1 acpi= [HW,ACPI,X86,ARM64]
2 Advanced Configuration and Power Interface
3 Format: { force | on | off | strict | noirq | rsdt |
5 force -- enable ACPI if default was off
6 on -- enable ACPI but allow fallback to DT [arm64]
7 off -- disable ACPI if default was on
8 noirq -- do not use ACPI for IRQ routing
9 strict -- Be less tolerant of platforms that are not
10 strictly ACPI specification compliant.
11 rsdt -- prefer RSDT over (default) XSDT
12 copy_dsdt -- copy DSDT to memory
13 For ARM64, ONLY "acpi=off", "acpi=on" or "acpi=force"
16 See also Documentation/power/runtime_pm.rst, pci=noacpi
18 acpi_apic_instance= [ACPI, IOAPIC]
20 2: use 2nd APIC table, if available
21 1,0: use 1st APIC table
24 acpi_backlight= [HW,ACPI]
25 { vendor | video | native | none }
26 If set to vendor, prefer vendor-specific driver
27 (e.g. thinkpad_acpi, sony_acpi, etc.) instead
28 of the ACPI video.ko driver.
29 If set to video, use the ACPI video.ko driver.
30 If set to native, use the device's native backlight mode.
31 If set to none, disable the ACPI backlight interface.
33 acpi_force_32bit_fadt_addr
34 force FADT to use 32 bit addresses rather than the
35 64 bit X_* addresses. Some firmware have broken 64
36 bit addresses for force ACPI ignore these and use
37 the older legacy 32 bit addresses.
39 acpica_no_return_repair [HW, ACPI]
40 Disable AML predefined validation mechanism
41 This mechanism can repair the evaluation result to make
42 the return objects more ACPI specification compliant.
43 This option is useful for developers to identify the
44 root cause of an AML interpreter issue when the issue
45 has something to do with the repair mechanism.
47 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG]
48 acpi.debug_level= [HW,ACPI,ACPI_DEBUG]
50 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
51 debug output. Bits in debug_layer correspond to a
52 _COMPONENT in an ACPI source file, e.g.,
53 #define _COMPONENT ACPI_EVENTS
54 Bits in debug_level correspond to a level in
55 ACPI_DEBUG_PRINT statements, e.g.,
56 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
57 The debug_level mask defaults to "info". See
58 Documentation/firmware-guide/acpi/debug.rst for more information about
59 debug layers and levels.
61 Enable processor driver info messages:
62 acpi.debug_layer=0x20000000
63 Enable 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
116 Format: <byte> or <bitmap-list>
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
304 force_enable - Force enable the IOMMU on platforms known
305 to be buggy with IOMMU enabled. Use this
308 amd_iommu_dump= [HW,X86-64]
309 Enable AMD IOMMU driver option to dump the ACPI table
310 for AMD IOMMU. With this option enabled, AMD IOMMU
311 driver will print ACPI tables for AMD IOMMU during
312 IOMMU initialization.
314 amd_iommu_intr= [HW,X86-64]
315 Specifies one of the following AMD IOMMU interrupt
317 legacy - Use legacy interrupt remapping mode.
318 vapic - Use virtual APIC mode, which allows IOMMU
319 to inject interrupts directly into guest.
320 This mode requires kvm-amd.avic=1.
321 (Default when IOMMU HW support is present.)
323 amijoy.map= [HW,JOY] Amiga joystick support
324 Map of devices attached to JOY0DAT and JOY1DAT
326 See also Documentation/input/joydev/joystick.rst
328 analog.map= [HW,JOY] Analog joystick and gamepad support
329 Specifies type or capabilities of an analog joystick
330 connected to one of 16 gameports
331 Format: <type1>,<type2>,..<type16>
334 Power management functions (SPARCstation-4/5 + deriv.)
336 Disable APC CPU standby support. SPARCstation-Fox does
337 not play well with APC CPU idle - disable it if you have
338 APC and your system crashes randomly.
340 apic= [APIC,X86] Advanced Programmable Interrupt Controller
341 Change the output verbosity while booting
342 Format: { quiet (default) | verbose | debug }
343 Change the amount of debugging information output
344 when initialising the APIC and IO-APIC components.
345 For X86-32, this can also be used to specify an APIC
347 Format: apic=driver_name
348 Examples: apic=bigsmp
350 apic_extnmi= [APIC,X86] External NMI delivery setting
351 Format: { bsp (default) | all | none }
352 bsp: External NMI is delivered only to CPU 0
353 all: External NMIs are broadcast to all CPUs as a
355 none: External NMI is masked for all CPUs. This is
356 useful so that a dump capture kernel won't be
360 See Documentation/networking/ipv6.rst.
362 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
363 Limit apic dumping. The parameter defines the maximal
364 number of local apics being dumped. Also it is possible
365 to set it to "all" by meaning -- no limit here.
366 Format: { 1 (default) | 2 | ... | all }.
367 The parameter valid if only apic=debug or
368 apic=verbose is specified.
369 Example: apic=debug show_lapic=all
371 apm= [APM] Advanced Power Management
372 See header of arch/x86/kernel/apm_32.c.
374 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
375 Format: <io>,<irq>,<nodeID>
377 arm64.nobti [ARM64] Unconditionally disable Branch Target
378 Identification support
380 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication
385 atarimouse= [HW,MOUSE] Atari Mouse
387 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
388 EzKey and similar keyboards
390 atkbd.reset= [HW] Reset keyboard during initialization
392 atkbd.set= [HW] Select keyboard code set
393 Format: <int> (2 = AT (default), 3 = PS/2)
395 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
398 atkbd.softraw= [HW] Choose between synthetic and real raw mode
399 Format: <bool> (0 = real, 1 = synthetic (default))
401 atkbd.softrepeat= [HW]
402 Use software keyboard repeat
404 audit= [KNL] Enable the audit sub-system
405 Format: { "0" | "1" | "off" | "on" }
406 0 | off - kernel audit is disabled and can not be
407 enabled until the next reboot
408 unset - kernel audit is initialized but disabled and
409 will be fully enabled by the userspace auditd.
410 1 | on - kernel audit is initialized and partially
411 enabled, storing at most audit_backlog_limit
412 messages in RAM until it is fully enabled by the
416 audit_backlog_limit= [KNL] Set the audit queue size limit.
417 Format: <int> (must be >=0)
420 bau= [X86_UV] Enable the BAU on SGI UV. The default
421 behavior is to disable the BAU (i.e. bau=0).
422 Format: { "0" | "1" }
425 unset - Disable the BAU.
427 baycom_epp= [HW,AX25]
430 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
432 See header of drivers/net/hamradio/baycom_par.c.
434 baycom_ser_fdx= [HW,AX25]
435 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
436 Format: <io>,<irq>,<mode>[,<baud>]
437 See header of drivers/net/hamradio/baycom_ser_fdx.c.
439 baycom_ser_hdx= [HW,AX25]
440 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
441 Format: <io>,<irq>,<mode>
442 See header of drivers/net/hamradio/baycom_ser_hdx.c.
444 blkdevparts= Manual partition parsing of block device(s) for
445 embedded devices based on command line input.
446 See Documentation/block/cmdline-partition.rst
448 boot_delay= Milliseconds to delay each printk during boot.
449 Values larger than 10 seconds (10000) are changed to
454 Extended command line options can be added to an initrd
455 and this will cause the kernel to look for it.
457 See Documentation/admin-guide/bootconfig.rst
460 Disable BERT OS support on buggy BIOSes.
462 bgrt_disable [ACPI][X86]
463 Disable BGRT to avoid flickering OEM logo.
465 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
466 bttv.radio= Most important insmod options are available as
468 bttv.pll= See Documentation/admin-guide/media/bttv.rst
471 bulk_remove=off [PPC] This parameter disables the use of the pSeries
472 firmware feature for flushing multiple hpte entries
475 c101= [NET] Moxa C101 synchronous serial card
477 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
478 Sometimes CPU hardware bugs make them report the cache
479 size incorrectly. The kernel will attempt work arounds
480 to fix known problems, but for some CPUs it is not
481 possible to determine what the correct size should be.
482 This option provides an override for these situations.
485 [NET] Specifies amount of time (in seconds) that
486 the kernel should wait for a network carrier. By default
487 it waits 120 seconds.
489 ca_keys= [KEYS] This parameter identifies a specific key(s) on
490 the system trusted keyring to be used for certificate
492 format: { id:<keyid> | builtin }
494 cca= [MIPS] Override the kernel pages' cache coherency
495 algorithm. Accepted values range from 0 to 7
496 inclusive. See arch/mips/include/asm/pgtable-bits.h
497 for platform specific values (SB1, Loongson3 and
500 ccw_timeout_log [S390]
501 See Documentation/s390/common_io.rst for details.
503 cgroup_disable= [KNL] Disable a particular controller or optional feature
504 Format: {name of the controller(s) or feature(s) to disable}
505 The effects of cgroup_disable=foo are:
506 - foo isn't auto-mounted if you mount all cgroups in
508 - foo isn't visible as an individually mountable
510 - if foo is an optional feature then the feature is
511 disabled and corresponding cgroup files are not
513 {Currently only "memory" controller deal with this and
514 cut the overhead, others just disable the usage. So
515 only cgroup_disable=memory is actually worthy}
516 Specifying "pressure" disables per-cgroup pressure
517 stall information accounting feature
519 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1
520 Format: { { controller | "all" | "named" }
521 [,{ controller | "all" | "named" }...] }
522 Like cgroup_disable, but only applies to cgroup v1;
523 the blacklisted controllers remain available in cgroup2.
524 "all" blacklists all controllers and "named" disables
525 named mounts. Specifying both "all" and "named" disables
528 cgroup.memory= [KNL] Pass options to the cgroup memory controller.
530 nosocket -- Disable socket memory accounting.
531 nokmem -- Disable kernel memory accounting.
533 checkreqprot [SELINUX] Set initial checkreqprot flag value.
534 Format: { "0" | "1" }
535 See security/selinux/Kconfig help text.
536 0 -- check protection applied by kernel (includes
537 any implied execute protection).
538 1 -- check protection requested by application.
539 Default value is set via a kernel config option.
540 Value can be changed at runtime via
541 /sys/fs/selinux/checkreqprot.
542 Setting checkreqprot to 1 is deprecated.
545 See Documentation/s390/common_io.rst for details.
548 Prevents the clock framework from automatically gating
549 clocks that have not been explicitly enabled by a Linux
550 device driver but are enabled in hardware at reset or
551 by the bootloader/firmware. Note that this does not
552 force such clocks to be always-on nor does it reserve
553 those clocks in any way. This parameter is useful for
554 debug and development, but should not be needed on a
555 platform with proper driver support. For more
556 information, see Documentation/driver-api/clk.rst.
558 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
560 Forces specified clocksource (if available) to be used
561 when calculating gettimeofday(). If specified
562 clocksource is not available, it defaults to PIT.
563 Format: { pit | tsc | cyclone | pmtmr }
565 clocksource= Override the default clocksource
567 Override the default clocksource and use the clocksource
568 with the name specified.
569 Some clocksource names to choose from, depending on
571 [all] jiffies (this is the base, fallback clocksource)
573 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
574 pxa_timer,timer3,32k_counter,timer0_1
575 [X86-32] pit,hpet,tsc;
576 scx200_hrt on Geode; cyclone on IBM x440
584 clocksource.arm_arch_timer.evtstrm=
587 Enable/disable the eventstream feature of the ARM
588 architected timer so that code using WFE-based polling
589 loops can be debugged more effectively on production
592 clocksource.max_cswd_read_retries= [KNL]
593 Number of clocksource_watchdog() retries due to
594 external delays before the clock will be marked
595 unstable. Defaults to three retries, that is,
596 four attempts to read the clock under test.
598 clocksource.verify_n_cpus= [KNL]
599 Limit the number of CPUs checked for clocksources
600 marked with CLOCK_SOURCE_VERIFY_PERCPU that
601 are marked unstable due to excessive skew.
602 A negative value says to check all CPUs, while
603 zero says not to check any. Values larger than
604 nr_cpu_ids are silently truncated to nr_cpu_ids.
605 The actual CPUs are chosen randomly, with
606 no replacement if the same CPU is chosen twice.
608 clocksource-wdtest.holdoff= [KNL]
609 Set the time in seconds that the clocksource
610 watchdog test waits before commencing its tests.
611 Defaults to zero when built as a module and to
612 10 seconds when built into the kernel.
614 clearcpuid=BITNUM[,BITNUM...] [X86]
615 Disable CPUID feature X for the kernel. See
616 arch/x86/include/asm/cpufeatures.h for the valid bit
617 numbers. Note the Linux specific bits are not necessarily
618 stable over kernel options, but the vendor specific
620 Also note that user programs calling CPUID directly
621 or using the feature without checking anything
622 will still see it. This just prevents it from
623 being used by the kernel or shown in /proc/cpuinfo.
624 Also note the kernel might malfunction if you disable
627 cma=nn[MG]@[start[MG][-end[MG]]]
629 Sets the size of kernel global memory area for
630 contiguous memory allocations and optionally the
631 placement constraint by the physical address range of
632 memory allocations. A value of 0 disables CMA
633 altogether. For more information, see
634 kernel/dma/contiguous.c
638 Sets the size of kernel per-numa memory area for
639 contiguous memory allocations. A value of 0 disables
640 per-numa CMA altogether. And If this option is not
641 specificed, the default value is 0.
642 With per-numa CMA enabled, DMA users on node nid will
643 first try to allocate buffer from the pernuma area
644 which is located in node nid, if the allocation fails,
645 they will fallback to the global default memory area.
647 cmo_free_hint= [PPC] Format: { yes | no }
648 Specify whether pages are marked as being inactive
649 when they are freed. This is used in CMO environments
650 to determine OS memory pressure for page stealing by
654 coherent_pool=nn[KMG] [ARM,KNL]
655 Sets the size of memory pool for coherent, atomic dma
656 allocations, by default set to 256K.
658 com20020= [HW,NET] ARCnet - COM20020 chipset
660 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
662 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
666 ARCnet - COM90xx chipset (memory-mapped buffers)
667 Format: <io>[,<irq>[,<memstart>]]
669 condev= [HW,S390] console device
672 console= [KNL] Output console device and options.
674 tty<n> Use the virtual console device <n>.
678 Use the specified serial port. The options are of
679 the form "bbbbpnf", where "bbbb" is the baud rate,
680 "p" is parity ("n", "o", or "e"), "n" is number of
681 bits, and "f" is flow control ("r" for RTS or
682 omit it). Default is "9600n8".
684 See Documentation/admin-guide/serial-console.rst for more
686 Documentation/networking/netconsole.rst for an
689 uart[8250],io,<addr>[,options]
690 uart[8250],mmio,<addr>[,options]
691 uart[8250],mmio16,<addr>[,options]
692 uart[8250],mmio32,<addr>[,options]
693 uart[8250],0x<addr>[,options]
694 Start an early, polled-mode console on the 8250/16550
695 UART at the specified I/O port or MMIO address,
696 switching to the matching ttyS device later.
697 MMIO inter-register address stride is either 8-bit
698 (mmio), 16-bit (mmio16), or 32-bit (mmio32).
699 If none of [io|mmio|mmio16|mmio32], <addr> is assumed
700 to be equivalent to 'mmio'. 'options' are specified in
701 the same format described for ttyS above; if unspecified,
702 the h/w is not re-initialized.
704 hvc<n> Use the hypervisor console device <n>. This is for
705 both Xen and PowerPC hypervisors.
707 If the device connected to the port is not a TTY but a braille
708 device, prepend "brl," before the device type, for instance
710 For now, only VisioBraille is supported.
713 [KNL] Change console messages format
715 By default we print messages on consoles in
716 "[time stamp] text\n" format (time stamp may not be
717 printed, depending on CONFIG_PRINTK_TIME or
718 `printk_time' param).
720 Switch to syslog format: "<%u>[time stamp] text\n"
721 IOW, each message will have a facility and loglevel
722 prefix. The format is similar to one used by syslog()
723 syscall, or to executing "dmesg -S --raw" or to reading
726 consoleblank= [KNL] The console blank (screen saver) timeout in
727 seconds. A value of 0 disables the blank timer.
731 [KNL] Change the default value for
732 /proc/<pid>/coredump_filter.
733 See also Documentation/filesystems/proc.rst.
735 coresight_cpu_debug.enable
738 Enable/disable the CPU sampling based debugging.
739 0: default value, disable debugging
740 1: enable debugging at boot time
742 cpuidle.off=1 [CPU_IDLE]
743 disable the cpuidle sub-system
746 [CPU_IDLE] Name of the cpuidle governor to use.
748 cpufreq.off=1 [CPU_FREQ]
749 disable the cpufreq sub-system
751 cpufreq.default_governor=
752 [CPU_FREQ] Name of the default cpufreq governor or
753 policy to use. This governor must be registered in the
754 kernel before the cpufreq driver probes.
757 [X86] Delay for N microsec between assert and de-assert
758 of APIC INIT to start processors. This delay occurs
759 on every CPU online, such as boot, and resume from suspend.
762 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
764 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
766 crashkernel=size[KMG][@offset[KMG]]
767 [KNL] Using kexec, Linux can switch to a 'crash kernel'
768 upon panic. This parameter reserves the physical
769 memory region [offset, offset + size] for that kernel
770 image. If '@offset' is omitted, then a suitable offset
771 is selected automatically.
772 [KNL, X86-64] Select a region under 4G first, and
773 fall back to reserve region above 4G when '@offset'
774 hasn't been specified.
775 See Documentation/admin-guide/kdump/kdump.rst for further details.
777 crashkernel=range1:size1[,range2:size2,...][@offset]
778 [KNL] Same as above, but depends on the memory
779 in the running system. The syntax of range is
780 start-[end] where start and end are both
781 a memory unit (amount[KMG]). See also
782 Documentation/admin-guide/kdump/kdump.rst for an example.
784 crashkernel=size[KMG],high
785 [KNL, X86-64] range could be above 4G. Allow kernel
786 to allocate physical memory region from top, so could
787 be above 4G if system have more than 4G ram installed.
788 Otherwise memory region will be allocated below 4G, if
790 It will be ignored if crashkernel=X is specified.
791 crashkernel=size[KMG],low
792 [KNL, X86-64] range under 4G. When crashkernel=X,high
793 is passed, kernel could allocate physical memory region
794 above 4G, that cause second kernel crash on system
795 that require some amount of low memory, e.g. swiotlb
796 requires at least 64M+32K low memory, also enough extra
797 low memory is needed to make sure DMA buffers for 32-bit
798 devices won't run out. Kernel would try to allocate at
799 at least 256M below 4G automatically.
800 This one let user to specify own low range under 4G
801 for second kernel instead.
802 0: to disable low allocation.
803 It will be ignored when crashkernel=X,high is not used
804 or memory reserved is below 4G.
807 [KNL] Disable crypto self-tests
812 cs89x0_media= [HW,NET]
813 Format: { rj45 | aui | bnc }
815 csdlock_debug= [KNL] Enable debug add-ons of cross-CPU function call
816 handling. When switched on, additional debug data is
817 printed to the console in case a hanging CPU is
818 detected, and that CPU is pinged again in order to try
819 to resolve the hang situation.
820 0: disable csdlock debugging (default)
821 1: enable basic csdlock debugging (minor impact)
822 ext: enable extended csdlock debugging (more impact,
826 See header of drivers/s390/block/dasd_devmap.c.
828 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
829 (one device per port)
830 Format: <port#>,<type>
831 See also Documentation/input/devices/joystick-parport.rst
833 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
835 Documentation/admin-guide/dynamic-debug-howto.rst for
836 details. Deprecated, see dyndbg.
838 debug [KNL] Enable kernel debugging (events log level).
841 [KNL] Enable printing [hashed] pointers early in the
842 boot sequence. If enabled, we use a weak hash instead
843 of siphash to hash pointers. Use this option if you are
844 seeing instances of '(___ptrval___)') and need to see a
845 value (hashed pointer) instead. Cryptographically
846 insecure, please do not use on production kernels.
849 [KNL] verbose locking self-tests
851 Print debugging info while doing the locking API
853 Bitmask for the various LOCKTYPE_ tests. Defaults to 0
854 (no extra messages), setting it to -1 (all bits set)
855 will print _a_lot_ more information - normally only
856 useful to lockdep developers.
858 debug_objects [KNL] Enable object debugging
861 [KNL] Disable object debugging
863 debug_guardpage_minorder=
864 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
865 parameter allows control of the order of pages that will
866 be intentionally kept free (and hence protected) by the
867 buddy allocator. Bigger value increase the probability
868 of catching random memory corruption, but reduce the
869 amount of memory for normal system use. The maximum
870 possible value is MAX_ORDER/2. Setting this parameter
871 to 1 or 2 should be enough to identify most random
872 memory corruption problems caused by bugs in kernel or
873 driver code when a CPU writes to (or reads from) a
874 random memory location. Note that there exists a class
875 of memory corruptions problems caused by buggy H/W or
876 F/W or by drivers badly programing DMA (basically when
877 memory is written at bus level and the CPU MMU is
878 bypassed) which are not detectable by
879 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
880 tracking down these problems.
883 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter
884 enables the feature at boot time. By default, it is
885 disabled and the system will work mostly the same as a
886 kernel built without CONFIG_DEBUG_PAGEALLOC.
887 Note: to get most of debug_pagealloc error reports, it's
888 useful to also enable the page_owner functionality.
889 on: enable the feature
891 debugfs= [KNL] This parameter enables what is exposed to userspace
892 and debugfs internal clients.
893 Format: { on, no-mount, off }
894 on: All functions are enabled.
896 Filesystem is not registered but kernel clients can
897 access APIs and a crashkernel can be used to read
898 its content. There is nothing to mount.
899 off: Filesystem is not registered and clients
900 get a -EPERM as result when trying to register files
901 or directories within debugfs.
902 This is equivalent of the runtime functionality if
903 debugfs was not enabled in the kernel at all.
904 Default value is set in build-time with a kernel configuration.
906 debugpat [X86] Enable PAT debugging
908 decnet.addr= [HW,NET]
909 Format: <area>[,<node>]
910 See also Documentation/networking/decnet.rst.
913 [HW] The size of the default HugeTLB page. This is
914 the size represented by the legacy /proc/ hugepages
915 APIs. In addition, this is the default hugetlb size
916 used for shmget(), mmap() and mounting hugetlbfs
917 filesystems. If not specified, defaults to the
918 architecture's default huge page size. Huge page
919 sizes are architecture dependent. See also
920 Documentation/admin-guide/mm/hugetlbpage.rst.
923 deferred_probe_timeout=
924 [KNL] Debugging option to set a timeout in seconds for
925 deferred probe to give up waiting on dependencies to
926 probe. Only specific dependencies (subsystems or
927 drivers) that have opted in will be ignored. A timeout of 0
928 will timeout at the end of initcalls. This option will also
929 dump out devices still on the deferred probe list after
933 Format: { on | off | def_only | inf_only | always }
934 on: s390 zlib hardware support for compression on
935 level 1 and decompression (default)
936 off: No s390 zlib hardware support
937 def_only: s390 zlib hardware support for deflate
938 only (compression on level 1)
939 inf_only: s390 zlib hardware support for inflate
941 always: Same as 'on' but ignores the selected compression
942 level always using hardware support (used for debugging)
945 Set number of hash buckets for dentry cache.
947 disable_1tb_segments [PPC]
948 Disables the use of 1TB hash page table segments. This
949 causes the kernel to fall back to 256MB segments which
950 can be useful when debugging issues that require an SLB
954 Limits the number of kernel SLB entries, and flushes
955 them frequently to increase the rate of SLB faults
959 See Documentation/networking/ipv6.rst.
962 [KNL] Under CONFIG_HARDENED_USERCOPY, whether
963 hardening is enabled for this boot. Hardened
964 usercopy checking is used to protect the kernel
965 from reading or writing beyond known memory
966 allocation boundaries as a proactive defense
967 against bounds-checking flaws in the kernel's
968 copy_to_user()/copy_from_user() interface.
969 on Perform hardened usercopy checks (default).
970 off Disable hardened usercopy checks.
973 Disable RADIX MMU mode on POWER9
975 radix_hcall_invalidate=on [PPC/PSERIES]
976 Disable RADIX GTSE feature and use hcall for TLB
980 Disable TLBIE instruction. Currently does not work
981 with KVM, with HASH MMU, or with coherent accelerators.
983 disable_cpu_apicid= [X86,APIC,SMP]
985 The number of initial APIC ID for the
986 corresponding CPU to be disabled at boot,
987 mostly used for the kdump 2nd kernel to
988 disable BSP to wake up multiple CPUs without
989 causing system reset or hang due to sending
992 disable_ddw [PPC/PSERIES]
993 Disable Dynamic DMA Window support. Use this
994 to workaround buggy firmware.
997 See Documentation/networking/ipv6.rst.
999 disable_mtrr_cleanup [X86]
1000 The kernel tries to adjust MTRR layout from continuous
1001 to discrete, to make X server driver able to add WB
1002 entry later. This parameter disables that.
1004 disable_mtrr_trim [X86, Intel and AMD only]
1005 By default the kernel will trim any uncacheable
1006 memory out of your available memory pool based on
1007 MTRR settings. This parameter disables that behavior,
1008 possibly causing your machine to run very slowly.
1010 disable_timer_pin_1 [X86]
1011 Disable PIN 1 of APIC timer
1012 Can be useful to work around chipset bugs.
1014 dis_ucode_ldr [X86] Disable the microcode loader.
1016 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
1017 this option disables the debugging code at boot.
1019 dma_debug_entries=<number>
1020 This option allows to tune the number of preallocated
1021 entries for DMA-API debugging code. One entry is
1022 required per DMA-API allocation. Use this if the
1023 DMA-API debugging code disables itself because the
1024 architectural default is too low.
