=off can be prevented by
2562 an active TAA mitigation as both vulnerabilities are
2563 mitigated with the same mechanism so in order to disable
2564 this mitigation, you need to specify tsx_async_abort=off
2567 Not specifying this option is equivalent to
2570 For details see: Documentation/admin-guide/hw-vuln/mds.rst
2572 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
2573 Amount of memory to be used when the kernel is not able
2574 to see the whole system memory or for test.
2575 [X86] Work as limiting max address. Use together
2576 with memmap= to avoid physical address space collisions.
2577 Without memmap= PCI devices could be placed at addresses
2578 belonging to unused RAM.
2580 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
2584 [KNL,SH] Allow user to override the default size for
2585 per-device physically contiguous DMA buffers.
2587 memhp_default_state=online/offline
2588 [KNL] Set the initial state for the memory hotplug
2589 onlining policy. If not specified, the default value is
2590 set according to the
2591 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config
2593 See Documentation/admin-guide/mm/memory-hotplug.rst.
2595 memmap=exactmap [KNL,X86] Enable setting of an exact
2596 E820 memory map, as specified by the user.
2597 Such memmap=exactmap lines can be constructed based on
2598 BIOS output or other requirements. See the memmap=nn@ss
2601 memmap=nn[KMG]@ss[KMG]
2602 [KNL] Force usage of a specific region of memory.
2603 Region of memory to be used is from ss to ss+nn.
2604 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG],
2605 which limits max address to nn[KMG].
2606 Multiple different regions can be specified,
2609 memmap=100M@2G,100M#3G,1G!1024G
2611 memmap=nn[KMG]#ss[KMG]
2612 [KNL,ACPI] Mark specific memory as ACPI data.
2613 Region of memory to be marked is from ss to ss+nn.
2615 memmap=nn[KMG]$ss[KMG]
2616 [KNL,ACPI] Mark specific memory as reserved.
2617 Region of memory to be reserved is from ss to ss+nn.
2618 Example: Exclude memory from 0x18690000-0x1869ffff
2619 memmap=64K$0x18690000
2621 memmap=0x10000$0x18690000
2622 Some bootloaders may need an escape character before '$',
2623 like Grub2, otherwise '$' and the following number
2626 memmap=nn[KMG]!ss[KMG]
2627 [KNL,X86] Mark specific memory as protected.
2628 Region of memory to be used, from ss to ss+nn.
2629 The memory region may be marked as e820 type 12 (0xc)
2630 and is NVDIMM or ADR memory.
2632 memmap=<size>%<offset>-<oldtype>+<newtype>
2633 [KNL,ACPI] Convert memory within the specified region
2634 from <oldtype> to <newtype>. If "-<oldtype>" is left
2635 out, the whole region will be marked as <newtype>,
2636 even if previously unavailable. If "+<newtype>" is left
2637 out, matching memory will be removed. Types are
2638 specified as e820 types, e.g., 1 = RAM, 2 = reserved,
2639 3 = ACPI, 12 = PRAM.
2641 memory_corruption_check=0/1 [X86]
2642 Some BIOSes seem to corrupt the first 64k of
2643 memory when doing things like suspend/resume.
2644 Setting this option will scan the memory
2645 looking for corruption. Enabling this will
2646 both detect corruption and prevent the kernel
2647 from using the memory being corrupted.
2648 However, its intended as a diagnostic tool; if
2649 repeatable BIOS-originated corruption always
2650 affects the same memory, you can use memmap=
2651 to prevent the kernel from using that memory.
2653 memory_corruption_check_size=size [X86]
2654 By default it checks for corruption in the low
2655 64k, making this memory unavailable for normal
2656 use. Use this parameter to scan for
2657 corruption in more or less memory.
2659 memory_corruption_check_period=seconds [X86]
2660 By default it checks for corruption every 60
2661 seconds. Use this parameter to check at some
2662 other rate. 0 disables periodic checking.
2664 memtest= [KNL,X86,ARM,PPC] Enable memtest
2666 default : 0 <disable>
2667 Specifies the number of memtest passes to be
2668 performed. Each pass selects another test
2669 pattern from a given set of patterns. Memtest
2670 fills the memory with this pattern, validates
2671 memory contents and reserves bad memory
2672 regions that are detected.
2674 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control
2675 Valid arguments: on, off
2676 Default (depends on kernel configuration option):
2677 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y)
2678 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n)
2679 mem_encrypt=on: Activate SME
2680 mem_encrypt=off: Do not activate SME
2682 Refer to Documentation/virt/kvm/amd-memory-encryption.rst
2683 for details on when memory encryption can be activated.
2685 mem_sleep_default= [SUSPEND] Default system suspend mode:
2686 s2idle - Suspend-To-Idle
2687 shallow - Power-On Suspend or equivalent (if supported)
2688 deep - Suspend-To-RAM or equivalent (if supported)
2689 See Documentation/admin-guide/pm/sleep-states.rst.
2691 meye.*= [HW] Set MotionEye Camera parameters
2692 See Documentation/media/v4l-drivers/meye.rst.
2694 mfgpt_irq= [IA-32] Specify the IRQ to use for the
2695 Multi-Function General Purpose Timers on AMD Geode
2698 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
2699 the BIOS has incorrectly applied a workaround. TinyBIOS
2700 version 0.98 is known to be affected, 0.99 fixes the
2701 problem by letting the user disable the workaround.
2705 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
2706 physical address is ignored.
2708 mini2440= [ARM,HW,KNL]
2709 Format:[0..2][b][c][t]
2711 MINI2440 configuration specification:
2712 0 - The attached screen is the 3.5" TFT
2713 1 - The attached screen is the 7" TFT
2714 2 - The VGA Shield is attached (1024x768)
2715 Leaving out the screen size parameter will not load
2716 the TFT driver, and the framebuffer will be left
2718 b - Enable backlight. The TFT backlight pin will be
2719 linked to the kernel VESA blanking code and a GPIO
2720 LED. This parameter is not necessary when using the
2722 c - Enable the s3c camera interface.
2723 t - Reserved for enabling touchscreen support. The
2724 touchscreen support is not enabled in the mainstream
2725 kernel as of 2.6.30, a preliminary port can be found
2726 in the "bleeding edge" mini2440 support kernel at
2727 http://repo.or.cz/w/linux-2.6/mini2440.git
2730 [X86,PPC,S390,ARM64] Control optional mitigations for
2731 CPU vulnerabilities. This is a set of curated,
2732 arch-independent options, each of which is an
2733 aggregation of existing arch-specific options.
2736 Disable all optional CPU mitigations. This
2737 improves system performance, but it may also
2738 expose users to several CPU vulnerabilities.
2739 Equivalent to: nopti [X86,PPC]
2741 nospectre_v1 [X86,PPC]
2743 nospectre_v2 [X86,PPC,S390,ARM64]
2744 spectre_v2_user=off [X86]
2745 spec_store_bypass_disable=off [X86,PPC]
2746 ssbd=force-off [ARM64]
2749 tsx_async_abort=off [X86]
2750 kvm.nx_huge_pages=off [X86]
2753 This does not have any effect on
2754 kvm.nx_huge_pages when
2755 kvm.nx_huge_pages=force.
2758 Mitigate all CPU vulnerabilities, but leave SMT
2759 enabled, even if it's vulnerable. This is for
2760 users who don't want to be surprised by SMT
2761 getting disabled across kernel upgrades, or who
2762 have other ways of avoiding SMT-based attacks.
2763 Equivalent to: (default behavior)
2766 Mitigate all CPU vulnerabilities, disabling SMT
2767 if needed. This is for users who always want to
2768 be fully mitigated, even if it means losing SMT.
2769 Equivalent to: l1tf=flush,nosmt [X86]
2770 mds=full,nosmt [X86]
2771 tsx_async_abort=full,nosmt [X86]
2774 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
2775 parameter allows control of the logging verbosity for
2776 the additional memory initialisation checks. A value
2777 of 0 disables mminit logging and a level of 4 will
2778 log everything. Information is printed at KERN_DEBUG
2779 so loglevel=8 may also need to be specified.
2782 [KNL] When CONFIG_MODULE_SIG is set, this means that
2783 modules without (valid) signatures will fail to load.
2784 Note that if CONFIG_MODULE_SIG_FORCE is set, that
2785 is always true, so this option does nothing.
2787 module_blacklist= [KNL] Do not load a comma-separated list of
2788 modules. Useful for debugging problem modules.
2791 [MOUSE] Maximum time between finger touching and
2792 leaving touchpad surface for touch to be considered
2793 a tap and be reported as a left button click (for
2794 touchpads working in absolute mode only).
2796 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
2797 reporting absolute coordinates, such as tablets
2798 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
2799 reporting absolute coordinates, such as tablets
2801 movablecore= [KNL,X86,IA-64,PPC]
2802 Format: nn[KMGTPE] | nn%
2803 This parameter is the complement to kernelcore=, it
2804 specifies the amount of memory used for migratable
2805 allocations. If both kernelcore and movablecore is
2806 specified, then kernelcore will be at *least* the
2807 specified value but may be more. If movablecore on its
2808 own is specified, the administrator must be careful
2809 that the amount of memory usable for all allocations
2812 movable_node [KNL] Boot-time switch to make hotplugable memory
2813 NUMA nodes to be movable. This means that the memory
2814 of such nodes will be usable only for movable
2815 allocations which rules out almost all kernel
2816 allocations. Use with caution!
2818 MTD_Partition= [MTD]
2819 Format: <name>,<region-number>,<size>,<offset>
2821 MTD_Region= [MTD] Format:
2822 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
2825 See drivers/mtd/parsers/cmdlinepart.c
2827 multitce=off [PPC] This parameter disables the use of the pSeries
2828 firmware feature for updating multiple TCE entries
2831 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
2833 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
2835 boundary - index of last SLC block on Flex-OneNAND.
2836 The remaining blocks are configured as MLC blocks.
2837 lock - Configure if Flex-OneNAND boundary should be locked.
2838 Once locked, the boundary cannot be changed.
2839 1 indicates lock status, 0 indicates unlock status.
2842 ARM/S3C2412 JIVE boot control
2844 See arch/arm/mach-s3c2412/mach-jive.c
2846 mtouchusb.raw_coordinates=
2847 [HW] Make the MicroTouch USB driver use raw coordinates
2848 ('y', default) or cooked coordinates ('n')
2850 mtrr_chunk_size=nn[KMG] [X86]
2851 used for mtrr cleanup. It is largest continuous chunk
2852 that could hold holes aka. UC entries.
2854 mtrr_gran_size=nn[KMG] [X86]
2855 Used for mtrr cleanup. It is granularity of mtrr block.
2857 Large value could prevent small alignment from
2860 mtrr_spare_reg_nr=n [X86]
2862 Range: 0,7 : spare reg number
2864 Used for mtrr cleanup. It is spare mtrr entries number.
2865 Set to 2 or more if your graphical card needs more.
2867 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
2869 netdev= [NET] Network devices parameters
2870 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
2871 Note that mem_start is often overloaded to mean
2872 something different and driver-specific.
2873 This usage is only documented in each driver source
2877 [NETFILTER] Enable connection tracking flow accounting
2878 0 to disable accounting
2879 1 to enable accounting
2882 nfsaddrs= [NFS] Deprecated. Use ip= instead.
2883 See Documentation/admin-guide/nfs/nfsroot.rst.
2885 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
2886 See Documentation/admin-guide/nfs/nfsroot.rst.
2888 nfsrootdebug [NFS] enable nfsroot debugging messages.
2889 See Documentation/admin-guide/nfs/nfsroot.rst.
2891 nfs.callback_nr_threads=
2892 [NFSv4] set the total number of threads that the
2893 NFS client will assign to service NFSv4 callback
2896 nfs.callback_tcpport=
2897 [NFS] set the TCP port on which the NFSv4 callback
2898 channel should listen.
2901 [NFS] sets the pathname to the program which is used
2902 to update the NFS client cache entries.
2904 nfs.cache_getent_timeout=
2905 [NFS] sets the timeout after which an attempt to
2906 update a cache entry is deemed to have failed.
2908 nfs.idmap_cache_timeout=
2909 [NFS] set the maximum lifetime for idmapper cache
2913 [NFS] enable 64-bit inode numbers.
2914 If zero, the NFS client will fake up a 32-bit inode
2915 number for the readdir() and stat() syscalls instead
2916 of returning the full 64-bit number.
2917 The default is to return 64-bit inode numbers.
2919 nfs.max_session_cb_slots=
2920 [NFSv4.1] Sets the maximum number of session
2921 slots the client will assign to the callback
2922 channel. This determines the maximum number of
2923 callbacks the client will process in parallel for
2924 a particular server.
2926 nfs.max_session_slots=
2927 [NFSv4.1] Sets the maximum number of session slots
2928 the client will attempt to negotiate with the server.
2929 This limits the number of simultaneous RPC requests
2930 that the client can send to the NFSv4.1 server.
2931 Note that there is little point in setting this
2932 value higher than the max_tcp_slot_table_limit.
