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_hwsig,
229 s4_nohwsig, old_ordering, nonvs,
230 sci_force_enable, nobl }
231 See Documentation/power/video.rst for information on
233 s3_beep is for debugging; it makes the PC's speaker beep
234 as soon as the kernel's real-mode entry point is called.
235 s4_hwsig causes the kernel to check the ACPI hardware
236 signature during resume from hibernation, and gracefully
237 refuse to resume if it has changed. This complies with
238 the ACPI specification but not with reality, since
239 Windows does not do this and many laptops do change it
240 on docking. So the default behaviour is to allow resume
241 and simply warn when the signature changes, unless the
242 s4_hwsig option is enabled.
243 s4_nohwsig prevents ACPI hardware signature from being
244 used (or even warned about) during resume.
245 old_ordering causes the ACPI 1.0 ordering of the _PTS
246 control method, with respect to putting devices into
247 low power states, to be enforced (the ACPI 2.0 ordering
248 of _PTS is used by default).
249 nonvs prevents the kernel from saving/restoring the
250 ACPI NVS memory during suspend/hibernation and resume.
251 sci_force_enable causes the kernel to set SCI_EN directly
252 on resume from S1/S3 (which is against the ACPI spec,
253 but some broken systems don't work without it).
254 nobl causes the internal blacklist of systems known to
255 behave incorrectly in some ways with respect to system
256 suspend and resume to be ignored (use wisely).
258 acpi_use_timer_override [HW,ACPI]
259 Use timer override. For some broken Nvidia NF5 boards
260 that require a timer override, but don't have HPET
262 add_efi_memmap [EFI; X86] Include EFI memory map in
263 kernel's map of available physical RAM.
266 { off | try_unsupported }
267 off: disable AGP support
268 try_unsupported: try to drive unsupported chipsets
269 (may crash computer or cause data corruption)
272 See Documentation/sound/alsa-configuration.rst
275 Allow the default userspace alignment fault handler
276 behaviour to be specified. Bit 0 enables warnings,
277 bit 1 enables fixups, and bit 2 sends a segfault.
279 align_va_addr= [X86-64]
280 Align virtual addresses by clearing slice [14:12] when
281 allocating a VMA at process creation time. This option
282 gives you up to 3% performance improvement on AMD F15h
283 machines (where it is enabled by default) for a
284 CPU-intensive style benchmark, and it can vary highly in
285 a microbenchmark depending on workload and compiler.
287 32: only for 32-bit processes
288 64: only for 64-bit processes
289 on: enable for both 32- and 64-bit processes
290 off: disable for both 32- and 64-bit processes
292 alloc_snapshot [FTRACE]
293 Allocate the ftrace snapshot buffer on boot up when the
294 main buffer is allocated. This is handy if debugging
295 and you need to use tracing_snapshot() on boot up, and
296 do not want to use tracing_snapshot_alloc() as it needs
297 to be done where GFP_KERNEL allocations are allowed.
299 allow_mismatched_32bit_el0 [ARM64]
300 Allow execve() of 32-bit applications and setting of the
301 PER_LINUX32 personality on systems where only a strict
302 subset of the CPUs support 32-bit EL0. When this
303 parameter is present, the set of CPUs supporting 32-bit
304 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0
305 and hot-unplug operations may be restricted.
307 See Documentation/arm64/asymmetric-32bit.rst for more
310 amd_iommu= [HW,X86-64]
311 Pass parameters to the AMD IOMMU driver in the system.
313 fullflush - Deprecated, equivalent to iommu.strict=1
314 off - do not initialize any AMD IOMMU found in
316 force_isolation - Force device isolation for all
317 devices. The IOMMU driver is not
318 allowed anymore to lift isolation
319 requirements as needed. This option
320 does not override iommu=pt
321 force_enable - Force enable the IOMMU on platforms known
322 to be buggy with IOMMU enabled. Use this
324 pgtbl_v1 - Use v1 page table for DMA-API (Default).
325 pgtbl_v2 - Use v2 page table for DMA-API.
327 amd_iommu_dump= [HW,X86-64]
328 Enable AMD IOMMU driver option to dump the ACPI table
329 for AMD IOMMU. With this option enabled, AMD IOMMU
330 driver will print ACPI tables for AMD IOMMU during
331 IOMMU initialization.
333 amd_iommu_intr= [HW,X86-64]
334 Specifies one of the following AMD IOMMU interrupt
336 legacy - Use legacy interrupt remapping mode.
337 vapic - Use virtual APIC mode, which allows IOMMU
338 to inject interrupts directly into guest.
339 This mode requires kvm-amd.avic=1.
340 (Default when IOMMU HW support is present.)
342 amijoy.map= [HW,JOY] Amiga joystick support
343 Map of devices attached to JOY0DAT and JOY1DAT
345 See also Documentation/input/joydev/joystick.rst
347 analog.map= [HW,JOY] Analog joystick and gamepad support
348 Specifies type or capabilities of an analog joystick
349 connected to one of 16 gameports
350 Format: <type1>,<type2>,..<type16>
353 Power management functions (SPARCstation-4/5 + deriv.)
355 Disable APC CPU standby support. SPARCstation-Fox does
356 not play well with APC CPU idle - disable it if you have
357 APC and your system crashes randomly.
359 apic= [APIC,X86] Advanced Programmable Interrupt Controller
360 Change the output verbosity while booting
361 Format: { quiet (default) | verbose | debug }
362 Change the amount of debugging information output
363 when initialising the APIC and IO-APIC components.
364 For X86-32, this can also be used to specify an APIC
366 Format: apic=driver_name
367 Examples: apic=bigsmp
369 apic_extnmi= [APIC,X86] External NMI delivery setting
370 Format: { bsp (default) | all | none }
371 bsp: External NMI is delivered only to CPU 0
372 all: External NMIs are broadcast to all CPUs as a
374 none: External NMI is masked for all CPUs. This is
375 useful so that a dump capture kernel won't be
379 See Documentation/networking/ipv6.rst.
381 apm= [APM] Advanced Power Management
382 See header of arch/x86/kernel/apm_32.c.
384 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
385 Format: { "0" | "1" }
386 See security/apparmor/Kconfig help text
389 Default value is set via kernel config option.
391 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
392 Format: <io>,<irq>,<nodeID>
394 arm64.nobti [ARM64] Unconditionally disable Branch Target
395 Identification support
397 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication
400 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension
403 arm64.nosve [ARM64] Unconditionally disable Scalable Vector
406 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix
411 atarimouse= [HW,MOUSE] Atari Mouse
413 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
414 EzKey and similar keyboards
416 atkbd.reset= [HW] Reset keyboard during initialization
418 atkbd.set= [HW] Select keyboard code set
419 Format: <int> (2 = AT (default), 3 = PS/2)
421 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
424 atkbd.softraw= [HW] Choose between synthetic and real raw mode
425 Format: <bool> (0 = real, 1 = synthetic (default))
427 atkbd.softrepeat= [HW]
428 Use software keyboard repeat
430 audit= [KNL] Enable the audit sub-system
431 Format: { "0" | "1" | "off" | "on" }
432 0 | off - kernel audit is disabled and can not be
433 enabled until the next reboot
434 unset - kernel audit is initialized but disabled and
435 will be fully enabled by the userspace auditd.
436 1 | on - kernel audit is initialized and partially
437 enabled, storing at most audit_backlog_limit
438 messages in RAM until it is fully enabled by the
442 audit_backlog_limit= [KNL] Set the audit queue size limit.
443 Format: <int> (must be >=0)
446 bau= [X86_UV] Enable the BAU on SGI UV. The default
447 behavior is to disable the BAU (i.e. bau=0).
448 Format: { "0" | "1" }
451 unset - Disable the BAU.
453 baycom_epp= [HW,AX25]
456 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
458 See header of drivers/net/hamradio/baycom_par.c.
460 baycom_ser_fdx= [HW,AX25]
461 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
462 Format: <io>,<irq>,<mode>[,<baud>]
463 See header of drivers/net/hamradio/baycom_ser_fdx.c.
465 baycom_ser_hdx= [HW,AX25]
466 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
467 Format: <io>,<irq>,<mode>
468 See header of drivers/net/hamradio/baycom_ser_hdx.c.
471 Disable BERT OS support on buggy BIOSes.
473 bgrt_disable [ACPI][X86]
474 Disable BGRT to avoid flickering OEM logo.
476 blkdevparts= Manual partition parsing of block device(s) for
477 embedded devices based on command line input.
478 See Documentation/block/cmdline-partition.rst
480 boot_delay= Milliseconds to delay each printk during boot.
481 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled,
482 and you may also have to specify "lpj=". Boot_delay
483 values larger than 10 seconds (10000) are assumed
484 erroneous and ignored.
488 Extended command line options can be added to an initrd
489 and this will cause the kernel to look for it.
491 See Documentation/admin-guide/bootconfig.rst
493 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
494 bttv.radio= Most important insmod options are available as
496 bttv.pll= See Documentation/admin-guide/media/bttv.rst
499 bulk_remove=off [PPC] This parameter disables the use of the pSeries
500 firmware feature for flushing multiple hpte entries
503 c101= [NET] Moxa C101 synchronous serial card
505 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
506 Sometimes CPU hardware bugs make them report the cache
507 size incorrectly. The kernel will attempt work arounds
508 to fix known problems, but for some CPUs it is not
509 possible to determine what the correct size should be.
510 This option provides an override for these situations.
513 [NET] Specifies amount of time (in seconds) that
514 the kernel should wait for a network carrier. By default
515 it waits 120 seconds.
517 ca_keys= [KEYS] This parameter identifies a specific key(s) on
518 the system trusted keyring to be used for certificate
520 format: { id:<keyid> | builtin }
522 cca= [MIPS] Override the kernel pages' cache coherency
523 algorithm. Accepted values range from 0 to 7
524 inclusive. See arch/mips/include/asm/pgtable-bits.h
525 for platform specific values (SB1, Loongson3 and
528 ccw_timeout_log [S390]
529 See Documentation/s390/common_io.rst for details.
531 cgroup_disable= [KNL] Disable a particular controller or optional feature
532 Format: {name of the controller(s) or feature(s) to disable}
533 The effects of cgroup_disable=foo are:
534 - foo isn't auto-mounted if you mount all cgroups in
536 - foo isn't visible as an individually mountable
538 - if foo is an optional feature then the feature is
539 disabled and corresponding cgroup files are not
541 {Currently only "memory" controller deal with this and
542 cut the overhead, others just disable the usage. So
543 only cgroup_disable=memory is actually worthy}
544 Specifying "pressure" disables per-cgroup pressure
545 stall information accounting feature
547 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1
548 Format: { { controller | "all" | "named" }
549 [,{ controller | "all" | "named" }...] }
550 Like cgroup_disable, but only applies to cgroup v1;
551 the blacklisted controllers remain available in cgroup2.