1026 dma_debug_driver=<driver_name>
1027 With this option the DMA-API debugging driver
1028 filter feature can be enabled at boot time. Just
1029 pass the driver to filter for as the parameter.
1030 The filter can be disabled or changed to another
1031 driver later using sysfs.
1033 driver_async_probe= [KNL]
1034 List of driver names to be probed asynchronously.
1035 Format: <driver_name1>,<driver_name2>...
1037 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
1038 Broken monitors, graphic adapters, KVMs and EDIDless
1039 panels may send no or incorrect EDID data sets.
1040 This parameter allows to specify an EDID data sets
1041 in the /lib/firmware directory that are used instead.
1042 Generic built-in EDID data sets are used, if one of
1043 edid/1024x768.bin, edid/1280x1024.bin,
1044 edid/1680x1050.bin, or edid/1920x1080.bin is given
1045 and no file with the same name exists. Details and
1046 instructions how to build your own EDID data are
1047 available in Documentation/admin-guide/edid.rst. An EDID
1048 data set will only be used for a particular connector,
1049 if its name and a colon are prepended to the EDID
1050 name. Each connector may use a unique EDID data
1051 set by separating the files with a comma. An EDID
1052 data set with no connector name will be used for
1053 any connectors not explicitly specified.
1058 Format: {"off" | "known"}
1059 Control how the dt_cpu_ftrs device-tree binding is
1060 used for CPU feature discovery and setup (if it
1062 off: Do not use it, fall back to legacy cpu table.
1063 known: Do not pass through unknown features to guests
1064 or userspace, only those that the kernel is aware of.
1066 dump_apple_properties [X86]
1067 Dump name and content of EFI device properties on
1068 x86 Macs. Useful for driver authors to determine
1069 what data is available or for reverse-engineering.
1071 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
1072 <module>.dyndbg[="val"]
1073 Enable debug messages at boot time. See
1074 Documentation/admin-guide/dynamic-debug-howto.rst
1077 nopku [X86] Disable Memory Protection Keys CPU feature found
1080 <module>.async_probe [KNL]
1081 Enable asynchronous probe on this module.
1083 early_ioremap_debug [KNL]
1084 Enable debug messages in early_ioremap support. This
1085 is useful for tracking down temporary early mappings
1086 which are not unmapped.
1088 earlycon= [KNL] Output early console device and options.
1090 When used with no options, the early console is
1091 determined by stdout-path property in device tree's
1092 chosen node or the ACPI SPCR table if supported by
1095 cdns,<addr>[,options]
1096 Start an early, polled-mode console on a Cadence
1097 (xuartps) serial port at the specified address. Only
1098 supported option is baud rate. If baud rate is not
1099 specified, the serial port must already be setup and
1102 uart[8250],io,<addr>[,options]
1103 uart[8250],mmio,<addr>[,options]
1104 uart[8250],mmio32,<addr>[,options]
1105 uart[8250],mmio32be,<addr>[,options]
1106 uart[8250],0x<addr>[,options]
1107 Start an early, polled-mode console on the 8250/16550
1108 UART at the specified I/O port or MMIO address.
1109 MMIO inter-register address stride is either 8-bit
1110 (mmio) or 32-bit (mmio32 or mmio32be).
1111 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
1112 to be equivalent to 'mmio'. 'options' are specified
1113 in the same format described for "console=ttyS<n>"; if
1114 unspecified, the h/w is not initialized.
1118 Start an early, polled-mode console on a pl011 serial
1119 port at the specified address. The pl011 serial port
1120 must already be setup and configured. Options are not
1121 yet supported. If 'mmio32' is specified, then only
1122 the driver will use only 32-bit accessors to read/write
1123 the device registers.
1126 Start an early console on a litex serial port at the
1127 specified address. The serial port must already be
1128 setup and configured. Options are not yet supported.
1131 Start an early, polled-mode console on a meson serial
1132 port at the specified address. The serial port must
1133 already be setup and configured. Options are not yet
1137 Start an early, polled-mode console on an msm serial
1138 port at the specified address. The serial port
1139 must already be setup and configured. Options are not
1142 msm_serial_dm,<addr>
1143 Start an early, polled-mode console on an msm serial
1144 dm port at the specified address. The serial port
1145 must already be setup and configured. Options are not
1149 Start an early, polled-mode console on a serial port
1150 of an Actions Semi SoC, such as S500 or S900, at the
1151 specified address. The serial port must already be
1152 setup and configured. Options are not yet supported.
1155 Start an early, polled-mode console on a serial port
1156 of an RDA Micro SoC, such as RDA8810PL, at the
1157 specified address. The serial port must already be
1158 setup and configured. Options are not yet supported.
1161 Use RISC-V SBI (Supervisor Binary Interface) for early
1164 smh Use ARM semihosting calls for early console.
1172 Use early console provided by serial driver available
1173 on Samsung SoCs, requires selecting proper type and
1174 a correct base address of the selected UART port. The
1175 serial port must already be setup and configured.
1176 Options are not yet supported.
1179 Start an early, polled-mode console on a lantiq serial
1180 (lqasc) port at the specified address. The serial port
1181 must already be setup and configured. Options are not
1186 Use early console provided by Freescale LP UART driver
1187 found on Freescale Vybrid and QorIQ LS1021A processors.
1188 A valid base address must be provided, and the serial
1189 port must already be setup and configured.
1193 Start an early, polled-mode, output-only console on the
1194 Freescale i.MX UART at the specified address. The UART
1195 must already be setup and configured.
1198 Start an early, polled-mode console on the
1199 Armada 3700 serial port at the specified
1200 address. The serial port must already be setup
1201 and configured. Options are not yet supported.
1204 Start an early, polled-mode console on a Qualcomm
1205 Generic Interface (GENI) based serial port at the
1206 specified address. The serial port must already be
1207 setup and configured. Options are not yet supported.
1210 Start an early, unaccelerated console on the EFI
1211 memory mapped framebuffer (if available). On cache
1212 coherent non-x86 systems that use system memory for
1213 the framebuffer, pass the 'ram' option so that it is
1214 mapped with the correct attributes.
1217 Use early console provided by Freescale LINFlexD UART
1218 serial driver for NXP S32V234 SoCs. A valid base
1219 address must be provided, and the serial port must
1220 already be setup and configured.
1222 earlyprintk= [X86,SH,ARM,M68k,S390]
1226 earlyprintk=serial[,ttySn[,baudrate]]
1227 earlyprintk=serial[,0x...[,baudrate]]
1228 earlyprintk=ttySn[,baudrate]
1229 earlyprintk=dbgp[debugController#]
1230 earlyprintk=pciserial[,force],bus:device.function[,baudrate]
1231 earlyprintk=xdbc[xhciController#]
1233 earlyprintk is useful when the kernel crashes before
1234 the normal console is initialized. It is not enabled by
1235 default because it has some cosmetic problems.
1237 Append ",keep" to not disable it when the real console
1240 Only one of vga, efi, serial, or usb debug port can
1243 Currently only ttyS0 and ttyS1 may be specified by
1244 name. Other I/O ports may be explicitly specified
1245 on some architectures (x86 and arm at least) by
1246 replacing ttySn with an I/O port address, like this:
1247 earlyprintk=serial,0x1008,115200
1248 You can find the port for a given device in
1249 /proc/tty/driver/serial:
1250 2: uart:ST16650V2 port:00001008 irq:18 ...
1252 Interaction with the standard serial driver is not
1255 The VGA and EFI output is eventually overwritten by
1258 The xen output can only be used by Xen PV guests.
1260 The sclp output can only be used on s390.
1262 The optional "force" to "pciserial" enables use of a
1263 PCI device even when its classcode is not of the
1266 edac_report= [HW,EDAC] Control how to report EDAC event
1267 Format: {"on" | "off" | "force"}
1268 on: enable EDAC to report H/W event. May be overridden
1269 by other higher priority error reporting module.
1270 off: disable H/W event reporting through EDAC.
1271 force: enforce the use of EDAC to report H/W event.
1274 ekgdboc= [X86,KGDB] Allow early kernel console debugging
1277 This is designed to be used in conjunction with
1278 the boot argument: earlyprintk=vga
1280 This parameter works in place of the kgdboc parameter
1281 but can only be used if the backing tty is available
1282 very early in the boot process. For early debugging
1283 via a serial port see kgdboc_earlycon instead.
1286 Format: {"off" | "on" | "skip[mbr]"}
1289 Format: { "debug", "disable_early_pci_dma",
1290 "nochunk", "noruntime", "nosoftreserve",
1291 "novamap", "no_disable_early_pci_dma" }
1292 debug: enable misc debug output.
1293 disable_early_pci_dma: disable the busmaster bit on all
1294 PCI bridges while in the EFI boot stub.
1295 nochunk: disable reading files in "chunks" in the EFI
1296 boot stub, as chunking can cause problems with some
1297 firmware implementations.
1298 noruntime : disable EFI runtime services support
1299 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose)
1300 attribute may cause the kernel to reserve the
1301 memory range for a memory mapping driver to
1302 claim. Specify efi=nosoftreserve to disable this
1303 reservation and treat the memory by its base type
1304 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM").
1305 novamap: do not call SetVirtualAddressMap().
1306 no_disable_early_pci_dma: Leave the busmaster bit set
1307 on all PCI bridges while in the EFI boot stub
1309 efi_no_storage_paranoia [EFI; X86]
1310 Using this parameter you can use more than 50% of
1311 your efi variable storage. Use this parameter only if
1312 you are really sure that your UEFI does sane gc and
1313 fulfills the spec otherwise your board may brick.
1315 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1316 Add arbitrary attribute to specific memory range by
1317 updating original EFI memory map.
1318 Region of memory which aa attribute is added to is
1321 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1322 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1323 attribute is added to range 0x100000000-0x180000000 and
1324 0x10a0000000-0x1120000000.
1326 If efi_fake_mem=8G@9G:0x40000 is specified, the
1327 EFI_MEMORY_SP(0x40000) attribute is added to
1328 range 0x240000000-0x43fffffff.
1330 Using this parameter you can do debugging of EFI memmap
1331 related features. For example, you can do debugging of
1332 Address Range Mirroring feature even if your box
1333 doesn't support it, or mark specific memory as
1336 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT
1337 that is to be dynamically loaded by Linux. If there are
1338 multiple variables with the same name but with different
1339 vendor GUIDs, all of them will be loaded. See
1340 Documentation/admin-guide/acpi/ssdt-overlays.rst for details.
1343 eisa_irq_edge= [PARISC,HW]
1344 See header of drivers/parisc/eisa.c.
1347 See comment before function elanfreq_setup() in
1348 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1350 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1351 Specifies physical address of start of kernel core
1352 image elf header and optionally the size. Generally
1353 kexec loader will pass this option to capture kernel.
1354 See Documentation/admin-guide/kdump/kdump.rst for details.
1356 enable_mtrr_cleanup [X86]
1357 The kernel tries to adjust MTRR layout from continuous
1358 to discrete, to make X server driver able to add WB
1359 entry later. This parameter enables that.
1361 enable_timer_pin_1 [X86]
1362 Enable PIN 1 of APIC timer
1363 Can be useful to work around chipset bugs
1364 (in particular on some ATI chipsets).
1365 The kernel tries to set a reasonable default.
1367 enforcing [SELINUX] Set initial enforcing status.
1369 See security/selinux/Kconfig help text.
1370 0 -- permissive (log only, no denials).
1371 1 -- enforcing (deny and log).
1373 Value can be changed at runtime via
1374 /sys/fs/selinux/enforce.
1377 Disable Error Record Serialization Table (ERST)
1380 ether= [HW,NET] Ethernet cards parameters
1381 This option is obsoleted by the "netdev=" option, which
1382 has equivalent usage. See its documentation for details.
1386 Permit 'security.evm' to be updated regardless of
1387 current integrity status.
1392 fail_make_request=[KNL]
1393 General fault injection mechanism.
1394 Format: <interval>,<probability>,<space>,<times>
1395 See also Documentation/fault-injection/.
1398 Format: { initns | none }
1399 See Documentation/admin-guide/sysctl/net.rst for
1400 fb_tunnels_only_for_init_ns
1403 See Documentation/admin-guide/blockdev/floppy.rst.
1405 force_pal_cache_flush
1406 [IA-64] Avoid check_sal_cache_flush which may hang on
1407 buggy SAL_CACHE_FLUSH implementations. Using this
1408 parameter will force ia64_sal_cache_flush to call
1409 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1412 Forcefully enable Physical Address Extension (PAE).
1413 Many Pentium M systems disable PAE but may have a
1414 functionally usable PAE implementation.
1415 Warning: use of this parameter will taint the kernel
1416 and may cause unknown problems.
1419 [FTRACE] will set and start the specified tracer
1420 as early as possible in order to facilitate early
1423 ftrace_dump_on_oops[=orig_cpu]
1424 [FTRACE] will dump the trace buffers on oops.
1425 If no parameter is passed, ftrace will dump
1426 buffers of all CPUs, but if you pass orig_cpu, it will
1427 dump only the buffer of the CPU that triggered the
1430 ftrace_filter=[function-list]
1431 [FTRACE] Limit the functions traced by the function
1432 tracer at boot up. function-list is a comma-separated
1433 list of functions. This list can be changed at run
1434 time by the set_ftrace_filter file in the debugfs
1437 ftrace_notrace=[function-list]
1438 [FTRACE] Do not trace the functions specified in
1439 function-list. This list can be changed at run time
1440 by the set_ftrace_notrace file in the debugfs
1443 ftrace_graph_filter=[function-list]
1444 [FTRACE] Limit the top level callers functions traced
1445 by the function graph tracer at boot up.
1446 function-list is a comma-separated list of functions
1447 that can be changed at run time by the
1448 set_graph_function file in the debugfs tracing directory.
1450 ftrace_graph_notrace=[function-list]
1451 [FTRACE] Do not trace from the functions specified in
1452 function-list. This list is a comma-separated list of
1453 functions that can be changed at run time by the
1454 set_graph_notrace file in the debugfs tracing directory.
1456 ftrace_graph_max_depth=<uint>
1457 [FTRACE] Used with the function graph tracer. This is
1458 the max depth it will trace into a function. This value
1459 can be changed at run time by the max_graph_depth file
1460 in the tracefs tracing directory. default: 0 (no limit)
1462 fw_devlink= [KNL] Create device links between consumer and supplier
1463 devices by scanning the firmware to infer the
1464 consumer/supplier relationships. This feature is
1465 especially useful when drivers are loaded as modules as
1466 it ensures proper ordering of tasks like device probing
1467 (suppliers first, then consumers), supplier boot state
1468 clean up (only after all consumers have probed),
1469 suspend/resume & runtime PM (consumers first, then
1471 Format: { off | permissive | on | rpm }
1472 off -- Don't create device links from firmware info.
1473 permissive -- Create device links from firmware info
1474 but use it only for ordering boot state clean
1475 up (sync_state() calls).
1476 on -- Create device links from firmware info and use it
1477 to enforce probe and suspend/resume ordering.
1478 rpm -- Like "on", but also use to order runtime PM.
1480 fw_devlink.strict=<bool>
1481 [KNL] Treat all inferred dependencies as mandatory
1482 dependencies. This only applies for fw_devlink=on|rpm.
1486 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1487 support via parallel port (up to 5 devices per port)
1488 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1489 See also Documentation/input/devices/joystick-parport.rst
1493 gart_fix_e820= [X86-64] disable the fix e820 for K8 GART
1497 gcov_persist= [GCOV] When non-zero (default), profiling data for
1498 kernel modules is saved and remains accessible via
1499 debugfs, even when the module is unloaded/reloaded.
1500 When zero, profiling data is discarded and associated
1501 debugfs files are removed at module unload time.
1503 goldfish [X86] Enable the goldfish android emulator platform.
1504 Don't use this when you are not running on the
1507 gpio-mockup.gpio_mockup_ranges
1508 [HW] Sets the ranges of gpiochip of for this device.
1509 Format: <start1>,<end1>,<start2>,<end2>...
1510 gpio-mockup.gpio_mockup_named_lines
1511 [HW] Let the driver know GPIO lines should be named.
1513 gpt [EFI] Forces disk with valid GPT signature but
1514 invalid Protective MBR to be treated as GPT. If the
1515 primary GPT is corrupted, it enables the backup/alternate
1516 GPT to be used instead.
1518 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1519 the "Enable 0" bit of the configuration register.
1522 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1523 the "Enable 0" bit of the configuration register.
1526 grcan.select= [HW] Select which physical interface to use.
1529 grcan.txsize= [HW] Sets the size of the tx buffer.
1530 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1532 grcan.rxsize= [HW] Sets the size of the rx buffer.
1533 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1536 hardlockup_all_cpu_backtrace=
1537 [KNL] Should the hard-lockup detector generate
1538 backtraces on all cpus.
1541 hashdist= [KNL,NUMA] Large hashes allocated during boot
1542 are distributed across NUMA nodes. Defaults on
1543 for 64-bit NUMA, off otherwise.
1544 Format: 0 | 1 (for off | on)
1546 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1548 hd= [EIDE] (E)IDE hard drive subsystem geometry
1549 Format: <cyl>,<head>,<sect>
1552 Disable Hardware Error Source Table (HEST) support;
1553 corresponding firmware-first mode error processing
1554 logic will be disabled.
1556 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1557 size of <nn>. This works even on boxes that have no
1558 highmem otherwise. This also works to reduce highmem
1559 size on bigger boxes.
1561 highres= [KNL] Enable/disable high resolution timer mode.
1562 Valid parameters: "on", "off"
1567 hpet= [X86-32,HPET] option to control HPET usage
1568 Format: { enable (default) | disable | force |
1570 disable: disable HPET and use PIT instead
1571 force: allow force enabled of undocumented chips (ICH4,
1573 verbose: show contents of HPET registers during setup
1575 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1576 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1578 hugetlb_cma= [HW,CMA] The size of a CMA area used for allocation
1579 of gigantic hugepages.
1582 Reserve a CMA area of given size and allocate gigantic
1583 hugepages using the CMA allocator. If enabled, the
1584 boot-time allocation of gigantic hugepages is skipped.
1586 hugepages= [HW] Number of HugeTLB pages to allocate at boot.
1587 If this follows hugepagesz (below), it specifies
1588 the number of pages of hugepagesz to be allocated.
1589 If this is the first HugeTLB parameter on the command
1590 line, it specifies the number of pages to allocate for
1591 the default huge page size. See also
1592 Documentation/admin-guide/mm/hugetlbpage.rst.
1596 [HW] The size of the HugeTLB pages. This is used in
1597 conjunction with hugepages (above) to allocate huge
1598 pages of a specific size at boot. The pair
1599 hugepagesz=X hugepages=Y can be specified once for
1600 each supported huge page size. Huge page sizes are
1601 architecture dependent. See also
1602 Documentation/admin-guide/mm/hugetlbpage.rst.
1605 hugetlb_free_vmemmap=
1606 [KNL] Reguires CONFIG_HUGETLB_PAGE_FREE_VMEMMAP
1608 Allows heavy hugetlb users to free up some more
1609 memory (6 * PAGE_SIZE for each 2MB hugetlb page).
1610 Format: { on | off (default) }
1612 on: enable the feature
1613 off: disable the feature
1615 Built with CONFIG_HUGETLB_PAGE_FREE_VMEMMAP_DEFAULT_ON=y,
1618 This is not compatible with memory_hotplug.memmap_on_memory.
1619 If both parameters are enabled, hugetlb_free_vmemmap takes
1620 precedence over memory_hotplug.memmap_on_memory.
1623 [KNL] Should the hung task detector generate panics.
1626 A value of 1 instructs the kernel to panic when a
1627 hung task is detected. The default value is controlled
1628 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time
1629 option. The value selected by this boot parameter can
1630 be changed later by the kernel.hung_task_panic sysctl.
1632 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1633 terminal devices. Valid values: 0..8
1634 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1635 If specified, z/VM IUCV HVC accepts connections
1636 from listed z/VM user IDs only.
1638 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations
1639 which allow the hypervisor to 'idle' the
1640 guest on lock contention.
1643 Do not unregister boot console at start. This is only
1644 useful for debugging when something happens in the window
1645 between unregistering the boot console and initializing
1648 i2c_bus= [HW] Override the default board specific I2C bus speed
1649 or register an additional I2C bus that is not
1650 registered from board initialization code.
1654 i8042.debug [HW] Toggle i8042 debug mode
1655 i8042.unmask_kbd_data
1656 [HW] Enable printing of interrupt data from the KBD port
1657 (disabled by default, and as a pre-condition
1658 requires that i8042.debug=1 be enabled)
1659 i8042.direct [HW] Put keyboard port into non-translated mode
1660 i8042.dumbkbd [HW] Pretend that controller can only read data from
1661 keyboard and cannot control its state
1662 (Don't attempt to blink the leds)
1663 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1664 i8042.nokbd [HW] Don't check/create keyboard port
1665 i8042.noloop [HW] Disable the AUX Loopback command while probing
1667 i8042.nomux [HW] Don't check presence of an active multiplexing
1669 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1671 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1672 i8042.reset [HW] Reset the controller during init, cleanup and
1673 suspend-to-ram transitions, only during s2r
1674 transitions, or never reset
1675 Format: { 1 | Y | y | 0 | N | n }
1676 1, Y, y: always reset controller
1677 0, N, n: don't ever reset controller
1678 Default: only on s2r transitions on x86; most other
1679 architectures force reset to be always executed
1680 i8042.unlock [HW] Unlock (ignore) the keylock
1681 i8042.kbdreset [HW] Reset device connected to KBD port
1685 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
1686 indicates that the driver is running on unsupported
1688 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
1689 does not match list of supported models.
1691 [HW] Report power status in /proc/i8k
1692 (disabled by default)
1693 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
1696 i915.invert_brightness=
1697 [DRM] Invert the sense of the variable that is used to
1698 set the brightness of the panel backlight. Normally a
1699 brightness value of 0 indicates backlight switched off,
1700 and the maximum of the brightness value sets the backlight
1701 to maximum brightness. If this parameter is set to 0
1702 (default) and the machine requires it, or this parameter
1703 is set to 1, a brightness value of 0 sets the backlight
1704 to maximum brightness, and the maximum of the brightness
1705 value switches the backlight off.
1706 -1 -- never invert brightness
1707 0 -- machine default
1708 1 -- force brightness inversion
1711 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1713 ide-core.nodma= [HW] (E)IDE subsystem
1714 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1715 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1716 .cdrom .chs .ignore_cable are additional options
1717 See Documentation/ide/ide.rst.
1719 ide-generic.probe-mask= [HW] (E)IDE subsystem
1721 Probe mask for legacy ISA IDE ports. Depending on
1722 platform up to 6 ports are supported, enabled by
1723 setting corresponding bits in the mask to 1. The
1724 default value is 0x0, which has a special meaning.
1725 On systems that have PCI, it triggers scanning the
1726 PCI bus for the first and the second port, which
1727 are then probed. On systems without PCI the value
1728 of 0x0 enables probing the two first ports as if it
1731 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1732 Claim all unknown PCI IDE storage controllers.
1735 Format: idle=poll, idle=halt, idle=nomwait
1736 Poll forces a polling idle loop that can slightly
1737 improve the performance of waking up a idle CPU, but
1738 will use a lot of power and make the system run hot.
1740 idle=halt: Halt is forced to be used for CPU idle.
1741 In such case C2/C3 won't be used again.
1742 idle=nomwait: Disable mwait for CPU C-states
1746 Allow force disabling of Shared Virtual Memory (SVA)
1747 support for the idxd driver. By default it is set to
1750 ieee754= [MIPS] Select IEEE Std 754 conformance mode
1751 Format: { strict | legacy | 2008 | relaxed }
1754 Choose which programs will be accepted for execution
1755 based on the IEEE 754 NaN encoding(s) supported by
1756 the FPU and the NaN encoding requested with the value
1757 of an ELF file header flag individually set by each
1758 binary. Hardware implementations are permitted to
1759 support either or both of the legacy and the 2008 NaN
1762 Available settings are as follows:
1763 strict accept binaries that request a NaN encoding
1764 supported by the FPU
1765 legacy only accept legacy-NaN binaries, if supported
1767 2008 only accept 2008-NaN binaries, if supported
1769 relaxed accept any binaries regardless of whether
1770 supported by the FPU
1772 The FPU emulator is always able to support both NaN
1773 encodings, so if no FPU hardware is present or it has
1774 been disabled with 'nofpu', then the settings of
1775 'legacy' and '2008' strap the emulator accordingly,
1776 'relaxed' straps the emulator for both legacy-NaN and
1777 2008-NaN, whereas 'strict' enables legacy-NaN only on
1778 legacy processors and both NaN encodings on MIPS32 or
1781 The setting for ABS.fmt/NEG.fmt instruction execution
1782 mode generally follows that for the NaN encoding,
1783 except where unsupported by hardware.
1785 ignore_loglevel [KNL]
1786 Ignore loglevel setting - this will print /all/
1787 kernel messages to the console. Useful for debugging.
1788 We also add it as printk module parameter, so users
1789 could change it dynamically, usually by
1790 /sys/module/printk/parameters/ignore_loglevel.
1793 Ignore RLIMIT_DATA setting for data mappings,
1794 print warning at first misuse. Can be changed via
1795 /sys/module/kernel/parameters/ignore_rlimit_data.
1797 ihash_entries= [KNL]
1798 Set number of hash buckets for inode cache.
1800 ima_appraise= [IMA] appraise integrity measurements
1801 Format: { "off" | "enforce" | "fix" | "log" }
1804 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead.