2934 nfs.nfs4_disable_idmapping=
2935 [NFSv4] When set to the default of '1', this option
2936 ensures that both the RPC level authentication
2937 scheme and the NFS level operations agree to use
2938 numeric uids/gids if the mount is using the
2939 'sec=sys' security flavour. In effect it is
2940 disabling idmapping, which can make migration from
2941 legacy NFSv2/v3 systems to NFSv4 easier.
2942 Servers that do not support this mode of operation
2943 will be autodetected by the client, and it will fall
2944 back to using the idmapper.
2945 To turn off this behaviour, set the value to '0'.
2947 [NFS4] Specify an additional fixed unique ident-
2948 ification string that NFSv4 clients can insert into
2949 their nfs_client_id4 string. This is typically a
2950 UUID that is generated at system install time.
2952 nfs.send_implementation_id =
2953 [NFSv4.1] Send client implementation identification
2954 information in exchange_id requests.
2955 If zero, no implementation identification information
2957 The default is to send the implementation identification
2960 nfs.recover_lost_locks =
2961 [NFSv4] Attempt to recover locks that were lost due
2962 to a lease timeout on the server. Please note that
2963 doing this risks data corruption, since there are
2964 no guarantees that the file will remain unchanged
2965 after the locks are lost.
2966 If you want to enable the kernel legacy behaviour of
2967 attempting to recover these locks, then set this
2969 The default parameter value of '0' causes the kernel
2970 not to attempt recovery of lost locks.
2972 nfs4.layoutstats_timer =
2973 [NFSv4.2] Change the rate at which the kernel sends
2974 layoutstats to the pNFS metadata server.
2976 Setting this to value to 0 causes the kernel to use
2977 whatever value is the default set by the layout
2978 driver. A non-zero value sets the minimum interval
2979 in seconds between layoutstats transmissions.
2981 nfsd.nfs4_disable_idmapping=
2982 [NFSv4] When set to the default of '1', the NFSv4
2983 server will return only numeric uids and gids to
2984 clients using auth_sys, and will accept numeric uids
2985 and gids from such clients. This is intended to ease
2986 migration from NFSv2/v3.
2988 nmi_debug= [KNL,SH] Specify one or more actions to take
2989 when a NMI is triggered.
2990 Format: [state][,regs][,debounce][,die]
2992 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
2993 Format: [panic,][nopanic,][num]
2995 0 - turn hardlockup detector in nmi_watchdog off
2996 1 - turn hardlockup detector in nmi_watchdog on
2997 When panic is specified, panic when an NMI watchdog
2998 timeout occurs (or 'nopanic' to not panic on an NMI
2999 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set)
3000 To disable both hard and soft lockup detectors,
3001 please see 'nowatchdog'.
3002 This is useful when you use a panic=... timeout and
3003 need the box quickly up again.
3005 These settings can be accessed at runtime via
3006 the nmi_watchdog and hardlockup_panic sysctls.
3008 netpoll.carrier_timeout=
3009 [NET] Specifies amount of time (in seconds) that
3010 netpoll should wait for a carrier. By default netpoll
3013 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
3014 emulation library even if a 387 maths coprocessor
3017 no5lvl [X86-64] Disable 5-level paging mode. Forces
3018 kernel to use 4-level paging instead.
3021 [HW] Never suspend the console
3022 Disable suspending of consoles during suspend and
3023 hibernate operations. Once disabled, debugging
3024 messages can reach various consoles while the rest
3025 of the system is being put to sleep (ie, while
3026 debugging driver suspend/resume hooks). This may
3027 not work reliably with all consoles, but is known
3028 to work with serial and VGA consoles.
3029 To facilitate more flexible debugging, we also add
3030 console_suspend, a printk module parameter to control
3031 it. Users could use console_suspend (usually
3032 /sys/module/printk/parameters/console_suspend) to
3033 turn on/off it dynamically.
3035 novmcoredd [KNL,KDUMP]
3036 Disable device dump. Device dump allows drivers to
3037 append dump data to vmcore so you can collect driver
3038 specified debug info. Drivers can append the data
3039 without any limit and this data is stored in memory,
3040 so this may cause significant memory stress. Disabling
3041 device dump can help save memory but the driver debug
3042 data will be no longer available. This parameter
3043 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP
3046 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
3047 caches in the slab allocator. Saves per-node memory,
3048 but will impact performance.
3052 noaltinstr [S390] Disables alternative instructions patching
3053 (CPU alternatives feature).
3055 noapic [SMP,APIC] Tells the kernel to not make use of any
3056 IOAPICs that may be present in the system.
3058 noautogroup Disable scheduler automatic task group creation.
3060 nobats [PPC] Do not use BATs for mapping kernel lowmem
3061 on "Classic" PPC cores.
3065 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
3067 nodelayacct [KNL] Disable per-task delay accounting
3069 nodsp [SH] Disable hardware DSP at boot time.
3071 noefi Disable EFI runtime services support.
3076 On X86-32 available only on PAE configured kernels.
3077 noexec=on: enable non-executable mappings (default)
3078 noexec=off: disable non-executable mappings
3081 Disable SMAP (Supervisor Mode Access Prevention)
3082 even if it is supported by processor.
3085 Disable SMEP (Supervisor Mode Execution Prevention)
3086 even if it is supported by processor.
3089 This affects only 32-bit executables.
3090 noexec32=on: enable non-executable mappings (default)
3091 read doesn't imply executable mappings
3092 noexec32=off: disable non-executable mappings
3093 read implies executable mappings
3095 nofpu [MIPS,SH] Disable hardware FPU at boot time.
3097 nofxsr [BUGS=X86-32] Disables x86 floating point extended
3098 register save and restore. The kernel will only save
3099 legacy floating-point registers on task switch.
3101 nohugeiomap [KNL,x86,PPC] Disable kernel huge I/O mappings.
3103 nosmt [KNL,S390] Disable symmetric multithreading (SMT).
3104 Equivalent to smt=1.
3106 [KNL,x86] Disable symmetric multithreading (SMT).
3107 nosmt=force: Force disable SMT, cannot be undone
3108 via the sysfs control file.
3110 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1
3111 (bounds check bypass). With this option data leaks are
3112 possible in the system.
3114 nospectre_v2 [X86,PPC_FSL_BOOK3E,ARM64] Disable all mitigations for
3115 the Spectre variant 2 (indirect branch prediction)
3116 vulnerability. System may allow data leaks with this
3119 nospec_store_bypass_disable
3120 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability
3122 noxsave [BUGS=X86] Disables x86 extended register state save
3123 and restore using xsave. The kernel will fallback to
3124 enabling legacy floating-point and sse state.
3126 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
3127 register states. The kernel will fall back to use
3128 xsave to save the states. By using this parameter,
3129 performance of saving the states is degraded because
3130 xsave doesn't support modified optimization while
3131 xsaveopt supports it on xsaveopt enabled systems.
3133 noxsaves [X86] Disables xsaves and xrstors used in saving and
3134 restoring x86 extended register state in compacted
3135 form of xsave area. The kernel will fall back to use
3136 xsaveopt and xrstor to save and restore the states
3137 in standard form of xsave area. By using this
3138 parameter, xsave area per process might occupy more
3139 memory on xsaves enabled systems.
3141 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
3142 wfi(ARM) instruction doesn't work correctly and not to
3143 use it. This is also useful when using JTAG debugger.
3145 no_file_caps Tells the kernel not to honor file capabilities. The
3146 only way then for a file to be executed with privilege
3147 is to be setuid root or executed by root.
3149 nohalt [IA-64] Tells the kernel not to use the power saving
3150 function PAL_HALT_LIGHT when idle. This increases
3151 power-consumption. On the positive side, it reduces
3152 interrupt wake-up latency, which may improve performance
3153 in certain environments such as networked servers or
3156 nohibernate [HIBERNATION] Disable hibernation and resume.
3158 nohz= [KNL] Boottime enable/disable dynamic ticks
3159 Valid arguments: on, off
3162 nohz_full= [KNL,BOOT,SMP,ISOL]
3163 The argument is a cpu list, as described above.
3164 In kernels built with CONFIG_NO_HZ_FULL=y, set
3165 the specified list of CPUs whose tick will be stopped
3166 whenever possible. The boot CPU will be forced outside
3167 the range to maintain the timekeeping. Any CPUs
3168 in this list will have their RCU callbacks offloaded,
3169 just as if they had also been called out in the
3170 rcu_nocbs= boot parameter.
3172 noiotrap [SH] Disables trapped I/O port accesses.
3174 noirqdebug [X86-32] Disables the code which attempts to detect and
3175 disable unhandled interrupt sources.
3177 no_timer_check [X86,APIC] Disables the code which tests for
3178 broken timer IRQ sources.
3180 noisapnp [ISAPNP] Disables ISA PnP code.
3182 noinitrd [RAM] Tells the kernel not to load any configured
3185 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
3187 [Deprecated - use intremap=off]
3191 noinvpcid [X86] Disable the INVPCID cpu feature.
3193 nojitter [IA-64] Disables jitter checking for ITC timers.
3195 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
3197 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
3201 [X86,PV_OPS] Disable paravirtualized VMware scheduler
3202 clock and use the default one.
3204 no-steal-acc [X86,KVM,ARM64] Disable paravirtualized steal time
3205 accounting. steal time is computed, but won't
3206 influence scheduler behaviour
3208 nolapic [X86-32,APIC] Do not enable or use the local APIC.
3210 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
3212 noltlbs [PPC] Do not use large page/tlb entries for kernel
3213 lowmem mapping on PPC40x and PPC8xx
3215 nomca [IA-64] Disable machine check abort handling
3217 nomce [X86-32] Disable Machine Check Exception
3219 nomfgpt [X86-32] Disable Multi-Function General Purpose
3220 Timer usage (for AMD Geode machines).
3222 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
3223 shutdown the other cpus. Instead use the REBOOT_VECTOR
3226 nomodule Disable module load
3228 nopat [X86] Disable PAT (page attribute table extension of
3229 pagetables) support.
3231 nopcid [X86-64] Disable the PCID cpu feature.
3233 norandmaps Don't use address space randomization. Equivalent to
3234 echo 0 > /proc/sys/kernel/randomize_va_space
3236 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
3237 with UP alternatives
3239 nordrand [X86] Disable kernel use of the RDRAND and
3240 RDSEED instructions even if they are supported
3241 by the processor. RDRAND and RDSEED are still
3242 available to user space applications.
3244 noresume [SWSUSP] Disables resume and restores original swap
3247 no-scroll [VGA] Disables scrollback.
3248 This is required for the Braillex ib80-piezo Braille
3249 reader made by F.H. Papenmeier (Germany).
3253 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
3255 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
3256 and disable the IO APIC. legacy for "maxcpus=0".
3258 nosoftlockup [KNL] Disable the soft-lockup detector.
3260 nosync [HW,M68K] Disables sync negotiation for all devices.
3262 nowatchdog [KNL] Disable both lockup detectors, i.e.
3263 soft-lockup and NMI watchdog (hard-lockup).
3267 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
3269 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
3270 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
3271 Some features depend on CPU0. Known dependencies are:
3272 1. Resume from suspend/hibernate depends on CPU0.
3273 Suspend/hibernate will fail if CPU0 is offline and you
3274 need to online CPU0 before suspend/hibernate.
3275 2. PIC interrupts also depend on CPU0. CPU0 can't be
3276 removed if a PIC interrupt is detected.
3277 It's said poweroff/reboot may depend on CPU0 on some
3278 machines although I haven't seen such issues so far
3279 after CPU0 is offline on a few tested machines.
3280 If the dependencies are under your control, you can
3281 turn on cpu0_hotplug.
3283 nps_mtm_hs_ctr= [KNL,ARC]
3284 This parameter sets the maximum duration, in
3285 cycles, each HW thread of the CTOP can run
3286 without interruptions, before HW switches it.
3287 The actual maximum duration is 16 times this
3289 Format: integer between 1 and 255
3292 nptcg= [IA-64] Override max number of concurrent global TLB
3293 purges which is reported from either PAL_VM_SUMMARY or
3296 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
3297 could support. nr_cpus=n : n >= 1 limits the kernel to
3298 support 'n' processors. It could be larger than the
3299 number of already plugged CPU during bootup, later in
3300 runtime you can physically add extra cpu until it reaches
3301 n. So during boot up some boot time memory for per-cpu
3302 variables need be pre-allocated for later physical cpu
3305 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
3307 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing.
3308 Allowed values are enable and disable
3310 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
3311 'node', 'default' can be specified
3312 This can be set from sysctl after boot.
3313 See Documentation/admin-guide/sysctl/vm.rst for details.
3315 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
3316 See Documentation/debugging-via-ohci1394.txt for more
3319 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
3320 Rather than timing out after 20 ms if an EC
3321 command is not properly ACKed, override the length
3322 of the timeout. We have interrupts disabled while
3323 waiting for the ACK, so if this is set too high
3324 interrupts *may* be lost!
3326 omap_mux= [OMAP] Override bootloader pin multiplexing.
3327 Format: <mux_mode0.mode_name=value>...
3328 For example, to override I2C bus2:
3329 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
3331 oprofile.timer= [HW]
3332 Use timer interrupt instead of performance counters
3334 oprofile.cpu_type= Force an oprofile cpu type
3335 This might be useful if you have an older oprofile
3336 userland or if you want common events.