552 "all" blacklists all controllers and "named" disables
553 named mounts. Specifying both "all" and "named" disables
556 cgroup.memory= [KNL] Pass options to the cgroup memory controller.
558 nosocket -- Disable socket memory accounting.
559 nokmem -- Disable kernel memory accounting.
560 nobpf -- Disable BPF memory accounting.
562 checkreqprot= [SELINUX] Set initial checkreqprot flag value.
563 Format: { "0" | "1" }
564 See security/selinux/Kconfig help text.
565 0 -- check protection applied by kernel (includes
566 any implied execute protection).
567 1 -- check protection requested by application.
568 Default value is set via a kernel config option.
569 Value can be changed at runtime via
570 /sys/fs/selinux/checkreqprot.
571 Setting checkreqprot to 1 is deprecated.
574 See Documentation/s390/common_io.rst for details.
576 clearcpuid=X[,X...] [X86]
577 Disable CPUID feature X for the kernel. See
578 arch/x86/include/asm/cpufeatures.h for the valid bit
579 numbers X. Note the Linux-specific bits are not necessarily
580 stable over kernel options, but the vendor-specific
582 X can also be a string as appearing in the flags: line
583 in /proc/cpuinfo which does not have the above
584 instability issue. However, not all features have names
586 Note that using this option will taint your kernel.
587 Also note that user programs calling CPUID directly
588 or using the feature without checking anything
589 will still see it. This just prevents it from
590 being used by the kernel or shown in /proc/cpuinfo.
591 Also note the kernel might malfunction if you disable
596 Prevents the clock framework from automatically gating
597 clocks that have not been explicitly enabled by a Linux
598 device driver but are enabled in hardware at reset or
599 by the bootloader/firmware. Note that this does not
600 force such clocks to be always-on nor does it reserve
601 those clocks in any way. This parameter is useful for
602 debug and development, but should not be needed on a
603 platform with proper driver support. For more
604 information, see Documentation/driver-api/clk.rst.
606 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
608 Forces specified clocksource (if available) to be used
609 when calculating gettimeofday(). If specified
610 clocksource is not available, it defaults to PIT.
611 Format: { pit | tsc | cyclone | pmtmr }
613 clocksource= Override the default clocksource
615 Override the default clocksource and use the clocksource
616 with the name specified.
617 Some clocksource names to choose from, depending on
619 [all] jiffies (this is the base, fallback clocksource)
621 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
622 pxa_timer,timer3,32k_counter,timer0_1
623 [X86-32] pit,hpet,tsc;
624 scx200_hrt on Geode; cyclone on IBM x440
632 clocksource.arm_arch_timer.evtstrm=
635 Enable/disable the eventstream feature of the ARM
636 architected timer so that code using WFE-based polling
637 loops can be debugged more effectively on production
640 clocksource.max_cswd_read_retries= [KNL]
641 Number of clocksource_watchdog() retries due to
642 external delays before the clock will be marked
643 unstable. Defaults to two retries, that is,
644 three attempts to read the clock under test.
646 clocksource.verify_n_cpus= [KNL]
647 Limit the number of CPUs checked for clocksources
648 marked with CLOCK_SOURCE_VERIFY_PERCPU that
649 are marked unstable due to excessive skew.
650 A negative value says to check all CPUs, while
651 zero says not to check any. Values larger than
652 nr_cpu_ids are silently truncated to nr_cpu_ids.
653 The actual CPUs are chosen randomly, with
654 no replacement if the same CPU is chosen twice.
656 clocksource-wdtest.holdoff= [KNL]
657 Set the time in seconds that the clocksource
658 watchdog test waits before commencing its tests.
659 Defaults to zero when built as a module and to
660 10 seconds when built into the kernel.
662 cma=nn[MG]@[start[MG][-end[MG]]]
664 Sets the size of kernel global memory area for
665 contiguous memory allocations and optionally the
666 placement constraint by the physical address range of
667 memory allocations. A value of 0 disables CMA
668 altogether. For more information, see
669 kernel/dma/contiguous.c
673 Sets the size of kernel per-numa memory area for
674 contiguous memory allocations. A value of 0 disables
675 per-numa CMA altogether. And If this option is not
676 specified, the default value is 0.
677 With per-numa CMA enabled, DMA users on node nid will
678 first try to allocate buffer from the pernuma area
679 which is located in node nid, if the allocation fails,
680 they will fallback to the global default memory area.
682 cmo_free_hint= [PPC] Format: { yes | no }
683 Specify whether pages are marked as being inactive
684 when they are freed. This is used in CMO environments
685 to determine OS memory pressure for page stealing by
689 coherent_pool=nn[KMG] [ARM,KNL]
690 Sets the size of memory pool for coherent, atomic dma
691 allocations, by default set to 256K.
693 com20020= [HW,NET] ARCnet - COM20020 chipset
695 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
697 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
701 ARCnet - COM90xx chipset (memory-mapped buffers)
702 Format: <io>[,<irq>[,<memstart>]]
704 condev= [HW,S390] console device
707 con3215_drop= [S390] 3215 console drop mode.
709 When set to true, drop data on the 3215 console when
710 the console buffer is full. In this case the
711 operator using a 3270 terminal emulator (for example
712 x3270) does not have to enter the clear key for the
713 console output to advance and the kernel to continue.
714 This leads to a much faster boot time when a 3270
715 terminal emulator is active. If no 3270 terminal
716 emulator is used, this parameter has no effect.
718 console= [KNL] Output console device and options.
720 tty<n> Use the virtual console device <n>.
724 Use the specified serial port. The options are of
725 the form "bbbbpnf", where "bbbb" is the baud rate,
726 "p" is parity ("n", "o", or "e"), "n" is number of
727 bits, and "f" is flow control ("r" for RTS or
728 omit it). Default is "9600n8".
730 See Documentation/admin-guide/serial-console.rst for more
732 Documentation/networking/netconsole.rst for an
735 uart[8250],io,<addr>[,options]
736 uart[8250],mmio,<addr>[,options]
737 uart[8250],mmio16,<addr>[,options]
738 uart[8250],mmio32,<addr>[,options]
739 uart[8250],0x<addr>[,options]
740 Start an early, polled-mode console on the 8250/16550
741 UART at the specified I/O port or MMIO address,
742 switching to the matching ttyS device later.
743 MMIO inter-register address stride is either 8-bit
744 (mmio), 16-bit (mmio16), or 32-bit (mmio32).
745 If none of [io|mmio|mmio16|mmio32], <addr> is assumed
746 to be equivalent to 'mmio'. 'options' are specified in
747 the same format described for ttyS above; if unspecified,
748 the h/w is not re-initialized.
750 hvc<n> Use the hypervisor console device <n>. This is for
751 both Xen and PowerPC hypervisors.
754 Use to disable console output, i.e., to have kernel
755 console messages discarded.
756 This must be the only console= parameter used on the
759 If the device connected to the port is not a TTY but a braille
760 device, prepend "brl," before the device type, for instance
762 For now, only VisioBraille is supported.
765 [KNL] Change console messages format
767 By default we print messages on consoles in
768 "[time stamp] text\n" format (time stamp may not be
769 printed, depending on CONFIG_PRINTK_TIME or
770 `printk_time' param).
772 Switch to syslog format: "<%u>[time stamp] text\n"
773 IOW, each message will have a facility and loglevel
774 prefix. The format is similar to one used by syslog()
775 syscall, or to executing "dmesg -S --raw" or to reading
778 consoleblank= [KNL] The console blank (screen saver) timeout in
779 seconds. A value of 0 disables the blank timer.
783 [KNL] Change the default value for
784 /proc/<pid>/coredump_filter.
785 See also Documentation/filesystems/proc.rst.
787 coresight_cpu_debug.enable
790 Enable/disable the CPU sampling based debugging.
791 0: default value, disable debugging
792 1: enable debugging at boot time
794 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
796 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
798 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
799 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
800 Some features depend on CPU0. Known dependencies are:
801 1. Resume from suspend/hibernate depends on CPU0.
802 Suspend/hibernate will fail if CPU0 is offline and you
803 need to online CPU0 before suspend/hibernate.
804 2. PIC interrupts also depend on CPU0. CPU0 can't be
805 removed if a PIC interrupt is detected.
806 It's said poweroff/reboot may depend on CPU0 on some
807 machines although I haven't seen such issues so far
808 after CPU0 is offline on a few tested machines.
809 If the dependencies are under your control, you can
810 turn on cpu0_hotplug.
812 cpuidle.off=1 [CPU_IDLE]
813 disable the cpuidle sub-system
816 [CPU_IDLE] Name of the cpuidle governor to use.
818 cpufreq.off=1 [CPU_FREQ]
819 disable the cpufreq sub-system
821 cpufreq.default_governor=
822 [CPU_FREQ] Name of the default cpufreq governor or
823 policy to use. This governor must be registered in the
824 kernel before the cpufreq driver probes.
827 [X86] Delay for N microsec between assert and de-assert
828 of APIC INIT to start processors. This delay occurs
829 on every CPU online, such as boot, and resume from suspend.
832 crash_kexec_post_notifiers
833 Run kdump after running panic-notifiers and dumping
834 kmsg. This only for the users who doubt kdump always
835 succeeds in any situation.
836 Note that this also increases risks of kdump failure,
837 because some panic notifiers can make the crashed
838 kernel more unstable.
840 crashkernel=size[KMG][@offset[KMG]]
841 [KNL] Using kexec, Linux can switch to a 'crash kernel'
842 upon panic. This parameter reserves the physical
843 memory region [offset, offset + size] for that kernel
844 image. If '@offset' is omitted, then a suitable offset
845 is selected automatically.
846 [KNL, X86-64, ARM64] Select a region under 4G first, and
847 fall back to reserve region above 4G when '@offset'
848 hasn't been specified.
849 See Documentation/admin-guide/kdump/kdump.rst for further details.
851 crashkernel=range1:size1[,range2:size2,...][@offset]
852 [KNL] Same as above, but depends on the memory
853 in the running system. The syntax of range is
854 start-[end] where start and end are both
855 a memory unit (amount[KMG]). See also
856 Documentation/admin-guide/kdump/kdump.rst for an example.
858 crashkernel=size[KMG],high
859 [KNL, X86-64, ARM64] range could be above 4G. Allow kernel
860 to allocate physical memory region from top, so could
861 be above 4G if system have more than 4G ram installed.