1805 The builtin appraise policy appraises all files
1808 ima_canonical_fmt [IMA]
1809 Use the canonical format for the binary runtime
1810 measurements, instead of host native format.
1813 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1817 The list of supported hash algorithms is defined
1818 in crypto/hash_info.h.
1821 The builtin policies to load during IMA setup.
1822 Format: "tcb | appraise_tcb | secure_boot |
1823 fail_securely | critical_data"
1825 The "tcb" policy measures all programs exec'd, files
1826 mmap'd for exec, and all files opened with the read
1827 mode bit set by either the effective uid (euid=0) or
1830 The "appraise_tcb" policy appraises the integrity of
1831 all files owned by root.
1833 The "secure_boot" policy appraises the integrity
1834 of files (eg. kexec kernel image, kernel modules,
1835 firmware, policy, etc) based on file signatures.
1837 The "fail_securely" policy forces file signature
1838 verification failure also on privileged mounted
1839 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE
1842 The "critical_data" policy measures kernel integrity
1845 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
1846 Load a policy which meets the needs of the Trusted
1847 Computing Base. This means IMA will measure all
1848 programs exec'd, files mmap'd for exec, and all files
1849 opened for read by uid=0.
1852 Select one of defined IMA measurements template formats.
1853 Formats: { "ima" | "ima-ng" | "ima-sig" }
1857 [IMA] Define a custom template format.
1858 Format: { "field1|...|fieldN" }
1860 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1861 Format: <min_file_size>
1862 Set the minimal file size for using asynchronous hash.
1863 If left unspecified, ahash usage is disabled.
1865 ahash performance varies for different data sizes on
1866 different crypto accelerators. This option can be used
1867 to achieve the best performance for a particular HW.
1869 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1871 Set hashing buffer size. Default: 4k.
1873 ahash performance varies for different chunk sizes on
1874 different crypto accelerators. This option can be used
1875 to achieve best performance for particular HW.
1879 Run specified binary instead of /sbin/init as init
1882 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1883 for working out where the kernel is dying during
1886 initcall_blacklist= [KNL] Do not execute a comma-separated list of
1887 initcall functions. Useful for debugging built-in
1888 modules and initcalls.
1890 initramfs_async= [KNL]
1893 This parameter controls whether the initramfs
1894 image is unpacked asynchronously, concurrently
1895 with devices being probed and
1896 initialized. This should normally just work,
1897 but as a debugging aid, one can get the
1898 historical behaviour of the initramfs
1899 unpacking being completed before device_ and
1902 initrd= [BOOT] Specify the location of the initial ramdisk
1904 initrdmem= [KNL] Specify a physical address and size from which to
1905 load the initrd. If an initrd is compiled in or
1906 specified in the bootparams, it takes priority over this
1908 Format: ss[KMG],nn[KMG]
1911 init_on_alloc= [MM] Fill newly allocated pages and heap objects with
1914 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON.
1916 init_on_free= [MM] Fill freed pages and heap objects with zeroes.
1918 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON.
1920 init_pkru= [X86] Specify the default memory protection keys rights
1921 register contents for all processes. 0x55555554 by
1922 default (disallow access to all but pkey 0). Can
1923 override in debugfs after boot.
1925 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1928 int_pln_enable [X86] Enable power limit notification interrupt
1930 integrity_audit=[IMA]
1931 Format: { "0" | "1" }
1932 0 -- basic integrity auditing messages. (Default)
1933 1 -- additional integrity auditing messages.
1935 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
1937 Enable intel iommu driver.
1939 Disable intel iommu driver.
1940 igfx_off [Default Off]
1941 By default, gfx is mapped as normal device. If a gfx
1942 device has a dedicated DMAR unit, the DMAR unit is
1943 bypassed by not enabling DMAR with this option. In
1944 this case, gfx device will use physical address for
1946 strict [Default Off]
1947 With this option on every unmap_single operation will
1948 result in a hardware IOTLB flush operation as opposed
1949 to batching them for performance.
1950 sp_off [Default Off]
1951 By default, super page will be supported if Intel IOMMU
1952 has the capability. With this option, super page will
1955 By default, scalable mode will be disabled even if the
1956 hardware advertises that it has support for the scalable
1957 mode translation. With this option set, scalable mode
1958 will be used on hardware which claims to support it.
1959 tboot_noforce [Default Off]
1960 Do not force the Intel IOMMU enabled under tboot.
1961 By default, tboot will force Intel IOMMU on, which
1962 could harm performance of some high-throughput
1963 devices like 40GBit network cards, even if identity
1965 Note that using this option lowers the security
1966 provided by tboot because it makes the system
1967 vulnerable to DMA attacks.
1969 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
1970 0 disables intel_idle and fall back on acpi_idle.
1971 1 to 9 specify maximum depth of C-state.
1975 Do not enable intel_pstate as the default
1976 scaling driver for the supported processors
1978 Use intel_pstate as a scaling driver, but configure it
1979 to work with generic cpufreq governors (instead of
1980 enabling its internal governor). This mode cannot be
1981 used along with the hardware-managed P-states (HWP)
1984 Enable intel_pstate on systems that prohibit it by default
1985 in favor of acpi-cpufreq. Forcing the intel_pstate driver
1986 instead of acpi-cpufreq may disable platform features, such
1987 as thermal controls and power capping, that rely on ACPI
1988 P-States information being indicated to OSPM and therefore
1989 should be used with caution. This option does not work with
1990 processors that aren't supported by the intel_pstate driver
1991 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
1993 Do not enable hardware P state control (HWP)
1996 Only load intel_pstate on systems which support
1997 hardware P state control (HWP) if available.
1999 Enforce ACPI _PPC performance limits. If the Fixed ACPI
2000 Description Table, specifies preferred power management
2001 profile as "Enterprise Server" or "Performance Server",
2002 then this feature is turned on by default.
2004 Allow per-logical-CPU P-State performance control limits using
2005 cpufreq sysfs interface
2007 intremap= [X86-64, Intel-IOMMU]
2008 on enable Interrupt Remapping (default)
2009 off disable Interrupt Remapping
2010 nosid disable Source ID checking
2012 BIOS x2APIC opt-out request will be ignored
2013 nopost disable Interrupt Posting
2015 iomem= Disable strict checking of access to MMIO memory
2016 strict regions from userspace.
2031 nobypass [PPC/POWERNV]
2032 Disable IOMMU bypass, using IOMMU for PCI devices.
2034 iommu.forcedac= [ARM64, X86] Control IOVA allocation for PCI devices.
2035 Format: { "0" | "1" }
2036 0 - Try to allocate a 32-bit DMA address first, before
2037 falling back to the full range if needed.
2038 1 - Allocate directly from the full usable range,
2039 forcing Dual Address Cycle for PCI cards supporting
2040 greater than 32-bit addressing.
2042 iommu.strict= [ARM64, X86] Configure TLB invalidation behaviour
2043 Format: { "0" | "1" }
2045 Request that DMA unmap operations use deferred
2046 invalidation of hardware TLBs, for increased
2047 throughput at the cost of reduced device isolation.
2048 Will fall back to strict mode if not supported by
2049 the relevant IOMMU driver.
2050 1 - Strict mode (default).
2051 DMA unmap operations invalidate IOMMU hardware TLBs
2053 Note: on x86, the default behaviour depends on the
2054 equivalent driver-specific parameters, but a strict
2055 mode explicitly specified by either method takes
2059 [ARM64, X86] Configure DMA to bypass the IOMMU by default.
2060 Format: { "0" | "1" }
2061 0 - Use IOMMU translation for DMA.
2062 1 - Bypass the IOMMU for DMA.
2063 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH.
2065 io7= [HW] IO7 for Marvel-based Alpha systems
2066 See comment before marvel_specify_io7 in
2067 arch/alpha/kernel/core_marvel.c.
2069 io_delay= [X86] I/O delay method
2071 Standard port 0x80 based delay
2073 Alternate port 0xed based delay (needed on some systems)
2075 Simple two microseconds delay
2080 See Documentation/admin-guide/nfs/nfsroot.rst.
2082 ipcmni_extend [KNL] Extend the maximum number of unique System V
2083 IPC identifiers from 32,768 to 16,777,216.
2085 irqaffinity= [SMP] Set the default irq affinity mask
2086 The argument is a cpu list, as described above.
2088 irqchip.gicv2_force_probe=
2091 Force the kernel to look for the second 4kB page
2092 of a GICv2 controller even if the memory range
2093 exposed by the device tree is too small.
2095 irqchip.gicv3_nolpi=
2097 Force the kernel to ignore the availability of
2098 LPIs (and by consequence ITSs). Intended for system
2099 that use the kernel as a bootloader, and thus want
2100 to let secondary kernels in charge of setting up
2103 irqchip.gicv3_pseudo_nmi= [ARM64]
2104 Enables support for pseudo-NMIs in the kernel. This
2105 requires the kernel to be built with
2106 CONFIG_ARM64_PSEUDO_NMI.
2109 When an interrupt is not handled search all handlers
2110 for it. Intended to get systems with badly broken
2114 When an interrupt is not handled search all handlers
2115 for it. Also check all handlers each timer
2116 interrupt. Intended to get systems with badly broken
2120 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
2122 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance.
2123 [Deprecated - use cpusets instead]
2124 Format: [flag-list,]<cpu-list>
2126 Specify one or more CPUs to isolate from disturbances
2127 specified in the flag list (default: domain):
2130 Disable the tick when a single task runs.
2132 A residual 1Hz tick is offloaded to workqueues, which you
2133 need to affine to housekeeping through the global
2134 workqueue's affinity configured via the
2135 /sys/devices/virtual/workqueue/cpumask sysfs file, or
2136 by using the 'domain' flag described below.
2138 NOTE: by default the global workqueue runs on all CPUs,
2139 so to protect individual CPUs the 'cpumask' file has to
2140 be configured manually after bootup.
2143 Isolate from the general SMP balancing and scheduling
2144 algorithms. Note that performing domain isolation this way
2145 is irreversible: it's not possible to bring back a CPU to
2146 the domains once isolated through isolcpus. It's strongly
2147 advised to use cpusets instead to disable scheduler load
2148 balancing through the "cpuset.sched_load_balance" file.
2149 It offers a much more flexible interface where CPUs can
2150 move in and out of an isolated set anytime.
2152 You can move a process onto or off an "isolated" CPU via
2153 the CPU affinity syscalls or cpuset.
2154 <cpu number> begins at 0 and the maximum value is
2155 "number of CPUs in system - 1".
2159 Isolate from being targeted by managed interrupts
2160 which have an interrupt mask containing isolated
2161 CPUs. The affinity of managed interrupts is
2162 handled by the kernel and cannot be changed via
2163 the /proc/irq/* interfaces.
2165 This isolation is best effort and only effective
2166 if the automatically assigned interrupt mask of a
2167 device queue contains isolated and housekeeping
2168 CPUs. If housekeeping CPUs are online then such
2169 interrupts are directed to the housekeeping CPU
2170 so that IO submitted on the housekeeping CPU
2171 cannot disturb the isolated CPU.
2173 If a queue's affinity mask contains only isolated
2174 CPUs then this parameter has no effect on the
2175 interrupt routing decision, though interrupts are
2176 only delivered when tasks running on those
2177 isolated CPUs submit IO. IO submitted on
2178 housekeeping CPUs has no influence on those
2181 The format of <cpu-list> is described above.
2185 ivrs_ioapic [HW,X86-64]
2186 Provide an override to the IOAPIC-ID<->DEVICE-ID
2187 mapping provided in the IVRS ACPI table. For
2188 example, to map IOAPIC-ID decimal 10 to
2189 PCI device 00:14.0 write the parameter as:
2190 ivrs_ioapic[10]=00:14.0
2192 ivrs_hpet [HW,X86-64]
2193 Provide an override to the HPET-ID<->DEVICE-ID
2194 mapping provided in the IVRS ACPI table. For
2195 example, to map HPET-ID decimal 0 to
2196 PCI device 00:14.0 write the parameter as:
2197 ivrs_hpet[0]=00:14.0
2199 ivrs_acpihid [HW,X86-64]
2200 Provide an override to the ACPI-HID:UID<->DEVICE-ID
2201 mapping provided in the IVRS ACPI table. For
2202 example, to map UART-HID:UID AMD0020:0 to
2203 PCI device 00:14.5 write the parameter as:
2204 ivrs_acpihid[00:14.5]=AMD0020:0
2206 js= [HW,JOY] Analog joystick
2207 See Documentation/input/joydev/joystick.rst.
2210 When CONFIG_RANDOMIZE_BASE is set, this disables
2211 kernel and module base offset ASLR (Address Space
2212 Layout Randomization).
2215 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print
2216 report on every invalid memory access. Without this
2217 parameter KASAN will print report only for the first
2222 kernelcore= [KNL,X86,IA-64,PPC]
2223 Format: nn[KMGTPE] | nn% | "mirror"
2224 This parameter specifies the amount of memory usable by
2225 the kernel for non-movable allocations. The requested
2226 amount is spread evenly throughout all nodes in the
2227 system as ZONE_NORMAL. The remaining memory is used for
2228 movable memory in its own zone, ZONE_MOVABLE. In the
2229 event, a node is too small to have both ZONE_NORMAL and
2230 ZONE_MOVABLE, kernelcore memory will take priority and
2231 other nodes will have a larger ZONE_MOVABLE.
2233 ZONE_MOVABLE is used for the allocation of pages that
2234 may be reclaimed or moved by the page migration
2235 subsystem. Note that allocations like PTEs-from-HighMem
2236 still use the HighMem zone if it exists, and the Normal
2237 zone if it does not.
2239 It is possible to specify the exact amount of memory in
2240 the form of "nn[KMGTPE]", a percentage of total system
2241 memory in the form of "nn%", or "mirror". If "mirror"
2242 option is specified, mirrored (reliable) memory is used
2243 for non-movable allocations and remaining memory is used
2244 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror"
2245 are exclusive, so you cannot specify multiple forms.
2247 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
2248 Format: <Controller#>[,poll interval]
2249 The controller # is the number of the ehci usb debug
2250 port as it is probed via PCI. The poll interval is
2251 optional and is the number seconds in between
2252 each poll cycle to the debug port in case you need
2253 the functionality for interrupting the kernel with
2254 gdb or control-c on the dbgp connection. When
2255 not using this parameter you use sysrq-g to break into
2256 the kernel debugger.
2258 kgdboc= [KGDB,HW] kgdb over consoles.
2259 Requires a tty driver that supports console polling,
2260 or a supported polling keyboard driver (non-usb).
2261 Serial only format: <serial_device>[,baud]
2262 keyboard only format: kbd
2263 keyboard and serial format: kbd,<serial_device>[,baud]
2264 Optional Kernel mode setting:
2265 kms, kbd format: kms,kbd
2266 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
2268 kgdboc_earlycon= [KGDB,HW]
2269 If the boot console provides the ability to read
2270 characters and can work in polling mode, you can use
2271 this parameter to tell kgdb to use it as a backend
2272 until the normal console is registered. Intended to
2273 be used together with the kgdboc parameter which
2274 specifies the normal console to transition to.
2276 The name of the early console should be specified
2277 as the value of this parameter. Note that the name of
2278 the early console might be different than the tty
2279 name passed to kgdboc. It's OK to leave the value
2280 blank and the first boot console that implements
2281 read() will be picked.
2283 kgdbwait [KGDB] Stop kernel execution and enter the
2284 kernel debugger at the earliest opportunity.
2286 kmac= [MIPS] Korina ethernet MAC address.
2287 Configure the RouterBoard 532 series on-chip
2288 Ethernet adapter MAC address.
2290 kmemleak= [KNL] Boot-time kmemleak enable/disable
2291 Valid arguments: on, off
2293 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
2296 kprobe_event=[probe-list]
2297 [FTRACE] Add kprobe events and enable at boot time.
2298 The probe-list is a semicolon delimited list of probe
2299 definitions. Each definition is same as kprobe_events
2300 interface, but the parameters are comma delimited.
2301 For example, to add a kprobe event on vfs_read with
2302 arg1 and arg2, add to the command line;
2304 kprobe_event=p,vfs_read,$arg1,$arg2
2306 See also Documentation/trace/kprobetrace.rst "Kernel
2307 Boot Parameter" section.
2309 kpti= [ARM64] Control page table isolation of user
2310 and kernel address spaces.
2311 Default: enabled on cores which need mitigation.
2315 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
2316 Default is 0 (don't ignore, but inject #GP)
2318 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface.
2319 Default is false (don't support).
2321 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit
2326 [KVM] Controls the software workaround for the
2327 X86_BUG_ITLB_MULTIHIT bug.
2328 force : Always deploy workaround.
2329 off : Never deploy workaround.
2330 auto : Deploy workaround based on the presence of
2331 X86_BUG_ITLB_MULTIHIT.
2335 If the software workaround is enabled for the host,
2336 guests do need not to enable it for nested guests.
2338 kvm.nx_huge_pages_recovery_ratio=
2339 [KVM] Controls how many 4KiB pages are periodically zapped
2340 back to huge pages. 0 disables the recovery, otherwise if
2341 the value is N KVM will zap 1/Nth of the 4KiB pages every
2342 minute. The default is 60.
2344 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
2345 Default is 1 (enabled)
2347 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
2349 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
2352 [KVM,ARM] Select one of KVM/arm64's modes of operation.
2354 nvhe: Standard nVHE-based mode, without support for
2357 protected: nVHE-based mode with support for guests whose
2358 state is kept private from the host.
2359 Not valid if the kernel is running in EL2.
2361 Defaults to VHE/nVHE based on hardware support.
2363 kvm-arm.vgic_v3_group0_trap=
2364 [KVM,ARM] Trap guest accesses to GICv3 group-0
2367 kvm-arm.vgic_v3_group1_trap=
2368 [KVM,ARM] Trap guest accesses to GICv3 group-1
2371 kvm-arm.vgic_v3_common_trap=
2372 [KVM,ARM] Trap guest accesses to GICv3 common
2375 kvm-arm.vgic_v4_enable=
2376 [KVM,ARM] Allow use of GICv4 for direct injection of
2379 kvm_cma_resv_ratio=n [PPC]
2380 Reserves given percentage from system memory area for
2381 contiguous memory allocation for KVM hash pagetable
2383 By default it reserves 5% of total system memory.
2387 kvm-intel.ept= [KVM,Intel] Disable extended page tables
2388 (virtualized MMU) support on capable Intel chips.
2389 Default is 1 (enabled)
2391 kvm-intel.emulate_invalid_guest_state=
2392 [KVM,Intel] Enable emulation of invalid guest states
2393 Default is 0 (disabled)
2395 kvm-intel.flexpriority=
2396 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
2397 Default is 1 (enabled)
2400 [KVM,Intel] Enable VMX nesting (nVMX).
2401 Default is 0 (disabled)
2403 kvm-intel.unrestricted_guest=
2404 [KVM,Intel] Disable unrestricted guest feature
2405 (virtualized real and unpaged mode) on capable
2406 Intel chips. Default is 1 (enabled)
2408 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault
2411 Valid arguments: never, cond, always
2413 always: L1D cache flush on every VMENTER.
2414 cond: Flush L1D on VMENTER only when the code between
2415 VMEXIT and VMENTER can leak host memory.
2416 never: Disables the mitigation
2418 Default is cond (do L1 cache flush in specific instances)
2420 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
2421 feature (tagged TLBs) on capable Intel chips.
2422 Default is 1 (enabled)
2424 l1d_flush= [X86,INTEL]
2425 Control mitigation for L1D based snooping vulnerability.
2427 Certain CPUs are vulnerable to an exploit against CPU
2428 internal buffers which can forward information to a
2429 disclosure gadget under certain conditions.
2431 In vulnerable processors, the speculatively
2432 forwarded data can be used in a cache side channel
2433 attack, to access data to which the attacker does
2434 not have direct access.
2436 This parameter controls the mitigation. The
2439 on - enable the interface for the mitigation
2441 l1tf= [X86] Control mitigation of the L1TF vulnerability on
2444 The kernel PTE inversion protection is unconditionally
2445 enabled and cannot be disabled.
2448 Provides all available mitigations for the
2449 L1TF vulnerability. Disables SMT and
2450 enables all mitigations in the
2451 hypervisors, i.e. unconditional L1D flush.
2453 SMT control and L1D flush control via the
2454 sysfs interface is still possible after
2455 boot. Hypervisors will issue a warning
2456 when the first VM is started in a
2457 potentially insecure configuration,
2458 i.e. SMT enabled or L1D flush disabled.
2461 Same as 'full', but disables SMT and L1D
2462 flush runtime control. Implies the
2463 'nosmt=force' command line option.
2464 (i.e. sysfs control of SMT is disabled.)
2467 Leaves SMT enabled and enables the default
2468 hypervisor mitigation, i.e. conditional
2471 SMT control and L1D flush control via the
2472 sysfs interface is still possible after
2473 boot. Hypervisors will issue a warning
2474 when the first VM is started in a
2475 potentially insecure configuration,
2476 i.e. SMT enabled or L1D flush disabled.
2480 Disables SMT and enables the default
2481 hypervisor mitigation.
2483 SMT control and L1D flush control via the
2484 sysfs interface is still possible after
2485 boot. Hypervisors will issue a warning
2486 when the first VM is started in a
2487 potentially insecure configuration,
2488 i.e. SMT enabled or L1D flush disabled.
2491 Same as 'flush', but hypervisors will not
2492 warn when a VM is started in a potentially
2493 insecure configuration.
2496 Disables hypervisor mitigations and doesn't
2498 It also drops the swap size and available
2499 RAM limit restriction on both hypervisor and
2504 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst
2510 lapic [X86-32,APIC] Enable the local APIC even if BIOS
2513 lapic= [X86,APIC] Do not use TSC deadline
2514 value for LAPIC timer one-shot implementation. Default
2515 back to the programmable timer unit in the LAPIC.
2516 Format: notscdeadline
2518 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
2521 libata.dma= [LIBATA] DMA control
2522 libata.dma=0 Disable all PATA and SATA DMA
2523 libata.dma=1 PATA and SATA Disk DMA only
2524 libata.dma=2 ATAPI (CDROM) DMA only
2525 libata.dma=4 Compact Flash DMA only
2526 Combinations also work, so libata.dma=3 enables DMA
2527 for disks and CDROMs, but not CFs.
2529 libata.ignore_hpa= [LIBATA] Ignore HPA limit
2530 libata.ignore_hpa=0 keep BIOS limits (default)
2531 libata.ignore_hpa=1 ignore limits, using full disk
2533 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
2537 libata.force= [LIBATA] Force configurations. The format is comma-
2538 separated list of "[ID:]VAL" where ID is
2539 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
2540 matching port, link or device. Basically, it matches
2541 the ATA ID string printed on console by libata. If
2542 the whole ID part is omitted, the last PORT and DEVICE
2543 values are used. If ID hasn't been specified yet, the
2544 configuration applies to all ports, links and devices.
2546 If only DEVICE is omitted, the parameter applies to
2547 the port and all links and devices behind it. DEVICE
2548 number of 0 either selects the first device or the
2549 first fan-out link behind PMP device. It does not
2550 select the host link. DEVICE number of 15 selects the
2551 host link and device attached to it.
2553 The VAL specifies the configuration to force. As long
2554 as there's no ambiguity shortcut notation is allowed.
2555 For example, both 1.5 and 1.5G would work for 1.5Gbps.
2556 The following configurations can be forced.
2558 * Cable type: 40c, 80c, short40c, unk, ign or sata.
2559 Any ID with matching PORT is used.
2561 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
2563 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
2564 udma[/][16,25,33,44,66,100,133] notation is also
2567 * [no]ncq: Turn on or off NCQ.
2569 * [no]ncqtrim: Turn off queued DSM TRIM.
2571 * nohrst, nosrst, norst: suppress hard, soft
2574 * rstonce: only attempt one reset during
2575 hot-unplug link recovery
2577 * dump_id: dump IDENTIFY data.
2579 * atapi_dmadir: Enable ATAPI DMADIR bridge support
2581 * disable: Disable this device.
2583 If there are multiple matching configurations changing
2584 the same attribute, the last one is used.
2586 memblock=debug [KNL] Enable memblock debug messages.
2588 load_ramdisk= [RAM] [Deprecated]
2590 lockd.nlm_grace_period=P [NFS] Assign grace period.
2593 lockd.nlm_tcpport=N [NFS] Assign TCP port.
2596 lockd.nlm_timeout=T [NFS] Assign timeout value.
2599 lockd.nlm_udpport=M [NFS] Assign UDP port.
2602 lockdown= [SECURITY]
2603 { integrity | confidentiality }
2604 Enable the kernel lockdown feature. If set to
2605 integrity, kernel features that allow userland to
2606 modify the running kernel are disabled. If set to
2607 confidentiality, kernel features that allow userland
2608 to extract confidential information from the kernel
2611 locktorture.nreaders_stress= [KNL]
2612 Set the number of locking read-acquisition kthreads.
2613 Defaults to being automatically set based on the
2614 number of online CPUs.
2616 locktorture.nwriters_stress= [KNL]
2617 Set the number of locking write-acquisition kthreads.
2619 locktorture.onoff_holdoff= [KNL]
2620 Set time (s) after boot for CPU-hotplug testing.
2622 locktorture.onoff_interval= [KNL]
2623 Set time (s) between CPU-hotplug operations, or
2624 zero to disable CPU-hotplug testing.
2626 locktorture.shuffle_interval= [KNL]
2627 Set task-shuffle interval (jiffies). Shuffling
2628 tasks allows some CPUs to go into dyntick-idle
2629 mode during the locktorture test.
2631 locktorture.shutdown_secs= [KNL]
2632 Set time (s) after boot system shutdown. This
2633 is useful for hands-off automated testing.