3337 Format: { arch_perfmon }
3338 arch_perfmon: [X86] Force use of architectural
3339 perfmon on Intel CPUs instead of the
3340 CPU specific event set.
3341 timer: [X86] Force use of architectural NMI
3342 timer mode (see also oprofile.timer
3343 for generic hr timer mode)
3345 oops=panic Always panic on oopses. Default is to just kill the
3346 process, but there is a small probability of
3347 deadlocking the machine.
3348 This will also cause panics on machine check exceptions.
3349 Useful together with panic=30 to trigger a reboot.
3352 [KNL] Boolean flag to control whether the page allocator
3353 should randomize its free lists. The randomization may
3354 be automatically enabled if the kernel detects it is
3355 running on a platform with a direct-mapped memory-side
3356 cache, and this parameter can be used to
3357 override/disable that behavior. The state of the flag
3358 can be read from sysfs at:
3359 /sys/module/page_alloc/parameters/shuffle.
3361 page_owner= [KNL] Boot-time page_owner enabling option.
3362 Storage of the information about who allocated
3363 each page is disabled in default. With this switch,
3365 on: enable the feature
3367 page_poison= [KNL] Boot-time parameter changing the state of
3368 poisoning on the buddy allocator, available with
3369 CONFIG_PAGE_POISONING=y.
3370 off: turn off poisoning (default)
3371 on: turn on poisoning
3373 panic= [KNL] Kernel behaviour on panic: delay <timeout>
3374 timeout > 0: seconds before rebooting
3375 timeout = 0: wait forever
3376 timeout < 0: reboot immediately
3379 panic_print= Bitmask for printing system info when panic happens.
3380 User can chose combination of the following bits:
3381 bit 0: print all tasks info
3382 bit 1: print system memory info
3383 bit 2: print timer info
3384 bit 3: print locks info if CONFIG_LOCKDEP is on
3385 bit 4: print ftrace buffer
3386 bit 5: print all printk messages in buffer
3388 panic_on_warn panic() instead of WARN(). Useful to cause kdump
3391 crash_kexec_post_notifiers
3392 Run kdump after running panic-notifiers and dumping
3393 kmsg. This only for the users who doubt kdump always
3394 succeeds in any situation.
3395 Note that this also increases risks of kdump failure,
3396 because some panic notifiers can make the crashed
3397 kernel more unstable.
3399 parkbd.port= [HW] Parallel port number the keyboard adapter is
3400 connected to, default is 0.
3402 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
3403 0 for XT, 1 for AT (default is AT).
3406 parport= [HW,PPT] Specify parallel ports. 0 disables.
3407 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
3408 Use 'auto' to force the driver to use any
3409 IRQ/DMA settings detected (the default is to
3410 ignore detected IRQ/DMA settings because of
3411 possible conflicts). You can specify the base
3412 address, IRQ, and DMA settings; IRQ and DMA
3413 should be numbers, or 'auto' (for using detected
3414 settings on that particular port), or 'nofifo'
3415 (to avoid using a FIFO even if it is detected).
3416 Parallel ports are assigned in the order they
3417 are specified on the command line, starting
3420 parport_init_mode= [HW,PPT]
3421 Configure VIA parallel port to operate in
3422 a specific mode. This is necessary on Pegasos
3423 computer where firmware has no options for setting
3424 up parallel port mode and sets it to spp.
3425 Currently this function knows 686a and 8231 chips.
3426 Format: [spp|ps2|epp|ecp|ecpepp]
3429 Halt all CPUs after the first oops has been printed for
3430 the specified number of seconds. This is to be used if
3431 your oopses keep scrolling off the screen.
3436 See header of drivers/block/paride/pcd.c.
3437 See also Documentation/admin-guide/blockdev/paride.rst.
3439 pci=option[,option...] [PCI] various PCI subsystem options.
3441 Some options herein operate on a specific device
3442 or a set of devices (<pci_dev>). These are
3443 specified in one of the following formats:
3445 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]*
3446 pci:<vendor>:<device>[:<subvendor>:<subdevice>]
3448 Note: the first format specifies a PCI
3449 bus/device/function address which may change
3450 if new hardware is inserted, if motherboard
3451 firmware changes, or due to changes caused
3452 by other kernel parameters. If the
3453 domain is left unspecified, it is
3454 taken to be zero. Optionally, a path
3455 to a device through multiple device/function
3456 addresses can be specified after the base
3457 address (this is more robust against
3458 renumbering issues). The second format
3459 selects devices using IDs from the
3460 configuration space which may match multiple
3461 devices in the system.
3463 earlydump dump PCI config space before the kernel
3465 off [X86] don't probe for the PCI bus
3466 bios [X86-32] force use of PCI BIOS, don't access
3467 the hardware directly. Use this if your machine
3468 has a non-standard PCI host bridge.
3469 nobios [X86-32] disallow use of PCI BIOS, only direct
3470 hardware access methods are allowed. Use this
3471 if you experience crashes upon bootup and you
3472 suspect they are caused by the BIOS.
3473 conf1 [X86] Force use of PCI Configuration Access
3474 Mechanism 1 (config address in IO port 0xCF8,
3475 data in IO port 0xCFC, both 32-bit).
3476 conf2 [X86] Force use of PCI Configuration Access
3477 Mechanism 2 (IO port 0xCF8 is an 8-bit port for
3478 the function, IO port 0xCFA, also 8-bit, sets
3479 bus number. The config space is then accessed
3480 through ports 0xC000-0xCFFF).
3481 See http://wiki.osdev.org/PCI for more info
3482 on the configuration access mechanisms.
3483 noaer [PCIE] If the PCIEAER kernel config parameter is
3484 enabled, this kernel boot option can be used to
3485 disable the use of PCIE advanced error reporting.
3486 nodomains [PCI] Disable support for multiple PCI
3487 root domains (aka PCI segments, in ACPI-speak).
3488 nommconf [X86] Disable use of MMCONFIG for PCI
3490 check_enable_amd_mmconf [X86] check for and enable
3491 properly configured MMIO access to PCI
3492 config space on AMD family 10h CPU
3493 nomsi [MSI] If the PCI_MSI kernel config parameter is
3494 enabled, this kernel boot option can be used to
3495 disable the use of MSI interrupts system-wide.
3496 noioapicquirk [APIC] Disable all boot interrupt quirks.
3497 Safety option to keep boot IRQs enabled. This
3498 should never be necessary.
3499 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
3500 primary IO-APIC for bridges that cannot disable
3501 boot IRQs. This fixes a source of spurious IRQs
3502 when the system masks IRQs.
3503 noioapicreroute [APIC] Disable workaround that uses the
3504 boot IRQ equivalent of an IRQ that connects to
3505 a chipset where boot IRQs cannot be disabled.
3506 The opposite of ioapicreroute.
3507 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
3508 routing table. These calls are known to be buggy
3509 on several machines and they hang the machine
3510 when used, but on other computers it's the only
3511 way to get the interrupt routing table. Try
3512 this option if the kernel is unable to allocate
3513 IRQs or discover secondary PCI buses on your
3515 rom [X86] Assign address space to expansion ROMs.
3516 Use with caution as certain devices share
3517 address decoders between ROMs and other
3519 norom [X86] Do not assign address space to
3520 expansion ROMs that do not already have
3521 BIOS assigned address ranges.
3522 nobar [X86] Do not assign address space to the
3523 BARs that weren't assigned by the BIOS.
3524 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
3525 assigned automatically to PCI devices. You can
3526 make the kernel exclude IRQs of your ISA cards
3528 pirqaddr=0xAAAAA [X86] Specify the physical address
3529 of the PIRQ table (normally generated
3530 by the BIOS) if it is outside the
3531 F0000h-100000h range.
3532 lastbus=N [X86] Scan all buses thru bus #N. Can be
3533 useful if the kernel is unable to find your
3534 secondary buses and you want to tell it
3535 explicitly which ones they are.
3536 assign-busses [X86] Always assign all PCI bus
3537 numbers ourselves, overriding
3538 whatever the firmware may have done.
3539 usepirqmask [X86] Honor the possible IRQ mask stored
3540 in the BIOS $PIR table. This is needed on
3541 some systems with broken BIOSes, notably
3542 some HP Pavilion N5400 and Omnibook XE3
3543 notebooks. This will have no effect if ACPI
3544 IRQ routing is enabled.
3545 noacpi [X86] Do not use ACPI for IRQ routing
3546 or for PCI scanning.
3547 use_crs [X86] Use PCI host bridge window information
3548 from ACPI. On BIOSes from 2008 or later, this
3549 is enabled by default. If you need to use this,
3550 please report a bug.
3551 nocrs [X86] Ignore PCI host bridge windows from ACPI.
3552 If you need to use this, please report a bug.
3553 routeirq Do IRQ routing for all PCI devices.
3554 This is normally done in pci_enable_device(),
3555 so this option is a temporary workaround
3556 for broken drivers that don't call it.
3557 skip_isa_align [X86] do not align io start addr, so can
3558 handle more pci cards
3559 noearly [X86] Don't do any early type 1 scanning.
3560 This might help on some broken boards which
3561 machine check when some devices' config space
3562 is read. But various workarounds are disabled
3563 and some IOMMU drivers will not work.
3564 bfsort Sort PCI devices into breadth-first order.
3565 This sorting is done to get a device
3566 order compatible with older (<= 2.4) kernels.
3567 nobfsort Don't sort PCI devices into breadth-first order.
3568 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
3569 tuning and use the BIOS-configured MPS defaults.
3570 pcie_bus_safe Set every device's MPS to the largest value
3571 supported by all devices below the root complex.
3572 pcie_bus_perf Set device MPS to the largest allowable MPS
3573 based on its parent bus. Also set MRRS (Max
3574 Read Request Size) to the largest supported
3575 value (no larger than the MPS that the device
3576 or bus can support) for best performance.
3577 pcie_bus_peer2peer Set every device's MPS to 128B, which
3578 every device is guaranteed to support. This
3579 configuration allows peer-to-peer DMA between
3580 any pair of devices, possibly at the cost of
3581 reduced performance. This also guarantees
3582 that hot-added devices will work.
3583 cbiosize=nn[KMG] The fixed amount of bus space which is
3584 reserved for the CardBus bridge's IO window.
3585 The default value is 256 bytes.
3586 cbmemsize=nn[KMG] The fixed amount of bus space which is
3587 reserved for the CardBus bridge's memory
3588 window. The default value is 64 megabytes.
3591 [<order of align>@]<pci_dev>[; ...]
3592 Specifies alignment and device to reassign
3593 aligned memory resources. How to
3594 specify the device is described above.
3595 If <order of align> is not specified,
3596 PAGE_SIZE is used as alignment.
3597 A PCI-PCI bridge can be specified if resource
3598 windows need to be expanded.
3599 To specify the alignment for several
3600 instances of a device, the PCI vendor,
3601 device, subvendor, and subdevice may be
3602 specified, e.g., 12@pci:8086:9c22:103c:198f
3603 for 4096-byte alignment.
3604 ecrc= Enable/disable PCIe ECRC (transaction layer
3605 end-to-end CRC checking).
3606 bios: Use BIOS/firmware settings. This is the
3610 hpiosize=nn[KMG] The fixed amount of bus space which is
3611 reserved for hotplug bridge's IO window.
3612 Default size is 256 bytes.
3613 hpmmiosize=nn[KMG] The fixed amount of bus space which is
3614 reserved for hotplug bridge's MMIO window.
3615 Default size is 2 megabytes.
3616 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is
3617 reserved for hotplug bridge's MMIO_PREF window.
3618 Default size is 2 megabytes.
3619 hpmemsize=nn[KMG] The fixed amount of bus space which is
3620 reserved for hotplug bridge's MMIO and
3622 Default size is 2 megabytes.
3623 hpbussize=nn The minimum amount of additional bus numbers
3624 reserved for buses below a hotplug bridge.
3626 realloc= Enable/disable reallocating PCI bridge resources
3627 if allocations done by BIOS are too small to
3628 accommodate resources required by all child
3630 off: Turn realloc off
3632 realloc same as realloc=on
3633 noari do not use PCIe ARI.
3634 noats [PCIE, Intel-IOMMU, AMD-IOMMU]
3635 do not use PCIe ATS (and IOMMU device IOTLB).
3636 pcie_scan_all Scan all possible PCIe devices. Otherwise we
3637 only look for one device below a PCIe downstream
3639 big_root_window Try to add a big 64bit memory window to the PCIe
3640 root complex on AMD CPUs. Some GFX hardware
3641 can resize a BAR to allow access to all VRAM.
3642 Adding the window is slightly risky (it may
3643 conflict with unreported devices), so this
3645 disable_acs_redir=<pci_dev>[; ...]
3646 Specify one or more PCI devices (in the format
3647 specified above) separated by semicolons.
3648 Each device specified will have the PCI ACS
3649 redirect capabilities forced off which will
3650 allow P2P traffic between devices through
3651 bridges without forcing it upstream. Note:
3652 this removes isolation between devices and
3653 may put more devices in an IOMMU group.