862 Otherwise memory region will be allocated below 4G, if
864 It will be ignored if crashkernel=X is specified.
865 crashkernel=size[KMG],low
866 [KNL, X86-64, ARM64] range under 4G. When crashkernel=X,high
867 is passed, kernel could allocate physical memory region
868 above 4G, that cause second kernel crash on system
869 that require some amount of low memory, e.g. swiotlb
870 requires at least 64M+32K low memory, also enough extra
871 low memory is needed to make sure DMA buffers for 32-bit
872 devices won't run out. Kernel would try to allocate
873 default size of memory below 4G automatically. The default
874 size is platform dependent.
875 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB)
877 This one lets the user specify own low range under 4G
878 for second kernel instead.
879 0: to disable low allocation.
880 It will be ignored when crashkernel=X,high is not used
881 or memory reserved is below 4G.
884 [KNL] Disable crypto self-tests
889 cs89x0_media= [HW,NET]
890 Format: { rj45 | aui | bnc }
892 csdlock_debug= [KNL] Enable debug add-ons of cross-CPU function call
893 handling. When switched on, additional debug data is
894 printed to the console in case a hanging CPU is
895 detected, and that CPU is pinged again in order to try
896 to resolve the hang situation.
897 0: disable csdlock debugging (default)
898 1: enable basic csdlock debugging (minor impact)
899 ext: enable extended csdlock debugging (more impact,
903 See header of drivers/s390/block/dasd_devmap.c.
905 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
906 (one device per port)
907 Format: <port#>,<type>
908 See also Documentation/input/devices/joystick-parport.rst
910 debug [KNL] Enable kernel debugging (events log level).
913 [KNL] Enable printing [hashed] pointers early in the
914 boot sequence. If enabled, we use a weak hash instead
915 of siphash to hash pointers. Use this option if you are
916 seeing instances of '(___ptrval___)') and need to see a
917 value (hashed pointer) instead. Cryptographically
918 insecure, please do not use on production kernels.
921 [KNL] verbose locking self-tests
923 Print debugging info while doing the locking API
925 Bitmask for the various LOCKTYPE_ tests. Defaults to 0
926 (no extra messages), setting it to -1 (all bits set)
927 will print _a_lot_ more information - normally only
928 useful to lockdep developers.
930 debug_objects [KNL] Enable object debugging
933 [KNL] Disable object debugging
935 debug_guardpage_minorder=
936 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
937 parameter allows control of the order of pages that will
938 be intentionally kept free (and hence protected) by the
939 buddy allocator. Bigger value increase the probability
940 of catching random memory corruption, but reduce the
941 amount of memory for normal system use. The maximum
942 possible value is MAX_ORDER/2. Setting this parameter
943 to 1 or 2 should be enough to identify most random
944 memory corruption problems caused by bugs in kernel or
945 driver code when a CPU writes to (or reads from) a
946 random memory location. Note that there exists a class
947 of memory corruptions problems caused by buggy H/W or
948 F/W or by drivers badly programming DMA (basically when
949 memory is written at bus level and the CPU MMU is
950 bypassed) which are not detectable by
951 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
952 tracking down these problems.
955 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter
956 enables the feature at boot time. By default, it is
957 disabled and the system will work mostly the same as a
958 kernel built without CONFIG_DEBUG_PAGEALLOC.
959 Note: to get most of debug_pagealloc error reports, it's
960 useful to also enable the page_owner functionality.
961 on: enable the feature
963 debugfs= [KNL] This parameter enables what is exposed to userspace
964 and debugfs internal clients.
965 Format: { on, no-mount, off }
966 on: All functions are enabled.
968 Filesystem is not registered but kernel clients can
969 access APIs and a crashkernel can be used to read
970 its content. There is nothing to mount.
971 off: Filesystem is not registered and clients
972 get a -EPERM as result when trying to register files
973 or directories within debugfs.
974 This is equivalent of the runtime functionality if
975 debugfs was not enabled in the kernel at all.
976 Default value is set in build-time with a kernel configuration.
978 debugpat [X86] Enable PAT debugging
981 [HW] The size of the default HugeTLB page. This is
982 the size represented by the legacy /proc/ hugepages
983 APIs. In addition, this is the default hugetlb size
984 used for shmget(), mmap() and mounting hugetlbfs
985 filesystems. If not specified, defaults to the
986 architecture's default huge page size. Huge page
987 sizes are architecture dependent. See also
988 Documentation/admin-guide/mm/hugetlbpage.rst.
991 deferred_probe_timeout=
992 [KNL] Debugging option to set a timeout in seconds for
993 deferred probe to give up waiting on dependencies to
994 probe. Only specific dependencies (subsystems or
995 drivers) that have opted in will be ignored. A timeout
996 of 0 will timeout at the end of initcalls. If the time
997 out hasn't expired, it'll be restarted by each
998 successful driver registration. This option will also
999 dump out devices still on the deferred probe list after
1002 delayacct [KNL] Enable per-task delay accounting
1004 dell_smm_hwmon.ignore_dmi=
1005 [HW] Continue probing hardware even if DMI data
1006 indicates that the driver is running on unsupported
1009 dell_smm_hwmon.force=
1010 [HW] Activate driver even if SMM BIOS signature does
1011 not match list of supported models and enable otherwise
1012 blacklisted features.
1014 dell_smm_hwmon.power_status=
1015 [HW] Report power status in /proc/i8k
1016 (disabled by default).
1018 dell_smm_hwmon.restricted=
1019 [HW] Allow controlling fans only if SYS_ADMIN
1022 dell_smm_hwmon.fan_mult=
1023 [HW] Factor to multiply fan speed with.
1025 dell_smm_hwmon.fan_max=
1026 [HW] Maximum configurable fan speed.
1029 Format: { on | off | def_only | inf_only | always }
1030 on: s390 zlib hardware support for compression on
1031 level 1 and decompression (default)
1032 off: No s390 zlib hardware support
1033 def_only: s390 zlib hardware support for deflate
1034 only (compression on level 1)
1035 inf_only: s390 zlib hardware support for inflate
1036 only (decompression)
1037 always: Same as 'on' but ignores the selected compression
1038 level always using hardware support (used for debugging)
1040 dhash_entries= [KNL]
1041 Set number of hash buckets for dentry cache.
1043 disable_1tb_segments [PPC]
1044 Disables the use of 1TB hash page table segments. This
1045 causes the kernel to fall back to 256MB segments which
1046 can be useful when debugging issues that require an SLB
1050 See Documentation/networking/ipv6.rst.
1053 Disable RADIX MMU mode on POWER9
1056 Disable TLBIE instruction. Currently does not work
1057 with KVM, with HASH MMU, or with coherent accelerators.
1059 disable_cpu_apicid= [X86,APIC,SMP]
1061 The number of initial APIC ID for the
1062 corresponding CPU to be disabled at boot,
1063 mostly used for the kdump 2nd kernel to
1064 disable BSP to wake up multiple CPUs without
1065 causing system reset or hang due to sending
1066 INIT from AP to BSP.
1068 disable_ddw [PPC/PSERIES]
1069 Disable Dynamic DMA Window support. Use this
1070 to workaround buggy firmware.
1072 disable_ipv6= [IPV6]
1073 See Documentation/networking/ipv6.rst.
1075 disable_mtrr_cleanup [X86]
1076 The kernel tries to adjust MTRR layout from continuous
1077 to discrete, to make X server driver able to add WB
1078 entry later. This parameter disables that.
1080 disable_mtrr_trim [X86, Intel and AMD only]
1081 By default the kernel will trim any uncacheable
1082 memory out of your available memory pool based on
1083 MTRR settings. This parameter disables that behavior,
1084 possibly causing your machine to run very slowly.
1086 disable_timer_pin_1 [X86]
1087 Disable PIN 1 of APIC timer
1088 Can be useful to work around chipset bugs.
1090 dis_ucode_ldr [X86] Disable the microcode loader.
1092 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
1093 this option disables the debugging code at boot.
1095 dma_debug_entries=<number>
1096 This option allows to tune the number of preallocated
1097 entries for DMA-API debugging code. One entry is
1098 required per DMA-API allocation. Use this if the
1099 DMA-API debugging code disables itself because the
1100 architectural default is too low.
1102 dma_debug_driver=<driver_name>
1103 With this option the DMA-API debugging driver
1104 filter feature can be enabled at boot time. Just
1105 pass the driver to filter for as the parameter.
1106 The filter can be disabled or changed to another
1107 driver later using sysfs.
1109 driver_async_probe= [KNL]
1110 List of driver names to be probed asynchronously. *
1111 matches with all driver names. If * is specified, the
1112 rest of the listed driver names are those that will NOT
1114 Format: <driver_name1>,<driver_name2>...
1116 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
1117 Broken monitors, graphic adapters, KVMs and EDIDless
1118 panels may send no or incorrect EDID data sets.
1119 This parameter allows to specify an EDID data sets
1120 in the /lib/firmware directory that are used instead.
1121 Generic built-in EDID data sets are used, if one of
1122 edid/1024x768.bin, edid/1280x1024.bin,
1123 edid/1680x1050.bin, or edid/1920x1080.bin is given
1124 and no file with the same name exists. Details and
1125 instructions how to build your own EDID data are
1126 available in Documentation/admin-guide/edid.rst. An EDID
1127 data set will only be used for a particular connector,
1128 if its name and a colon are prepended to the EDID
1129 name. Each connector may use a unique EDID data
1130 set by separating the files with a comma. An EDID
1131 data set with no connector name will be used for
1132 any connectors not explicitly specified.
1137 Format: {"off" | "known"}
1138 Control how the dt_cpu_ftrs device-tree binding is
1139 used for CPU feature discovery and setup (if it
1141 off: Do not use it, fall back to legacy cpu table.
1142 known: Do not pass through unknown features to guests
1143 or userspace, only those that the kernel is aware of.
1145 dump_apple_properties [X86]
1146 Dump name and content of EFI device properties on
1147 x86 Macs. Useful for driver authors to determine
1148 what data is available or for reverse-engineering.
1150 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
1151 <module>.dyndbg[="val"]
1152 Enable debug messages at boot time. See
1153 Documentation/admin-guide/dynamic-debug-howto.rst
1156 early_ioremap_debug [KNL]
1157 Enable debug messages in early_ioremap support. This
1158 is useful for tracking down temporary early mappings
1159 which are not unmapped.