2635 locktorture.stat_interval= [KNL]
2636 Time (s) between statistics printk()s.
2638 locktorture.stutter= [KNL]
2639 Time (s) to stutter testing, for example,
2640 specifying five seconds causes the test to run for
2641 five seconds, wait for five seconds, and so on.
2642 This tests the locking primitive's ability to
2643 transition abruptly to and from idle.
2645 locktorture.torture_type= [KNL]
2646 Specify the locking implementation to test.
2648 locktorture.verbose= [KNL]
2649 Enable additional printk() statements.
2651 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
2654 loglevel= All Kernel Messages with a loglevel smaller than the
2655 console loglevel will be printed to the console. It can
2656 also be changed with klogd or other programs. The
2657 loglevels are defined as follows:
2659 0 (KERN_EMERG) system is unusable
2660 1 (KERN_ALERT) action must be taken immediately
2661 2 (KERN_CRIT) critical conditions
2662 3 (KERN_ERR) error conditions
2663 4 (KERN_WARNING) warning conditions
2664 5 (KERN_NOTICE) normal but significant condition
2665 6 (KERN_INFO) informational
2666 7 (KERN_DEBUG) debug-level messages
2668 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
2669 in bytes. n must be a power of two and greater
2670 than the minimal size. The minimal size is defined
2671 by LOG_BUF_SHIFT kernel config parameter. There is
2672 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
2673 that allows to increase the default size depending on
2674 the number of CPUs. See init/Kconfig for more details.
2676 logo.nologo [FB] Disables display of the built-in Linux logo.
2677 This may be used to provide more screen space for
2678 kernel log messages and is useful when debugging
2679 kernel boot problems.
2681 lp=0 [LP] Specify parallel ports to use, e.g,
2682 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
2683 lp=reset first parallel port). 'lp=0' disables the
2684 lp=auto printer driver. 'lp=reset' (which can be
2685 specified in addition to the ports) causes
2686 attached printers to be reset. Using
2687 lp=port1,port2,... specifies the parallel ports
2688 to associate lp devices with, starting with
2689 lp0. A port specification may be 'none' to skip
2690 that lp device, or a parport name such as
2691 'parport0'. Specifying 'lp=auto' instead of a
2692 port specification list means that device IDs
2693 from each port should be examined, to see if
2694 an IEEE 1284-compliant printer is attached; if
2695 so, the driver will manage that printer.
2696 See also header of drivers/char/lp.c.
2699 Sets loops_per_jiffy to given constant, thus avoiding
2700 time-consuming boot-time autodetection (up to 250 ms per
2701 CPU). 0 enables autodetection (default). To determine
2702 the correct value for your kernel, boot with normal
2703 autodetection and see what value is printed. Note that
2704 on SMP systems the preset will be applied to all CPUs,
2705 which is likely to cause problems if your CPUs need
2706 significantly divergent settings. An incorrect value
2707 will cause delays in the kernel to be wrong, leading to
2708 unpredictable I/O errors and other breakage. Although
2709 unlikely, in the extreme case this might damage your
2713 Format: <io>,<irq>,<dma>
2715 lsm.debug [SECURITY] Enable LSM initialization debugging output.
2718 [SECURITY] Choose order of LSM initialization. This
2719 overrides CONFIG_LSM, and the "security=" parameter.
2721 machvec= [IA-64] Force the use of a particular machine-vector
2722 (machvec) in a generic kernel.
2723 Example: machvec=hpzx1
2725 machtype= [Loongson] Share the same kernel image file between
2726 different yeeloong laptops.
2727 Example: machtype=lemote-yeeloong-2f-7inch
2729 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
2730 than or equal to this physical address is ignored.
2732 maxcpus= [SMP] Maximum number of processors that an SMP kernel
2733 will bring up during bootup. maxcpus=n : n >= 0 limits
2734 the kernel to bring up 'n' processors. Surely after
2735 bootup you can bring up the other plugged cpu by executing
2736 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus
2737 only takes effect during system bootup.
2738 While n=0 is a special case, it is equivalent to "nosmp",
2739 which also disables the IO APIC.
2741 max_loop= [LOOP] The number of loop block devices that get
2742 (loop.max_loop) unconditionally pre-created at init time. The default
2743 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
2744 of statically allocating a predefined number, loop
2745 devices can be requested on-demand with the
2746 /dev/loop-control interface.
2748 mce [X86-32] Machine Check Exception
2750 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.rst
2752 md= [HW] RAID subsystems devices and level
2753 See Documentation/admin-guide/md.rst.
2756 Format: <first>,<last>
2757 Specifies range of consoles to be captured by the MDA.
2760 Control mitigation for the Micro-architectural Data
2761 Sampling (MDS) vulnerability.
2763 Certain CPUs are vulnerable to an exploit against CPU
2764 internal buffers which can forward information to a
2765 disclosure gadget under certain conditions.
2767 In vulnerable processors, the speculatively
2768 forwarded data can be used in a cache side channel
2769 attack, to access data to which the attacker does
2770 not have direct access.
2772 This parameter controls the MDS mitigation. The
2775 full - Enable MDS mitigation on vulnerable CPUs
2776 full,nosmt - Enable MDS mitigation and disable
2777 SMT on vulnerable CPUs
2778 off - Unconditionally disable MDS mitigation
2780 On TAA-affected machines, mds=off can be prevented by
2781 an active TAA mitigation as both vulnerabilities are
2782 mitigated with the same mechanism so in order to disable
2783 this mitigation, you need to specify tsx_async_abort=off
2786 Not specifying this option is equivalent to
2789 For details see: Documentation/admin-guide/hw-vuln/mds.rst
2791 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
2792 Amount of memory to be used in cases as follows:
2795 2 when the kernel is not able to see the whole system memory;
2796 3 memory that lies after 'mem=' boundary is excluded from
2797 the hypervisor, then assigned to KVM guests.
2799 [X86] Work as limiting max address. Use together
2800 with memmap= to avoid physical address space collisions.
2801 Without memmap= PCI devices could be placed at addresses
2802 belonging to unused RAM.
2804 Note that this only takes effects during boot time since
2805 in above case 3, memory may need be hot added after boot
2806 if system memory of hypervisor is not sufficient.
2808 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
2812 [KNL,SH] Allow user to override the default size for
2813 per-device physically contiguous DMA buffers.
2815 memhp_default_state=online/offline
2816 [KNL] Set the initial state for the memory hotplug
2817 onlining policy. If not specified, the default value is
2818 set according to the
2819 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config
2821 See Documentation/admin-guide/mm/memory-hotplug.rst.
2823 memmap=exactmap [KNL,X86] Enable setting of an exact
2824 E820 memory map, as specified by the user.
2825 Such memmap=exactmap lines can be constructed based on
2826 BIOS output or other requirements. See the memmap=nn@ss
2829 memmap=nn[KMG]@ss[KMG]
2830 [KNL, X86, MIPS, XTENSA] Force usage of a specific region of memory.
2831 Region of memory to be used is from ss to ss+nn.
2832 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG],
2833 which limits max address to nn[KMG].
2834 Multiple different regions can be specified,
2837 memmap=100M@2G,100M#3G,1G!1024G
2839 memmap=nn[KMG]#ss[KMG]
2840 [KNL,ACPI] Mark specific memory as ACPI data.
2841 Region of memory to be marked is from ss to ss+nn.
2843 memmap=nn[KMG]$ss[KMG]
2844 [KNL,ACPI] Mark specific memory as reserved.
2845 Region of memory to be reserved is from ss to ss+nn.
2846 Example: Exclude memory from 0x18690000-0x1869ffff
2847 memmap=64K$0x18690000
2849 memmap=0x10000$0x18690000
2850 Some bootloaders may need an escape character before '$',
2851 like Grub2, otherwise '$' and the following number
2854 memmap=nn[KMG]!ss[KMG]
2855 [KNL,X86] Mark specific memory as protected.
2856 Region of memory to be used, from ss to ss+nn.
2857 The memory region may be marked as e820 type 12 (0xc)
2858 and is NVDIMM or ADR memory.
2860 memmap=<size>%<offset>-<oldtype>+<newtype>
2861 [KNL,ACPI] Convert memory within the specified region
2862 from <oldtype> to <newtype>. If "-<oldtype>" is left
2863 out, the whole region will be marked as <newtype>,
2864 even if previously unavailable. If "+<newtype>" is left
2865 out, matching memory will be removed. Types are
2866 specified as e820 types, e.g., 1 = RAM, 2 = reserved,
2867 3 = ACPI, 12 = PRAM.
2869 memory_corruption_check=0/1 [X86]
2870 Some BIOSes seem to corrupt the first 64k of
2871 memory when doing things like suspend/resume.
2872 Setting this option will scan the memory
2873 looking for corruption. Enabling this will
2874 both detect corruption and prevent the kernel
2875 from using the memory being corrupted.
2876 However, its intended as a diagnostic tool; if
2877 repeatable BIOS-originated corruption always
2878 affects the same memory, you can use memmap=
2879 to prevent the kernel from using that memory.
2881 memory_corruption_check_size=size [X86]
2882 By default it checks for corruption in the low
2883 64k, making this memory unavailable for normal
2884 use. Use this parameter to scan for
2885 corruption in more or less memory.
2887 memory_corruption_check_period=seconds [X86]
2888 By default it checks for corruption every 60
2889 seconds. Use this parameter to check at some
2890 other rate. 0 disables periodic checking.
2892 memory_hotplug.memmap_on_memory
2893 [KNL,X86,ARM] Boolean flag to enable this feature.
2894 Format: {on | off (default)}
2895 When enabled, runtime hotplugged memory will
2896 allocate its internal metadata (struct pages)
2897 from the hotadded memory which will allow to
2898 hotadd a lot of memory without requiring
2899 additional memory to do so.
2900 This feature is disabled by default because it
2901 has some implication on large (e.g. GB)
2902 allocations in some configurations (e.g. small
2904 The state of the flag can be read in
2905 /sys/module/memory_hotplug/parameters/memmap_on_memory.
2906 Note that even when enabled, there are a few cases where
2907 the feature is not effective.
2909 This is not compatible with hugetlb_free_vmemmap. If
2910 both parameters are enabled, hugetlb_free_vmemmap takes
2911 precedence over memory_hotplug.memmap_on_memory.
2913 memtest= [KNL,X86,ARM,PPC,RISCV] Enable memtest
2915 default : 0 <disable>
2916 Specifies the number of memtest passes to be
2917 performed. Each pass selects another test
2918 pattern from a given set of patterns. Memtest
2919 fills the memory with this pattern, validates
2920 memory contents and reserves bad memory
2921 regions that are detected.
2923 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control
2924 Valid arguments: on, off
2925 Default (depends on kernel configuration option):
2926 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y)
2927 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n)
2928 mem_encrypt=on: Activate SME
2929 mem_encrypt=off: Do not activate SME
2931 Refer to Documentation/virt/kvm/amd-memory-encryption.rst
2932 for details on when memory encryption can be activated.
2934 mem_sleep_default= [SUSPEND] Default system suspend mode:
2935 s2idle - Suspend-To-Idle
2936 shallow - Power-On Suspend or equivalent (if supported)
2937 deep - Suspend-To-RAM or equivalent (if supported)
2938 See Documentation/admin-guide/pm/sleep-states.rst.
2940 meye.*= [HW] Set MotionEye Camera parameters
2941 See Documentation/admin-guide/media/meye.rst.
2943 mfgpt_irq= [IA-32] Specify the IRQ to use for the
2944 Multi-Function General Purpose Timers on AMD Geode
2947 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
2948 the BIOS has incorrectly applied a workaround. TinyBIOS
2949 version 0.98 is known to be affected, 0.99 fixes the
2950 problem by letting the user disable the workaround.
2954 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
2955 physical address is ignored.
2957 mini2440= [ARM,HW,KNL]
2958 Format:[0..2][b][c][t]
2960 MINI2440 configuration specification:
2961 0 - The attached screen is the 3.5" TFT
2962 1 - The attached screen is the 7" TFT
2963 2 - The VGA Shield is attached (1024x768)
2964 Leaving out the screen size parameter will not load
2965 the TFT driver, and the framebuffer will be left
2967 b - Enable backlight. The TFT backlight pin will be
2968 linked to the kernel VESA blanking code and a GPIO
2969 LED. This parameter is not necessary when using the
2971 c - Enable the s3c camera interface.
2972 t - Reserved for enabling touchscreen support. The
2973 touchscreen support is not enabled in the mainstream
2974 kernel as of 2.6.30, a preliminary port can be found
2975 in the "bleeding edge" mini2440 support kernel at
2976 https://repo.or.cz/w/linux-2.6/mini2440.git
2979 [X86,PPC,S390,ARM64] Control optional mitigations for
2980 CPU vulnerabilities. This is a set of curated,
2981 arch-independent options, each of which is an
2982 aggregation of existing arch-specific options.
2985 Disable all optional CPU mitigations. This
2986 improves system performance, but it may also
2987 expose users to several CPU vulnerabilities.
2988 Equivalent to: nopti [X86,PPC]
2990 nospectre_v1 [X86,PPC]
2992 nospectre_v2 [X86,PPC,S390,ARM64]
2993 spectre_v2_user=off [X86]
2994 spec_store_bypass_disable=off [X86,PPC]
2995 ssbd=force-off [ARM64]
2998 tsx_async_abort=off [X86]
2999 kvm.nx_huge_pages=off [X86]
3000 no_entry_flush [PPC]
3001 no_uaccess_flush [PPC]
3004 This does not have any effect on
3005 kvm.nx_huge_pages when
3006 kvm.nx_huge_pages=force.
3009 Mitigate all CPU vulnerabilities, but leave SMT
3010 enabled, even if it's vulnerable. This is for
3011 users who don't want to be surprised by SMT
3012 getting disabled across kernel upgrades, or who
3013 have other ways of avoiding SMT-based attacks.
3014 Equivalent to: (default behavior)
3017 Mitigate all CPU vulnerabilities, disabling SMT
3018 if needed. This is for users who always want to
3019 be fully mitigated, even if it means losing SMT.
3020 Equivalent to: l1tf=flush,nosmt [X86]
3021 mds=full,nosmt [X86]
3022 tsx_async_abort=full,nosmt [X86]
3025 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
3026 parameter allows control of the logging verbosity for
3027 the additional memory initialisation checks. A value
3028 of 0 disables mminit logging and a level of 4 will
3029 log everything. Information is printed at KERN_DEBUG
3030 so loglevel=8 may also need to be specified.
3033 [KNL] When CONFIG_MODULE_SIG is set, this means that
3034 modules without (valid) signatures will fail to load.
3035 Note that if CONFIG_MODULE_SIG_FORCE is set, that
3036 is always true, so this option does nothing.
3038 module_blacklist= [KNL] Do not load a comma-separated list of
3039 modules. Useful for debugging problem modules.
3042 [MOUSE] Maximum time between finger touching and
3043 leaving touchpad surface for touch to be considered
3044 a tap and be reported as a left button click (for
3045 touchpads working in absolute mode only).
3047 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
3048 reporting absolute coordinates, such as tablets
3049 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
3050 reporting absolute coordinates, such as tablets
3052 movablecore= [KNL,X86,IA-64,PPC]
3053 Format: nn[KMGTPE] | nn%
3054 This parameter is the complement to kernelcore=, it
3055 specifies the amount of memory used for migratable
3056 allocations. If both kernelcore and movablecore is
3057 specified, then kernelcore will be at *least* the
3058 specified value but may be more. If movablecore on its
3059 own is specified, the administrator must be careful
3060 that the amount of memory usable for all allocations
3063 movable_node [KNL] Boot-time switch to make hotplugable memory
3064 NUMA nodes to be movable. This means that the memory
3065 of such nodes will be usable only for movable
3066 allocations which rules out almost all kernel
3067 allocations. Use with caution!
3069 MTD_Partition= [MTD]
3070 Format: <name>,<region-number>,<size>,<offset>
3072 MTD_Region= [MTD] Format:
3073 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
3076 See drivers/mtd/parsers/cmdlinepart.c
3078 multitce=off [PPC] This parameter disables the use of the pSeries
3079 firmware feature for updating multiple TCE entries
3082 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
3084 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
3086 boundary - index of last SLC block on Flex-OneNAND.
3087 The remaining blocks are configured as MLC blocks.
3088 lock - Configure if Flex-OneNAND boundary should be locked.
3089 Once locked, the boundary cannot be changed.
3090 1 indicates lock status, 0 indicates unlock status.
3093 ARM/S3C2412 JIVE boot control
3095 See arch/arm/mach-s3c/mach-jive.c
3097 mtouchusb.raw_coordinates=
3098 [HW] Make the MicroTouch USB driver use raw coordinates
3099 ('y', default) or cooked coordinates ('n')
3101 mtrr_chunk_size=nn[KMG] [X86]
3102 used for mtrr cleanup. It is largest continuous chunk
3103 that could hold holes aka. UC entries.
3105 mtrr_gran_size=nn[KMG] [X86]
3106 Used for mtrr cleanup. It is granularity of mtrr block.
3108 Large value could prevent small alignment from
3111 mtrr_spare_reg_nr=n [X86]
3113 Range: 0,7 : spare reg number
3115 Used for mtrr cleanup. It is spare mtrr entries number.
3116 Set to 2 or more if your graphical card needs more.
3118 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
3120 netdev= [NET] Network devices parameters
3121 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
3122 Note that mem_start is often overloaded to mean
3123 something different and driver-specific.
3124 This usage is only documented in each driver source
3128 [NETFILTER] Enable connection tracking flow accounting
3129 0 to disable accounting
3130 1 to enable accounting
3133 nfsaddrs= [NFS] Deprecated. Use ip= instead.
3134 See Documentation/admin-guide/nfs/nfsroot.rst.
3136 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
3137 See Documentation/admin-guide/nfs/nfsroot.rst.
3139 nfsrootdebug [NFS] enable nfsroot debugging messages.
3140 See Documentation/admin-guide/nfs/nfsroot.rst.
3142 nfs.callback_nr_threads=
3143 [NFSv4] set the total number of threads that the
3144 NFS client will assign to service NFSv4 callback
3147 nfs.callback_tcpport=
3148 [NFS] set the TCP port on which the NFSv4 callback
3149 channel should listen.
3152 [NFS] sets the pathname to the program which is used
3153 to update the NFS client cache entries.
3155 nfs.cache_getent_timeout=
3156 [NFS] sets the timeout after which an attempt to
3157 update a cache entry is deemed to have failed.
3159 nfs.idmap_cache_timeout=
3160 [NFS] set the maximum lifetime for idmapper cache
3164 [NFS] enable 64-bit inode numbers.
3165 If zero, the NFS client will fake up a 32-bit inode
3166 number for the readdir() and stat() syscalls instead
3167 of returning the full 64-bit number.
3168 The default is to return 64-bit inode numbers.
3170 nfs.max_session_cb_slots=
3171 [NFSv4.1] Sets the maximum number of session
3172 slots the client will assign to the callback
3173 channel. This determines the maximum number of
3174 callbacks the client will process in parallel for
3175 a particular server.
3177 nfs.max_session_slots=
3178 [NFSv4.1] Sets the maximum number of session slots
3179 the client will attempt to negotiate with the server.
3180 This limits the number of simultaneous RPC requests
3181 that the client can send to the NFSv4.1 server.
3182 Note that there is little point in setting this
3183 value higher than the max_tcp_slot_table_limit.
3185 nfs.nfs4_disable_idmapping=
3186 [NFSv4] When set to the default of '1', this option
3187 ensures that both the RPC level authentication
3188 scheme and the NFS level operations agree to use
3189 numeric uids/gids if the mount is using the
3190 'sec=sys' security flavour. In effect it is
3191 disabling idmapping, which can make migration from
3192 legacy NFSv2/v3 systems to NFSv4 easier.
3193 Servers that do not support this mode of operation
3194 will be autodetected by the client, and it will fall
3195 back to using the idmapper.
3196 To turn off this behaviour, set the value to '0'.
3198 [NFS4] Specify an additional fixed unique ident-
3199 ification string that NFSv4 clients can insert into
3200 their nfs_client_id4 string. This is typically a
3201 UUID that is generated at system install time.
3203 nfs.send_implementation_id =
3204 [NFSv4.1] Send client implementation identification
3205 information in exchange_id requests.
3206 If zero, no implementation identification information
3208 The default is to send the implementation identification
3211 nfs.recover_lost_locks =
3212 [NFSv4] Attempt to recover locks that were lost due
3213 to a lease timeout on the server. Please note that
3214 doing this risks data corruption, since there are
3215 no guarantees that the file will remain unchanged
3216 after the locks are lost.
3217 If you want to enable the kernel legacy behaviour of
3218 attempting to recover these locks, then set this
3220 The default parameter value of '0' causes the kernel
3221 not to attempt recovery of lost locks.
3223 nfs4.layoutstats_timer =
3224 [NFSv4.2] Change the rate at which the kernel sends
3225 layoutstats to the pNFS metadata server.
3227 Setting this to value to 0 causes the kernel to use
3228 whatever value is the default set by the layout
3229 driver. A non-zero value sets the minimum interval
3230 in seconds between layoutstats transmissions.
3232 nfsd.nfs4_disable_idmapping=
3233 [NFSv4] When set to the default of '1', the NFSv4
3234 server will return only numeric uids and gids to
3235 clients using auth_sys, and will accept numeric uids
3236 and gids from such clients. This is intended to ease
3237 migration from NFSv2/v3.
3239 nmi_backtrace.backtrace_idle [KNL]
3240 Dump stacks even of idle CPUs in response to an
3241 NMI stack-backtrace request.
3243 nmi_debug= [KNL,SH] Specify one or more actions to take
3244 when a NMI is triggered.
3245 Format: [state][,regs][,debounce][,die]
3247 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
3248 Format: [panic,][nopanic,][num]
3250 0 - turn hardlockup detector in nmi_watchdog off
3251 1 - turn hardlockup detector in nmi_watchdog on
3252 When panic is specified, panic when an NMI watchdog
3253 timeout occurs (or 'nopanic' to not panic on an NMI
3254 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set)
3255 To disable both hard and soft lockup detectors,
3256 please see 'nowatchdog'.
3257 This is useful when you use a panic=... timeout and
3258 need the box quickly up again.
3260 These settings can be accessed at runtime via
3261 the nmi_watchdog and hardlockup_panic sysctls.
3263 netpoll.carrier_timeout=
3264 [NET] Specifies amount of time (in seconds) that
3265 netpoll should wait for a carrier. By default netpoll
3268 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
3269 emulation library even if a 387 maths coprocessor
3272 no5lvl [X86-64] Disable 5-level paging mode. Forces
3273 kernel to use 4-level paging instead.
3275 nofsgsbase [X86] Disables FSGSBASE instructions.
3278 [HW] Never suspend the console
3279 Disable suspending of consoles during suspend and
3280 hibernate operations. Once disabled, debugging
3281 messages can reach various consoles while the rest
3282 of the system is being put to sleep (ie, while
3283 debugging driver suspend/resume hooks). This may
3284 not work reliably with all consoles, but is known
3285 to work with serial and VGA consoles.
3286 To facilitate more flexible debugging, we also add
3287 console_suspend, a printk module parameter to control
3288 it. Users could use console_suspend (usually
3289 /sys/module/printk/parameters/console_suspend) to
3290 turn on/off it dynamically.
3292 novmcoredd [KNL,KDUMP]
3293 Disable device dump. Device dump allows drivers to
3294 append dump data to vmcore so you can collect driver
3295 specified debug info. Drivers can append the data
3296 without any limit and this data is stored in memory,
3297 so this may cause significant memory stress. Disabling
3298 device dump can help save memory but the driver debug
3299 data will be no longer available. This parameter
3300 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP
3303 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
3304 caches in the slab allocator. Saves per-node memory,
3305 but will impact performance.
3309 noaltinstr [S390] Disables alternative instructions patching
3310 (CPU alternatives feature).
3312 noapic [SMP,APIC] Tells the kernel to not make use of any
3313 IOAPICs that may be present in the system.
3315 noautogroup Disable scheduler automatic task group creation.
3317 nobats [PPC] Do not use BATs for mapping kernel lowmem
3318 on "Classic" PPC cores.
3322 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
3324 delayacct [KNL] Enable per-task delay accounting
3326 nodsp [SH] Disable hardware DSP at boot time.
3328 noefi Disable EFI runtime services support.
3330 no_entry_flush [PPC] Don't flush the L1-D cache when entering the kernel.
3335 On X86-32 available only on PAE configured kernels.
3336 noexec=on: enable non-executable mappings (default)
3337 noexec=off: disable non-executable mappings
3340 Disable SMAP (Supervisor Mode Access Prevention)
3341 even if it is supported by processor.
3344 Disable SMEP (Supervisor Mode Execution Prevention)
3345 even if it is supported by processor.
3348 This affects only 32-bit executables.
3349 noexec32=on: enable non-executable mappings (default)
3350 read doesn't imply executable mappings
3351 noexec32=off: disable non-executable mappings
3352 read implies executable mappings
3354 nofpu [MIPS,SH] Disable hardware FPU at boot time.
3356 nofxsr [BUGS=X86-32] Disables x86 floating point extended
3357 register save and restore. The kernel will only save
3358 legacy floating-point registers on task switch.
3360 nohugeiomap [KNL,X86,PPC,ARM64] Disable kernel huge I/O mappings.
3362 nohugevmalloc [PPC] Disable kernel huge vmalloc mappings.
3364 nosmt [KNL,S390] Disable symmetric multithreading (SMT).
3365 Equivalent to smt=1.
3367 [KNL,X86] Disable symmetric multithreading (SMT).
3368 nosmt=force: Force disable SMT, cannot be undone
3369 via the sysfs control file.