3654 force_floating [S390] Force usage of floating interrupts.
3655 nomio [S390] Do not use MIO instructions.
3657 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
3660 force Enable ASPM even on devices that claim not to support it.
3661 WARNING: Forcing ASPM on may cause system lockups.
3663 pcie_ports= [PCIE] PCIe port services handling:
3664 native Use native PCIe services (PME, AER, DPC, PCIe hotplug)
3665 even if the platform doesn't give the OS permission to
3666 use them. This may cause conflicts if the platform
3667 also tries to use these services.
3668 dpc-native Use native PCIe service for DPC only. May
3669 cause conflicts if firmware uses AER or DPC.
3670 compat Disable native PCIe services (PME, AER, DPC, PCIe
3673 pcie_port_pm= [PCIE] PCIe port power management handling:
3674 off Disable power management of all PCIe ports
3675 force Forcibly enable power management of all PCIe ports
3677 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
3678 nomsi Do not use MSI for native PCIe PME signaling (this makes
3679 all PCIe root ports use INTx for all services).
3681 pcmv= [HW,PCMCIA] BadgePAD 4
3685 Keep all power-domains already enabled by bootloader on,
3686 even if no driver has claimed them. This is useful
3687 for debug and development, but should not be
3688 needed on a platform with proper driver support.
3691 See Documentation/admin-guide/blockdev/paride.rst.
3693 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
3696 See arch/parisc/kernel/pdc_chassis.c
3698 percpu_alloc= Select which percpu first chunk allocator to use.
3699 Currently supported values are "embed" and "page".
3700 Archs may support subset or none of the selections.
3701 See comments in mm/percpu.c for details on each
3702 allocator. This parameter is primarily for debugging
3703 and performance comparison.
3706 See Documentation/admin-guide/blockdev/paride.rst.
3709 See Documentation/admin-guide/blockdev/paride.rst.
3711 pirq= [SMP,APIC] Manual mp-table setup
3712 See Documentation/x86/i386/IO-APIC.rst.
3714 plip= [PPT,NET] Parallel port network link
3715 Format: { parport<nr> | timid | 0 }
3716 See also Documentation/admin-guide/parport.rst.
3718 pmtmr= [X86] Manual setup of pmtmr I/O Port.
3719 Override pmtimer IOPort with a hex value.
3723 Enable PNP debug messages (depends on the
3724 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
3725 via /sys/module/pnp/parameters/debug. We always show
3726 current resource usage; turning this on also shows
3727 possible settings and some assignment information.
3733 { on | off | curr | res | no-curr | no-res }
3736 [ISAPNP] Exclude IRQs for the autoconfiguration
3739 [ISAPNP] Exclude DMAs for the autoconfiguration
3741 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
3742 Ranges are in pairs (I/O port base and size).
3745 [ISAPNP] Exclude memory regions for the
3747 Ranges are in pairs (memory base and size).
3749 ports= [IP_VS_FTP] IPVS ftp helper module
3751 Up to 8 (IP_VS_APP_MAX_PORTS) ports
3753 Format: <port>,<port>....
3755 powersave=off [PPC] This option disables power saving features.
3756 It specifically disables cpuidle and sets the
3757 platform machine description specific power_save
3758 function to NULL. On Idle the CPU just reduces
3761 ppc_strict_facility_enable
3762 [PPC] This option catches any kernel floating point,
3763 Altivec, VSX and SPE outside of regions specifically
3764 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
3765 There is some performance impact when enabling this.
3769 Disable Hardware Transactional Memory
3771 print-fatal-signals=
3772 [KNL] debug: print fatal signals
3774 If enabled, warn about various signal handling
3775 related application anomalies: too many signals,
3776 too many POSIX.1 timers, fatal signals causing a
3779 If you hit the warning due to signal overflow,
3780 you might want to try "ulimit -i unlimited".
3784 printk.always_kmsg_dump=
3785 Trigger kmsg_dump for cases other than kernel oops or
3787 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3790 printk.devkmsg={on,off,ratelimit}
3791 Control writing to /dev/kmsg.
3792 on - unlimited logging to /dev/kmsg from userspace
3793 off - logging to /dev/kmsg disabled
3794 ratelimit - ratelimit the logging
3797 printk.time= Show timing data prefixed to each printk message line
3798 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3800 processor.max_cstate= [HW,ACPI]
3801 Limit processor to maximum C-state
3802 max_cstate=9 overrides any DMI blacklist limit.
3804 processor.nocst [HW,ACPI]
3805 Ignore the _CST method to determine C-states,
3806 instead using the legacy FADT method
3808 profile= [KNL] Enable kernel profiling via /proc/profile
3809 Format: [<profiletype>,]<number>
3810 Param: <profiletype>: "schedule", "sleep", or "kvm"
3811 [defaults to kernel profiling]
3812 Param: "schedule" - profile schedule points.
3813 Param: "sleep" - profile D-state sleeping (millisecs).
3814 Requires CONFIG_SCHEDSTATS
3815 Param: "kvm" - profile VM exits.
3816 Param: <number> - step/bucket size as a power of 2 for
3817 statistical time based profiling.
3819 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
3821 See Documentation/admin-guide/blockdev/ramdisk.rst.
3823 psi= [KNL] Enable or disable pressure stall information
3827 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
3828 probe for; one of (bare|imps|exps|lifebook|any).
3829 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
3831 psmouse.resetafter= [HW,MOUSE]
3832 Try to reset the device after so many bad packets
3835 [HW,MOUSE] Set desired mouse resolution, in dpi.
3836 psmouse.smartscroll=
3837 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
3838 0 = disabled, 1 = enabled (default).
3840 pstore.backend= Specify the name of the pstore backend to use
3843 See Documentation/admin-guide/blockdev/paride.rst.
3845 pti= [X86_64] Control Page Table Isolation of user and
3846 kernel address spaces. Disabling this feature
3847 removes hardening, but improves performance of
3848 system calls and interrupts.
3850 on - unconditionally enable
3851 off - unconditionally disable
3852 auto - kernel detects whether your CPU model is
3853 vulnerable to issues that PTI mitigates
3855 Not specifying this option is equivalent to pti=auto.
3858 Equivalent to pti=off
3861 [KNL] Number of legacy pty's. Overwrites compiled-in
3864 quiet [KNL] Disable most log messages
3869 See Documentation/admin-guide/md.rst.
3871 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
3872 See Documentation/admin-guide/blockdev/ramdisk.rst.
3874 random.trust_cpu={on,off}
3875 [KNL] Enable or disable trusting the use of the
3876 CPU's random number generator (if available) to
3877 fully seed the kernel's CRNG. Default is controlled
3878 by CONFIG_RANDOM_TRUST_CPU.
3880 ras=option[,option,...] [KNL] RAS-specific options
3883 Disable the Correctable Errors Collector,
3884 see CONFIG_RAS_CEC help text.
3887 The argument is a cpu list, as described above,
3888 except that the string "all" can be used to
3889 specify every CPU on the system.
3891 In kernels built with CONFIG_RCU_NOCB_CPU=y, set
3892 the specified list of CPUs to be no-callback CPUs.
3893 Invocation of these CPUs' RCU callbacks will be
3894 offloaded to "rcuox/N" kthreads created for that
3895 purpose, where "x" is "p" for RCU-preempt, and
3896 "s" for RCU-sched, and "N" is the CPU number.
3897 This reduces OS jitter on the offloaded CPUs,
3898 which can be useful for HPC and real-time
3899 workloads. It can also improve energy efficiency
3900 for asymmetric multiprocessors.
3903 Rather than requiring that offloaded CPUs
3904 (specified by rcu_nocbs= above) explicitly
3905 awaken the corresponding "rcuoN" kthreads,
3906 make these kthreads poll for callbacks.
3907 This improves the real-time response for the
3908 offloaded CPUs by relieving them of the need to
3909 wake up the corresponding kthread, but degrades
3910 energy efficiency by requiring that the kthreads
3911 periodically wake up to do the polling.
3913 rcutree.blimit= [KNL]
3914 Set maximum number of finished RCU callbacks to
3915 process in one batch.
3917 rcutree.dump_tree= [KNL]
3918 Dump the structure of the rcu_node combining tree
3919 out at early boot. This is used for diagnostic
3920 purposes, to verify correct tree setup.
3922 rcutree.gp_cleanup_delay= [KNL]
3923 Set the number of jiffies to delay each step of
3924 RCU grace-period cleanup.
3926 rcutree.gp_init_delay= [KNL]
3927 Set the number of jiffies to delay each step of
3928 RCU grace-period initialization.
3930 rcutree.gp_preinit_delay= [KNL]
3931 Set the number of jiffies to delay each step of
3932 RCU grace-period pre-initialization, that is,
3933 the propagation of recent CPU-hotplug changes up
3934 the rcu_node combining tree.
3936 rcutree.use_softirq= [KNL]
3937 If set to zero, move all RCU_SOFTIRQ processing to
3938 per-CPU rcuc kthreads. Defaults to a non-zero
3939 value, meaning that RCU_SOFTIRQ is used by default.
3940 Specify rcutree.use_softirq=0 to use rcuc kthreads.
3942 rcutree.rcu_fanout_exact= [KNL]
3943 Disable autobalancing of the rcu_node combining
3944 tree. This is used by rcutorture, and might
3945 possibly be useful for architectures having high
3946 cache-to-cache transfer latencies.
3948 rcutree.rcu_fanout_leaf= [KNL]
3949 Change the number of CPUs assigned to each
3950 leaf rcu_node structure. Useful for very
3951 large systems, which will choose the value 64,
3952 and for NUMA systems with large remote-access
3953 latencies, which will choose a value aligned
3954 with the appropriate hardware boundaries.
3956 rcutree.jiffies_till_first_fqs= [KNL]
3957 Set delay from grace-period initialization to
3958 first attempt to force quiescent states.
3959 Units are jiffies, minimum value is zero,
3960 and maximum value is HZ.
3962 rcutree.jiffies_till_next_fqs= [KNL]
3963 Set delay between subsequent attempts to force
3964 quiescent states. Units are jiffies, minimum
3965 value is one, and maximum value is HZ.
3967 rcutree.jiffies_till_sched_qs= [KNL]
3968 Set required age in jiffies for a
3969 given grace period before RCU starts
3970 soliciting quiescent-state help from
3971 rcu_note_context_switch() and cond_resched().
3972 If not specified, the kernel will calculate
3973 a value based on the most recent settings
3974 of rcutree.jiffies_till_first_fqs
3975 and rcutree.jiffies_till_next_fqs.
3976 This calculated value may be viewed in
3977 rcutree.jiffies_to_sched_qs. Any attempt to set
3978 rcutree.jiffies_to_sched_qs will be cheerfully
3981 rcutree.kthread_prio= [KNL,BOOT]
3982 Set the SCHED_FIFO priority of the RCU per-CPU
3983 kthreads (rcuc/N). This value is also used for
3984 the priority of the RCU boost threads (rcub/N)
3985 and for the RCU grace-period kthreads (rcu_bh,
3986 rcu_preempt, and rcu_sched). If RCU_BOOST is
3987 set, valid values are 1-99 and the default is 1
3988 (the least-favored priority). Otherwise, when
3989 RCU_BOOST is not set, valid values are 0-99 and
3990 the default is zero (non-realtime operation).
3992 rcutree.rcu_nocb_gp_stride= [KNL]
3993 Set the number of NOCB callback kthreads in
3994 each group, which defaults to the square root
3995 of the number of CPUs. Larger numbers reduce
3996 the wakeup overhead on the global grace-period
3997 kthread, but increases that same overhead on
3998 each group's NOCB grace-period kthread.
4000 rcutree.qhimark= [KNL]
4001 Set threshold of queued RCU callbacks beyond which
4002 batch limiting is disabled.
4004 rcutree.qlowmark= [KNL]
4005 Set threshold of queued RCU callbacks below which
4006 batch limiting is re-enabled.
4008 rcutree.qovld= [KNL]
4009 Set threshold of queued RCU callbacks beyond which
4010 RCU's force-quiescent-state scan will aggressively
4011 enlist help from cond_resched() and sched IPIs to
4012 help CPUs more quickly reach quiescent states.
4013 Set to less than zero to make this be set based
4014 on rcutree.qhimark at boot time and to zero to
4015 disable more aggressive help enlistment.
4017 rcutree.rcu_idle_gp_delay= [KNL]
4018 Set wakeup interval for idle CPUs that have
4019 RCU callbacks (RCU_FAST_NO_HZ=y).
4021 rcutree.rcu_idle_lazy_gp_delay= [KNL]
4022 Set wakeup interval for idle CPUs that have
4023 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
4024 Lazy RCU callbacks are those which RCU can
4025 prove do nothing more than free memory.
4027 rcutree.rcu_kick_kthreads= [KNL]
4028 Cause the grace-period kthread to get an extra
4029 wake_up() if it sleeps three times longer than
4030 it should at force-quiescent-state time.