1161 earlycon= [KNL] Output early console device and options.
1163 When used with no options, the early console is
1164 determined by stdout-path property in device tree's
1165 chosen node or the ACPI SPCR table if supported by
1168 cdns,<addr>[,options]
1169 Start an early, polled-mode console on a Cadence
1170 (xuartps) serial port at the specified address. Only
1171 supported option is baud rate. If baud rate is not
1172 specified, the serial port must already be setup and
1175 uart[8250],io,<addr>[,options]
1176 uart[8250],mmio,<addr>[,options]
1177 uart[8250],mmio32,<addr>[,options]
1178 uart[8250],mmio32be,<addr>[,options]
1179 uart[8250],0x<addr>[,options]
1180 Start an early, polled-mode console on the 8250/16550
1181 UART at the specified I/O port or MMIO address.
1182 MMIO inter-register address stride is either 8-bit
1183 (mmio) or 32-bit (mmio32 or mmio32be).
1184 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
1185 to be equivalent to 'mmio'. 'options' are specified
1186 in the same format described for "console=ttyS<n>"; if
1187 unspecified, the h/w is not initialized.
1191 Start an early, polled-mode console on a pl011 serial
1192 port at the specified address. The pl011 serial port
1193 must already be setup and configured. Options are not
1194 yet supported. If 'mmio32' is specified, then only
1195 the driver will use only 32-bit accessors to read/write
1196 the device registers.
1199 Start an early console on a litex serial port at the
1200 specified address. The serial port must already be
1201 setup and configured. Options are not yet supported.
1204 Start an early, polled-mode console on a meson serial
1205 port at the specified address. The serial port must
1206 already be setup and configured. Options are not yet
1210 Start an early, polled-mode console on an msm serial
1211 port at the specified address. The serial port
1212 must already be setup and configured. Options are not
1215 msm_serial_dm,<addr>
1216 Start an early, polled-mode console on an msm serial
1217 dm port at the specified address. The serial port
1218 must already be setup and configured. Options are not
1222 Start an early, polled-mode console on a serial port
1223 of an Actions Semi SoC, such as S500 or S900, at the
1224 specified address. The serial port must already be
1225 setup and configured. Options are not yet supported.
1228 Start an early, polled-mode console on a serial port
1229 of an RDA Micro SoC, such as RDA8810PL, at the
1230 specified address. The serial port must already be
1231 setup and configured. Options are not yet supported.
1234 Use RISC-V SBI (Supervisor Binary Interface) for early
1237 smh Use ARM semihosting calls for early console.
1245 Use early console provided by serial driver available
1246 on Samsung SoCs, requires selecting proper type and
1247 a correct base address of the selected UART port. The
1248 serial port must already be setup and configured.
1249 Options are not yet supported.
1252 Start an early, polled-mode console on a lantiq serial
1253 (lqasc) port at the specified address. The serial port
1254 must already be setup and configured. Options are not
1259 Use early console provided by Freescale LP UART driver
1260 found on Freescale Vybrid and QorIQ LS1021A processors.
1261 A valid base address must be provided, and the serial
1262 port must already be setup and configured.
1266 Start an early, polled-mode, output-only console on the
1267 Freescale i.MX UART at the specified address. The UART
1268 must already be setup and configured.
1271 Start an early, polled-mode console on the
1272 Armada 3700 serial port at the specified
1273 address. The serial port must already be setup
1274 and configured. Options are not yet supported.
1277 Start an early, polled-mode console on a Qualcomm
1278 Generic Interface (GENI) based serial port at the
1279 specified address. The serial port must already be
1280 setup and configured. Options are not yet supported.
1283 Start an early, unaccelerated console on the EFI
1284 memory mapped framebuffer (if available). On cache
1285 coherent non-x86 systems that use system memory for
1286 the framebuffer, pass the 'ram' option so that it is
1287 mapped with the correct attributes.
1290 Use early console provided by Freescale LINFlexD UART
1291 serial driver for NXP S32V234 SoCs. A valid base
1292 address must be provided, and the serial port must
1293 already be setup and configured.
1295 earlyprintk= [X86,SH,ARM,M68k,S390]
1299 earlyprintk=serial[,ttySn[,baudrate]]
1300 earlyprintk=serial[,0x...[,baudrate]]
1301 earlyprintk=ttySn[,baudrate]
1302 earlyprintk=dbgp[debugController#]
1303 earlyprintk=pciserial[,force],bus:device.function[,baudrate]
1304 earlyprintk=xdbc[xhciController#]
1306 earlyprintk is useful when the kernel crashes before
1307 the normal console is initialized. It is not enabled by
1308 default because it has some cosmetic problems.
1310 Append ",keep" to not disable it when the real console
1313 Only one of vga, serial, or usb debug port can
1316 Currently only ttyS0 and ttyS1 may be specified by
1317 name. Other I/O ports may be explicitly specified
1318 on some architectures (x86 and arm at least) by
1319 replacing ttySn with an I/O port address, like this:
1320 earlyprintk=serial,0x1008,115200
1321 You can find the port for a given device in
1322 /proc/tty/driver/serial:
1323 2: uart:ST16650V2 port:00001008 irq:18 ...
1325 Interaction with the standard serial driver is not
1328 The VGA output is eventually overwritten by
1331 The xen option can only be used in Xen domains.
1333 The sclp output can only be used on s390.
1335 The optional "force" to "pciserial" enables use of a
1336 PCI device even when its classcode is not of the
1339 edac_report= [HW,EDAC] Control how to report EDAC event
1340 Format: {"on" | "off" | "force"}
1341 on: enable EDAC to report H/W event. May be overridden
1342 by other higher priority error reporting module.
1343 off: disable H/W event reporting through EDAC.
1344 force: enforce the use of EDAC to report H/W event.
1348 Format: {"off" | "on" | "skip[mbr]"}
1351 Format: { "debug", "disable_early_pci_dma",
1352 "nochunk", "noruntime", "nosoftreserve",
1353 "novamap", "no_disable_early_pci_dma" }
1354 debug: enable misc debug output.
1355 disable_early_pci_dma: disable the busmaster bit on all
1356 PCI bridges while in the EFI boot stub.
1357 nochunk: disable reading files in "chunks" in the EFI
1358 boot stub, as chunking can cause problems with some
1359 firmware implementations.
1360 noruntime : disable EFI runtime services support
1361 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose)
1362 attribute may cause the kernel to reserve the
1363 memory range for a memory mapping driver to
1364 claim. Specify efi=nosoftreserve to disable this
1365 reservation and treat the memory by its base type
1366 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM").
1367 novamap: do not call SetVirtualAddressMap().
1368 no_disable_early_pci_dma: Leave the busmaster bit set
1369 on all PCI bridges while in the EFI boot stub
1371 efi_no_storage_paranoia [EFI; X86]
1372 Using this parameter you can use more than 50% of
1373 your efi variable storage. Use this parameter only if
1374 you are really sure that your UEFI does sane gc and
1375 fulfills the spec otherwise your board may brick.
1377 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1378 Add arbitrary attribute to specific memory range by
1379 updating original EFI memory map.
1380 Region of memory which aa attribute is added to is
1383 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1384 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1385 attribute is added to range 0x100000000-0x180000000 and
1386 0x10a0000000-0x1120000000.
1388 If efi_fake_mem=8G@9G:0x40000 is specified, the
1389 EFI_MEMORY_SP(0x40000) attribute is added to
1390 range 0x240000000-0x43fffffff.
1392 Using this parameter you can do debugging of EFI memmap
1393 related features. For example, you can do debugging of
1394 Address Range Mirroring feature even if your box
1395 doesn't support it, or mark specific memory as
1398 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT
1399 that is to be dynamically loaded by Linux. If there are
1400 multiple variables with the same name but with different
1401 vendor GUIDs, all of them will be loaded. See
1402 Documentation/admin-guide/acpi/ssdt-overlays.rst for details.
1405 eisa_irq_edge= [PARISC,HW]
1406 See header of drivers/parisc/eisa.c.
1408 ekgdboc= [X86,KGDB] Allow early kernel console debugging
1411 This is designed to be used in conjunction with
1412 the boot argument: earlyprintk=vga
1414 This parameter works in place of the kgdboc parameter
1415 but can only be used if the backing tty is available
1416 very early in the boot process. For early debugging
1417 via a serial port see kgdboc_earlycon instead.
1420 See comment before function elanfreq_setup() in
1421 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1423 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1424 Specifies physical address of start of kernel core
1425 image elf header and optionally the size. Generally
1426 kexec loader will pass this option to capture kernel.
1427 See Documentation/admin-guide/kdump/kdump.rst for details.
1429 enable_mtrr_cleanup [X86]
1430 The kernel tries to adjust MTRR layout from continuous
1431 to discrete, to make X server driver able to add WB
1432 entry later. This parameter enables that.
1434 enable_timer_pin_1 [X86]
1435 Enable PIN 1 of APIC timer
1436 Can be useful to work around chipset bugs
1437 (in particular on some ATI chipsets).
1438 The kernel tries to set a reasonable default.
1440 enforcing= [SELINUX] Set initial enforcing status.
1442 See security/selinux/Kconfig help text.
1443 0 -- permissive (log only, no denials).
1444 1 -- enforcing (deny and log).
1446 Value can be changed at runtime via
1447 /sys/fs/selinux/enforce.
1450 Disable Error Record Serialization Table (ERST)
1453 ether= [HW,NET] Ethernet cards parameters
1454 This option is obsoleted by the "netdev=" option, which
1455 has equivalent usage. See its documentation for details.
1459 Permit 'security.evm' to be updated regardless of
1460 current integrity status.
1462 early_page_ext [KNL] Enforces page_ext initialization to earlier
1463 stages so cover more early boot allocations.
1464 Please note that as side effect some optimizations
1465 might be disabled to achieve that (e.g. parallelized
1466 memory initialization is disabled) so the boot process
1467 might take longer, especially on systems with a lot of
1468 memory. Available with CONFIG_PAGE_EXTENSION=y.
1473 fail_make_request=[KNL]
1474 General fault injection mechanism.
1475 Format: <interval>,<probability>,<space>,<times>
1476 See also Documentation/fault-injection/.
1479 Format: { initns | none }
1480 See Documentation/admin-guide/sysctl/net.rst for
1481 fb_tunnels_only_for_init_ns
1484 See Documentation/admin-guide/blockdev/floppy.rst.
1486 force_pal_cache_flush
1487 [IA-64] Avoid check_sal_cache_flush which may hang on
1488 buggy SAL_CACHE_FLUSH implementations. Using this
1489 parameter will force ia64_sal_cache_flush to call
1490 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1493 Forcefully enable Physical Address Extension (PAE).