3371 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1
3372 (bounds check bypass). With this option data leaks are
3373 possible in the system.
3375 nospectre_v2 [X86,PPC_FSL_BOOK3E,ARM64] Disable all mitigations for
3376 the Spectre variant 2 (indirect branch prediction)
3377 vulnerability. System may allow data leaks with this
3380 nospec_store_bypass_disable
3381 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability
3384 [PPC] Don't flush the L1-D cache after accessing user data.
3386 noxsave [BUGS=X86] Disables x86 extended register state save
3387 and restore using xsave. The kernel will fallback to
3388 enabling legacy floating-point and sse state.
3390 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
3391 register states. The kernel will fall back to use
3392 xsave to save the states. By using this parameter,
3393 performance of saving the states is degraded because
3394 xsave doesn't support modified optimization while
3395 xsaveopt supports it on xsaveopt enabled systems.
3397 noxsaves [X86] Disables xsaves and xrstors used in saving and
3398 restoring x86 extended register state in compacted
3399 form of xsave area. The kernel will fall back to use
3400 xsaveopt and xrstor to save and restore the states
3401 in standard form of xsave area. By using this
3402 parameter, xsave area per process might occupy more
3403 memory on xsaves enabled systems.
3405 nohlt [ARM,ARM64,MICROBLAZE,SH] Forces the kernel to busy wait
3406 in do_idle() and not use the arch_cpu_idle()
3407 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP
3408 to be effective. This is useful on platforms where the
3409 sleep(SH) or wfi(ARM,ARM64) instructions do not work
3410 correctly or when doing power measurements to evalute
3411 the impact of the sleep instructions. This is also
3412 useful when using JTAG debugger.
3414 no_file_caps Tells the kernel not to honor file capabilities. The
3415 only way then for a file to be executed with privilege
3416 is to be setuid root or executed by root.
3418 nohalt [IA-64] Tells the kernel not to use the power saving
3419 function PAL_HALT_LIGHT when idle. This increases
3420 power-consumption. On the positive side, it reduces
3421 interrupt wake-up latency, which may improve performance
3422 in certain environments such as networked servers or
3426 Force pointers printed to the console or buffers to be
3427 unhashed. By default, when a pointer is printed via %p
3428 format string, that pointer is "hashed", i.e. obscured
3429 by hashing the pointer value. This is a security feature
3430 that hides actual kernel addresses from unprivileged
3431 users, but it also makes debugging the kernel more
3432 difficult since unequal pointers can no longer be
3433 compared. However, if this command-line option is
3434 specified, then all normal pointers will have their true
3435 value printed. Pointers printed via %pK may still be
3436 hashed. This option should only be specified when
3437 debugging the kernel. Please do not use on production
3440 nohibernate [HIBERNATION] Disable hibernation and resume.
3442 nohz= [KNL] Boottime enable/disable dynamic ticks
3443 Valid arguments: on, off
3446 nohz_full= [KNL,BOOT,SMP,ISOL]
3447 The argument is a cpu list, as described above.
3448 In kernels built with CONFIG_NO_HZ_FULL=y, set
3449 the specified list of CPUs whose tick will be stopped
3450 whenever possible. The boot CPU will be forced outside
3451 the range to maintain the timekeeping. Any CPUs
3452 in this list will have their RCU callbacks offloaded,
3453 just as if they had also been called out in the
3454 rcu_nocbs= boot parameter.
3456 noiotrap [SH] Disables trapped I/O port accesses.
3458 noirqdebug [X86-32] Disables the code which attempts to detect and
3459 disable unhandled interrupt sources.
3461 no_timer_check [X86,APIC] Disables the code which tests for
3462 broken timer IRQ sources.
3464 noisapnp [ISAPNP] Disables ISA PnP code.
3466 noinitrd [RAM] Tells the kernel not to load any configured
3469 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
3471 [Deprecated - use intremap=off]
3475 noinvpcid [X86] Disable the INVPCID cpu feature.
3477 nojitter [IA-64] Disables jitter checking for ITC timers.
3479 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
3481 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
3485 [X86,PV_OPS] Disable paravirtualized VMware scheduler
3486 clock and use the default one.
3488 no-steal-acc [X86,PV_OPS,ARM64] Disable paravirtualized steal time
3489 accounting. steal time is computed, but won't
3490 influence scheduler behaviour
3492 nolapic [X86-32,APIC] Do not enable or use the local APIC.
3494 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
3496 noltlbs [PPC] Do not use large page/tlb entries for kernel
3497 lowmem mapping on PPC40x and PPC8xx
3499 nomca [IA-64] Disable machine check abort handling
3501 nomce [X86-32] Disable Machine Check Exception
3503 nomfgpt [X86-32] Disable Multi-Function General Purpose
3504 Timer usage (for AMD Geode machines).
3506 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
3507 shutdown the other cpus. Instead use the REBOOT_VECTOR
3510 nomodule Disable module load
3512 nopat [X86] Disable PAT (page attribute table extension of
3513 pagetables) support.
3515 nopcid [X86-64] Disable the PCID cpu feature.
3517 norandmaps Don't use address space randomization. Equivalent to
3518 echo 0 > /proc/sys/kernel/randomize_va_space
3520 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
3521 with UP alternatives
3523 nordrand [X86] Disable kernel use of the RDRAND and
3524 RDSEED instructions even if they are supported
3525 by the processor. RDRAND and RDSEED are still
3526 available to user space applications.
3528 noresume [SWSUSP] Disables resume and restores original swap
3531 no-scroll [VGA] Disables scrollback.
3532 This is required for the Braillex ib80-piezo Braille
3533 reader made by F.H. Papenmeier (Germany).
3537 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
3539 nosgx [X86-64,SGX] Disables Intel SGX kernel support.
3541 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
3542 and disable the IO APIC. legacy for "maxcpus=0".
3544 nosoftlockup [KNL] Disable the soft-lockup detector.
3546 nosync [HW,M68K] Disables sync negotiation for all devices.
3548 nowatchdog [KNL] Disable both lockup detectors, i.e.
3549 soft-lockup and NMI watchdog (hard-lockup).
3553 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
3555 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
3556 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
3557 Some features depend on CPU0. Known dependencies are:
3558 1. Resume from suspend/hibernate depends on CPU0.
3559 Suspend/hibernate will fail if CPU0 is offline and you
3560 need to online CPU0 before suspend/hibernate.
3561 2. PIC interrupts also depend on CPU0. CPU0 can't be
3562 removed if a PIC interrupt is detected.
3563 It's said poweroff/reboot may depend on CPU0 on some
3564 machines although I haven't seen such issues so far
3565 after CPU0 is offline on a few tested machines.
3566 If the dependencies are under your control, you can
3567 turn on cpu0_hotplug.
3569 nps_mtm_hs_ctr= [KNL,ARC]
3570 This parameter sets the maximum duration, in
3571 cycles, each HW thread of the CTOP can run
3572 without interruptions, before HW switches it.
3573 The actual maximum duration is 16 times this
3575 Format: integer between 1 and 255
3578 nptcg= [IA-64] Override max number of concurrent global TLB
3579 purges which is reported from either PAL_VM_SUMMARY or
3582 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
3583 could support. nr_cpus=n : n >= 1 limits the kernel to
3584 support 'n' processors. It could be larger than the
3585 number of already plugged CPU during bootup, later in
3586 runtime you can physically add extra cpu until it reaches
3587 n. So during boot up some boot time memory for per-cpu
3588 variables need be pre-allocated for later physical cpu
3591 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
3593 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86] Disable NUMA, Only
3594 set up a single NUMA node spanning all memory.
3596 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic
3598 Allowed values are enable and disable
3600 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
3601 'node', 'default' can be specified
3602 This can be set from sysctl after boot.
3603 See Documentation/admin-guide/sysctl/vm.rst for details.
3605 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
3606 See Documentation/core-api/debugging-via-ohci1394.rst for more
3609 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
3610 Rather than timing out after 20 ms if an EC
3611 command is not properly ACKed, override the length
3612 of the timeout. We have interrupts disabled while
3613 waiting for the ACK, so if this is set too high
3614 interrupts *may* be lost!
3616 omap_mux= [OMAP] Override bootloader pin multiplexing.
3617 Format: <mux_mode0.mode_name=value>...
3618 For example, to override I2C bus2:
3619 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
3621 oops=panic Always panic on oopses. Default is to just kill the
3622 process, but there is a small probability of
3623 deadlocking the machine.
3624 This will also cause panics on machine check exceptions.
3625 Useful together with panic=30 to trigger a reboot.
3628 [KNL] Boolean flag to control whether the page allocator
3629 should randomize its free lists. The randomization may
3630 be automatically enabled if the kernel detects it is
3631 running on a platform with a direct-mapped memory-side
3632 cache, and this parameter can be used to
3633 override/disable that behavior. The state of the flag
3634 can be read from sysfs at:
3635 /sys/module/page_alloc/parameters/shuffle.
3637 page_owner= [KNL] Boot-time page_owner enabling option.
3638 Storage of the information about who allocated
3639 each page is disabled in default. With this switch,
3641 on: enable the feature
3643 page_poison= [KNL] Boot-time parameter changing the state of
3644 poisoning on the buddy allocator, available with
3645 CONFIG_PAGE_POISONING=y.
3646 off: turn off poisoning (default)
3647 on: turn on poisoning
3649 page_reporting.page_reporting_order=
3650 [KNL] Minimal page reporting order
3652 Adjust the minimal page reporting order. The page
3653 reporting is disabled when it exceeds (MAX_ORDER-1).
3655 panic= [KNL] Kernel behaviour on panic: delay <timeout>
3656 timeout > 0: seconds before rebooting
3657 timeout = 0: wait forever
3658 timeout < 0: reboot immediately
3661 panic_print= Bitmask for printing system info when panic happens.
3662 User can chose combination of the following bits:
3663 bit 0: print all tasks info
3664 bit 1: print system memory info
3665 bit 2: print timer info
3666 bit 3: print locks info if CONFIG_LOCKDEP is on
3667 bit 4: print ftrace buffer
3668 bit 5: print all printk messages in buffer
3670 panic_on_taint= Bitmask for conditionally calling panic() in add_taint()
3671 Format: <hex>[,nousertaint]
3672 Hexadecimal bitmask representing the set of TAINT flags
3673 that will cause the kernel to panic when add_taint() is
3674 called with any of the flags in this set.
3675 The optional switch "nousertaint" can be utilized to
3676 prevent userspace forced crashes by writing to sysctl
3677 /proc/sys/kernel/tainted any flagset matching with the
3678 bitmask set on panic_on_taint.
3679 See Documentation/admin-guide/tainted-kernels.rst for
3680 extra details on the taint flags that users can pick
3681 to compose the bitmask to assign to panic_on_taint.
3683 panic_on_warn panic() instead of WARN(). Useful to cause kdump
3686 crash_kexec_post_notifiers
3687 Run kdump after running panic-notifiers and dumping
3688 kmsg. This only for the users who doubt kdump always
3689 succeeds in any situation.
3690 Note that this also increases risks of kdump failure,
3691 because some panic notifiers can make the crashed
3692 kernel more unstable.
3694 parkbd.port= [HW] Parallel port number the keyboard adapter is
3695 connected to, default is 0.
3697 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
3698 0 for XT, 1 for AT (default is AT).
3701 parport= [HW,PPT] Specify parallel ports. 0 disables.
3702 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
3703 Use 'auto' to force the driver to use any
3704 IRQ/DMA settings detected (the default is to
3705 ignore detected IRQ/DMA settings because of
3706 possible conflicts). You can specify the base
3707 address, IRQ, and DMA settings; IRQ and DMA
3708 should be numbers, or 'auto' (for using detected
3709 settings on that particular port), or 'nofifo'
3710 (to avoid using a FIFO even if it is detected).
3711 Parallel ports are assigned in the order they
3712 are specified on the command line, starting
3715 parport_init_mode= [HW,PPT]
3716 Configure VIA parallel port to operate in
3717 a specific mode. This is necessary on Pegasos
3718 computer where firmware has no options for setting
3719 up parallel port mode and sets it to spp.
3720 Currently this function knows 686a and 8231 chips.
3721 Format: [spp|ps2|epp|ecp|ecpepp]
3723 pata_legacy.all= [HW,LIBATA]
3725 Set to non-zero to probe primary and secondary ISA
3726 port ranges on PCI systems where no PCI PATA device
3727 has been found at either range. Disabled by default.
3729 pata_legacy.autospeed= [HW,LIBATA]
3731 Set to non-zero if a chip is present that snoops speed
3732 changes. Disabled by default.
3734 pata_legacy.ht6560a= [HW,LIBATA]
3736 Set to 1, 2, or 3 for HT 6560A on the primary channel,
3737 the secondary channel, or both channels respectively.
3738 Disabled by default.
3740 pata_legacy.ht6560b= [HW,LIBATA]
3742 Set to 1, 2, or 3 for HT 6560B on the primary channel,
3743 the secondary channel, or both channels respectively.
3744 Disabled by default.
3746 pata_legacy.iordy_mask= [HW,LIBATA]
3748 IORDY enable mask. Set individual bits to allow IORDY
3749 for the respective channel. Bit 0 is for the first
3750 legacy channel handled by this driver, bit 1 is for
3751 the second channel, and so on. The sequence will often
3752 correspond to the primary legacy channel, the secondary
3753 legacy channel, and so on, but the handling of a PCI
3754 bus and the use of other driver options may interfere
3755 with the sequence. By default IORDY is allowed across
3758 pata_legacy.opti82c46x= [HW,LIBATA]
3760 Set to 1, 2, or 3 for Opti 82c611A on the primary
3761 channel, the secondary channel, or both channels
3762 respectively. Disabled by default.
3764 pata_legacy.opti82c611a= [HW,LIBATA]
3766 Set to 1, 2, or 3 for Opti 82c465MV on the primary
3767 channel, the secondary channel, or both channels
3768 respectively. Disabled by default.
3770 pata_legacy.pio_mask= [HW,LIBATA]
3772 PIO mode mask for autospeed devices. Set individual
3773 bits to allow the use of the respective PIO modes.
3774 Bit 0 is for mode 0, bit 1 is for mode 1, and so on.
3775 All modes allowed by default.
3777 pata_legacy.probe_all= [HW,LIBATA]
3779 Set to non-zero to probe tertiary and further ISA
3780 port ranges on PCI systems. Disabled by default.
3782 pata_legacy.probe_mask= [HW,LIBATA]
3784 Probe mask for legacy ISA PATA ports. Depending on
3785 platform configuration and the use of other driver
3786 options up to 6 legacy ports are supported: 0x1f0,
3787 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing
3788 of individual ports can be disabled by setting the
3789 corresponding bits in the mask to 1. Bit 0 is for
3790 the first port in the list above (0x1f0), and so on.
3791 By default all supported ports are probed.
3793 pata_legacy.qdi= [HW,LIBATA]
3795 Set to non-zero to probe QDI controllers. By default
3796 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise.
3798 pata_legacy.winbond= [HW,LIBATA]
3800 Set to non-zero to probe Winbond controllers. Use
3801 the standard I/O port (0x130) if 1, otherwise the
3802 value given is the I/O port to use (typically 0x1b0).
3803 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE,
3806 pata_platform.pio_mask= [HW,LIBATA]
3808 Supported PIO mode mask. Set individual bits to allow
3809 the use of the respective PIO modes. Bit 0 is for
3810 mode 0, bit 1 is for mode 1, and so on. Mode 0 only
3814 Halt all CPUs after the first oops has been printed for
3815 the specified number of seconds. This is to be used if
3816 your oopses keep scrolling off the screen.
3821 See header of drivers/block/paride/pcd.c.
3822 See also Documentation/admin-guide/blockdev/paride.rst.
3824 pci=option[,option...] [PCI] various PCI subsystem options.
3826 Some options herein operate on a specific device
3827 or a set of devices (<pci_dev>). These are
3828 specified in one of the following formats:
3830 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]*
3831 pci:<vendor>:<device>[:<subvendor>:<subdevice>]
3833 Note: the first format specifies a PCI
3834 bus/device/function address which may change
3835 if new hardware is inserted, if motherboard
3836 firmware changes, or due to changes caused
3837 by other kernel parameters. If the
3838 domain is left unspecified, it is
3839 taken to be zero. Optionally, a path
3840 to a device through multiple device/function
3841 addresses can be specified after the base
3842 address (this is more robust against
3843 renumbering issues). The second format
3844 selects devices using IDs from the
3845 configuration space which may match multiple
3846 devices in the system.
3848 earlydump dump PCI config space before the kernel
3850 off [X86] don't probe for the PCI bus
3851 bios [X86-32] force use of PCI BIOS, don't access
3852 the hardware directly. Use this if your machine
3853 has a non-standard PCI host bridge.
3854 nobios [X86-32] disallow use of PCI BIOS, only direct
3855 hardware access methods are allowed. Use this
3856 if you experience crashes upon bootup and you
3857 suspect they are caused by the BIOS.
3858 conf1 [X86] Force use of PCI Configuration Access
3859 Mechanism 1 (config address in IO port 0xCF8,
3860 data in IO port 0xCFC, both 32-bit).
3861 conf2 [X86] Force use of PCI Configuration Access
3862 Mechanism 2 (IO port 0xCF8 is an 8-bit port for
3863 the function, IO port 0xCFA, also 8-bit, sets
3864 bus number. The config space is then accessed
3865 through ports 0xC000-0xCFFF).
3866 See http://wiki.osdev.org/PCI for more info
3867 on the configuration access mechanisms.
3868 noaer [PCIE] If the PCIEAER kernel config parameter is
3869 enabled, this kernel boot option can be used to
3870 disable the use of PCIE advanced error reporting.
3871 nodomains [PCI] Disable support for multiple PCI
3872 root domains (aka PCI segments, in ACPI-speak).
3873 nommconf [X86] Disable use of MMCONFIG for PCI
3875 check_enable_amd_mmconf [X86] check for and enable
3876 properly configured MMIO access to PCI
3877 config space on AMD family 10h CPU
3878 nomsi [MSI] If the PCI_MSI kernel config parameter is
3879 enabled, this kernel boot option can be used to
3880 disable the use of MSI interrupts system-wide.
3881 noioapicquirk [APIC] Disable all boot interrupt quirks.
3882 Safety option to keep boot IRQs enabled. This
3883 should never be necessary.
3884 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
3885 primary IO-APIC for bridges that cannot disable
3886 boot IRQs. This fixes a source of spurious IRQs
3887 when the system masks IRQs.
3888 noioapicreroute [APIC] Disable workaround that uses the
3889 boot IRQ equivalent of an IRQ that connects to
3890 a chipset where boot IRQs cannot be disabled.
3891 The opposite of ioapicreroute.
3892 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
3893 routing table. These calls are known to be buggy
3894 on several machines and they hang the machine
3895 when used, but on other computers it's the only
3896 way to get the interrupt routing table. Try
3897 this option if the kernel is unable to allocate
3898 IRQs or discover secondary PCI buses on your
3900 rom [X86] Assign address space to expansion ROMs.
3901 Use with caution as certain devices share
3902 address decoders between ROMs and other
3904 norom [X86] Do not assign address space to
3905 expansion ROMs that do not already have
3906 BIOS assigned address ranges.
3907 nobar [X86] Do not assign address space to the
3908 BARs that weren't assigned by the BIOS.
3909 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
3910 assigned automatically to PCI devices. You can
3911 make the kernel exclude IRQs of your ISA cards
3913 pirqaddr=0xAAAAA [X86] Specify the physical address
3914 of the PIRQ table (normally generated
3915 by the BIOS) if it is outside the
3916 F0000h-100000h range.
3917 lastbus=N [X86] Scan all buses thru bus #N. Can be
3918 useful if the kernel is unable to find your
3919 secondary buses and you want to tell it
3920 explicitly which ones they are.
3921 assign-busses [X86] Always assign all PCI bus
3922 numbers ourselves, overriding
3923 whatever the firmware may have done.
3924 usepirqmask [X86] Honor the possible IRQ mask stored
3925 in the BIOS $PIR table. This is needed on
3926 some systems with broken BIOSes, notably
3927 some HP Pavilion N5400 and Omnibook XE3
3928 notebooks. This will have no effect if ACPI
3929 IRQ routing is enabled.
3930 noacpi [X86] Do not use ACPI for IRQ routing
3931 or for PCI scanning.
3932 use_crs [X86] Use PCI host bridge window information
3933 from ACPI. On BIOSes from 2008 or later, this
3934 is enabled by default. If you need to use this,
3935 please report a bug.
3936 nocrs [X86] Ignore PCI host bridge windows from ACPI.
3937 If you need to use this, please report a bug.
3938 routeirq Do IRQ routing for all PCI devices.
3939 This is normally done in pci_enable_device(),
3940 so this option is a temporary workaround
3941 for broken drivers that don't call it.
3942 skip_isa_align [X86] do not align io start addr, so can
3943 handle more pci cards
3944 noearly [X86] Don't do any early type 1 scanning.
3945 This might help on some broken boards which
3946 machine check when some devices' config space
3947 is read. But various workarounds are disabled
3948 and some IOMMU drivers will not work.
3949 bfsort Sort PCI devices into breadth-first order.
3950 This sorting is done to get a device
3951 order compatible with older (<= 2.4) kernels.
3952 nobfsort Don't sort PCI devices into breadth-first order.
3953 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
3954 tuning and use the BIOS-configured MPS defaults.
3955 pcie_bus_safe Set every device's MPS to the largest value
3956 supported by all devices below the root complex.
3957 pcie_bus_perf Set device MPS to the largest allowable MPS
3958 based on its parent bus. Also set MRRS (Max
3959 Read Request Size) to the largest supported
3960 value (no larger than the MPS that the device
3961 or bus can support) for best performance.
3962 pcie_bus_peer2peer Set every device's MPS to 128B, which
3963 every device is guaranteed to support. This
3964 configuration allows peer-to-peer DMA between
3965 any pair of devices, possibly at the cost of
3966 reduced performance. This also guarantees
3967 that hot-added devices will work.
3968 cbiosize=nn[KMG] The fixed amount of bus space which is
3969 reserved for the CardBus bridge's IO window.
3970 The default value is 256 bytes.
3971 cbmemsize=nn[KMG] The fixed amount of bus space which is
3972 reserved for the CardBus bridge's memory
3973 window. The default value is 64 megabytes.
3976 [<order of align>@]<pci_dev>[; ...]
3977 Specifies alignment and device to reassign
3978 aligned memory resources. How to
3979 specify the device is described above.
3980 If <order of align> is not specified,
3981 PAGE_SIZE is used as alignment.
3982 A PCI-PCI bridge can be specified if resource
3983 windows need to be expanded.
3984 To specify the alignment for several
3985 instances of a device, the PCI vendor,
3986 device, subvendor, and subdevice may be
3987 specified, e.g., 12@pci:8086:9c22:103c:198f
3988 for 4096-byte alignment.
3989 ecrc= Enable/disable PCIe ECRC (transaction layer
3990 end-to-end CRC checking).
3991 bios: Use BIOS/firmware settings. This is the
3995 hpiosize=nn[KMG] The fixed amount of bus space which is
3996 reserved for hotplug bridge's IO window.
3997 Default size is 256 bytes.
3998 hpmmiosize=nn[KMG] The fixed amount of bus space which is
3999 reserved for hotplug bridge's MMIO window.
4000 Default size is 2 megabytes.
4001 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is
4002 reserved for hotplug bridge's MMIO_PREF window.
4003 Default size is 2 megabytes.
4004 hpmemsize=nn[KMG] The fixed amount of bus space which is
4005 reserved for hotplug bridge's MMIO and
4007 Default size is 2 megabytes.
4008 hpbussize=nn The minimum amount of additional bus numbers
4009 reserved for buses below a hotplug bridge.
4011 realloc= Enable/disable reallocating PCI bridge resources
4012 if allocations done by BIOS are too small to
4013 accommodate resources required by all child
4015 off: Turn realloc off
4017 realloc same as realloc=on
4018 noari do not use PCIe ARI.
4019 noats [PCIE, Intel-IOMMU, AMD-IOMMU]
4020 do not use PCIe ATS (and IOMMU device IOTLB).
4021 pcie_scan_all Scan all possible PCIe devices. Otherwise we
4022 only look for one device below a PCIe downstream
4024 big_root_window Try to add a big 64bit memory window to the PCIe
4025 root complex on AMD CPUs. Some GFX hardware
4026 can resize a BAR to allow access to all VRAM.
4027 Adding the window is slightly risky (it may
4028 conflict with unreported devices), so this
4030 disable_acs_redir=<pci_dev>[; ...]
4031 Specify one or more PCI devices (in the format
4032 specified above) separated by semicolons.
4033 Each device specified will have the PCI ACS
4034 redirect capabilities forced off which will
4035 allow P2P traffic between devices through
4036 bridges without forcing it upstream. Note:
4037 this removes isolation between devices and
4038 may put more devices in an IOMMU group.
4039 force_floating [S390] Force usage of floating interrupts.
4040 nomio [S390] Do not use MIO instructions.
4041 norid [S390] ignore the RID field and force use of
4042 one PCI domain per PCI function
4044 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
4047 force Enable ASPM even on devices that claim not to support it.
4048 WARNING: Forcing ASPM on may cause system lockups.
4050 pcie_ports= [PCIE] PCIe port services handling:
4051 native Use native PCIe services (PME, AER, DPC, PCIe hotplug)
4052 even if the platform doesn't give the OS permission to
4053 use them. This may cause conflicts if the platform
4054 also tries to use these services.
4055 dpc-native Use native PCIe service for DPC only. May
4056 cause conflicts if firmware uses AER or DPC.