4031 This wake_up() will be accompanied by a
4032 WARN_ONCE() splat and an ftrace_dump().
4034 rcutree.sysrq_rcu= [KNL]
4035 Commandeer a sysrq key to dump out Tree RCU's
4036 rcu_node tree with an eye towards determining
4037 why a new grace period has not yet started.
4039 rcuperf.gp_async= [KNL]
4040 Measure performance of asynchronous
4041 grace-period primitives such as call_rcu().
4043 rcuperf.gp_async_max= [KNL]
4044 Specify the maximum number of outstanding
4045 callbacks per writer thread. When a writer
4046 thread exceeds this limit, it invokes the
4047 corresponding flavor of rcu_barrier() to allow
4048 previously posted callbacks to drain.
4050 rcuperf.gp_exp= [KNL]
4051 Measure performance of expedited synchronous
4052 grace-period primitives.
4054 rcuperf.holdoff= [KNL]
4055 Set test-start holdoff period. The purpose of
4056 this parameter is to delay the start of the
4057 test until boot completes in order to avoid
4060 rcuperf.kfree_rcu_test= [KNL]
4061 Set to measure performance of kfree_rcu() flooding.
4063 rcuperf.kfree_nthreads= [KNL]
4064 The number of threads running loops of kfree_rcu().
4066 rcuperf.kfree_alloc_num= [KNL]
4067 Number of allocations and frees done in an iteration.
4069 rcuperf.kfree_loops= [KNL]
4070 Number of loops doing rcuperf.kfree_alloc_num number
4071 of allocations and frees.
4073 rcuperf.nreaders= [KNL]
4074 Set number of RCU readers. The value -1 selects
4075 N, where N is the number of CPUs. A value
4076 "n" less than -1 selects N-n+1, where N is again
4077 the number of CPUs. For example, -2 selects N
4078 (the number of CPUs), -3 selects N+1, and so on.
4079 A value of "n" less than or equal to -N selects
4082 rcuperf.nwriters= [KNL]
4083 Set number of RCU writers. The values operate
4084 the same as for rcuperf.nreaders.
4085 N, where N is the number of CPUs
4087 rcuperf.perf_type= [KNL]
4088 Specify the RCU implementation to test.
4090 rcuperf.shutdown= [KNL]
4091 Shut the system down after performance tests
4092 complete. This is useful for hands-off automated
4095 rcuperf.verbose= [KNL]
4096 Enable additional printk() statements.
4098 rcuperf.writer_holdoff= [KNL]
4099 Write-side holdoff between grace periods,
4100 in microseconds. The default of zero says
4103 rcutorture.fqs_duration= [KNL]
4104 Set duration of force_quiescent_state bursts
4107 rcutorture.fqs_holdoff= [KNL]
4108 Set holdoff time within force_quiescent_state bursts
4111 rcutorture.fqs_stutter= [KNL]
4112 Set wait time between force_quiescent_state bursts
4115 rcutorture.fwd_progress= [KNL]
4116 Enable RCU grace-period forward-progress testing
4117 for the types of RCU supporting this notion.
4119 rcutorture.fwd_progress_div= [KNL]
4120 Specify the fraction of a CPU-stall-warning
4121 period to do tight-loop forward-progress testing.
4123 rcutorture.fwd_progress_holdoff= [KNL]
4124 Number of seconds to wait between successive
4125 forward-progress tests.
4127 rcutorture.fwd_progress_need_resched= [KNL]
4128 Enclose cond_resched() calls within checks for
4129 need_resched() during tight-loop forward-progress
4132 rcutorture.gp_cond= [KNL]
4133 Use conditional/asynchronous update-side
4134 primitives, if available.
4136 rcutorture.gp_exp= [KNL]
4137 Use expedited update-side primitives, if available.
4139 rcutorture.gp_normal= [KNL]
4140 Use normal (non-expedited) asynchronous
4141 update-side primitives, if available.
4143 rcutorture.gp_sync= [KNL]
4144 Use normal (non-expedited) synchronous
4145 update-side primitives, if available. If all
4146 of rcutorture.gp_cond=, rcutorture.gp_exp=,
4147 rcutorture.gp_normal=, and rcutorture.gp_sync=
4148 are zero, rcutorture acts as if is interpreted
4149 they are all non-zero.
4151 rcutorture.n_barrier_cbs= [KNL]
4152 Set callbacks/threads for rcu_barrier() testing.
4154 rcutorture.nfakewriters= [KNL]
4155 Set number of concurrent RCU writers. These just
4156 stress RCU, they don't participate in the actual
4157 test, hence the "fake".
4159 rcutorture.nreaders= [KNL]
4160 Set number of RCU readers. The value -1 selects
4161 N-1, where N is the number of CPUs. A value
4162 "n" less than -1 selects N-n-2, where N is again
4163 the number of CPUs. For example, -2 selects N
4164 (the number of CPUs), -3 selects N+1, and so on.
4166 rcutorture.object_debug= [KNL]
4167 Enable debug-object double-call_rcu() testing.
4169 rcutorture.onoff_holdoff= [KNL]
4170 Set time (s) after boot for CPU-hotplug testing.
4172 rcutorture.onoff_interval= [KNL]
4173 Set time (jiffies) between CPU-hotplug operations,
4174 or zero to disable CPU-hotplug testing.
4176 rcutorture.shuffle_interval= [KNL]
4177 Set task-shuffle interval (s). Shuffling tasks
4178 allows some CPUs to go into dyntick-idle mode
4179 during the rcutorture test.
4181 rcutorture.shutdown_secs= [KNL]
4182 Set time (s) after boot system shutdown. This
4183 is useful for hands-off automated testing.
4185 rcutorture.stall_cpu= [KNL]
4186 Duration of CPU stall (s) to test RCU CPU stall
4187 warnings, zero to disable.
4189 rcutorture.stall_cpu_holdoff= [KNL]
4190 Time to wait (s) after boot before inducing stall.
4192 rcutorture.stall_cpu_irqsoff= [KNL]
4193 Disable interrupts while stalling if set.
4195 rcutorture.stat_interval= [KNL]
4196 Time (s) between statistics printk()s.
4198 rcutorture.stutter= [KNL]
4199 Time (s) to stutter testing, for example, specifying
4200 five seconds causes the test to run for five seconds,
4201 wait for five seconds, and so on. This tests RCU's
4202 ability to transition abruptly to and from idle.
4204 rcutorture.test_boost= [KNL]
4205 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
4206 "Maybe" means test if the RCU implementation
4207 under test support RCU priority boosting.
4209 rcutorture.test_boost_duration= [KNL]
4210 Duration (s) of each individual boost test.
4212 rcutorture.test_boost_interval= [KNL]
4213 Interval (s) between each boost test.
4215 rcutorture.test_no_idle_hz= [KNL]
4216 Test RCU's dyntick-idle handling. See also the
4217 rcutorture.shuffle_interval parameter.
4219 rcutorture.torture_type= [KNL]
4220 Specify the RCU implementation to test.
4222 rcutorture.verbose= [KNL]
4223 Enable additional printk() statements.
4225 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL]
4226 Dump ftrace buffer after reporting RCU CPU
4229 rcupdate.rcu_cpu_stall_suppress= [KNL]
4230 Suppress RCU CPU stall warning messages.
4232 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL]
4233 Suppress RCU CPU stall warning messages and
4234 rcutorture writer stall warnings that occur
4235 during early boot, that is, during the time
4236 before the init task is spawned.
4238 rcupdate.rcu_cpu_stall_timeout= [KNL]
4239 Set timeout for RCU CPU stall warning messages.
4241 rcupdate.rcu_expedited= [KNL]
4242 Use expedited grace-period primitives, for
4243 example, synchronize_rcu_expedited() instead
4244 of synchronize_rcu(). This reduces latency,
4245 but can increase CPU utilization, degrade
4246 real-time latency, and degrade energy efficiency.
4247 No effect on CONFIG_TINY_RCU kernels.
4249 rcupdate.rcu_normal= [KNL]
4250 Use only normal grace-period primitives,
4251 for example, synchronize_rcu() instead of
4252 synchronize_rcu_expedited(). This improves
4253 real-time latency, CPU utilization, and
4254 energy efficiency, but can expose users to
4255 increased grace-period latency. This parameter
4256 overrides rcupdate.rcu_expedited. No effect on
4257 CONFIG_TINY_RCU kernels.
4259 rcupdate.rcu_normal_after_boot= [KNL]
4260 Once boot has completed (that is, after
4261 rcu_end_inkernel_boot() has been invoked), use
4262 only normal grace-period primitives. No effect
4263 on CONFIG_TINY_RCU kernels.
4265 rcupdate.rcu_task_stall_timeout= [KNL]
4266 Set timeout in jiffies for RCU task stall warning
4267 messages. Disable with a value less than or equal
4270 rcupdate.rcu_self_test= [KNL]
4271 Run the RCU early boot self tests
4275 Run specified binary instead of /init from the ramdisk,
4276 used for early userspace startup. See initrd.
4279 force - Override the decision by the kernel to hide the
4280 advertisement of RDRAND support (this affects
4281 certain AMD processors because of buggy BIOS
4282 support, specifically around the suspend/resume
4286 Turn on/off individual RDT features. List is:
4287 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp,
4289 E.g. to turn on cmt and turn off mba use:
4293 Format (x86 or x86_64):
4294 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
4296 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
4298 Where reboot_mode is one of warm (soft) or cold (hard) or gpio
4299 (prefix with 'panic_' to set mode for panic
4301 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
4302 reboot_force is either force or not specified,
4303 reboot_cpu is s[mp]#### with #### being the processor
4304 to be used for rebooting.
4307 [KNL, SMP] Set scheduler's default relax_domain_level.
4308 See Documentation/admin-guide/cgroup-v1/cpusets.rst.
4310 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory
4311 Format: <base1>,<size1>[,<base2>,<size2>,...]
4312 Reserve I/O ports or memory so the kernel won't use
4313 them. If <base> is less than 0x10000, the region
4314 is assumed to be I/O ports; otherwise it is memory.
4316 reservetop= [X86-32]
4318 Reserves a hole at the top of the kernel virtual
4323 Set the amount of memory to reserve for BIOS at
4324 the bottom of the address space.
4326 reset_devices [KNL] Force drivers to reset the underlying device
4327 during initialization.
4330 Specify the partition device for software suspend
4332 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
4334 resume_offset= [SWSUSP]
4335 Specify the offset from the beginning of the partition
4336 given by "resume=" at which the swap header is located,
4337 in <PAGE_SIZE> units (needed only for swap files).
4338 See Documentation/power/swsusp-and-swap-files.rst
4340 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
4341 read the resume files
4343 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
4344 Useful for devices that are detected asynchronously
4345 (e.g. USB and MMC devices).
4347 hibernate= [HIBERNATION]
4348 noresume Don't check if there's a hibernation image
4349 present during boot.
4350 nocompress Don't compress/decompress hibernation images.
4351 no Disable hibernation and resume.
4352 protect_image Turn on image protection during restoration
4353 (that will set all pages holding image data
4354 during restoration read-only).
4356 retain_initrd [RAM] Keep initrd memory after extraction
4358 rfkill.default_state=
4359 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
4360 etc. communication is blocked by default.
4363 rfkill.master_switch_mode=
4364 0 The "airplane mode" button does nothing.
4365 1 The "airplane mode" button toggles between everything
4366 blocked and the previous configuration.
4367 2 The "airplane mode" button toggles between everything
4368 blocked and everything unblocked.
4370 rhash_entries= [KNL,NET]
4371 Set number of hash buckets for route cache
4374 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported
4377 ro [KNL] Mount root device read-only on boot
4380 on Mark read-only kernel memory as read-only (default).
4381 off Leave read-only kernel memory writable for debugging.
4384 Enable the uart passthrough on the designated usb port
4385 on Rockchip SoCs. When active, the signals of the
4386 debug-uart get routed to the D+ and D- pins of the usb
4387 port and the regular usb controller gets disabled.
4389 root= [KNL] Root filesystem
4390 See name_to_dev_t comment in init/do_mounts.c.
4392 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
4393 mount the root filesystem
4395 rootflags= [KNL] Set root filesystem mount option string
4397 rootfstype= [KNL] Set root filesystem type
4399 rootwait [KNL] Wait (indefinitely) for root device to show up.
4400 Useful for devices that are detected asynchronously
4401 (e.g. USB and MMC devices).
4403 rproc_mem=nn[KMG][@address]
4404 [KNL,ARM,CMA] Remoteproc physical memory block.
4405 Memory area to be used by remote processor image,
4408 rw [KNL] Mount root device read-write on boot
4410 S [KNL] Run init in single mode
4412 s390_iommu= [HW,S390]
4413 Set s390 IOTLB flushing mode
4415 With strict flushing every unmap operation will result in
4416 an IOTLB flush. Default is lazy flushing before reuse,
4420 See drivers/net/irda/sa1100_ir.c.
4422 sbni= [NET] Granch SBNI12 leased line adapter
4424 sched_debug [KNL] Enables verbose scheduler debug messages.
4426 schedstats= [KNL,X86] Enable or disable scheduled statistics.
4427 Allowed values are enable and disable. This feature
4428 incurs a small amount of overhead in the scheduler
4429 but is useful for debugging and performance tuning.