1494 Many Pentium M systems disable PAE but may have a
1495 functionally usable PAE implementation.
1496 Warning: use of this parameter will taint the kernel
1497 and may cause unknown problems.
1500 [FTRACE] will set and start the specified tracer
1501 as early as possible in order to facilitate early
1504 ftrace_boot_snapshot
1505 [FTRACE] On boot up, a snapshot will be taken of the
1506 ftrace ring buffer that can be read at:
1507 /sys/kernel/tracing/snapshot.
1508 This is useful if you need tracing information from kernel
1509 boot up that is likely to be overridden by user space
1510 start up functionality.
1512 ftrace_dump_on_oops[=orig_cpu]
1513 [FTRACE] will dump the trace buffers on oops.
1514 If no parameter is passed, ftrace will dump
1515 buffers of all CPUs, but if you pass orig_cpu, it will
1516 dump only the buffer of the CPU that triggered the
1519 ftrace_filter=[function-list]
1520 [FTRACE] Limit the functions traced by the function
1521 tracer at boot up. function-list is a comma-separated
1522 list of functions. This list can be changed at run
1523 time by the set_ftrace_filter file in the debugfs
1526 ftrace_notrace=[function-list]
1527 [FTRACE] Do not trace the functions specified in
1528 function-list. This list can be changed at run time
1529 by the set_ftrace_notrace file in the debugfs
1532 ftrace_graph_filter=[function-list]
1533 [FTRACE] Limit the top level callers functions traced
1534 by the function graph tracer at boot up.
1535 function-list is a comma-separated list of functions
1536 that can be changed at run time by the
1537 set_graph_function file in the debugfs tracing directory.
1539 ftrace_graph_notrace=[function-list]
1540 [FTRACE] Do not trace from the functions specified in
1541 function-list. This list is a comma-separated list of
1542 functions that can be changed at run time by the
1543 set_graph_notrace file in the debugfs tracing directory.
1545 ftrace_graph_max_depth=<uint>
1546 [FTRACE] Used with the function graph tracer. This is
1547 the max depth it will trace into a function. This value
1548 can be changed at run time by the max_graph_depth file
1549 in the tracefs tracing directory. default: 0 (no limit)
1551 fw_devlink= [KNL] Create device links between consumer and supplier
1552 devices by scanning the firmware to infer the
1553 consumer/supplier relationships. This feature is
1554 especially useful when drivers are loaded as modules as
1555 it ensures proper ordering of tasks like device probing
1556 (suppliers first, then consumers), supplier boot state
1557 clean up (only after all consumers have probed),
1558 suspend/resume & runtime PM (consumers first, then
1560 Format: { off | permissive | on | rpm }
1561 off -- Don't create device links from firmware info.
1562 permissive -- Create device links from firmware info
1563 but use it only for ordering boot state clean
1564 up (sync_state() calls).
1565 on -- Create device links from firmware info and use it
1566 to enforce probe and suspend/resume ordering.
1567 rpm -- Like "on", but also use to order runtime PM.
1569 fw_devlink.strict=<bool>
1570 [KNL] Treat all inferred dependencies as mandatory
1571 dependencies. This only applies for fw_devlink=on|rpm.
1575 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1576 support via parallel port (up to 5 devices per port)
1577 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1578 See also Documentation/input/devices/joystick-parport.rst
1582 gart_fix_e820= [X86-64] disable the fix e820 for K8 GART
1586 gcov_persist= [GCOV] When non-zero (default), profiling data for
1587 kernel modules is saved and remains accessible via
1588 debugfs, even when the module is unloaded/reloaded.
1589 When zero, profiling data is discarded and associated
1590 debugfs files are removed at module unload time.
1592 goldfish [X86] Enable the goldfish android emulator platform.
1593 Don't use this when you are not running on the
1596 gpio-mockup.gpio_mockup_ranges
1597 [HW] Sets the ranges of gpiochip of for this device.
1598 Format: <start1>,<end1>,<start2>,<end2>...
1599 gpio-mockup.gpio_mockup_named_lines
1600 [HW] Let the driver know GPIO lines should be named.
1602 gpt [EFI] Forces disk with valid GPT signature but
1603 invalid Protective MBR to be treated as GPT. If the
1604 primary GPT is corrupted, it enables the backup/alternate
1605 GPT to be used instead.
1607 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1608 the "Enable 0" bit of the configuration register.
1611 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1612 the "Enable 0" bit of the configuration register.
1615 grcan.select= [HW] Select which physical interface to use.
1618 grcan.txsize= [HW] Sets the size of the tx buffer.
1619 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1621 grcan.rxsize= [HW] Sets the size of the rx buffer.
1622 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1626 [KNL] Under CONFIG_HARDENED_USERCOPY, whether
1627 hardening is enabled for this boot. Hardened
1628 usercopy checking is used to protect the kernel
1629 from reading or writing beyond known memory
1630 allocation boundaries as a proactive defense
1631 against bounds-checking flaws in the kernel's
1632 copy_to_user()/copy_from_user() interface.
1633 on Perform hardened usercopy checks (default).
1634 off Disable hardened usercopy checks.
1636 hardlockup_all_cpu_backtrace=
1637 [KNL] Should the hard-lockup detector generate
1638 backtraces on all cpus.
1641 hashdist= [KNL,NUMA] Large hashes allocated during boot
1642 are distributed across NUMA nodes. Defaults on
1643 for 64-bit NUMA, off otherwise.
1644 Format: 0 | 1 (for off | on)
1646 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1648 hd= [EIDE] (E)IDE hard drive subsystem geometry
1649 Format: <cyl>,<head>,<sect>
1652 Disable Hardware Error Source Table (HEST) support;
1653 corresponding firmware-first mode error processing
1654 logic will be disabled.
1656 hibernate= [HIBERNATION]
1657 noresume Don't check if there's a hibernation image
1658 present during boot.
1659 nocompress Don't compress/decompress hibernation images.
1660 no Disable hibernation and resume.
1661 protect_image Turn on image protection during restoration
1662 (that will set all pages holding image data
1663 during restoration read-only).
1665 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1666 size of <nn>. This works even on boxes that have no
1667 highmem otherwise. This also works to reduce highmem
1668 size on bigger boxes.
1670 highres= [KNL] Enable/disable high resolution timer mode.
1671 Valid parameters: "on", "off"
1676 hostname= [KNL] Set the hostname (aka UTS nodename).
1678 This allows setting the system's hostname during early
1679 startup. This sets the name returned by gethostname.
1680 Using this parameter to set the hostname makes it
1681 possible to ensure the hostname is correctly set before
1682 any userspace processes run, avoiding the possibility
1683 that a process may call gethostname before the hostname
1684 has been explicitly set, resulting in the calling
1685 process getting an incorrect result. The string must
1686 not exceed the maximum allowed hostname length (usually
1687 64 characters) and will be truncated otherwise.
1689 hpet= [X86-32,HPET] option to control HPET usage
1690 Format: { enable (default) | disable | force |
1692 disable: disable HPET and use PIT instead
1693 force: allow force enabled of undocumented chips (ICH4,
1695 verbose: show contents of HPET registers during setup
1697 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1698 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1700 hugepages= [HW] Number of HugeTLB pages to allocate at boot.
1701 If this follows hugepagesz (below), it specifies
1702 the number of pages of hugepagesz to be allocated.
1703 If this is the first HugeTLB parameter on the command
1704 line, it specifies the number of pages to allocate for
1705 the default huge page size. If using node format, the
1706 number of pages to allocate per-node can be specified.
1707 See also Documentation/admin-guide/mm/hugetlbpage.rst.
1708 Format: <integer> or (node format)
1709 <node>:<integer>[,<node>:<integer>]
1712 [HW] The size of the HugeTLB pages. This is used in
1713 conjunction with hugepages (above) to allocate huge
1714 pages of a specific size at boot. The pair
1715 hugepagesz=X hugepages=Y can be specified once for
1716 each supported huge page size. Huge page sizes are
1717 architecture dependent. See also
1718 Documentation/admin-guide/mm/hugetlbpage.rst.
1721 hugetlb_cma= [HW,CMA] The size of a CMA area used for allocation
1722 of gigantic hugepages. Or using node format, the size
1723 of a CMA area per node can be specified.
1724 Format: nn[KMGTPE] or (node format)
1725 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]]
1727 Reserve a CMA area of given size and allocate gigantic
1728 hugepages using the CMA allocator. If enabled, the
1729 boot-time allocation of gigantic hugepages is skipped.
1731 hugetlb_free_vmemmap=
1732 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP
1734 Control if HugeTLB Vmemmap Optimization (HVO) is enabled.
1735 Allows heavy hugetlb users to free up some more
1736 memory (7 * PAGE_SIZE for each 2MB hugetlb page).
1737 Format: { on | off (default) }
1742 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y,
1745 Note that the vmemmap pages may be allocated from the added
1746 memory block itself when memory_hotplug.memmap_on_memory is
1747 enabled, those vmemmap pages cannot be optimized even if this
1748 feature is enabled. Other vmemmap pages not allocated from
1749 the added memory block itself do not be affected.
1752 [KNL] Should the hung task detector generate panics.
1755 A value of 1 instructs the kernel to panic when a
1756 hung task is detected. The default value is controlled
1757 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time
1758 option. The value selected by this boot parameter can
1759 be changed later by the kernel.hung_task_panic sysctl.
1761 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1762 terminal devices. Valid values: 0..8
1763 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1764 If specified, z/VM IUCV HVC accepts connections
1765 from listed z/VM user IDs only.
1767 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations
1768 which allow the hypervisor to 'idle' the
1769 guest on lock contention.
1771 i2c_bus= [HW] Override the default board specific I2C bus speed
1772 or register an additional I2C bus that is not
1773 registered from board initialization code.