4057 compat Disable native PCIe services (PME, AER, DPC, PCIe
4060 pcie_port_pm= [PCIE] PCIe port power management handling:
4061 off Disable power management of all PCIe ports
4062 force Forcibly enable power management of all PCIe ports
4064 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
4065 nomsi Do not use MSI for native PCIe PME signaling (this makes
4066 all PCIe root ports use INTx for all services).
4068 pcmv= [HW,PCMCIA] BadgePAD 4
4072 Keep all power-domains already enabled by bootloader on,
4073 even if no driver has claimed them. This is useful
4074 for debug and development, but should not be
4075 needed on a platform with proper driver support.
4078 See Documentation/admin-guide/blockdev/paride.rst.
4080 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
4083 See arch/parisc/kernel/pdc_chassis.c
4085 percpu_alloc= Select which percpu first chunk allocator to use.
4086 Currently supported values are "embed" and "page".
4087 Archs may support subset or none of the selections.
4088 See comments in mm/percpu.c for details on each
4089 allocator. This parameter is primarily for debugging
4090 and performance comparison.
4093 See Documentation/admin-guide/blockdev/paride.rst.
4096 See Documentation/admin-guide/blockdev/paride.rst.
4098 pirq= [SMP,APIC] Manual mp-table setup
4099 See Documentation/x86/i386/IO-APIC.rst.
4101 plip= [PPT,NET] Parallel port network link
4102 Format: { parport<nr> | timid | 0 }
4103 See also Documentation/admin-guide/parport.rst.
4105 pmtmr= [X86] Manual setup of pmtmr I/O Port.
4106 Override pmtimer IOPort with a hex value.
4109 pm_debug_messages [SUSPEND,KNL]
4110 Enable suspend/resume debug messages during boot up.
4113 Enable PNP debug messages (depends on the
4114 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
4115 via /sys/module/pnp/parameters/debug. We always show
4116 current resource usage; turning this on also shows
4117 possible settings and some assignment information.
4123 { on | off | curr | res | no-curr | no-res }
4126 [ISAPNP] Exclude IRQs for the autoconfiguration
4129 [ISAPNP] Exclude DMAs for the autoconfiguration
4131 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
4132 Ranges are in pairs (I/O port base and size).
4135 [ISAPNP] Exclude memory regions for the
4137 Ranges are in pairs (memory base and size).
4139 ports= [IP_VS_FTP] IPVS ftp helper module
4141 Up to 8 (IP_VS_APP_MAX_PORTS) ports
4143 Format: <port>,<port>....
4145 powersave=off [PPC] This option disables power saving features.
4146 It specifically disables cpuidle and sets the
4147 platform machine description specific power_save
4148 function to NULL. On Idle the CPU just reduces
4151 ppc_strict_facility_enable
4152 [PPC] This option catches any kernel floating point,
4153 Altivec, VSX and SPE outside of regions specifically
4154 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
4155 There is some performance impact when enabling this.
4159 Disable Hardware Transactional Memory
4162 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC
4163 none - Limited to cond_resched() calls
4164 voluntary - Limited to cond_resched() and might_sleep() calls
4165 full - Any section that isn't explicitly preempt disabled
4166 can be preempted anytime.
4168 print-fatal-signals=
4169 [KNL] debug: print fatal signals
4171 If enabled, warn about various signal handling
4172 related application anomalies: too many signals,
4173 too many POSIX.1 timers, fatal signals causing a
4176 If you hit the warning due to signal overflow,
4177 you might want to try "ulimit -i unlimited".
4181 printk.always_kmsg_dump=
4182 Trigger kmsg_dump for cases other than kernel oops or
4184 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
4187 printk.devkmsg={on,off,ratelimit}
4188 Control writing to /dev/kmsg.
4189 on - unlimited logging to /dev/kmsg from userspace
4190 off - logging to /dev/kmsg disabled
4191 ratelimit - ratelimit the logging
4194 printk.time= Show timing data prefixed to each printk message line
4195 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
4197 processor.max_cstate= [HW,ACPI]
4198 Limit processor to maximum C-state
4199 max_cstate=9 overrides any DMI blacklist limit.
4201 processor.nocst [HW,ACPI]
4202 Ignore the _CST method to determine C-states,
4203 instead using the legacy FADT method
4205 profile= [KNL] Enable kernel profiling via /proc/profile
4206 Format: [<profiletype>,]<number>
4207 Param: <profiletype>: "schedule", "sleep", or "kvm"
4208 [defaults to kernel profiling]
4209 Param: "schedule" - profile schedule points.
4210 Param: "sleep" - profile D-state sleeping (millisecs).
4211 Requires CONFIG_SCHEDSTATS
4212 Param: "kvm" - profile VM exits.
4213 Param: <number> - step/bucket size as a power of 2 for
4214 statistical time based profiling.
4216 prompt_ramdisk= [RAM] [Deprecated]
4218 prot_virt= [S390] enable hosting protected virtual machines
4219 isolated from the hypervisor (if hardware supports
4223 psi= [KNL] Enable or disable pressure stall information
4227 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
4228 probe for; one of (bare|imps|exps|lifebook|any).
4229 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
4231 psmouse.resetafter= [HW,MOUSE]
4232 Try to reset the device after so many bad packets
4235 [HW,MOUSE] Set desired mouse resolution, in dpi.
4236 psmouse.smartscroll=
4237 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
4238 0 = disabled, 1 = enabled (default).
4240 pstore.backend= Specify the name of the pstore backend to use
4243 See Documentation/admin-guide/blockdev/paride.rst.
4245 pti= [X86-64] Control Page Table Isolation of user and
4246 kernel address spaces. Disabling this feature
4247 removes hardening, but improves performance of
4248 system calls and interrupts.
4250 on - unconditionally enable
4251 off - unconditionally disable
4252 auto - kernel detects whether your CPU model is
4253 vulnerable to issues that PTI mitigates
4255 Not specifying this option is equivalent to pti=auto.
4258 Equivalent to pti=off
4261 [KNL] Number of legacy pty's. Overwrites compiled-in
4264 quiet [KNL] Disable most log messages
4269 See Documentation/admin-guide/md.rst.
4271 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
4272 See Documentation/admin-guide/blockdev/ramdisk.rst.
4274 ramdisk_start= [RAM] RAM disk image start address
4276 random.trust_cpu={on,off}
4277 [KNL] Enable or disable trusting the use of the
4278 CPU's random number generator (if available) to
4279 fully seed the kernel's CRNG. Default is controlled
4280 by CONFIG_RANDOM_TRUST_CPU.
4282 randomize_kstack_offset=
4283 [KNL] Enable or disable kernel stack offset
4284 randomization, which provides roughly 5 bits of
4285 entropy, frustrating memory corruption attacks
4286 that depend on stack address determinism or
4287 cross-syscall address exposures. This is only
4288 available on architectures that have defined
4289 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET.
4290 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
4291 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT.
4293 ras=option[,option,...] [KNL] RAS-specific options
4296 Disable the Correctable Errors Collector,
4297 see CONFIG_RAS_CEC help text.
4300 The argument is a cpu list, as described above.
4302 In kernels built with CONFIG_RCU_NOCB_CPU=y, set
4303 the specified list of CPUs to be no-callback CPUs.
4304 Invocation of these CPUs' RCU callbacks will be
4305 offloaded to "rcuox/N" kthreads created for that
4306 purpose, where "x" is "p" for RCU-preempt, and
4307 "s" for RCU-sched, and "N" is the CPU number.
4308 This reduces OS jitter on the offloaded CPUs,
4309 which can be useful for HPC and real-time
4310 workloads. It can also improve energy efficiency
4311 for asymmetric multiprocessors.
4314 Rather than requiring that offloaded CPUs
4315 (specified by rcu_nocbs= above) explicitly
4316 awaken the corresponding "rcuoN" kthreads,
4317 make these kthreads poll for callbacks.
4318 This improves the real-time response for the
4319 offloaded CPUs by relieving them of the need to
4320 wake up the corresponding kthread, but degrades
4321 energy efficiency by requiring that the kthreads
4322 periodically wake up to do the polling.
4324 rcutree.blimit= [KNL]
4325 Set maximum number of finished RCU callbacks to
4326 process in one batch.
4328 rcutree.dump_tree= [KNL]
4329 Dump the structure of the rcu_node combining tree
4330 out at early boot. This is used for diagnostic
4331 purposes, to verify correct tree setup.
4333 rcutree.gp_cleanup_delay= [KNL]
4334 Set the number of jiffies to delay each step of
4335 RCU grace-period cleanup.
4337 rcutree.gp_init_delay= [KNL]
4338 Set the number of jiffies to delay each step of
4339 RCU grace-period initialization.
4341 rcutree.gp_preinit_delay= [KNL]
4342 Set the number of jiffies to delay each step of
4343 RCU grace-period pre-initialization, that is,
4344 the propagation of recent CPU-hotplug changes up
4345 the rcu_node combining tree.
4347 rcutree.use_softirq= [KNL]
4348 If set to zero, move all RCU_SOFTIRQ processing to
4349 per-CPU rcuc kthreads. Defaults to a non-zero
4350 value, meaning that RCU_SOFTIRQ is used by default.
4351 Specify rcutree.use_softirq=0 to use rcuc kthreads.
4353 But note that CONFIG_PREEMPT_RT=y kernels disable
4354 this kernel boot parameter, forcibly setting it
4357 rcutree.rcu_fanout_exact= [KNL]
4358 Disable autobalancing of the rcu_node combining
4359 tree. This is used by rcutorture, and might
4360 possibly be useful for architectures having high
4361 cache-to-cache transfer latencies.
4363 rcutree.rcu_fanout_leaf= [KNL]
4364 Change the number of CPUs assigned to each
4365 leaf rcu_node structure. Useful for very
4366 large systems, which will choose the value 64,
4367 and for NUMA systems with large remote-access
4368 latencies, which will choose a value aligned
4369 with the appropriate hardware boundaries.
4371 rcutree.rcu_min_cached_objs= [KNL]
4372 Minimum number of objects which are cached and
4373 maintained per one CPU. Object size is equal
4374 to PAGE_SIZE. The cache allows to reduce the
4375 pressure to page allocator, also it makes the
4376 whole algorithm to behave better in low memory
4379 rcutree.rcu_delay_page_cache_fill_msec= [KNL]
4380 Set the page-cache refill delay (in milliseconds)
4381 in response to low-memory conditions. The range
4382 of permitted values is in the range 0:100000.
4384 rcutree.jiffies_till_first_fqs= [KNL]
4385 Set delay from grace-period initialization to
4386 first attempt to force quiescent states.
4387 Units are jiffies, minimum value is zero,
4388 and maximum value is HZ.
4390 rcutree.jiffies_till_next_fqs= [KNL]
4391 Set delay between subsequent attempts to force
4392 quiescent states. Units are jiffies, minimum
4393 value is one, and maximum value is HZ.
4395 rcutree.jiffies_till_sched_qs= [KNL]
4396 Set required age in jiffies for a
4397 given grace period before RCU starts
4398 soliciting quiescent-state help from
4399 rcu_note_context_switch() and cond_resched().
4400 If not specified, the kernel will calculate
4401 a value based on the most recent settings
4402 of rcutree.jiffies_till_first_fqs
4403 and rcutree.jiffies_till_next_fqs.
4404 This calculated value may be viewed in
4405 rcutree.jiffies_to_sched_qs. Any attempt to set
4406 rcutree.jiffies_to_sched_qs will be cheerfully
4409 rcutree.kthread_prio= [KNL,BOOT]
4410 Set the SCHED_FIFO priority of the RCU per-CPU
4411 kthreads (rcuc/N). This value is also used for
4412 the priority of the RCU boost threads (rcub/N)
4413 and for the RCU grace-period kthreads (rcu_bh,
4414 rcu_preempt, and rcu_sched). If RCU_BOOST is
4415 set, valid values are 1-99 and the default is 1
4416 (the least-favored priority). Otherwise, when
4417 RCU_BOOST is not set, valid values are 0-99 and
4418 the default is zero (non-realtime operation).
4420 rcutree.rcu_nocb_gp_stride= [KNL]
4421 Set the number of NOCB callback kthreads in
4422 each group, which defaults to the square root
4423 of the number of CPUs. Larger numbers reduce
4424 the wakeup overhead on the global grace-period
4425 kthread, but increases that same overhead on
4426 each group's NOCB grace-period kthread.
4428 rcutree.qhimark= [KNL]
4429 Set threshold of queued RCU callbacks beyond which
4430 batch limiting is disabled.
4432 rcutree.qlowmark= [KNL]
4433 Set threshold of queued RCU callbacks below which
4434 batch limiting is re-enabled.
4436 rcutree.qovld= [KNL]
4437 Set threshold of queued RCU callbacks beyond which
4438 RCU's force-quiescent-state scan will aggressively
4439 enlist help from cond_resched() and sched IPIs to
4440 help CPUs more quickly reach quiescent states.
4441 Set to less than zero to make this be set based
4442 on rcutree.qhimark at boot time and to zero to
4443 disable more aggressive help enlistment.
4445 rcutree.rcu_idle_gp_delay= [KNL]
4446 Set wakeup interval for idle CPUs that have
4447 RCU callbacks (RCU_FAST_NO_HZ=y).
4449 rcutree.rcu_kick_kthreads= [KNL]
4450 Cause the grace-period kthread to get an extra
4451 wake_up() if it sleeps three times longer than
4452 it should at force-quiescent-state time.
4453 This wake_up() will be accompanied by a
4454 WARN_ONCE() splat and an ftrace_dump().
4456 rcutree.rcu_unlock_delay= [KNL]
4457 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels,
4458 this specifies an rcu_read_unlock()-time delay
4459 in microseconds. This defaults to zero.
4460 Larger delays increase the probability of
4461 catching RCU pointer leaks, that is, buggy use
4462 of RCU-protected pointers after the relevant
4463 rcu_read_unlock() has completed.
4465 rcutree.sysrq_rcu= [KNL]
4466 Commandeer a sysrq key to dump out Tree RCU's
4467 rcu_node tree with an eye towards determining
4468 why a new grace period has not yet started.
4470 rcuscale.gp_async= [KNL]
4471 Measure performance of asynchronous
4472 grace-period primitives such as call_rcu().
4474 rcuscale.gp_async_max= [KNL]
4475 Specify the maximum number of outstanding
4476 callbacks per writer thread. When a writer
4477 thread exceeds this limit, it invokes the
4478 corresponding flavor of rcu_barrier() to allow
4479 previously posted callbacks to drain.
4481 rcuscale.gp_exp= [KNL]
4482 Measure performance of expedited synchronous
4483 grace-period primitives.
4485 rcuscale.holdoff= [KNL]
4486 Set test-start holdoff period. The purpose of
4487 this parameter is to delay the start of the
4488 test until boot completes in order to avoid
4491 rcuscale.kfree_rcu_test= [KNL]
4492 Set to measure performance of kfree_rcu() flooding.
4494 rcuscale.kfree_rcu_test_double= [KNL]
4495 Test the double-argument variant of kfree_rcu().
4496 If this parameter has the same value as
4497 rcuscale.kfree_rcu_test_single, both the single-
4498 and double-argument variants are tested.
4500 rcuscale.kfree_rcu_test_single= [KNL]
4501 Test the single-argument variant of kfree_rcu().
4502 If this parameter has the same value as
4503 rcuscale.kfree_rcu_test_double, both the single-
4504 and double-argument variants are tested.
4506 rcuscale.kfree_nthreads= [KNL]
4507 The number of threads running loops of kfree_rcu().
4509 rcuscale.kfree_alloc_num= [KNL]
4510 Number of allocations and frees done in an iteration.
4512 rcuscale.kfree_loops= [KNL]
4513 Number of loops doing rcuscale.kfree_alloc_num number
4514 of allocations and frees.
4516 rcuscale.nreaders= [KNL]
4517 Set number of RCU readers. The value -1 selects
4518 N, where N is the number of CPUs. A value
4519 "n" less than -1 selects N-n+1, where N is again
4520 the number of CPUs. For example, -2 selects N
4521 (the number of CPUs), -3 selects N+1, and so on.
4522 A value of "n" less than or equal to -N selects
4525 rcuscale.nwriters= [KNL]
4526 Set number of RCU writers. The values operate
4527 the same as for rcuscale.nreaders.
4528 N, where N is the number of CPUs
4530 rcuscale.perf_type= [KNL]
4531 Specify the RCU implementation to test.
4533 rcuscale.shutdown= [KNL]
4534 Shut the system down after performance tests
4535 complete. This is useful for hands-off automated
4538 rcuscale.verbose= [KNL]
4539 Enable additional printk() statements.
4541 rcuscale.writer_holdoff= [KNL]
4542 Write-side holdoff between grace periods,
4543 in microseconds. The default of zero says
4546 rcutorture.fqs_duration= [KNL]
4547 Set duration of force_quiescent_state bursts
4550 rcutorture.fqs_holdoff= [KNL]
4551 Set holdoff time within force_quiescent_state bursts
4554 rcutorture.fqs_stutter= [KNL]
4555 Set wait time between force_quiescent_state bursts
4558 rcutorture.fwd_progress= [KNL]
4559 Enable RCU grace-period forward-progress testing
4560 for the types of RCU supporting this notion.
4562 rcutorture.fwd_progress_div= [KNL]
4563 Specify the fraction of a CPU-stall-warning
4564 period to do tight-loop forward-progress testing.
4566 rcutorture.fwd_progress_holdoff= [KNL]
4567 Number of seconds to wait between successive
4568 forward-progress tests.
4570 rcutorture.fwd_progress_need_resched= [KNL]
4571 Enclose cond_resched() calls within checks for
4572 need_resched() during tight-loop forward-progress
4575 rcutorture.gp_cond= [KNL]
4576 Use conditional/asynchronous update-side
4577 primitives, if available.
4579 rcutorture.gp_exp= [KNL]
4580 Use expedited update-side primitives, if available.
4582 rcutorture.gp_normal= [KNL]
4583 Use normal (non-expedited) asynchronous
4584 update-side primitives, if available.
4586 rcutorture.gp_sync= [KNL]
4587 Use normal (non-expedited) synchronous
4588 update-side primitives, if available. If all
4589 of rcutorture.gp_cond=, rcutorture.gp_exp=,
4590 rcutorture.gp_normal=, and rcutorture.gp_sync=
4591 are zero, rcutorture acts as if is interpreted
4592 they are all non-zero.
4594 rcutorture.irqreader= [KNL]
4595 Run RCU readers from irq handlers, or, more
4596 accurately, from a timer handler. Not all RCU
4597 flavors take kindly to this sort of thing.
4599 rcutorture.leakpointer= [KNL]
4600 Leak an RCU-protected pointer out of the reader.
4601 This can of course result in splats, and is
4602 intended to test the ability of things like
4603 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect
4606 rcutorture.n_barrier_cbs= [KNL]
4607 Set callbacks/threads for rcu_barrier() testing.
4609 rcutorture.nfakewriters= [KNL]
4610 Set number of concurrent RCU writers. These just
4611 stress RCU, they don't participate in the actual
4612 test, hence the "fake".
4614 rcutorture.nocbs_nthreads= [KNL]
4615 Set number of RCU callback-offload togglers.
4616 Zero (the default) disables toggling.
4618 rcutorture.nocbs_toggle= [KNL]
4619 Set the delay in milliseconds between successive
4620 callback-offload toggling attempts.
4622 rcutorture.nreaders= [KNL]
4623 Set number of RCU readers. The value -1 selects
4624 N-1, where N is the number of CPUs. A value
4625 "n" less than -1 selects N-n-2, where N is again
4626 the number of CPUs. For example, -2 selects N
4627 (the number of CPUs), -3 selects N+1, and so on.
4629 rcutorture.object_debug= [KNL]
4630 Enable debug-object double-call_rcu() testing.
4632 rcutorture.onoff_holdoff= [KNL]
4633 Set time (s) after boot for CPU-hotplug testing.
4635 rcutorture.onoff_interval= [KNL]
4636 Set time (jiffies) between CPU-hotplug operations,
4637 or zero to disable CPU-hotplug testing.
4639 rcutorture.read_exit= [KNL]
4640 Set the number of read-then-exit kthreads used
4641 to test the interaction of RCU updaters and
4642 task-exit processing.
4644 rcutorture.read_exit_burst= [KNL]
4645 The number of times in a given read-then-exit
4646 episode that a set of read-then-exit kthreads
4649 rcutorture.read_exit_delay= [KNL]
4650 The delay, in seconds, between successive
4651 read-then-exit testing episodes.
4653 rcutorture.shuffle_interval= [KNL]
4654 Set task-shuffle interval (s). Shuffling tasks
4655 allows some CPUs to go into dyntick-idle mode
4656 during the rcutorture test.
4658 rcutorture.shutdown_secs= [KNL]
4659 Set time (s) after boot system shutdown. This
4660 is useful for hands-off automated testing.
4662 rcutorture.stall_cpu= [KNL]
4663 Duration of CPU stall (s) to test RCU CPU stall
4664 warnings, zero to disable.
4666 rcutorture.stall_cpu_block= [KNL]
4667 Sleep while stalling if set. This will result
4668 in warnings from preemptible RCU in addition
4669 to any other stall-related activity.
4671 rcutorture.stall_cpu_holdoff= [KNL]
4672 Time to wait (s) after boot before inducing stall.
4674 rcutorture.stall_cpu_irqsoff= [KNL]
4675 Disable interrupts while stalling if set.
4677 rcutorture.stall_gp_kthread= [KNL]
4678 Duration (s) of forced sleep within RCU
4679 grace-period kthread to test RCU CPU stall
4680 warnings, zero to disable. If both stall_cpu
4681 and stall_gp_kthread are specified, the
4682 kthread is starved first, then the CPU.
4684 rcutorture.stat_interval= [KNL]
4685 Time (s) between statistics printk()s.
4687 rcutorture.stutter= [KNL]
4688 Time (s) to stutter testing, for example, specifying
4689 five seconds causes the test to run for five seconds,
4690 wait for five seconds, and so on. This tests RCU's
4691 ability to transition abruptly to and from idle.
4693 rcutorture.test_boost= [KNL]
4694 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
4695 "Maybe" means test if the RCU implementation
4696 under test support RCU priority boosting.
4698 rcutorture.test_boost_duration= [KNL]
4699 Duration (s) of each individual boost test.
4701 rcutorture.test_boost_interval= [KNL]
4702 Interval (s) between each boost test.
4704 rcutorture.test_no_idle_hz= [KNL]
4705 Test RCU's dyntick-idle handling. See also the
4706 rcutorture.shuffle_interval parameter.
4708 rcutorture.torture_type= [KNL]
4709 Specify the RCU implementation to test.
4711 rcutorture.verbose= [KNL]
4712 Enable additional printk() statements.
4714 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL]
4715 Dump ftrace buffer after reporting RCU CPU
4718 rcupdate.rcu_cpu_stall_suppress= [KNL]
4719 Suppress RCU CPU stall warning messages.
4721 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL]
4722 Suppress RCU CPU stall warning messages and
4723 rcutorture writer stall warnings that occur
4724 during early boot, that is, during the time
4725 before the init task is spawned.
4727 rcupdate.rcu_cpu_stall_timeout= [KNL]
4728 Set timeout for RCU CPU stall warning messages.
4730 rcupdate.rcu_expedited= [KNL]
4731 Use expedited grace-period primitives, for
4732 example, synchronize_rcu_expedited() instead
4733 of synchronize_rcu(). This reduces latency,
4734 but can increase CPU utilization, degrade
4735 real-time latency, and degrade energy efficiency.
4736 No effect on CONFIG_TINY_RCU kernels.
4738 rcupdate.rcu_normal= [KNL]
4739 Use only normal grace-period primitives,
4740 for example, synchronize_rcu() instead of
4741 synchronize_rcu_expedited(). This improves
4742 real-time latency, CPU utilization, and
4743 energy efficiency, but can expose users to
4744 increased grace-period latency. This parameter
4745 overrides rcupdate.rcu_expedited. No effect on
4746 CONFIG_TINY_RCU kernels.
4748 rcupdate.rcu_normal_after_boot= [KNL]
4749 Once boot has completed (that is, after
4750 rcu_end_inkernel_boot() has been invoked), use
4751 only normal grace-period primitives. No effect
4752 on CONFIG_TINY_RCU kernels.
4754 But note that CONFIG_PREEMPT_RT=y kernels enables
4755 this kernel boot parameter, forcibly setting
4756 it to the value one, that is, converting any
4757 post-boot attempt at an expedited RCU grace
4758 period to instead use normal non-expedited
4759 grace-period processing.
4761 rcupdate.rcu_task_ipi_delay= [KNL]
4762 Set time in jiffies during which RCU tasks will
4763 avoid sending IPIs, starting with the beginning
4764 of a given grace period. Setting a large
4765 number avoids disturbing real-time workloads,
4766 but lengthens grace periods.
4768 rcupdate.rcu_task_stall_timeout= [KNL]
4769 Set timeout in jiffies for RCU task stall warning
4770 messages. Disable with a value less than or equal
4773 rcupdate.rcu_self_test= [KNL]
4774 Run the RCU early boot self tests
4778 Run specified binary instead of /init from the ramdisk,
4779 used for early userspace startup. See initrd.
4782 force - Override the decision by the kernel to hide the
4783 advertisement of RDRAND support (this affects
4784 certain AMD processors because of buggy BIOS
4785 support, specifically around the suspend/resume
4789 Turn on/off individual RDT features. List is:
4790 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp,
4792 E.g. to turn on cmt and turn off mba use:
4796 Format (x86 or x86_64):
4797 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \
4799 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
4801 Where reboot_mode is one of warm (soft) or cold (hard) or gpio
4802 (prefix with 'panic_' to set mode for panic
4804 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
4805 reboot_force is either force or not specified,
4806 reboot_cpu is s[mp]#### with #### being the processor
4807 to be used for rebooting.