4431 sched_thermal_decay_shift=
4432 [KNL, SMP] Set a decay shift for scheduler thermal
4433 pressure signal. Thermal pressure signal follows the
4434 default decay period of other scheduler pelt
4435 signals(usually 32 ms but configurable). Setting
4436 sched_thermal_decay_shift will left shift the decay
4437 period for the thermal pressure signal by the shift
4439 i.e. with the default pelt decay period of 32 ms
4440 sched_thermal_decay_shift thermal pressure decay pr
4444 Format: integer between 0 and 10
4447 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
4448 xtime_lock contention on larger systems, and/or RCU lock
4449 contention on all systems with CONFIG_MAXSMP set.
4450 Format: { "0" | "1" }
4451 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
4453 Note: increases power consumption, thus should only be
4454 enabled if running jitter sensitive (HPC/RT) workloads.
4456 security= [SECURITY] Choose a legacy "major" security module to
4457 enable at boot. This has been deprecated by the
4460 selinux= [SELINUX] Disable or enable SELinux at boot time.
4461 Format: { "0" | "1" }
4462 See security/selinux/Kconfig help text.
4467 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
4468 Format: { "0" | "1" }
4469 See security/apparmor/Kconfig help text
4472 Default value is set via kernel config option.
4474 serialnumber [BUGS=X86-32]
4477 Maximal number of shapers.
4485 Disable merging of slabs with similar size. May be
4486 necessary if there is some reason to distinguish
4487 allocs to different slabs, especially in hardened
4488 environments where the risk of heap overflows and
4489 layout control by attackers can usually be
4490 frustrated by disabling merging. This will reduce
4491 most of the exposure of a heap attack to a single
4492 cache (risks via metadata attacks are mostly
4493 unchanged). Debug options disable merging on their
4495 For more information see Documentation/vm/slub.rst.
4497 slab_max_order= [MM, SLAB]
4498 Determines the maximum allowed order for slabs.
4499 A high setting may cause OOMs due to memory
4500 fragmentation. Defaults to 1 for systems with
4501 more than 32MB of RAM, 0 otherwise.
4503 slub_debug[=options[,slabs]] [MM, SLUB]
4504 Enabling slub_debug allows one to determine the
4505 culprit if slab objects become corrupted. Enabling
4506 slub_debug can create guard zones around objects and
4507 may poison objects when not in use. Also tracks the
4508 last alloc / free. For more information see
4509 Documentation/vm/slub.rst.
4511 slub_memcg_sysfs= [MM, SLUB]
4512 Determines whether to enable sysfs directories for
4513 memory cgroup sub-caches. 1 to enable, 0 to disable.
4514 The default is determined by CONFIG_SLUB_MEMCG_SYSFS_ON.
4515 Enabling this can lead to a very high number of debug
4516 directories and files being created under
4519 slub_max_order= [MM, SLUB]
4520 Determines the maximum allowed order for slabs.
4521 A high setting may cause OOMs due to memory
4522 fragmentation. For more information see
4523 Documentation/vm/slub.rst.
4525 slub_min_objects= [MM, SLUB]
4526 The minimum number of objects per slab. SLUB will
4527 increase the slab order up to slub_max_order to
4528 generate a sufficiently large slab able to contain
4529 the number of objects indicated. The higher the number
4530 of objects the smaller the overhead of tracking slabs
4531 and the less frequently locks need to be acquired.
4532 For more information see Documentation/vm/slub.rst.
4534 slub_min_order= [MM, SLUB]
4535 Determines the minimum page order for slabs. Must be
4536 lower than slub_max_order.
4537 For more information see Documentation/vm/slub.rst.
4539 slub_nomerge [MM, SLUB]
4540 Same with slab_nomerge. This is supported for legacy.
4541 See slab_nomerge for more information.
4544 Format: <io1>[,<io2>[,...,<io8>]]
4546 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
4547 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
4548 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
4549 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
4550 smsc-ircc2.ircc_irq= [HW] IRQ line
4551 smsc-ircc2.ircc_dma= [HW] DMA channel
4552 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
4553 0: Toshiba Satellite 1800 (GP data pin select)
4554 1: Fast pin select (default)
4557 smt [KNL,S390] Set the maximum number of threads (logical
4558 CPUs) to use per physical CPU on systems capable of
4559 symmetric multithreading (SMT). Will be capped to the
4560 actual hardware limit.
4562 Default: -1 (no limit)
4565 [KNL] Should the soft-lockup detector generate panics.
4568 A nonzero value instructs the soft-lockup detector
4569 to panic the machine when a soft-lockup occurs. It is
4570 also controlled by the kernel.softlockup_panic sysctl
4571 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the
4572 respective build-time switch to that functionality.
4574 softlockup_all_cpu_backtrace=
4575 [KNL] Should the soft-lockup detector generate
4576 backtraces on all cpus.
4579 sonypi.*= [HW] Sony Programmable I/O Control Device driver
4580 See Documentation/admin-guide/laptops/sonypi.rst
4582 spectre_v2= [X86] Control mitigation of Spectre variant 2
4583 (indirect branch speculation) vulnerability.
4584 The default operation protects the kernel from
4587 on - unconditionally enable, implies
4589 off - unconditionally disable, implies
4591 auto - kernel detects whether your CPU model is
4594 Selecting 'on' will, and 'auto' may, choose a
4595 mitigation method at run time according to the
4596 CPU, the available microcode, the setting of the
4597 CONFIG_RETPOLINE configuration option, and the
4598 compiler with which the kernel was built.
4600 Selecting 'on' will also enable the mitigation
4601 against user space to user space task attacks.
4603 Selecting 'off' will disable both the kernel and
4604 the user space protections.
4606 Specific mitigations can also be selected manually:
4608 retpoline - replace indirect branches
4609 retpoline,generic - google's original retpoline
4610 retpoline,amd - AMD-specific minimal thunk
4612 Not specifying this option is equivalent to
4616 [X86] Control mitigation of Spectre variant 2
4617 (indirect branch speculation) vulnerability between
4620 on - Unconditionally enable mitigations. Is
4621 enforced by spectre_v2=on
4623 off - Unconditionally disable mitigations. Is
4624 enforced by spectre_v2=off
4626 prctl - Indirect branch speculation is enabled,
4627 but mitigation can be enabled via prctl
4628 per thread. The mitigation control state
4629 is inherited on fork.
4632 - Like "prctl" above, but only STIBP is
4633 controlled per thread. IBPB is issued
4634 always when switching between different user
4638 - Same as "prctl" above, but all seccomp
4639 threads will enable the mitigation unless
4640 they explicitly opt out.
4643 - Like "seccomp" above, but only STIBP is
4644 controlled per thread. IBPB is issued
4645 always when switching between different
4646 user space processes.
4648 auto - Kernel selects the mitigation depending on
4649 the available CPU features and vulnerability.
4652 If CONFIG_SECCOMP=y then "seccomp", otherwise "prctl"
4654 Not specifying this option is equivalent to
4655 spectre_v2_user=auto.
4657 spec_store_bypass_disable=
4658 [HW] Control Speculative Store Bypass (SSB) Disable mitigation
4659 (Speculative Store Bypass vulnerability)
4661 Certain CPUs are vulnerable to an exploit against a
4662 a common industry wide performance optimization known
4663 as "Speculative Store Bypass" in which recent stores
4664 to the same memory location may not be observed by
4665 later loads during speculative execution. The idea
4666 is that such stores are unlikely and that they can
4667 be detected prior to instruction retirement at the
4668 end of a particular speculation execution window.
4670 In vulnerable processors, the speculatively forwarded
4671 store can be used in a cache side channel attack, for
4672 example to read memory to which the attacker does not
4673 directly have access (e.g. inside sandboxed code).
4675 This parameter controls whether the Speculative Store
4676 Bypass optimization is used.
4678 On x86 the options are:
4680 on - Unconditionally disable Speculative Store Bypass
4681 off - Unconditionally enable Speculative Store Bypass
4682 auto - Kernel detects whether the CPU model contains an
4683 implementation of Speculative Store Bypass and
4684 picks the most appropriate mitigation. If the
4685 CPU is not vulnerable, "off" is selected. If the
4686 CPU is vulnerable the default mitigation is
4687 architecture and Kconfig dependent. See below.
4688 prctl - Control Speculative Store Bypass per thread
4689 via prctl. Speculative Store Bypass is enabled
4690 for a process by default. The state of the control
4691 is inherited on fork.
4692 seccomp - Same as "prctl" above, but all seccomp threads
4693 will disable SSB unless they explicitly opt out.
4695 Default mitigations:
4696 X86: If CONFIG_SECCOMP=y "seccomp", otherwise "prctl"
4698 On powerpc the options are:
4700 on,auto - On Power8 and Power9 insert a store-forwarding
4701 barrier on kernel entry and exit. On Power7
4702 perform a software flush on kernel entry and
4706 Not specifying this option is equivalent to
4707 spec_store_bypass_disable=auto.
4709 spia_io_base= [HW,MTD]
4714 srcutree.counter_wrap_check [KNL]
4715 Specifies how frequently to check for
4716 grace-period sequence counter wrap for the
4717 srcu_data structure's ->srcu_gp_seq_needed field.
4718 The greater the number of bits set in this kernel
4719 parameter, the less frequently counter wrap will
4720 be checked for. Note that the bottom two bits
4723 srcutree.exp_holdoff [KNL]
4724 Specifies how many nanoseconds must elapse
4725 since the end of the last SRCU grace period for
4726 a given srcu_struct until the next normal SRCU
4727 grace period will be considered for automatic
4728 expediting. Set to zero to disable automatic
4732 Speculative Store Bypass Disable control
4734 On CPUs that are vulnerable to the Speculative
4735 Store Bypass vulnerability and offer a
4736 firmware based mitigation, this parameter
4737 indicates how the mitigation should be used:
4739 force-on: Unconditionally enable mitigation for
4740 for both kernel and userspace
4741 force-off: Unconditionally disable mitigation for
4742 for both kernel and userspace
4743 kernel: Always enable mitigation in the
4744 kernel, and offer a prctl interface
4745 to allow userspace to register its
4746 interest in being mitigated too.
4748 stack_guard_gap= [MM]
4749 override the default stack gap protection. The value
4750 is in page units and it defines how many pages prior
4751 to (for stacks growing down) resp. after (for stacks
4752 growing up) the main stack are reserved for no other
4753 mapping. Default value is 256 pages.
4756 Enabled the stack tracer on boot up.
4758 stacktrace_filter=[function-list]
4759 [FTRACE] Limit the functions that the stack tracer
4760 will trace at boot up. function-list is a comma separated
4761 list of functions. This list can be changed at run
4762 time by the stack_trace_filter file in the debugfs
4763 tracing directory. Note, this enables stack tracing
4764 and the stacktrace above is not needed.
4768 Set the STI (builtin display/keyboard on the HP-PARISC
4769 machines) console (graphic card) which should be used
4770 as the initial boot-console.
4771 See also comment in drivers/video/console/sticore.c.
4774 See comment in drivers/video/console/sticore.c.
4777 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
4779 sunrpc.min_resvport=
4780 sunrpc.max_resvport=
4782 SunRPC servers often require that client requests
4783 originate from a privileged port (i.e. a port in the
4784 range 0 < portnr < 1024).
4785 An administrator who wishes to reserve some of these
4786 ports for other uses may adjust the range that the
4787 kernel's sunrpc client considers to be privileged
4788 using these two parameters to set the minimum and
4789 maximum port values.
4791 sunrpc.svc_rpc_per_connection_limit=
4793 Limit the number of requests that the server will
4794 process in parallel from a single connection.
4795 The default value is 0 (no limit).
4799 Control how the NFS server code allocates CPUs to
4800 service thread pools. Depending on how many NICs
4801 you have and where their interrupts are bound, this
4802 option will affect which CPUs will do NFS serving.
4803 Note: this parameter cannot be changed while the
4804 NFS server is running.
4806 auto the server chooses an appropriate mode
4807 automatically using heuristics
4808 global a single global pool contains all CPUs
4809 percpu one pool for each CPU
4810 pernode one pool for each NUMA node (equivalent
4811 to global on non-NUMA machines)
4813 sunrpc.tcp_slot_table_entries=
4814 sunrpc.udp_slot_table_entries=
4816 Sets the upper limit on the number of simultaneous
4817 RPC calls that can be sent from the client to a
4818 server. Increasing these values may allow you to
4819 improve throughput, but will also increase the
4820 amount of memory reserved for use by the client.
4822 suspend.pm_test_delay=
4824 Sets the number of seconds to remain in a suspend test
4825 mode before resuming the system (see
4826 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
4827 is set. Default value is 5.