1777 i8042.debug [HW] Toggle i8042 debug mode
1778 i8042.unmask_kbd_data
1779 [HW] Enable printing of interrupt data from the KBD port
1780 (disabled by default, and as a pre-condition
1781 requires that i8042.debug=1 be enabled)
1782 i8042.direct [HW] Put keyboard port into non-translated mode
1783 i8042.dumbkbd [HW] Pretend that controller can only read data from
1784 keyboard and cannot control its state
1785 (Don't attempt to blink the leds)
1786 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1787 i8042.nokbd [HW] Don't check/create keyboard port
1788 i8042.noloop [HW] Disable the AUX Loopback command while probing
1790 i8042.nomux [HW] Don't check presence of an active multiplexing
1792 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1794 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1795 i8042.reset [HW] Reset the controller during init, cleanup and
1796 suspend-to-ram transitions, only during s2r
1797 transitions, or never reset
1798 Format: { 1 | Y | y | 0 | N | n }
1799 1, Y, y: always reset controller
1800 0, N, n: don't ever reset controller
1801 Default: only on s2r transitions on x86; most other
1802 architectures force reset to be always executed
1803 i8042.unlock [HW] Unlock (ignore) the keylock
1804 i8042.kbdreset [HW] Reset device connected to KBD port
1806 [HW] Allow deferred probing upon i8042 probe errors
1810 i915.invert_brightness=
1811 [DRM] Invert the sense of the variable that is used to
1812 set the brightness of the panel backlight. Normally a
1813 brightness value of 0 indicates backlight switched off,
1814 and the maximum of the brightness value sets the backlight
1815 to maximum brightness. If this parameter is set to 0
1816 (default) and the machine requires it, or this parameter
1817 is set to 1, a brightness value of 0 sets the backlight
1818 to maximum brightness, and the maximum of the brightness
1819 value switches the backlight off.
1820 -1 -- never invert brightness
1821 0 -- machine default
1822 1 -- force brightness inversion
1825 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1829 Format: idle=poll, idle=halt, idle=nomwait
1830 Poll forces a polling idle loop that can slightly
1831 improve the performance of waking up a idle CPU, but
1832 will use a lot of power and make the system run hot.
1834 idle=halt: Halt is forced to be used for CPU idle.
1835 In such case C2/C3 won't be used again.
1836 idle=nomwait: Disable mwait for CPU C-states
1840 Allow force disabling of Shared Virtual Memory (SVA)
1841 support for the idxd driver. By default it is set to
1844 idxd.tc_override= [HW]
1846 Allow override of default traffic class configuration
1847 for the device. By default it is set to false (0).
1849 ieee754= [MIPS] Select IEEE Std 754 conformance mode
1850 Format: { strict | legacy | 2008 | relaxed }
1853 Choose which programs will be accepted for execution
1854 based on the IEEE 754 NaN encoding(s) supported by
1855 the FPU and the NaN encoding requested with the value
1856 of an ELF file header flag individually set by each
1857 binary. Hardware implementations are permitted to
1858 support either or both of the legacy and the 2008 NaN
1861 Available settings are as follows:
1862 strict accept binaries that request a NaN encoding
1863 supported by the FPU
1864 legacy only accept legacy-NaN binaries, if supported
1866 2008 only accept 2008-NaN binaries, if supported
1868 relaxed accept any binaries regardless of whether
1869 supported by the FPU
1871 The FPU emulator is always able to support both NaN
1872 encodings, so if no FPU hardware is present or it has
1873 been disabled with 'nofpu', then the settings of
1874 'legacy' and '2008' strap the emulator accordingly,
1875 'relaxed' straps the emulator for both legacy-NaN and
1876 2008-NaN, whereas 'strict' enables legacy-NaN only on
1877 legacy processors and both NaN encodings on MIPS32 or
1880 The setting for ABS.fmt/NEG.fmt instruction execution
1881 mode generally follows that for the NaN encoding,
1882 except where unsupported by hardware.
1884 ignore_loglevel [KNL]
1885 Ignore loglevel setting - this will print /all/
1886 kernel messages to the console. Useful for debugging.
1887 We also add it as printk module parameter, so users
1888 could change it dynamically, usually by
1889 /sys/module/printk/parameters/ignore_loglevel.
1892 Ignore RLIMIT_DATA setting for data mappings,
1893 print warning at first misuse. Can be changed via
1894 /sys/module/kernel/parameters/ignore_rlimit_data.
1896 ihash_entries= [KNL]
1897 Set number of hash buckets for inode cache.
1899 ima_appraise= [IMA] appraise integrity measurements
1900 Format: { "off" | "enforce" | "fix" | "log" }
1903 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead.
1904 The builtin appraise policy appraises all files
1907 ima_canonical_fmt [IMA]
1908 Use the canonical format for the binary runtime
1909 measurements, instead of host native format.
1912 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1916 The list of supported hash algorithms is defined
1917 in crypto/hash_info.h.
1920 The builtin policies to load during IMA setup.
1921 Format: "tcb | appraise_tcb | secure_boot |
1922 fail_securely | critical_data"
1924 The "tcb" policy measures all programs exec'd, files
1925 mmap'd for exec, and all files opened with the read
1926 mode bit set by either the effective uid (euid=0) or
1929 The "appraise_tcb" policy appraises the integrity of
1930 all files owned by root.
1932 The "secure_boot" policy appraises the integrity
1933 of files (eg. kexec kernel image, kernel modules,
1934 firmware, policy, etc) based on file signatures.
1936 The "fail_securely" policy forces file signature
1937 verification failure also on privileged mounted
1938 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE
1941 The "critical_data" policy measures kernel integrity
1944 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
1945 Load a policy which meets the needs of the Trusted
1946 Computing Base. This means IMA will measure all
1947 programs exec'd, files mmap'd for exec, and all files
1948 opened for read by uid=0.
1951 Select one of defined IMA measurements template formats.
1952 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" |
1957 [IMA] Define a custom template format.
1958 Format: { "field1|...|fieldN" }
1960 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1961 Format: <min_file_size>
1962 Set the minimal file size for using asynchronous hash.
1963 If left unspecified, ahash usage is disabled.
1965 ahash performance varies for different data sizes on
1966 different crypto accelerators. This option can be used
1967 to achieve the best performance for a particular HW.
1969 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1971 Set hashing buffer size. Default: 4k.
1973 ahash performance varies for different chunk sizes on
1974 different crypto accelerators. This option can be used
1975 to achieve best performance for particular HW.
1979 Run specified binary instead of /sbin/init as init
1982 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1983 for working out where the kernel is dying during
1986 initcall_blacklist= [KNL] Do not execute a comma-separated list of
1987 initcall functions. Useful for debugging built-in
1988 modules and initcalls.
1990 initramfs_async= [KNL]
1993 This parameter controls whether the initramfs
1994 image is unpacked asynchronously, concurrently
1995 with devices being probed and
1996 initialized. This should normally just work,
1997 but as a debugging aid, one can get the
1998 historical behaviour of the initramfs
1999 unpacking being completed before device_ and
2002 initrd= [BOOT] Specify the location of the initial ramdisk
2004 initrdmem= [KNL] Specify a physical address and size from which to
2005 load the initrd. If an initrd is compiled in or
2006 specified in the bootparams, it takes priority over this
2008 Format: ss[KMG],nn[KMG]
2011 init_on_alloc= [MM] Fill newly allocated pages and heap objects with
2014 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON.
2016 init_on_free= [MM] Fill freed pages and heap objects with zeroes.
2018 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON.
2020 init_pkru= [X86] Specify the default memory protection keys rights
2021 register contents for all processes. 0x55555554 by
2022 default (disallow access to all but pkey 0). Can
2023 override in debugfs after boot.
2025 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
2028 int_pln_enable [X86] Enable power limit notification interrupt
2030 integrity_audit=[IMA]
2031 Format: { "0" | "1" }
2032 0 -- basic integrity auditing messages. (Default)
2033 1 -- additional integrity auditing messages.
2035 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
2037 Enable intel iommu driver.
2039 Disable intel iommu driver.
2040 igfx_off [Default Off]
2041 By default, gfx is mapped as normal device. If a gfx
2042 device has a dedicated DMAR unit, the DMAR unit is
2043 bypassed by not enabling DMAR with this option. In
2044 this case, gfx device will use physical address for
2046 strict [Default Off]
2047 Deprecated, equivalent to iommu.strict=1.
2048 sp_off [Default Off]
2049 By default, super page will be supported if Intel IOMMU
2050 has the capability. With this option, super page will
2053 Enable the Intel IOMMU scalable mode if the hardware
2054 advertises that it has support for the scalable mode
2057 Disallow use of the Intel IOMMU scalable mode.
2058 tboot_noforce [Default Off]
2059 Do not force the Intel IOMMU enabled under tboot.
2060 By default, tboot will force Intel IOMMU on, which
2061 could harm performance of some high-throughput
2062 devices like 40GBit network cards, even if identity
2064 Note that using this option lowers the security
2065 provided by tboot because it makes the system
2066 vulnerable to DMA attacks.
2068 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
2069 0 disables intel_idle and fall back on acpi_idle.
2070 1 to 9 specify maximum depth of C-state.
2074 Do not enable intel_pstate as the default
2075 scaling driver for the supported processors
2077 Use intel_pstate as a scaling driver, but configure it
2078 to work with generic cpufreq governors (instead of
2079 enabling its internal governor). This mode cannot be
2080 used along with the hardware-managed P-states (HWP)
2083 Enable intel_pstate on systems that prohibit it by default
2084 in favor of acpi-cpufreq. Forcing the intel_pstate driver
2085 instead of acpi-cpufreq may disable platform features, such
2086 as thermal controls and power capping, that rely on ACPI
2087 P-States information being indicated to OSPM and therefore
2088 should be used with caution. This option does not work with
2089 processors that aren't supported by the intel_pstate driver
2090 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
2092 Do not enable hardware P state control (HWP)
2095 Only load intel_pstate on systems which support
2096 hardware P state control (HWP) if available.
2098 Enforce ACPI _PPC performance limits. If the Fixed ACPI
2099 Description Table, specifies preferred power management
2100 profile as "Enterprise Server" or "Performance Server",
2101 then this feature is turned on by default.
2103 Allow per-logical-CPU P-State performance control limits using
2104 cpufreq sysfs interface
2106 intremap= [X86-64, Intel-IOMMU]
2107 on enable Interrupt Remapping (default)
2108 off disable Interrupt Remapping
2109 nosid disable Source ID checking
2111 BIOS x2APIC opt-out request will be ignored
2112 nopost disable Interrupt Posting
2114 iomem= Disable strict checking of access to MMIO memory
2115 strict regions from userspace.
2130 nobypass [PPC/POWERNV]
2131 Disable IOMMU bypass, using IOMMU for PCI devices.
2133 iommu.forcedac= [ARM64, X86] Control IOVA allocation for PCI devices.
2134 Format: { "0" | "1" }
2135 0 - Try to allocate a 32-bit DMA address first, before
2136 falling back to the full range if needed.