4809 refscale.holdoff= [KNL]
4810 Set test-start holdoff period. The purpose of
4811 this parameter is to delay the start of the
4812 test until boot completes in order to avoid
4815 refscale.loops= [KNL]
4816 Set the number of loops over the synchronization
4817 primitive under test. Increasing this number
4818 reduces noise due to loop start/end overhead,
4819 but the default has already reduced the per-pass
4820 noise to a handful of picoseconds on ca. 2020
4823 refscale.nreaders= [KNL]
4824 Set number of readers. The default value of -1
4825 selects N, where N is roughly 75% of the number
4826 of CPUs. A value of zero is an interesting choice.
4828 refscale.nruns= [KNL]
4829 Set number of runs, each of which is dumped onto
4832 refscale.readdelay= [KNL]
4833 Set the read-side critical-section duration,
4834 measured in microseconds.
4836 refscale.scale_type= [KNL]
4837 Specify the read-protection implementation to test.
4839 refscale.shutdown= [KNL]
4840 Shut down the system at the end of the performance
4841 test. This defaults to 1 (shut it down) when
4842 refscale is built into the kernel and to 0 (leave
4843 it running) when refscale is built as a module.
4845 refscale.verbose= [KNL]
4846 Enable additional printk() statements.
4848 refscale.verbose_batched= [KNL]
4849 Batch the additional printk() statements. If zero
4850 (the default) or negative, print everything. Otherwise,
4851 print every Nth verbose statement, where N is the value
4855 [KNL, SMP] Set scheduler's default relax_domain_level.
4856 See Documentation/admin-guide/cgroup-v1/cpusets.rst.
4858 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory
4859 Format: <base1>,<size1>[,<base2>,<size2>,...]
4860 Reserve I/O ports or memory so the kernel won't use
4861 them. If <base> is less than 0x10000, the region
4862 is assumed to be I/O ports; otherwise it is memory.
4864 reservetop= [X86-32]
4866 Reserves a hole at the top of the kernel virtual
4869 reset_devices [KNL] Force drivers to reset the underlying device
4870 during initialization.
4873 Specify the partition device for software suspend
4875 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
4877 resume_offset= [SWSUSP]
4878 Specify the offset from the beginning of the partition
4879 given by "resume=" at which the swap header is located,
4880 in <PAGE_SIZE> units (needed only for swap files).
4881 See Documentation/power/swsusp-and-swap-files.rst
4883 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
4884 read the resume files
4886 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
4887 Useful for devices that are detected asynchronously
4888 (e.g. USB and MMC devices).
4890 hibernate= [HIBERNATION]
4891 noresume Don't check if there's a hibernation image
4892 present during boot.
4893 nocompress Don't compress/decompress hibernation images.
4894 no Disable hibernation and resume.
4895 protect_image Turn on image protection during restoration
4896 (that will set all pages holding image data
4897 during restoration read-only).
4899 retain_initrd [RAM] Keep initrd memory after extraction
4901 rfkill.default_state=
4902 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
4903 etc. communication is blocked by default.
4906 rfkill.master_switch_mode=
4907 0 The "airplane mode" button does nothing.
4908 1 The "airplane mode" button toggles between everything
4909 blocked and the previous configuration.
4910 2 The "airplane mode" button toggles between everything
4911 blocked and everything unblocked.
4913 rhash_entries= [KNL,NET]
4914 Set number of hash buckets for route cache
4917 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported
4920 ro [KNL] Mount root device read-only on boot
4923 on Mark read-only kernel memory as read-only (default).
4924 off Leave read-only kernel memory writable for debugging.
4927 Enable the uart passthrough on the designated usb port
4928 on Rockchip SoCs. When active, the signals of the
4929 debug-uart get routed to the D+ and D- pins of the usb
4930 port and the regular usb controller gets disabled.
4932 root= [KNL] Root filesystem
4933 See name_to_dev_t comment in init/do_mounts.c.
4935 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
4936 mount the root filesystem
4938 rootflags= [KNL] Set root filesystem mount option string
4940 rootfstype= [KNL] Set root filesystem type
4942 rootwait [KNL] Wait (indefinitely) for root device to show up.
4943 Useful for devices that are detected asynchronously
4944 (e.g. USB and MMC devices).
4946 rproc_mem=nn[KMG][@address]
4947 [KNL,ARM,CMA] Remoteproc physical memory block.
4948 Memory area to be used by remote processor image,
4951 rw [KNL] Mount root device read-write on boot
4953 S [KNL] Run init in single mode
4955 s390_iommu= [HW,S390]
4956 Set s390 IOTLB flushing mode
4958 With strict flushing every unmap operation will result in
4959 an IOTLB flush. Default is lazy flushing before reuse,
4963 See drivers/net/irda/sa1100_ir.c.
4965 sbni= [NET] Granch SBNI12 leased line adapter
4967 sched_verbose [KNL] Enables verbose scheduler debug messages.
4969 schedstats= [KNL,X86] Enable or disable scheduled statistics.
4970 Allowed values are enable and disable. This feature
4971 incurs a small amount of overhead in the scheduler
4972 but is useful for debugging and performance tuning.
4974 sched_thermal_decay_shift=
4975 [KNL, SMP] Set a decay shift for scheduler thermal
4976 pressure signal. Thermal pressure signal follows the
4977 default decay period of other scheduler pelt
4978 signals(usually 32 ms but configurable). Setting
4979 sched_thermal_decay_shift will left shift the decay
4980 period for the thermal pressure signal by the shift
4982 i.e. with the default pelt decay period of 32 ms
4983 sched_thermal_decay_shift thermal pressure decay pr
4987 Format: integer between 0 and 10
4990 scftorture.holdoff= [KNL]
4991 Number of seconds to hold off before starting
4992 test. Defaults to zero for module insertion and
4993 to 10 seconds for built-in smp_call_function()
4996 scftorture.longwait= [KNL]
4997 Request ridiculously long waits randomly selected
4998 up to the chosen limit in seconds. Zero (the
4999 default) disables this feature. Please note
5000 that requesting even small non-zero numbers of
5001 seconds can result in RCU CPU stall warnings,
5002 softlockup complaints, and so on.
5004 scftorture.nthreads= [KNL]
5005 Number of kthreads to spawn to invoke the
5006 smp_call_function() family of functions.
5007 The default of -1 specifies a number of kthreads
5008 equal to the number of CPUs.
5010 scftorture.onoff_holdoff= [KNL]
5011 Number seconds to wait after the start of the
5012 test before initiating CPU-hotplug operations.
5014 scftorture.onoff_interval= [KNL]
5015 Number seconds to wait between successive
5016 CPU-hotplug operations. Specifying zero (which
5017 is the default) disables CPU-hotplug operations.
5019 scftorture.shutdown_secs= [KNL]
5020 The number of seconds following the start of the
5021 test after which to shut down the system. The
5022 default of zero avoids shutting down the system.
5023 Non-zero values are useful for automated tests.
5025 scftorture.stat_interval= [KNL]
5026 The number of seconds between outputting the
5027 current test statistics to the console. A value
5028 of zero disables statistics output.
5030 scftorture.stutter_cpus= [KNL]
5031 The number of jiffies to wait between each change
5032 to the set of CPUs under test.
5034 scftorture.use_cpus_read_lock= [KNL]
5035 Use use_cpus_read_lock() instead of the default
5036 preempt_disable() to disable CPU hotplug
5037 while invoking one of the smp_call_function*()
5040 scftorture.verbose= [KNL]
5041 Enable additional printk() statements.
5043 scftorture.weight_single= [KNL]
5044 The probability weighting to use for the
5045 smp_call_function_single() function with a zero
5046 "wait" parameter. A value of -1 selects the
5047 default if all other weights are -1. However,
5048 if at least one weight has some other value, a
5049 value of -1 will instead select a weight of zero.
5051 scftorture.weight_single_wait= [KNL]
5052 The probability weighting to use for the
5053 smp_call_function_single() function with a
5054 non-zero "wait" parameter. See weight_single.
5056 scftorture.weight_many= [KNL]
5057 The probability weighting to use for the
5058 smp_call_function_many() function with a zero
5059 "wait" parameter. See weight_single.
5060 Note well that setting a high probability for
5061 this weighting can place serious IPI load
5064 scftorture.weight_many_wait= [KNL]
5065 The probability weighting to use for the
5066 smp_call_function_many() function with a
5067 non-zero "wait" parameter. See weight_single
5070 scftorture.weight_all= [KNL]
5071 The probability weighting to use for the
5072 smp_call_function_all() function with a zero
5073 "wait" parameter. See weight_single and
5076 scftorture.weight_all_wait= [KNL]
5077 The probability weighting to use for the
5078 smp_call_function_all() function with a
5079 non-zero "wait" parameter. See weight_single
5082 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
5083 xtime_lock contention on larger systems, and/or RCU lock
5084 contention on all systems with CONFIG_MAXSMP set.
5085 Format: { "0" | "1" }
5086 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
5088 Note: increases power consumption, thus should only be
5089 enabled if running jitter sensitive (HPC/RT) workloads.
5091 security= [SECURITY] Choose a legacy "major" security module to
5092 enable at boot. This has been deprecated by the
5095 selinux= [SELINUX] Disable or enable SELinux at boot time.
5096 Format: { "0" | "1" }
5097 See security/selinux/Kconfig help text.
5102 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
5103 Format: { "0" | "1" }
5104 See security/apparmor/Kconfig help text
5107 Default value is set via kernel config option.
5109 serialnumber [BUGS=X86-32]
5112 Maximal number of shapers.
5120 Enable merging of slabs with similar size when the
5121 kernel is built without CONFIG_SLAB_MERGE_DEFAULT.
5124 Disable merging of slabs with similar size. May be
5125 necessary if there is some reason to distinguish
5126 allocs to different slabs, especially in hardened
5127 environments where the risk of heap overflows and
5128 layout control by attackers can usually be
5129 frustrated by disabling merging. This will reduce
5130 most of the exposure of a heap attack to a single
5131 cache (risks via metadata attacks are mostly
5132 unchanged). Debug options disable merging on their
5134 For more information see Documentation/vm/slub.rst.
5136 slab_max_order= [MM, SLAB]
5137 Determines the maximum allowed order for slabs.
5138 A high setting may cause OOMs due to memory
5139 fragmentation. Defaults to 1 for systems with
5140 more than 32MB of RAM, 0 otherwise.
5142 slub_debug[=options[,slabs][;[options[,slabs]]...] [MM, SLUB]
5143 Enabling slub_debug allows one to determine the
5144 culprit if slab objects become corrupted. Enabling
5145 slub_debug can create guard zones around objects and
5146 may poison objects when not in use. Also tracks the
5147 last alloc / free. For more information see
5148 Documentation/vm/slub.rst.
5150 slub_max_order= [MM, SLUB]
5151 Determines the maximum allowed order for slabs.
5152 A high setting may cause OOMs due to memory
5153 fragmentation. For more information see
5154 Documentation/vm/slub.rst.
5156 slub_min_objects= [MM, SLUB]
5157 The minimum number of objects per slab. SLUB will
5158 increase the slab order up to slub_max_order to
5159 generate a sufficiently large slab able to contain
5160 the number of objects indicated. The higher the number
5161 of objects the smaller the overhead of tracking slabs
5162 and the less frequently locks need to be acquired.
5163 For more information see Documentation/vm/slub.rst.
5165 slub_min_order= [MM, SLUB]
5166 Determines the minimum page order for slabs. Must be
5167 lower than slub_max_order.
5168 For more information see Documentation/vm/slub.rst.
5170 slub_merge [MM, SLUB]
5171 Same with slab_merge.
5173 slub_nomerge [MM, SLUB]
5174 Same with slab_nomerge. This is supported for legacy.
5175 See slab_nomerge for more information.
5178 Format: <io1>[,<io2>[,...,<io8>]]
5180 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
5181 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
5182 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
5183 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
5184 smsc-ircc2.ircc_irq= [HW] IRQ line
5185 smsc-ircc2.ircc_dma= [HW] DMA channel
5186 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
5187 0: Toshiba Satellite 1800 (GP data pin select)
5188 1: Fast pin select (default)
5191 smt [KNL,S390] Set the maximum number of threads (logical
5192 CPUs) to use per physical CPU on systems capable of
5193 symmetric multithreading (SMT). Will be capped to the
5194 actual hardware limit.
5196 Default: -1 (no limit)
5199 [KNL] Should the soft-lockup detector generate panics.
5202 A value of 1 instructs the soft-lockup detector
5203 to panic the machine when a soft-lockup occurs. It is
5204 also controlled by the kernel.softlockup_panic sysctl
5205 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the
5206 respective build-time switch to that functionality.
5208 softlockup_all_cpu_backtrace=
5209 [KNL] Should the soft-lockup detector generate
5210 backtraces on all cpus.
5213 sonypi.*= [HW] Sony Programmable I/O Control Device driver
5214 See Documentation/admin-guide/laptops/sonypi.rst
5216 spectre_v2= [X86] Control mitigation of Spectre variant 2
5217 (indirect branch speculation) vulnerability.
5218 The default operation protects the kernel from
5221 on - unconditionally enable, implies
5223 off - unconditionally disable, implies
5225 auto - kernel detects whether your CPU model is
5228 Selecting 'on' will, and 'auto' may, choose a
5229 mitigation method at run time according to the
5230 CPU, the available microcode, the setting of the
5231 CONFIG_RETPOLINE configuration option, and the
5232 compiler with which the kernel was built.
5234 Selecting 'on' will also enable the mitigation
5235 against user space to user space task attacks.
5237 Selecting 'off' will disable both the kernel and
5238 the user space protections.
5240 Specific mitigations can also be selected manually:
5242 retpoline - replace indirect branches
5243 retpoline,generic - google's original retpoline
5244 retpoline,amd - AMD-specific minimal thunk
5246 Not specifying this option is equivalent to
5250 [X86] Control mitigation of Spectre variant 2
5251 (indirect branch speculation) vulnerability between
5254 on - Unconditionally enable mitigations. Is
5255 enforced by spectre_v2=on
5257 off - Unconditionally disable mitigations. Is
5258 enforced by spectre_v2=off
5260 prctl - Indirect branch speculation is enabled,
5261 but mitigation can be enabled via prctl
5262 per thread. The mitigation control state
5263 is inherited on fork.
5266 - Like "prctl" above, but only STIBP is
5267 controlled per thread. IBPB is issued
5268 always when switching between different user
5272 - Same as "prctl" above, but all seccomp
5273 threads will enable the mitigation unless
5274 they explicitly opt out.
5277 - Like "seccomp" above, but only STIBP is
5278 controlled per thread. IBPB is issued
5279 always when switching between different
5280 user space processes.
5282 auto - Kernel selects the mitigation depending on
5283 the available CPU features and vulnerability.
5286 If CONFIG_SECCOMP=y then "seccomp", otherwise "prctl"
5288 Not specifying this option is equivalent to
5289 spectre_v2_user=auto.
5291 spec_store_bypass_disable=
5292 [HW] Control Speculative Store Bypass (SSB) Disable mitigation
5293 (Speculative Store Bypass vulnerability)
5295 Certain CPUs are vulnerable to an exploit against a
5296 a common industry wide performance optimization known
5297 as "Speculative Store Bypass" in which recent stores
5298 to the same memory location may not be observed by
5299 later loads during speculative execution. The idea
5300 is that such stores are unlikely and that they can
5301 be detected prior to instruction retirement at the
5302 end of a particular speculation execution window.
5304 In vulnerable processors, the speculatively forwarded
5305 store can be used in a cache side channel attack, for
5306 example to read memory to which the attacker does not
5307 directly have access (e.g. inside sandboxed code).
5309 This parameter controls whether the Speculative Store
5310 Bypass optimization is used.
5312 On x86 the options are:
5314 on - Unconditionally disable Speculative Store Bypass
5315 off - Unconditionally enable Speculative Store Bypass
5316 auto - Kernel detects whether the CPU model contains an
5317 implementation of Speculative Store Bypass and
5318 picks the most appropriate mitigation. If the
5319 CPU is not vulnerable, "off" is selected. If the
5320 CPU is vulnerable the default mitigation is
5321 architecture and Kconfig dependent. See below.
5322 prctl - Control Speculative Store Bypass per thread
5323 via prctl. Speculative Store Bypass is enabled
5324 for a process by default. The state of the control
5325 is inherited on fork.
5326 seccomp - Same as "prctl" above, but all seccomp threads
5327 will disable SSB unless they explicitly opt out.
5329 Default mitigations:
5330 X86: If CONFIG_SECCOMP=y "seccomp", otherwise "prctl"
5332 On powerpc the options are:
5334 on,auto - On Power8 and Power9 insert a store-forwarding
5335 barrier on kernel entry and exit. On Power7
5336 perform a software flush on kernel entry and
5340 Not specifying this option is equivalent to
5341 spec_store_bypass_disable=auto.
5343 spia_io_base= [HW,MTD]
5349 [X86] Enable split lock detection or bus lock detection
5351 When enabled (and if hardware support is present), atomic
5352 instructions that access data across cache line
5353 boundaries will result in an alignment check exception
5354 for split lock detection or a debug exception for
5359 warn - the kernel will emit rate-limited warnings
5360 about applications triggering the #AC
5361 exception or the #DB exception. This mode is
5362 the default on CPUs that support split lock
5363 detection or bus lock detection. Default
5364 behavior is by #AC if both features are
5365 enabled in hardware.
5367 fatal - the kernel will send SIGBUS to applications
5368 that trigger the #AC exception or the #DB
5369 exception. Default behavior is by #AC if
5370 both features are enabled in hardware.
5373 Set system wide rate limit to N bus locks
5374 per second for bus lock detection.
5377 N/A for split lock detection.
5380 If an #AC exception is hit in the kernel or in
5381 firmware (i.e. not while executing in user mode)
5382 the kernel will oops in either "warn" or "fatal"
5385 #DB exception for bus lock is triggered only when
5389 Control the Special Register Buffer Data Sampling
5392 Certain CPUs are vulnerable to an MDS-like
5393 exploit which can leak bits from the random
5396 By default, this issue is mitigated by
5397 microcode. However, the microcode fix can cause
5398 the RDRAND and RDSEED instructions to become
5399 much slower. Among other effects, this will
5400 result in reduced throughput from /dev/urandom.
5402 The microcode mitigation can be disabled with
5403 the following option:
5405 off: Disable mitigation and remove
5406 performance impact to RDRAND and RDSEED
5408 srcutree.counter_wrap_check [KNL]
5409 Specifies how frequently to check for
5410 grace-period sequence counter wrap for the
5411 srcu_data structure's ->srcu_gp_seq_needed field.
5412 The greater the number of bits set in this kernel
5413 parameter, the less frequently counter wrap will
5414 be checked for. Note that the bottom two bits
5417 srcutree.exp_holdoff [KNL]
5418 Specifies how many nanoseconds must elapse
5419 since the end of the last SRCU grace period for
5420 a given srcu_struct until the next normal SRCU
5421 grace period will be considered for automatic
5422 expediting. Set to zero to disable automatic
5426 Speculative Store Bypass Disable control
5428 On CPUs that are vulnerable to the Speculative
5429 Store Bypass vulnerability and offer a
5430 firmware based mitigation, this parameter
5431 indicates how the mitigation should be used:
5433 force-on: Unconditionally enable mitigation for
5434 for both kernel and userspace
5435 force-off: Unconditionally disable mitigation for
5436 for both kernel and userspace
5437 kernel: Always enable mitigation in the
5438 kernel, and offer a prctl interface
5439 to allow userspace to register its
5440 interest in being mitigated too.
5442 stack_guard_gap= [MM]
5443 override the default stack gap protection. The value
5444 is in page units and it defines how many pages prior
5445 to (for stacks growing down) resp. after (for stacks
5446 growing up) the main stack are reserved for no other
5447 mapping. Default value is 256 pages.
5449 stack_depot_disable= [KNL]
5450 Setting this to true through kernel command line will
5451 disable the stack depot thereby saving the static memory
5452 consumed by the stack hash table. By default this is set
5456 Enabled the stack tracer on boot up.
5458 stacktrace_filter=[function-list]
5459 [FTRACE] Limit the functions that the stack tracer
5460 will trace at boot up. function-list is a comma-separated
5461 list of functions. This list can be changed at run
5462 time by the stack_trace_filter file in the debugfs
5463 tracing directory. Note, this enables stack tracing
5464 and the stacktrace above is not needed.
5468 Set the STI (builtin display/keyboard on the HP-PARISC
5469 machines) console (graphic card) which should be used
5470 as the initial boot-console.
5471 See also comment in drivers/video/console/sticore.c.
5474 See comment in drivers/video/console/sticore.c.
5477 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
5479 sunrpc.min_resvport=
5480 sunrpc.max_resvport=
5482 SunRPC servers often require that client requests
5483 originate from a privileged port (i.e. a port in the
5484 range 0 < portnr < 1024).
5485 An administrator who wishes to reserve some of these
5486 ports for other uses may adjust the range that the
5487 kernel's sunrpc client considers to be privileged
5488 using these two parameters to set the minimum and
5489 maximum port values.
5491 sunrpc.svc_rpc_per_connection_limit=
5493 Limit the number of requests that the server will
5494 process in parallel from a single connection.
5495 The default value is 0 (no limit).
5499 Control how the NFS server code allocates CPUs to
5500 service thread pools. Depending on how many NICs
5501 you have and where their interrupts are bound, this
5502 option will affect which CPUs will do NFS serving.
5503 Note: this parameter cannot be changed while the
5504 NFS server is running.
5506 auto the server chooses an appropriate mode
5507 automatically using heuristics
5508 global a single global pool contains all CPUs
5509 percpu one pool for each CPU
5510 pernode one pool for each NUMA node (equivalent
5511 to global on non-NUMA machines)
5513 sunrpc.tcp_slot_table_entries=
5514 sunrpc.udp_slot_table_entries=
5516 Sets the upper limit on the number of simultaneous
5517 RPC calls that can be sent from the client to a
5518 server. Increasing these values may allow you to
5519 improve throughput, but will also increase the
5520 amount of memory reserved for use by the client.
5522 suspend.pm_test_delay=
5524 Sets the number of seconds to remain in a suspend test
5525 mode before resuming the system (see
5526 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
5527 is set. Default value is 5.
5530 Format: { on | off | y | n | 1 | 0 }
5531 This parameter controls use of the Protected
5532 Execution Facility on pSeries.
5535 [KNL] Enable accounting of swap in memory resource
5536 controller if no parameter or 1 is given or disable
5537 it if 0 is given (See Documentation/admin-guide/cgroup-v1/memory.rst)
5539 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
5540 Format: { <int> | force | noforce }
5541 <int> -- Number of I/O TLB slabs
5542 force -- force using of bounce buffers even if they
5543 wouldn't be automatically used by the kernel
5544 noforce -- Never use bounce buffers (for debugging)
5549 Set a sysctl parameter, right before loading the init
5550 process, as if the value was written to the respective
5551 /proc/sys/... file. Both '.' and '/' are recognized as
5552 separators. Unrecognized parameters and invalid values
5553 are reported in the kernel log. Sysctls registered
5554 later by a loaded module cannot be set this way.
5555 Example: sysctl.vm.swappiness=40
5557 sysfs.deprecated=0|1 [KNL]
5558 Enable/disable old style sysfs layout for old udev
5559 on older distributions. When this option is enabled
5560 very new udev will not work anymore. When this option
5561 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
5562 in older udev will not work anymore.
5563 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
5564 the kernel configuration.
5566 sysrq_always_enabled
5568 Ignore sysrq setting - this boot parameter will
5569 neutralize any effect of /proc/sys/kernel/sysrq.
5570 Useful for debugging.
5572 tcpmhash_entries= [KNL,NET]
5573 Set the number of tcp_metrics_hash slots.
5574 Default value is 8192 or 16384 depending on total
5575 ram pages. This is used to specify the TCP metrics
5576 cache size. See Documentation/networking/ip-sysctl.rst
5577 "tcp_no_metrics_save" section for more details.
5581 test_suspend= [SUSPEND][,N]
5582 Specify "mem" (for Suspend-to-RAM) or "standby" (for
5583 standby suspend) or "freeze" (for suspend type freeze)
5584 as the system sleep state during system startup with
5585 the optional capability to repeat N number of times.
5586 The system is woken from this state using a
5587 wakeup-capable RTC alarm.
5589 thash_entries= [KNL,NET]
5590 Set number of hash buckets for TCP connection
5592 thermal.act= [HW,ACPI]
5593 -1: disable all active trip points in all thermal zones
5594 <degrees C>: override all lowest active trip points
5596 thermal.crt= [HW,ACPI]
5597 -1: disable all critical trip points in all thermal zones
5598 <degrees C>: override all critical trip points
5600 thermal.nocrt= [HW,ACPI]
5601 Set to disable actions on ACPI thermal zone
5602 critical and hot trip points.
5604 thermal.off= [HW,ACPI]
5605 1: disable ACPI thermal control
5607 thermal.psv= [HW,ACPI]
5608 -1: disable all passive trip points
5609 <degrees C>: override all passive trip points to this
5612 thermal.tzp= [HW,ACPI]
5613 Specify global default ACPI thermal zone polling rate
5614 <deci-seconds>: poll all this frequency
5615 0: no polling (default)
5618 Force threading of all interrupt handlers except those
5619 marked explicitly IRQF_NO_THREAD.
5623 Specify if the kernel should make use of the cpu
5624 topology information if the hardware supports this.
5625 The scheduler will make use of this information and
5626 e.g. base its process migration decisions on it.