4830 Format: { on | off | y | n | 1 | 0 }
4831 This parameter controls use of the Protected
4832 Execution Facility on pSeries.
4835 [KNL] Enable accounting of swap in memory resource
4836 controller if no parameter or 1 is given or disable
4837 it if 0 is given (See Documentation/admin-guide/cgroup-v1/memory.rst)
4839 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
4840 Format: { <int> | force | noforce }
4841 <int> -- Number of I/O TLB slabs
4842 force -- force using of bounce buffers even if they
4843 wouldn't be automatically used by the kernel
4844 noforce -- Never use bounce buffers (for debugging)
4848 sysfs.deprecated=0|1 [KNL]
4849 Enable/disable old style sysfs layout for old udev
4850 on older distributions. When this option is enabled
4851 very new udev will not work anymore. When this option
4852 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
4853 in older udev will not work anymore.
4854 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
4855 the kernel configuration.
4857 sysrq_always_enabled
4859 Ignore sysrq setting - this boot parameter will
4860 neutralize any effect of /proc/sys/kernel/sysrq.
4861 Useful for debugging.
4863 tcpmhash_entries= [KNL,NET]
4864 Set the number of tcp_metrics_hash slots.
4865 Default value is 8192 or 16384 depending on total
4866 ram pages. This is used to specify the TCP metrics
4867 cache size. See Documentation/networking/ip-sysctl.txt
4868 "tcp_no_metrics_save" section for more details.
4872 test_suspend= [SUSPEND][,N]
4873 Specify "mem" (for Suspend-to-RAM) or "standby" (for
4874 standby suspend) or "freeze" (for suspend type freeze)
4875 as the system sleep state during system startup with
4876 the optional capability to repeat N number of times.
4877 The system is woken from this state using a
4878 wakeup-capable RTC alarm.
4880 thash_entries= [KNL,NET]
4881 Set number of hash buckets for TCP connection
4883 thermal.act= [HW,ACPI]
4884 -1: disable all active trip points in all thermal zones
4885 <degrees C>: override all lowest active trip points
4887 thermal.crt= [HW,ACPI]
4888 -1: disable all critical trip points in all thermal zones
4889 <degrees C>: override all critical trip points
4891 thermal.nocrt= [HW,ACPI]
4892 Set to disable actions on ACPI thermal zone
4893 critical and hot trip points.
4895 thermal.off= [HW,ACPI]
4896 1: disable ACPI thermal control
4898 thermal.psv= [HW,ACPI]
4899 -1: disable all passive trip points
4900 <degrees C>: override all passive trip points to this
4903 thermal.tzp= [HW,ACPI]
4904 Specify global default ACPI thermal zone polling rate
4905 <deci-seconds>: poll all this frequency
4906 0: no polling (default)
4909 Force threading of all interrupt handlers except those
4910 marked explicitly IRQF_NO_THREAD.
4914 Specify if the kernel should make use of the cpu
4915 topology information if the hardware supports this.
4916 The scheduler will make use of this information and
4917 e.g. base its process migration decisions on it.
4920 topology_updates= [KNL, PPC, NUMA]
4922 Specify if the kernel should ignore (off)
4923 topology updates sent by the hypervisor to this
4926 torture.disable_onoff_at_boot= [KNL]
4927 Prevent the CPU-hotplug component of torturing
4928 until after init has spawned.
4932 tpm_suspend_pcr=[HW,TPM]
4933 Format: integer pcr id
4934 Specify that at suspend time, the tpm driver
4935 should extend the specified pcr with zeros,
4936 as a workaround for some chips which fail to
4937 flush the last written pcr on TPM_SaveState.
4938 This will guarantee that all the other pcrs
4941 trace_buf_size=nn[KMG]
4942 [FTRACE] will set tracing buffer size on each cpu.
4944 trace_event=[event-list]
4945 [FTRACE] Set and start specified trace events in order
4946 to facilitate early boot debugging. The event-list is a
4947 comma separated list of trace events to enable. See
4948 also Documentation/trace/events.rst
4950 trace_options=[option-list]
4951 [FTRACE] Enable or disable tracer options at boot.
4952 The option-list is a comma delimited list of options
4953 that can be enabled or disabled just as if you were
4954 to echo the option name into
4956 /sys/kernel/debug/tracing/trace_options
4958 For example, to enable stacktrace option (to dump the
4959 stack trace of each event), add to the command line:
4961 trace_options=stacktrace
4963 See also Documentation/trace/ftrace.rst "trace options"
4967 Have the tracepoints sent to printk as well as the
4968 tracing ring buffer. This is useful for early boot up
4969 where the system hangs or reboots and does not give the
4970 option for reading the tracing buffer or performing a
4971 ftrace_dump_on_oops.
4973 To turn off having tracepoints sent to printk,
4974 echo 0 > /proc/sys/kernel/tracepoint_printk
4975 Note, echoing 1 into this file without the
4976 tracepoint_printk kernel cmdline option has no effect.
4980 Having tracepoints sent to printk() and activating high
4981 frequency tracepoints such as irq or sched, can cause
4982 the system to live lock.
4985 [FTRACE] enable this option to disable tracing when a
4986 warning is hit. This turns off "tracing_on". Tracing can
4987 be enabled again by echoing '1' into the "tracing_on"
4988 file located in /sys/kernel/debug/tracing/
4990 This option is useful, as it disables the trace before
4991 the WARNING dump is called, which prevents the trace to
4992 be filled with content caused by the warning output.
4994 This option can also be set at run time via the sysctl
4995 option: kernel/traceoff_on_warning
4997 transparent_hugepage=
4999 Format: [always|madvise|never]
5000 Can be used to control the default behavior of the system
5001 with respect to transparent hugepages.
5002 See Documentation/admin-guide/mm/transhuge.rst
5005 tsc= Disable clocksource stability checks for TSC.
5007 [x86] reliable: mark tsc clocksource as reliable, this
5008 disables clocksource verification at runtime, as well
5009 as the stability checks done at bootup. Used to enable
5010 high-resolution timer mode on older hardware, and in
5011 virtualized environment.
5012 [x86] noirqtime: Do not use TSC to do irq accounting.
5013 Used to run time disable IRQ_TIME_ACCOUNTING on any
5014 platforms where RDTSC is slow and this accounting
5016 [x86] unstable: mark the TSC clocksource as unstable, this
5017 marks the TSC unconditionally unstable at bootup and
5018 avoids any further wobbles once the TSC watchdog notices.
5019 [x86] nowatchdog: disable clocksource watchdog. Used
5020 in situations with strict latency requirements (where
5021 interruptions from clocksource watchdog are not
5024 tsx= [X86] Control Transactional Synchronization
5025 Extensions (TSX) feature in Intel processors that
5026 support TSX control.
5028 This parameter controls the TSX feature. The options are:
5030 on - Enable TSX on the system. Although there are
5031 mitigations for all known security vulnerabilities,
5032 TSX has been known to be an accelerator for
5033 several previous speculation-related CVEs, and
5034 so there may be unknown security risks associated
5035 with leaving it enabled.
5037 off - Disable TSX on the system. (Note that this
5038 option takes effect only on newer CPUs which are
5039 not vulnerable to MDS, i.e., have
5040 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get
5041 the new IA32_TSX_CTRL MSR through a microcode
5042 update. This new MSR allows for the reliable
5043 deactivation of the TSX functionality.)
5045 auto - Disable TSX if X86_BUG_TAA is present,
5046 otherwise enable TSX on the system.
5048 Not specifying this option is equivalent to tsx=off.
5050 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
5053 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async
5054 Abort (TAA) vulnerability.
5056 Similar to Micro-architectural Data Sampling (MDS)
5057 certain CPUs that support Transactional
5058 Synchronization Extensions (TSX) are vulnerable to an
5059 exploit against CPU internal buffers which can forward
5060 information to a disclosure gadget under certain
5063 In vulnerable processors, the speculatively forwarded
5064 data can be used in a cache side channel attack, to
5065 access data to which the attacker does not have direct
5068 This parameter controls the TAA mitigation. The
5071 full - Enable TAA mitigation on vulnerable CPUs
5074 full,nosmt - Enable TAA mitigation and disable SMT on
5075 vulnerable CPUs. If TSX is disabled, SMT
5076 is not disabled because CPU is not
5077 vulnerable to cross-thread TAA attacks.
5078 off - Unconditionally disable TAA mitigation
5080 On MDS-affected machines, tsx_async_abort=off can be
5081 prevented by an active MDS mitigation as both vulnerabilities
5082 are mitigated with the same mechanism so in order to disable
5083 this mitigation, you need to specify mds=off too.
5085 Not specifying this option is equivalent to
5086 tsx_async_abort=full. On CPUs which are MDS affected
5087 and deploy MDS mitigation, TAA mitigation is not
5088 required and doesn't provide any additional
5092 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
5094 turbografx.map[2|3]= [HW,JOY]
5095 TurboGraFX parallel port interface
5097 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
5098 See also Documentation/input/devices/joystick-parport.rst
5100 udbg-immortal [PPC] When debugging early kernel crashes that
5101 happen after console_init() and before a proper
5102 console driver takes over, this boot options might
5103 help "seeing" what's going on.
5105 uhash_entries= [KNL,NET]
5106 Set number of hash buckets for UDP/UDP-Lite connections
5109 [USB] Ignore overcurrent events (default N).
5110 Some badly-designed motherboards generate lots of
5111 bogus events, for ports that aren't wired to
5112 anything. Set this parameter to avoid log spamming.
5113 Note that genuine overcurrent events won't be
5117 [X86] Cause panic on unknown NMI.
5119 usbcore.authorized_default=
5120 [USB] Default USB device authorization:
5121 (default -1 = authorized except for wireless USB,
5122 0 = not authorized, 1 = authorized, 2 = authorized
5123 if device connected to internal port)
5125 usbcore.autosuspend=
5126 [USB] The autosuspend time delay (in seconds) used
5127 for newly-detected USB devices (default 2). This
5128 is the time required before an idle device will be
5129 autosuspended. Devices for which the delay is set
5130 to a negative value won't be autosuspended at all.
5132 usbcore.usbfs_snoop=
5133 [USB] Set to log all usbfs traffic (default 0 = off).
5135 usbcore.usbfs_snoop_max=
5136 [USB] Maximum number of bytes to snoop in each URB
5139 usbcore.blinkenlights=
5140 [USB] Set to cycle leds on hubs (default 0 = off).
5142 usbcore.old_scheme_first=
5143 [USB] Start with the old device initialization
5144 scheme, applies only to low and full-speed devices
5147 usbcore.usbfs_memory_mb=
5148 [USB] Memory limit (in MB) for buffers allocated by
5149 usbfs (default = 16, 0 = max = 2047).
5151 usbcore.use_both_schemes=
5152 [USB] Try the other device initialization scheme
5153 if the first one fails (default 1 = enabled).
5155 usbcore.initial_descriptor_timeout=
5156 [USB] Specifies timeout for the initial 64-byte
5157 USB_REQ_GET_DESCRIPTOR request in milliseconds
5158 (default 5000 = 5.0 seconds).
5160 usbcore.nousb [USB] Disable the USB subsystem
5163 [USB] A list of quirk entries to augment the built-in
5164 usb core quirk list. List entries are separated by
5165 commas. Each entry has the form
5166 VendorID:ProductID:Flags. The IDs are 4-digit hex
5167 numbers and Flags is a set of letters. Each letter
5168 will change the built-in quirk; setting it if it is
5169 clear and clearing it if it is set. The letters have
5170 the following meanings:
5171 a = USB_QUIRK_STRING_FETCH_255 (string
5172 descriptors must not be fetched using
5174 b = USB_QUIRK_RESET_RESUME (device can't resume
5175 correctly so reset it instead);
5176 c = USB_QUIRK_NO_SET_INTF (device can't handle
5177 Set-Interface requests);
5178 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't
5179 handle its Configuration or Interface
5181 e = USB_QUIRK_RESET (device can't be reset
5182 (e.g morph devices), don't use reset);
5183 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has
5184 more interface descriptions than the
5185 bNumInterfaces count, and can't handle
5186 talking to these interfaces);
5187 g = USB_QUIRK_DELAY_INIT (device needs a pause
5188 during initialization, after we read
5189 the device descriptor);
5190 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For
5191 high speed and super speed interrupt
5192 endpoints, the USB 2.0 and USB 3.0 spec
5193 require the interval in microframes (1
5194 microframe = 125 microseconds) to be
5195 calculated as interval = 2 ^
5197 Devices with this quirk report their
5198 bInterval as the result of this
5199 calculation instead of the exponent
5200 variable used in the calculation);
5201 i = USB_QUIRK_DEVICE_QUALIFIER (device can't
5202 handle device_qualifier descriptor
5204 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device
5205 generates spurious wakeup, ignore
5206 remote wakeup capability);
5207 k = USB_QUIRK_NO_LPM (device can't handle Link
5209 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL
5210 (Device reports its bInterval as linear
5211 frames instead of the USB 2.0
5213 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs
5214 to be disconnected before suspend to
5215 prevent spurious wakeup);
5216 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a
5217 pause after every control message);
5218 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra
5219 delay after resetting its port);
5220 Example: quirks=0781:5580:bk,0a5c:5834:gij
5223 [USBHID] The interval which mice are to be polled at.
5226 [USBHID] The interval which joysticks are to be polled at.
5229 [USBHID] The interval which keyboards are to be polled at.
5231 usb-storage.delay_use=
5232 [UMS] The delay in seconds before a new device is
5233 scanned for Logical Units (default 1).