2137 1 - Allocate directly from the full usable range,
2138 forcing Dual Address Cycle for PCI cards supporting
2139 greater than 32-bit addressing.
2141 iommu.strict= [ARM64, X86] Configure TLB invalidation behaviour
2142 Format: { "0" | "1" }
2144 Request that DMA unmap operations use deferred
2145 invalidation of hardware TLBs, for increased
2146 throughput at the cost of reduced device isolation.
2147 Will fall back to strict mode if not supported by
2148 the relevant IOMMU driver.
2150 DMA unmap operations invalidate IOMMU hardware TLBs
2152 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}.
2153 Note: on x86, strict mode specified via one of the
2154 legacy driver-specific options takes precedence.
2157 [ARM64, X86] Configure DMA to bypass the IOMMU by default.
2158 Format: { "0" | "1" }
2159 0 - Use IOMMU translation for DMA.
2160 1 - Bypass the IOMMU for DMA.
2161 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH.
2163 io7= [HW] IO7 for Marvel-based Alpha systems
2164 See comment before marvel_specify_io7 in
2165 arch/alpha/kernel/core_marvel.c.
2167 io_delay= [X86] I/O delay method
2169 Standard port 0x80 based delay
2171 Alternate port 0xed based delay (needed on some systems)
2173 Simple two microseconds delay
2178 See Documentation/admin-guide/nfs/nfsroot.rst.
2180 ipcmni_extend [KNL] Extend the maximum number of unique System V
2181 IPC identifiers from 32,768 to 16,777,216.
2183 irqaffinity= [SMP] Set the default irq affinity mask
2184 The argument is a cpu list, as described above.
2186 irqchip.gicv2_force_probe=
2189 Force the kernel to look for the second 4kB page
2190 of a GICv2 controller even if the memory range
2191 exposed by the device tree is too small.
2193 irqchip.gicv3_nolpi=
2195 Force the kernel to ignore the availability of
2196 LPIs (and by consequence ITSs). Intended for system
2197 that use the kernel as a bootloader, and thus want
2198 to let secondary kernels in charge of setting up
2201 irqchip.gicv3_pseudo_nmi= [ARM64]
2202 Enables support for pseudo-NMIs in the kernel. This
2203 requires the kernel to be built with
2204 CONFIG_ARM64_PSEUDO_NMI.
2207 When an interrupt is not handled search all handlers
2208 for it. Intended to get systems with badly broken
2212 When an interrupt is not handled search all handlers
2213 for it. Also check all handlers each timer
2214 interrupt. Intended to get systems with badly broken
2218 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
2220 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance.
2221 [Deprecated - use cpusets instead]
2222 Format: [flag-list,]<cpu-list>
2224 Specify one or more CPUs to isolate from disturbances
2225 specified in the flag list (default: domain):
2228 Disable the tick when a single task runs.
2230 A residual 1Hz tick is offloaded to workqueues, which you
2231 need to affine to housekeeping through the global
2232 workqueue's affinity configured via the
2233 /sys/devices/virtual/workqueue/cpumask sysfs file, or
2234 by using the 'domain' flag described below.
2236 NOTE: by default the global workqueue runs on all CPUs,
2237 so to protect individual CPUs the 'cpumask' file has to
2238 be configured manually after bootup.
2241 Isolate from the general SMP balancing and scheduling
2242 algorithms. Note that performing domain isolation this way
2243 is irreversible: it's not possible to bring back a CPU to
2244 the domains once isolated through isolcpus. It's strongly
2245 advised to use cpusets instead to disable scheduler load
2246 balancing through the "cpuset.sched_load_balance" file.
2247 It offers a much more flexible interface where CPUs can
2248 move in and out of an isolated set anytime.
2250 You can move a process onto or off an "isolated" CPU via
2251 the CPU affinity syscalls or cpuset.
2252 <cpu number> begins at 0 and the maximum value is
2253 "number of CPUs in system - 1".
2257 Isolate from being targeted by managed interrupts
2258 which have an interrupt mask containing isolated
2259 CPUs. The affinity of managed interrupts is
2260 handled by the kernel and cannot be changed via
2261 the /proc/irq/* interfaces.
2263 This isolation is best effort and only effective
2264 if the automatically assigned interrupt mask of a
2265 device queue contains isolated and housekeeping
2266 CPUs. If housekeeping CPUs are online then such
2267 interrupts are directed to the housekeeping CPU
2268 so that IO submitted on the housekeeping CPU
2269 cannot disturb the isolated CPU.
2271 If a queue's affinity mask contains only isolated
2272 CPUs then this parameter has no effect on the
2273 interrupt routing decision, though interrupts are
2274 only delivered when tasks running on those
2275 isolated CPUs submit IO. IO submitted on
2276 housekeeping CPUs has no influence on those
2279 The format of <cpu-list> is described above.
2283 ivrs_ioapic [HW,X86-64]
2284 Provide an override to the IOAPIC-ID<->DEVICE-ID
2285 mapping provided in the IVRS ACPI table.
2286 By default, PCI segment is 0, and can be omitted.
2288 For example, to map IOAPIC-ID decimal 10 to
2289 PCI segment 0x1 and PCI device 00:14.0,
2290 write the parameter as:
2291 ivrs_ioapic=10@0001:00:14.0
2294 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0
2295 write the parameter as:
2296 ivrs_ioapic[10]=00:14.0
2297 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and
2298 PCI device 00:14.0 write the parameter as:
2299 ivrs_ioapic[10]=0001:00:14.0
2301 ivrs_hpet [HW,X86-64]
2302 Provide an override to the HPET-ID<->DEVICE-ID
2303 mapping provided in the IVRS ACPI table.
2304 By default, PCI segment is 0, and can be omitted.
2306 For example, to map HPET-ID decimal 10 to
2307 PCI segment 0x1 and PCI device 00:14.0,
2308 write the parameter as:
2309 ivrs_hpet=10@0001:00:14.0
2312 * To map HPET-ID decimal 0 to PCI device 00:14.0
2313 write the parameter as:
2314 ivrs_hpet[0]=00:14.0
2315 * To map HPET-ID decimal 10 to PCI segment 0x1 and
2316 PCI device 00:14.0 write the parameter as:
2317 ivrs_ioapic[10]=0001:00:14.0
2319 ivrs_acpihid [HW,X86-64]
2320 Provide an override to the ACPI-HID:UID<->DEVICE-ID
2321 mapping provided in the IVRS ACPI table.
2322 By default, PCI segment is 0, and can be omitted.
2324 For example, to map UART-HID:UID AMD0020:0 to
2325 PCI segment 0x1 and PCI device ID 00:14.5,
2326 write the parameter as:
2327 ivrs_acpihid=AMD0020:0@0001:00:14.5
2330 * To map UART-HID:UID AMD0020:0 to PCI segment is 0,
2331 PCI device ID 00:14.5, write the parameter as:
2332 ivrs_acpihid[00:14.5]=AMD0020:0
2333 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and
2334 PCI device ID 00:14.5, write the parameter as:
2335 ivrs_acpihid[0001:00:14.5]=AMD0020:0
2337 js= [HW,JOY] Analog joystick
2338 See Documentation/input/joydev/joystick.rst.
2341 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print
2342 report on every invalid memory access. Without this
2343 parameter KASAN will print report only for the first
2347 Do not unregister boot console at start. This is only
2348 useful for debugging when something happens in the window
2349 between unregistering the boot console and initializing
2354 kernelcore= [KNL,X86,IA-64,PPC]
2355 Format: nn[KMGTPE] | nn% | "mirror"
2356 This parameter specifies the amount of memory usable by
2357 the kernel for non-movable allocations. The requested
2358 amount is spread evenly throughout all nodes in the
2359 system as ZONE_NORMAL. The remaining memory is used for
2360 movable memory in its own zone, ZONE_MOVABLE. In the
2361 event, a node is too small to have both ZONE_NORMAL and
2362 ZONE_MOVABLE, kernelcore memory will take priority and
2363 other nodes will have a larger ZONE_MOVABLE.
2365 ZONE_MOVABLE is used for the allocation of pages that
2366 may be reclaimed or moved by the page migration
2367 subsystem. Note that allocations like PTEs-from-HighMem
2368 still use the HighMem zone if it exists, and the Normal
2369 zone if it does not.
2371 It is possible to specify the exact amount of memory in
2372 the form of "nn[KMGTPE]", a percentage of total system
2373 memory in the form of "nn%", or "mirror". If "mirror"
2374 option is specified, mirrored (reliable) memory is used
2375 for non-movable allocations and remaining memory is used
2376 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror"
2377 are exclusive, so you cannot specify multiple forms.
2379 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
2380 Format: <Controller#>[,poll interval]
2381 The controller # is the number of the ehci usb debug
2382 port as it is probed via PCI. The poll interval is
2383 optional and is the number seconds in between
2384 each poll cycle to the debug port in case you need
2385 the functionality for interrupting the kernel with
2386 gdb or control-c on the dbgp connection. When
2387 not using this parameter you use sysrq-g to break into
2388 the kernel debugger.
2390 kgdboc= [KGDB,HW] kgdb over consoles.
2391 Requires a tty driver that supports console polling,
2392 or a supported polling keyboard driver (non-usb).
2393 Serial only format: <serial_device>[,baud]
2394 keyboard only format: kbd
2395 keyboard and serial format: kbd,<serial_device>[,baud]
2396 Optional Kernel mode setting:
2397 kms, kbd format: kms,kbd
2398 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
2400 kgdboc_earlycon= [KGDB,HW]
2401 If the boot console provides the ability to read
2402 characters and can work in polling mode, you can use
2403 this parameter to tell kgdb to use it as a backend
2404 until the normal console is registered. Intended to
2405 be used together with the kgdboc parameter which
2406 specifies the normal console to transition to.
2408 The name of the early console should be specified
2409 as the value of this parameter. Note that the name of
2410 the early console might be different than the tty
2411 name passed to kgdboc. It's OK to leave the value
2412 blank and the first boot console that implements
2413 read() will be picked.
2415 kgdbwait [KGDB] Stop kernel execution and enter the
2416 kernel debugger at the earliest opportunity.
2418 kmac= [MIPS] Korina ethernet MAC address.
2419 Configure the RouterBoard 532 series on-chip
2420 Ethernet adapter MAC address.
2422 kmemleak= [KNL] Boot-time kmemleak enable/disable
2423 Valid arguments: on, off
2425 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
2428 kprobe_event=[probe-list]
2429 [FTRACE] Add kprobe events and enable at boot time.