5629 topology_updates= [KNL, PPC, NUMA]
5631 Specify if the kernel should ignore (off)
5632 topology updates sent by the hypervisor to this
5635 torture.disable_onoff_at_boot= [KNL]
5636 Prevent the CPU-hotplug component of torturing
5637 until after init has spawned.
5639 torture.ftrace_dump_at_shutdown= [KNL]
5640 Dump the ftrace buffer at torture-test shutdown,
5641 even if there were no errors. This can be a
5642 very costly operation when many torture tests
5643 are running concurrently, especially on systems
5644 with rotating-rust storage.
5646 torture.verbose_sleep_frequency= [KNL]
5647 Specifies how many verbose printk()s should be
5648 emitted between each sleep. The default of zero
5649 disables verbose-printk() sleeping.
5651 torture.verbose_sleep_duration= [KNL]
5652 Duration of each verbose-printk() sleep in jiffies.
5656 tpm_suspend_pcr=[HW,TPM]
5657 Format: integer pcr id
5658 Specify that at suspend time, the tpm driver
5659 should extend the specified pcr with zeros,
5660 as a workaround for some chips which fail to
5661 flush the last written pcr on TPM_SaveState.
5662 This will guarantee that all the other pcrs
5665 trace_buf_size=nn[KMG]
5666 [FTRACE] will set tracing buffer size on each cpu.
5668 trace_event=[event-list]
5669 [FTRACE] Set and start specified trace events in order
5670 to facilitate early boot debugging. The event-list is a
5671 comma-separated list of trace events to enable. See
5672 also Documentation/trace/events.rst
5674 trace_options=[option-list]
5675 [FTRACE] Enable or disable tracer options at boot.
5676 The option-list is a comma delimited list of options
5677 that can be enabled or disabled just as if you were
5678 to echo the option name into
5680 /sys/kernel/debug/tracing/trace_options
5682 For example, to enable stacktrace option (to dump the
5683 stack trace of each event), add to the command line:
5685 trace_options=stacktrace
5687 See also Documentation/trace/ftrace.rst "trace options"
5691 Have the tracepoints sent to printk as well as the
5692 tracing ring buffer. This is useful for early boot up
5693 where the system hangs or reboots and does not give the
5694 option for reading the tracing buffer or performing a
5695 ftrace_dump_on_oops.
5697 To turn off having tracepoints sent to printk,
5698 echo 0 > /proc/sys/kernel/tracepoint_printk
5699 Note, echoing 1 into this file without the
5700 tracepoint_printk kernel cmdline option has no effect.
5702 The tp_printk_stop_on_boot (see below) can also be used
5703 to stop the printing of events to console at
5708 Having tracepoints sent to printk() and activating high
5709 frequency tracepoints such as irq or sched, can cause
5710 the system to live lock.
5712 tp_printk_stop_on_boot[FTRACE]
5713 When tp_printk (above) is set, it can cause a lot of noise
5714 on the console. It may be useful to only include the
5715 printing of events during boot up, as user space may
5716 make the system inoperable.
5718 This command line option will stop the printing of events
5719 to console at the late_initcall_sync() time frame.
5722 [FTRACE] enable this option to disable tracing when a
5723 warning is hit. This turns off "tracing_on". Tracing can
5724 be enabled again by echoing '1' into the "tracing_on"
5725 file located in /sys/kernel/debug/tracing/
5727 This option is useful, as it disables the trace before
5728 the WARNING dump is called, which prevents the trace to
5729 be filled with content caused by the warning output.
5731 This option can also be set at run time via the sysctl
5732 option: kernel/traceoff_on_warning
5734 transparent_hugepage=
5736 Format: [always|madvise|never]
5737 Can be used to control the default behavior of the system
5738 with respect to transparent hugepages.
5739 See Documentation/admin-guide/mm/transhuge.rst
5742 trusted.source= [KEYS]
5744 This parameter identifies the trust source as a backend
5745 for trusted keys implementation. Supported trust
5749 If not specified then it defaults to iterating through
5750 the trust source list starting with TPM and assigns the
5751 first trust source as a backend which is initialized
5752 successfully during iteration.
5754 tsc= Disable clocksource stability checks for TSC.
5756 [x86] reliable: mark tsc clocksource as reliable, this
5757 disables clocksource verification at runtime, as well
5758 as the stability checks done at bootup. Used to enable
5759 high-resolution timer mode on older hardware, and in
5760 virtualized environment.
5761 [x86] noirqtime: Do not use TSC to do irq accounting.
5762 Used to run time disable IRQ_TIME_ACCOUNTING on any
5763 platforms where RDTSC is slow and this accounting
5765 [x86] unstable: mark the TSC clocksource as unstable, this
5766 marks the TSC unconditionally unstable at bootup and
5767 avoids any further wobbles once the TSC watchdog notices.
5768 [x86] nowatchdog: disable clocksource watchdog. Used
5769 in situations with strict latency requirements (where
5770 interruptions from clocksource watchdog are not
5773 tsc_early_khz= [X86] Skip early TSC calibration and use the given
5774 value instead. Useful when the early TSC frequency discovery
5775 procedure is not reliable, such as on overclocked systems
5776 with CPUID.16h support and partial CPUID.15h support.
5777 Format: <unsigned int>
5779 tsx= [X86] Control Transactional Synchronization
5780 Extensions (TSX) feature in Intel processors that
5781 support TSX control.
5783 This parameter controls the TSX feature. The options are:
5785 on - Enable TSX on the system. Although there are
5786 mitigations for all known security vulnerabilities,
5787 TSX has been known to be an accelerator for
5788 several previous speculation-related CVEs, and
5789 so there may be unknown security risks associated
5790 with leaving it enabled.
5792 off - Disable TSX on the system. (Note that this
5793 option takes effect only on newer CPUs which are
5794 not vulnerable to MDS, i.e., have
5795 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get
5796 the new IA32_TSX_CTRL MSR through a microcode
5797 update. This new MSR allows for the reliable
5798 deactivation of the TSX functionality.)
5800 auto - Disable TSX if X86_BUG_TAA is present,
5801 otherwise enable TSX on the system.
5803 Not specifying this option is equivalent to tsx=off.
5805 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
5808 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async
5809 Abort (TAA) vulnerability.
5811 Similar to Micro-architectural Data Sampling (MDS)
5812 certain CPUs that support Transactional
5813 Synchronization Extensions (TSX) are vulnerable to an
5814 exploit against CPU internal buffers which can forward
5815 information to a disclosure gadget under certain
5818 In vulnerable processors, the speculatively forwarded
5819 data can be used in a cache side channel attack, to
5820 access data to which the attacker does not have direct
5823 This parameter controls the TAA mitigation. The
5826 full - Enable TAA mitigation on vulnerable CPUs
5829 full,nosmt - Enable TAA mitigation and disable SMT on
5830 vulnerable CPUs. If TSX is disabled, SMT
5831 is not disabled because CPU is not
5832 vulnerable to cross-thread TAA attacks.
5833 off - Unconditionally disable TAA mitigation
5835 On MDS-affected machines, tsx_async_abort=off can be
5836 prevented by an active MDS mitigation as both vulnerabilities
5837 are mitigated with the same mechanism so in order to disable
5838 this mitigation, you need to specify mds=off too.
5840 Not specifying this option is equivalent to
5841 tsx_async_abort=full. On CPUs which are MDS affected
5842 and deploy MDS mitigation, TAA mitigation is not
5843 required and doesn't provide any additional
5847 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
5849 turbografx.map[2|3]= [HW,JOY]
5850 TurboGraFX parallel port interface
5852 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
5853 See also Documentation/input/devices/joystick-parport.rst
5855 udbg-immortal [PPC] When debugging early kernel crashes that
5856 happen after console_init() and before a proper
5857 console driver takes over, this boot options might
5858 help "seeing" what's going on.
5860 uhash_entries= [KNL,NET]
5861 Set number of hash buckets for UDP/UDP-Lite connections
5864 [USB] Ignore overcurrent events (default N).
5865 Some badly-designed motherboards generate lots of
5866 bogus events, for ports that aren't wired to
5867 anything. Set this parameter to avoid log spamming.
5868 Note that genuine overcurrent events won't be
5872 [X86] Cause panic on unknown NMI.
5874 usbcore.authorized_default=
5875 [USB] Default USB device authorization:
5876 (default -1 = authorized except for wireless USB,
5877 0 = not authorized, 1 = authorized, 2 = authorized
5878 if device connected to internal port)
5880 usbcore.autosuspend=
5881 [USB] The autosuspend time delay (in seconds) used
5882 for newly-detected USB devices (default 2). This
5883 is the time required before an idle device will be
5884 autosuspended. Devices for which the delay is set
5885 to a negative value won't be autosuspended at all.
5887 usbcore.usbfs_snoop=
5888 [USB] Set to log all usbfs traffic (default 0 = off).
5890 usbcore.usbfs_snoop_max=
5891 [USB] Maximum number of bytes to snoop in each URB
5894 usbcore.blinkenlights=
5895 [USB] Set to cycle leds on hubs (default 0 = off).
5897 usbcore.old_scheme_first=
5898 [USB] Start with the old device initialization
5899 scheme (default 0 = off).
5901 usbcore.usbfs_memory_mb=
5902 [USB] Memory limit (in MB) for buffers allocated by
5903 usbfs (default = 16, 0 = max = 2047).
5905 usbcore.use_both_schemes=
5906 [USB] Try the other device initialization scheme
5907 if the first one fails (default 1 = enabled).
5909 usbcore.initial_descriptor_timeout=
5910 [USB] Specifies timeout for the initial 64-byte
5911 USB_REQ_GET_DESCRIPTOR request in milliseconds
5912 (default 5000 = 5.0 seconds).
5914 usbcore.nousb [USB] Disable the USB subsystem
5917 [USB] A list of quirk entries to augment the built-in
5918 usb core quirk list. List entries are separated by
5919 commas. Each entry has the form
5920 VendorID:ProductID:Flags. The IDs are 4-digit hex
5921 numbers and Flags is a set of letters. Each letter
5922 will change the built-in quirk; setting it if it is
5923 clear and clearing it if it is set. The letters have
5924 the following meanings:
5925 a = USB_QUIRK_STRING_FETCH_255 (string
5926 descriptors must not be fetched using
5928 b = USB_QUIRK_RESET_RESUME (device can't resume
5929 correctly so reset it instead);
5930 c = USB_QUIRK_NO_SET_INTF (device can't handle
5931 Set-Interface requests);
5932 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't
5933 handle its Configuration or Interface
5935 e = USB_QUIRK_RESET (device can't be reset
5936 (e.g morph devices), don't use reset);
5937 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has
5938 more interface descriptions than the
5939 bNumInterfaces count, and can't handle
5940 talking to these interfaces);
5941 g = USB_QUIRK_DELAY_INIT (device needs a pause
5942 during initialization, after we read
5943 the device descriptor);
5944 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For
5945 high speed and super speed interrupt
5946 endpoints, the USB 2.0 and USB 3.0 spec
5947 require the interval in microframes (1
5948 microframe = 125 microseconds) to be
5949 calculated as interval = 2 ^
5951 Devices with this quirk report their
5952 bInterval as the result of this
5953 calculation instead of the exponent
5954 variable used in the calculation);
5955 i = USB_QUIRK_DEVICE_QUALIFIER (device can't
5956 handle device_qualifier descriptor
5958 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device
5959 generates spurious wakeup, ignore
5960 remote wakeup capability);
5961 k = USB_QUIRK_NO_LPM (device can't handle Link
5963 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL
5964 (Device reports its bInterval as linear
5965 frames instead of the USB 2.0
5967 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs
5968 to be disconnected before suspend to
5969 prevent spurious wakeup);
5970 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a
5971 pause after every control message);
5972 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra
5973 delay after resetting its port);
5974 Example: quirks=0781:5580:bk,0a5c:5834:gij
5977 [USBHID] The interval which mice are to be polled at.
5980 [USBHID] The interval which joysticks are to be polled at.
5983 [USBHID] The interval which keyboards are to be polled at.
5985 usb-storage.delay_use=
5986 [UMS] The delay in seconds before a new device is
5987 scanned for Logical Units (default 1).
5990 [UMS] A list of quirks entries to supplement or
5991 override the built-in unusual_devs list. List
5992 entries are separated by commas. Each entry has
5993 the form VID:PID:Flags where VID and PID are Vendor
5994 and Product ID values (4-digit hex numbers) and
5995 Flags is a set of characters, each corresponding
5996 to a common usb-storage quirk flag as follows:
5997 a = SANE_SENSE (collect more than 18 bytes
5998 of sense data, not on uas);
5999 b = BAD_SENSE (don't collect more than 18
6000 bytes of sense data, not on uas);
6001 c = FIX_CAPACITY (decrease the reported
6002 device capacity by one sector);
6003 d = NO_READ_DISC_INFO (don't use
6004 READ_DISC_INFO command, not on uas);
6005 e = NO_READ_CAPACITY_16 (don't use
6006 READ_CAPACITY_16 command);
6007 f = NO_REPORT_OPCODES (don't use report opcodes
6009 g = MAX_SECTORS_240 (don't transfer more than
6010 240 sectors at a time, uas only);
6011 h = CAPACITY_HEURISTICS (decrease the
6012 reported device capacity by one
6013 sector if the number is odd);
6014 i = IGNORE_DEVICE (don't bind to this
6016 j = NO_REPORT_LUNS (don't use report luns
6018 k = NO_SAME (do not use WRITE_SAME, uas only)
6019 l = NOT_LOCKABLE (don't try to lock and
6020 unlock ejectable media, not on uas);
6021 m = MAX_SECTORS_64 (don't transfer more
6022 than 64 sectors = 32 KB at a time,
6024 n = INITIAL_READ10 (force a retry of the
6025 initial READ(10) command, not on uas);
6026 o = CAPACITY_OK (accept the capacity
6027 reported by the device, not on uas);
6028 p = WRITE_CACHE (the device cache is ON
6029 by default, not on uas);
6030 r = IGNORE_RESIDUE (the device reports
6031 bogus residue values, not on uas);
6032 s = SINGLE_LUN (the device has only one
6034 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
6035 commands, uas only);
6036 u = IGNORE_UAS (don't bind to the uas driver);
6037 w = NO_WP_DETECT (don't test whether the
6038 medium is write-protected).
6039 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE
6040 even if the device claims no cache,
6042 Example: quirks=0419:aaf5:rl,0421:0433:rc
6044 user_debug= [KNL,ARM]
6046 See arch/arm/Kconfig.debug help text.
6047 1 - undefined instruction events
6049 4 - invalid data aborts
6052 Example: user_debug=31
6055 [X86] Flags controlling user PTE allocations.
6057 nohigh = do not allocate PTE pages in
6058 HIGHMEM regardless of setting
6062 On X86_32, this is an alias for vdso32=. Otherwise:
6064 vdso=1: enable VDSO (the default)
6065 vdso=0: disable VDSO mapping
6067 vdso32= [X86] Control the 32-bit vDSO
6068 vdso32=1: enable 32-bit VDSO
6069 vdso32=0 or vdso32=2: disable 32-bit VDSO
6071 See the help text for CONFIG_COMPAT_VDSO for more
6072 details. If CONFIG_COMPAT_VDSO is set, the default is
6073 vdso32=0; otherwise, the default is vdso32=1.
6075 For compatibility with older kernels, vdso32=2 is an
6078 Try vdso32=0 if you encounter an error that says:
6079 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
6082 vector=percpu: enable percpu vector domain
6084 video= [FB] Frame buffer configuration
6085 See Documentation/fb/modedb.rst.
6087 video.brightness_switch_enabled= [0,1]
6088 If set to 1, on receiving an ACPI notify event
6089 generated by hotkey, video driver will adjust brightness
6090 level and then send out the event to user space through
6091 the allocated input device; If set to 0, video driver
6092 will only send out the event without touching backlight
6097 [VMMIO] Memory mapped virtio (platform) device.
6099 <size>@<baseaddr>:<irq>[:<id>]
6101 <size> := size (can use standard suffixes
6103 <baseaddr> := physical base address
6104 <irq> := interrupt number (as passed to
6106 <id> := (optional) platform device id
6108 virtio_mmio.device=1K@0x100b0000:48:7
6110 Can be used multiple times for multiple devices.
6112 vga= [BOOT,X86-32] Select a particular video mode
6113 See Documentation/x86/boot.rst and
6114 Documentation/admin-guide/svga.rst.
6115 Use vga=ask for menu.
6116 This is actually a boot loader parameter; the value is
6117 passed to the kernel using a special protocol.
6119 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y.
6120 May slow down system boot speed, especially when
6121 enabled on systems with a large amount of memory.
6122 All options are enabled by default, and this
6123 interface is meant to allow for selectively
6124 enabling or disabling specific virtual memory
6127 Available options are:
6128 P Enable page structure init time poisoning
6129 - Disable all of the above options
6131 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
6132 size of <nn>. This can be used to increase the
6133 minimum size (128MB on x86). It can also be used to
6134 decrease the size and leave more room for directly
6137 vmcp_cma=nn[MG] [KNL,S390]
6138 Sets the memory size reserved for contiguous memory
6139 allocations for the vmcp device driver.
6141 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
6144 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
6147 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
6151 Controls the behavior of vsyscalls (i.e. calls to
6152 fixed addresses of 0xffffffffff600x00 from legacy
6153 code). Most statically-linked binaries and older
6154 versions of glibc use these calls. Because these
6155 functions are at fixed addresses, they make nice
6156 targets for exploits that can control RIP.
6158 emulate [default] Vsyscalls turn into traps and are
6159 emulated reasonably safely. The vsyscall
6162 xonly Vsyscalls turn into traps and are
6163 emulated reasonably safely. The vsyscall
6164 page is not readable.
6166 none Vsyscalls don't work at all. This makes
6167 them quite hard to use for exploits but
6168 might break your system.
6170 vt.color= [VT] Default text color.
6171 Format: 0xYX, X = foreground, Y = background.
6172 Default: 0x07 = light gray on black.
6174 vt.cur_default= [VT] Default cursor shape.
6175 Format: 0xCCBBAA, where AA, BB, and CC are the same as
6176 the parameters of the <Esc>[?A;B;Cc escape sequence;
6177 see VGA-softcursor.txt. Default: 2 = underline.
6179 vt.default_blu= [VT]
6180 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
6181 Change the default blue palette of the console.
6182 This is a 16-member array composed of values
6185 vt.default_grn= [VT]
6186 Format: <green0>,<green1>,<green2>,...,<green15>
6187 Change the default green palette of the console.
6188 This is a 16-member array composed of values
6191 vt.default_red= [VT]
6192 Format: <red0>,<red1>,<red2>,...,<red15>
6193 Change the default red palette of the console.
6194 This is a 16-member array composed of values
6200 Set system-wide default UTF-8 mode for all tty's.
6201 Default is 1, i.e. UTF-8 mode is enabled for all
6202 newly opened terminals.
6204 vt.global_cursor_default=
6207 Set system-wide default for whether a cursor
6208 is shown on new VTs. Default is -1,
6209 i.e. cursors will be created by default unless
6210 overridden by individual drivers. 0 will hide
6211 cursors, 1 will display them.
6213 vt.italic= [VT] Default color for italic text; 0-15.
6216 vt.underline= [VT] Default color for underlined text; 0-15.
6219 watchdog timers [HW,WDT] For information on watchdog timers,
6220 see Documentation/watchdog/watchdog-parameters.rst
6221 or other driver-specific files in the
6222 Documentation/watchdog/ directory.
6226 Set the hard lockup detector stall duration
6227 threshold in seconds. The soft lockup detector
6228 threshold is set to twice the value. A value of 0
6229 disables both lockup detectors. Default is 10
6232 workqueue.watchdog_thresh=
6233 If CONFIG_WQ_WATCHDOG is configured, workqueue can
6234 warn stall conditions and dump internal state to
6235 help debugging. 0 disables workqueue stall
6236 detection; otherwise, it's the stall threshold
6237 duration in seconds. The default value is 30 and
6238 it can be updated at runtime by writing to the
6239 corresponding sysfs file.
6241 workqueue.disable_numa
6242 By default, all work items queued to unbound
6243 workqueues are affine to the NUMA nodes they're
6244 issued on, which results in better behavior in
6245 general. If NUMA affinity needs to be disabled for
6246 whatever reason, this option can be used. Note
6247 that this also can be controlled per-workqueue for
6248 workqueues visible under /sys/bus/workqueue/.
6250 workqueue.power_efficient
6251 Per-cpu workqueues are generally preferred because
6252 they show better performance thanks to cache
6253 locality; unfortunately, per-cpu workqueues tend to
6254 be more power hungry than unbound workqueues.
6256 Enabling this makes the per-cpu workqueues which
6257 were observed to contribute significantly to power
6258 consumption unbound, leading to measurably lower
6259 power usage at the cost of small performance
6262 The default value of this parameter is determined by
6263 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
6265 workqueue.debug_force_rr_cpu
6266 Workqueue used to implicitly guarantee that work
6267 items queued without explicit CPU specified are put
6268 on the local CPU. This guarantee is no longer true
6269 and while local CPU is still preferred work items
6270 may be put on foreign CPUs. This debug option
6271 forces round-robin CPU selection to flush out
6272 usages which depend on the now broken guarantee.
6273 When enabled, memory and cache locality will be
6276 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
6277 default x2apic cluster mode on platforms
6280 xen_512gb_limit [KNL,X86-64,XEN]
6281 Restricts the kernel running paravirtualized under Xen
6282 to use only up to 512 GB of RAM. The reason to do so is
6283 crash analysis tools and Xen tools for doing domain
6284 save/restore/migration must be enabled to handle larger
6287 xen_emul_unplug= [HW,X86,XEN]
6288 Unplug Xen emulated devices
6289 Format: [unplug0,][unplug1]
6290 ide-disks -- unplug primary master IDE devices
6291 aux-ide-disks -- unplug non-primary-master IDE devices
6292 nics -- unplug network devices
6293 all -- unplug all emulated devices (NICs and IDE disks)
6294 unnecessary -- unplugging emulated devices is
6295 unnecessary even if the host did not respond to
6297 never -- do not unplug even if version check succeeds
6299 xen_legacy_crash [X86,XEN]
6300 Crash from Xen panic notifier, without executing late
6301 panic() code such as dumping handler.
6303 xen_nopvspin [X86,XEN]
6304 Disables the qspinlock slowpath using Xen PV optimizations.
6305 This parameter is obsoleted by "nopvspin" parameter, which
6306 has equivalent effect for XEN platform.
6309 Disables the PV optimizations forcing the HVM guest to
6310 run as generic HVM guest with no PV drivers.
6311 This option is obsoleted by the "nopv" option, which
6312 has equivalent effect for XEN platform.
6314 xen_no_vector_callback
6315 [KNL,X86,XEN] Disable the vector callback for Xen
6316 event channel interrupts.
6318 xen_scrub_pages= [XEN]
6319 Boolean option to control scrubbing pages before giving them back
6320 to Xen, for use by other domains. Can be also changed at runtime
6321 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages.
6322 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT.
6324 xen_timer_slop= [X86-64,XEN]
6325 Set the timer slop (in nanoseconds) for the virtual Xen
6326 timers (default is 100000). This adjusts the minimum
6327 delta of virtualized Xen timers, where lower values
6328 improve timer resolution at the expense of processing
6329 more timer interrupts.
6331 xen.event_eoi_delay= [XEN]
6332 How long to delay EOI handling in case of event
6333 storms (jiffies). Default is 10.
6335 xen.event_loop_timeout= [XEN]
6336 After which time (jiffies) the event handling loop
6337 should start to delay EOI handling. Default is 2.
6339 xen.fifo_events= [XEN]
6340 Boolean parameter to disable using fifo event handling
6341 even if available. Normally fifo event handling is
6342 preferred over the 2-level event handling, as it is
6343 fairer and the number of possible event channels is
6344 much higher. Default is on (use fifo events).
6346 nopv= [X86,XEN,KVM,HYPER_V,VMWARE]
6347 Disables the PV optimizations forcing the guest to run
6348 as generic guest with no PV drivers. Currently support
6349 XEN HVM, KVM, HYPER_V and VMWARE guest.
6351 nopvspin [X86,XEN,KVM]
6352 Disables the qspinlock slow path using PV optimizations
6353 which allow the hypervisor to 'idle' the guest on lock
6356 xirc2ps_cs= [NET,PCMCIA]
6358 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
6361 By default on POWER9 and above, the kernel will
6362 natively use the XIVE interrupt controller. This option
6363 allows the fallback firmware mode to be used:
6365 off Fallback to firmware control of XIVE interrupt
6366 controller on both pseries and powernv
6367 platforms. Only useful on POWER9 and above.
6369 xhci-hcd.quirks [USB,KNL]
6370 A hex value specifying bitmask with supplemental xhci
6371 host controller quirks. Meaning of each bit can be
6372 consulted in header drivers/usb/host/xhci.h.
6375 Format: { early | on | rw | ro | off }
6376 Controls if xmon debugger is enabled. Default is off.
6377 Passing only "xmon" is equivalent to "xmon=early".
6378 early Call xmon as early as possible on boot; xmon
6379 debugger is called from setup_arch().
6380 on xmon debugger hooks will be installed so xmon
6381 is only called on a kernel crash. Default mode,
6382 i.e. either "ro" or "rw" mode, is controlled
6383 with CONFIG_XMON_DEFAULT_RO_MODE.
6384 rw xmon debugger hooks will be installed so xmon
6385 is called only on a kernel crash, mode is write,
6386 meaning SPR registers, memory and, other data
6387 can be written using xmon commands.
6388 ro same as "rw" option above but SPR registers,
6389 memory, and other data can't be written using
6391 off xmon is disabled.