5236 [UMS] A list of quirks entries to supplement or
5237 override the built-in unusual_devs list. List
5238 entries are separated by commas. Each entry has
5239 the form VID:PID:Flags where VID and PID are Vendor
5240 and Product ID values (4-digit hex numbers) and
5241 Flags is a set of characters, each corresponding
5242 to a common usb-storage quirk flag as follows:
5243 a = SANE_SENSE (collect more than 18 bytes
5244 of sense data, not on uas);
5245 b = BAD_SENSE (don't collect more than 18
5246 bytes of sense data, not on uas);
5247 c = FIX_CAPACITY (decrease the reported
5248 device capacity by one sector);
5249 d = NO_READ_DISC_INFO (don't use
5250 READ_DISC_INFO command, not on uas);
5251 e = NO_READ_CAPACITY_16 (don't use
5252 READ_CAPACITY_16 command);
5253 f = NO_REPORT_OPCODES (don't use report opcodes
5255 g = MAX_SECTORS_240 (don't transfer more than
5256 240 sectors at a time, uas only);
5257 h = CAPACITY_HEURISTICS (decrease the
5258 reported device capacity by one
5259 sector if the number is odd);
5260 i = IGNORE_DEVICE (don't bind to this
5262 j = NO_REPORT_LUNS (don't use report luns
5264 l = NOT_LOCKABLE (don't try to lock and
5265 unlock ejectable media, not on uas);
5266 m = MAX_SECTORS_64 (don't transfer more
5267 than 64 sectors = 32 KB at a time,
5269 n = INITIAL_READ10 (force a retry of the
5270 initial READ(10) command, not on uas);
5271 o = CAPACITY_OK (accept the capacity
5272 reported by the device, not on uas);
5273 p = WRITE_CACHE (the device cache is ON
5274 by default, not on uas);
5275 r = IGNORE_RESIDUE (the device reports
5276 bogus residue values, not on uas);
5277 s = SINGLE_LUN (the device has only one
5279 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
5280 commands, uas only);
5281 u = IGNORE_UAS (don't bind to the uas driver);
5282 w = NO_WP_DETECT (don't test whether the
5283 medium is write-protected).
5284 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE
5285 even if the device claims no cache,
5287 Example: quirks=0419:aaf5:rl,0421:0433:rc
5289 user_debug= [KNL,ARM]
5291 See arch/arm/Kconfig.debug help text.
5292 1 - undefined instruction events
5294 4 - invalid data aborts
5297 Example: user_debug=31
5300 [X86] Flags controlling user PTE allocations.
5302 nohigh = do not allocate PTE pages in
5303 HIGHMEM regardless of setting
5307 On X86_32, this is an alias for vdso32=. Otherwise:
5309 vdso=1: enable VDSO (the default)
5310 vdso=0: disable VDSO mapping
5312 vdso32= [X86] Control the 32-bit vDSO
5313 vdso32=1: enable 32-bit VDSO
5314 vdso32=0 or vdso32=2: disable 32-bit VDSO
5316 See the help text for CONFIG_COMPAT_VDSO for more
5317 details. If CONFIG_COMPAT_VDSO is set, the default is
5318 vdso32=0; otherwise, the default is vdso32=1.
5320 For compatibility with older kernels, vdso32=2 is an
5323 Try vdso32=0 if you encounter an error that says:
5324 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
5327 vector=percpu: enable percpu vector domain
5329 video= [FB] Frame buffer configuration
5330 See Documentation/fb/modedb.rst.
5332 video.brightness_switch_enabled= [0,1]
5333 If set to 1, on receiving an ACPI notify event
5334 generated by hotkey, video driver will adjust brightness
5335 level and then send out the event to user space through
5336 the allocated input device; If set to 0, video driver
5337 will only send out the event without touching backlight
5342 [VMMIO] Memory mapped virtio (platform) device.
5344 <size>@<baseaddr>:<irq>[:<id>]
5346 <size> := size (can use standard suffixes
5348 <baseaddr> := physical base address
5349 <irq> := interrupt number (as passed to
5351 <id> := (optional) platform device id
5353 virtio_mmio.device=1K@0x100b0000:48:7
5355 Can be used multiple times for multiple devices.
5357 vga= [BOOT,X86-32] Select a particular video mode
5358 See Documentation/x86/boot.rst and
5359 Documentation/admin-guide/svga.rst.
5360 Use vga=ask for menu.
5361 This is actually a boot loader parameter; the value is
5362 passed to the kernel using a special protocol.
5364 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y.
5365 May slow down system boot speed, especially when
5366 enabled on systems with a large amount of memory.
5367 All options are enabled by default, and this
5368 interface is meant to allow for selectively
5369 enabling or disabling specific virtual memory
5372 Available options are:
5373 P Enable page structure init time poisoning
5374 - Disable all of the above options
5376 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
5377 size of <nn>. This can be used to increase the
5378 minimum size (128MB on x86). It can also be used to
5379 decrease the size and leave more room for directly
5382 vmcp_cma=nn[MG] [KNL,S390]
5383 Sets the memory size reserved for contiguous memory
5384 allocations for the vmcp device driver.
5386 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
5389 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
5392 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
5396 Controls the behavior of vsyscalls (i.e. calls to
5397 fixed addresses of 0xffffffffff600x00 from legacy
5398 code). Most statically-linked binaries and older
5399 versions of glibc use these calls. Because these
5400 functions are at fixed addresses, they make nice
5401 targets for exploits that can control RIP.
5403 emulate [default] Vsyscalls turn into traps and are
5404 emulated reasonably safely. The vsyscall
5407 xonly Vsyscalls turn into traps and are
5408 emulated reasonably safely. The vsyscall
5409 page is not readable.
5411 none Vsyscalls don't work at all. This makes
5412 them quite hard to use for exploits but
5413 might break your system.
5415 vt.color= [VT] Default text color.
5416 Format: 0xYX, X = foreground, Y = background.
5417 Default: 0x07 = light gray on black.
5419 vt.cur_default= [VT] Default cursor shape.
5420 Format: 0xCCBBAA, where AA, BB, and CC are the same as
5421 the parameters of the <Esc>[?A;B;Cc escape sequence;
5422 see VGA-softcursor.txt. Default: 2 = underline.
5424 vt.default_blu= [VT]
5425 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
5426 Change the default blue palette of the console.
5427 This is a 16-member array composed of values
5430 vt.default_grn= [VT]
5431 Format: <green0>,<green1>,<green2>,...,<green15>
5432 Change the default green palette of the console.
5433 This is a 16-member array composed of values
5436 vt.default_red= [VT]
5437 Format: <red0>,<red1>,<red2>,...,<red15>
5438 Change the default red palette of the console.
5439 This is a 16-member array composed of values
5445 Set system-wide default UTF-8 mode for all tty's.
5446 Default is 1, i.e. UTF-8 mode is enabled for all
5447 newly opened terminals.
5449 vt.global_cursor_default=
5452 Set system-wide default for whether a cursor
5453 is shown on new VTs. Default is -1,
5454 i.e. cursors will be created by default unless
5455 overridden by individual drivers. 0 will hide
5456 cursors, 1 will display them.
5458 vt.italic= [VT] Default color for italic text; 0-15.
5461 vt.underline= [VT] Default color for underlined text; 0-15.
5464 watchdog timers [HW,WDT] For information on watchdog timers,
5465 see Documentation/watchdog/watchdog-parameters.rst
5466 or other driver-specific files in the
5467 Documentation/watchdog/ directory.
5471 Set the hard lockup detector stall duration
5472 threshold in seconds. The soft lockup detector
5473 threshold is set to twice the value. A value of 0
5474 disables both lockup detectors. Default is 10
5477 workqueue.watchdog_thresh=
5478 If CONFIG_WQ_WATCHDOG is configured, workqueue can
5479 warn stall conditions and dump internal state to
5480 help debugging. 0 disables workqueue stall
5481 detection; otherwise, it's the stall threshold
5482 duration in seconds. The default value is 30 and
5483 it can be updated at runtime by writing to the
5484 corresponding sysfs file.
5486 workqueue.disable_numa
5487 By default, all work items queued to unbound
5488 workqueues are affine to the NUMA nodes they're
5489 issued on, which results in better behavior in
5490 general. If NUMA affinity needs to be disabled for
5491 whatever reason, this option can be used. Note
5492 that this also can be controlled per-workqueue for
5493 workqueues visible under /sys/bus/workqueue/.
5495 workqueue.power_efficient
5496 Per-cpu workqueues are generally preferred because
5497 they show better performance thanks to cache
5498 locality; unfortunately, per-cpu workqueues tend to
5499 be more power hungry than unbound workqueues.
5501 Enabling this makes the per-cpu workqueues which
5502 were observed to contribute significantly to power
5503 consumption unbound, leading to measurably lower
5504 power usage at the cost of small performance
5507 The default value of this parameter is determined by
5508 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
5510 workqueue.debug_force_rr_cpu
5511 Workqueue used to implicitly guarantee that work
5512 items queued without explicit CPU specified are put
5513 on the local CPU. This guarantee is no longer true
5514 and while local CPU is still preferred work items
5515 may be put on foreign CPUs. This debug option
5516 forces round-robin CPU selection to flush out
5517 usages which depend on the now broken guarantee.
5518 When enabled, memory and cache locality will be
5521 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
5522 default x2apic cluster mode on platforms
5525 x86_intel_mid_timer= [X86-32,APBT]
5526 Choose timer option for x86 Intel MID platform.
5527 Two valid options are apbt timer only and lapic timer
5528 plus one apbt timer for broadcast timer.
5529 x86_intel_mid_timer=apbt_only | lapic_and_apbt
5531 xen_512gb_limit [KNL,X86-64,XEN]
5532 Restricts the kernel running paravirtualized under Xen
5533 to use only up to 512 GB of RAM. The reason to do so is
5534 crash analysis tools and Xen tools for doing domain
5535 save/restore/migration must be enabled to handle larger
5538 xen_emul_unplug= [HW,X86,XEN]
5539 Unplug Xen emulated devices
5540 Format: [unplug0,][unplug1]
5541 ide-disks -- unplug primary master IDE devices
5542 aux-ide-disks -- unplug non-primary-master IDE devices
5543 nics -- unplug network devices
5544 all -- unplug all emulated devices (NICs and IDE disks)
5545 unnecessary -- unplugging emulated devices is
5546 unnecessary even if the host did not respond to
5548 never -- do not unplug even if version check succeeds
5550 xen_legacy_crash [X86,XEN]
5551 Crash from Xen panic notifier, without executing late
5552 panic() code such as dumping handler.
5554 xen_nopvspin [X86,XEN]
5555 Disables the ticketlock slowpath using Xen PV
5559 Disables the PV optimizations forcing the HVM guest to
5560 run as generic HVM guest with no PV drivers.
5561 This option is obsoleted by the "nopv" option, which
5562 has equivalent effect for XEN platform.
5564 xen_scrub_pages= [XEN]
5565 Boolean option to control scrubbing pages before giving them back
5566 to Xen, for use by other domains. Can be also changed at runtime
5567 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages.
5568 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT.
5570 xen_timer_slop= [X86-64,XEN]
5571 Set the timer slop (in nanoseconds) for the virtual Xen
5572 timers (default is 100000). This adjusts the minimum
5573 delta of virtualized Xen timers, where lower values
5574 improve timer resolution at the expense of processing
5575 more timer interrupts.
5577 nopv= [X86,XEN,KVM,HYPER_V,VMWARE]
5578 Disables the PV optimizations forcing the guest to run
5579 as generic guest with no PV drivers. Currently support
5580 XEN HVM, KVM, HYPER_V and VMWARE guest.
5582 xirc2ps_cs= [NET,PCMCIA]
5584 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
5587 By default on POWER9 and above, the kernel will
5588 natively use the XIVE interrupt controller. This option
5589 allows the fallback firmware mode to be used:
5591 off Fallback to firmware control of XIVE interrupt
5592 controller on both pseries and powernv
5593 platforms. Only useful on POWER9 and above.
5595 xhci-hcd.quirks [USB,KNL]
5596 A hex value specifying bitmask with supplemental xhci
5597 host controller quirks. Meaning of each bit can be
5598 consulted in header drivers/usb/host/xhci.h.
5601 Format: { early | on | rw | ro | off }
5602 Controls if xmon debugger is enabled. Default is off.
5603 Passing only "xmon" is equivalent to "xmon=early".
5604 early Call xmon as early as possible on boot; xmon
5605 debugger is called from setup_arch().
5606 on xmon debugger hooks will be installed so xmon
5607 is only called on a kernel crash. Default mode,
5608 i.e. either "ro" or "rw" mode, is controlled
5609 with CONFIG_XMON_DEFAULT_RO_MODE.
5610 rw xmon debugger hooks will be installed so xmon
5611 is called only on a kernel crash, mode is write,
5612 meaning SPR registers, memory and, other data
5613 can be written using xmon commands.
5614 ro same as "rw" option above but SPR registers,
5615 memory, and other data can't be written using
5617 off xmon is disabled.
projects / linux-2.6-microblaze.git / blob