2430 The probe-list is a semicolon delimited list of probe
2431 definitions. Each definition is same as kprobe_events
2432 interface, but the parameters are comma delimited.
2433 For example, to add a kprobe event on vfs_read with
2434 arg1 and arg2, add to the command line;
2436 kprobe_event=p,vfs_read,$arg1,$arg2
2438 See also Documentation/trace/kprobetrace.rst "Kernel
2439 Boot Parameter" section.
2441 kpti= [ARM64] Control page table isolation of user
2442 and kernel address spaces.
2443 Default: enabled on cores which need mitigation.
2447 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires
2448 CONFIG_KUNIT to be set to be fully enabled. The
2449 default value can be overridden via
2450 KUNIT_DEFAULT_ENABLED.
2451 Default is 1 (enabled)
2453 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
2454 Default is 0 (don't ignore, but inject #GP)
2456 kvm.eager_page_split=
2457 [KVM,X86] Controls whether or not KVM will try to
2458 proactively split all huge pages during dirty logging.
2459 Eager page splitting reduces interruptions to vCPU
2460 execution by eliminating the write-protection faults
2461 and MMU lock contention that would otherwise be
2462 required to split huge pages lazily.
2464 VM workloads that rarely perform writes or that write
2465 only to a small region of VM memory may benefit from
2466 disabling eager page splitting to allow huge pages to
2467 still be used for reads.
2469 The behavior of eager page splitting depends on whether
2470 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If
2471 disabled, all huge pages in a memslot will be eagerly
2472 split when dirty logging is enabled on that memslot. If
2473 enabled, eager page splitting will be performed during
2474 the KVM_CLEAR_DIRTY ioctl, and only for the pages being
2477 Eager page splitting is only supported when kvm.tdp_mmu=Y.
2481 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface.
2482 Default is false (don't support).
2485 [KVM] Controls the software workaround for the
2486 X86_BUG_ITLB_MULTIHIT bug.
2487 force : Always deploy workaround.
2488 off : Never deploy workaround.
2489 auto : Deploy workaround based on the presence of
2490 X86_BUG_ITLB_MULTIHIT.
2494 If the software workaround is enabled for the host,
2495 guests do need not to enable it for nested guests.
2497 kvm.nx_huge_pages_recovery_ratio=
2498 [KVM] Controls how many 4KiB pages are periodically zapped
2499 back to huge pages. 0 disables the recovery, otherwise if
2500 the value is N KVM will zap 1/Nth of the 4KiB pages every
2501 period (see below). The default is 60.
2503 kvm.nx_huge_pages_recovery_period_ms=
2504 [KVM] Controls the time period at which KVM zaps 4KiB pages
2505 back to huge pages. If the value is a non-zero N, KVM will
2506 zap a portion (see ratio above) of the pages every N msecs.
2507 If the value is 0 (the default), KVM will pick a period based
2508 on the ratio, such that a page is zapped after 1 hour on average.
2510 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
2511 Default is 1 (enabled)
2513 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
2515 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
2518 [KVM,ARM] Select one of KVM/arm64's modes of operation.
2520 none: Forcefully disable KVM.
2522 nvhe: Standard nVHE-based mode, without support for
2525 protected: nVHE-based mode with support for guests whose
2526 state is kept private from the host.
2528 Defaults to VHE/nVHE based on hardware support. Setting
2529 mode to "protected" will disable kexec and hibernation
2532 kvm-arm.vgic_v3_group0_trap=
2533 [KVM,ARM] Trap guest accesses to GICv3 group-0
2536 kvm-arm.vgic_v3_group1_trap=
2537 [KVM,ARM] Trap guest accesses to GICv3 group-1
2540 kvm-arm.vgic_v3_common_trap=
2541 [KVM,ARM] Trap guest accesses to GICv3 common
2544 kvm-arm.vgic_v4_enable=
2545 [KVM,ARM] Allow use of GICv4 for direct injection of
2548 kvm_cma_resv_ratio=n [PPC]
2549 Reserves given percentage from system memory area for
2550 contiguous memory allocation for KVM hash pagetable
2552 By default it reserves 5% of total system memory.
2556 kvm-intel.ept= [KVM,Intel] Disable extended page tables
2557 (virtualized MMU) support on capable Intel chips.
2558 Default is 1 (enabled)
2560 kvm-intel.emulate_invalid_guest_state=
2561 [KVM,Intel] Disable emulation of invalid guest state.
2562 Ignored if kvm-intel.enable_unrestricted_guest=1, as
2563 guest state is never invalid for unrestricted guests.
2564 This param doesn't apply to nested guests (L2), as KVM
2565 never emulates invalid L2 guest state.
2566 Default is 1 (enabled)
2568 kvm-intel.flexpriority=
2569 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
2570 Default is 1 (enabled)
2573 [KVM,Intel] Enable VMX nesting (nVMX).
2574 Default is 0 (disabled)
2576 kvm-intel.unrestricted_guest=
2577 [KVM,Intel] Disable unrestricted guest feature
2578 (virtualized real and unpaged mode) on capable
2579 Intel chips. Default is 1 (enabled)
2581 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault
2584 Valid arguments: never, cond, always
2586 always: L1D cache flush on every VMENTER.
2587 cond: Flush L1D on VMENTER only when the code between
2588 VMEXIT and VMENTER can leak host memory.
2589 never: Disables the mitigation
2591 Default is cond (do L1 cache flush in specific instances)
2593 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
2594 feature (tagged TLBs) on capable Intel chips.
2595 Default is 1 (enabled)
2597 l1d_flush= [X86,INTEL]
2598 Control mitigation for L1D based snooping vulnerability.
2600 Certain CPUs are vulnerable to an exploit against CPU
2601 internal buffers which can forward information to a
2602 disclosure gadget under certain conditions.
2604 In vulnerable processors, the speculatively
2605 forwarded data can be used in a cache side channel
2606 attack, to access data to which the attacker does
2607 not have direct access.
2609 This parameter controls the mitigation. The
2612 on - enable the interface for the mitigation
2614 l1tf= [X86] Control mitigation of the L1TF vulnerability on
2617 The kernel PTE inversion protection is unconditionally
2618 enabled and cannot be disabled.
2621 Provides all available mitigations for the
2622 L1TF vulnerability. Disables SMT and
2623 enables all mitigations in the
2624 hypervisors, i.e. unconditional L1D flush.
2626 SMT control and L1D flush control via the
2627 sysfs interface is still possible after
2628 boot. Hypervisors will issue a warning
2629 when the first VM is started in a
2630 potentially insecure configuration,
2631 i.e. SMT enabled or L1D flush disabled.
2634 Same as 'full', but disables SMT and L1D
2635 flush runtime control. Implies the
2636 'nosmt=force' command line option.
2637 (i.e. sysfs control of SMT is disabled.)
2640 Leaves SMT enabled and enables the default
2641 hypervisor mitigation, i.e. conditional
2644 SMT control and L1D flush control via the
2645 sysfs interface is still possible after
2646 boot. Hypervisors will issue a warning
2647 when the first VM is started in a
2648 potentially insecure configuration,
2649 i.e. SMT enabled or L1D flush disabled.
2653 Disables SMT and enables the default
2654 hypervisor mitigation.
2656 SMT control and L1D flush control via the
2657 sysfs interface is still possible after
2658 boot. Hypervisors will issue a warning
2659 when the first VM is started in a
2660 potentially insecure configuration,
2661 i.e. SMT enabled or L1D flush disabled.
2664 Same as 'flush', but hypervisors will not
2665 warn when a VM is started in a potentially
2666 insecure configuration.
2669 Disables hypervisor mitigations and doesn't
2671 It also drops the swap size and available
2672 RAM limit restriction on both hypervisor and
2677 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst
2683 lapic [X86-32,APIC] Enable the local APIC even if BIOS
2686 lapic= [X86,APIC] Do not use TSC deadline
2687 value for LAPIC timer one-shot implementation. Default
2688 back to the programmable timer unit in the LAPIC.
2689 Format: notscdeadline
2691 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
2694 libata.dma= [LIBATA] DMA control
2695 libata.dma=0 Disable all PATA and SATA DMA
2696 libata.dma=1 PATA and SATA Disk DMA only
2697 libata.dma=2 ATAPI (CDROM) DMA only
2698 libata.dma=4 Compact Flash DMA only
2699 Combinations also work, so libata.dma=3 enables DMA
2700 for disks and CDROMs, but not CFs.
2702 libata.ignore_hpa= [LIBATA] Ignore HPA limit
2703 libata.ignore_hpa=0 keep BIOS limits (default)
2704 libata.ignore_hpa=1 ignore limits, using full disk
2706 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
2710 libata.force= [LIBATA] Force configurations. The format is a comma-
2711 separated list of "[ID:]VAL" where ID is PORT[.DEVICE].
2712 PORT and DEVICE are decimal numbers matching port, link
2713 or device. Basically, it matches the ATA ID string
2714 printed on console by libata. If the whole ID part is
2715 omitted, the last PORT and DEVICE values are used. If
2716 ID hasn't been specified yet, the configuration applies
2717 to all ports, links and devices.
2719 If only DEVICE is omitted, the parameter applies to
2720 the port and all links and devices behind it. DEVICE
2721 number of 0 either selects the first device or the
2722 first fan-out link behind PMP device. It does not
2723 select the host link. DEVICE number of 15 selects the
2724 host link and device attached to it.
2726 The VAL specifies the configuration to force. As long
2727 as there is no ambiguity, shortcut notation is allowed.
2728 For example, both 1.5 and 1.5G would work for 1.5Gbps.
2729 The following configurations can be forced.
2731 * Cable type: 40c, 80c, short40c, unk, ign or sata.
2732 Any ID with matching PORT is used.
2734 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
2736 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
2737 udma[/][16,25,33,44,66,100,133] notation is also
2740 * nohrst, nosrst, norst: suppress hard, soft and both
2743 * rstonce: only attempt one reset during hot-unplug
2746 * [no]dbdelay: Enable or disable the extra 200ms delay
2747 before debouncing a link PHY and device presence
2750 * [no]ncq: Turn on or off NCQ.
2752 * [no]ncqtrim: Enable or disable queued DSM TRIM.
2754 * [no]ncqati: Enable or disable NCQ trim on ATI chipset.
2756 * [no]trim: Enable or disable (unqueued) TRIM.
2758 * trim_zero: Indicate that TRIM command zeroes data.
2760 * max_trim_128m: Set 128M maximum trim size limit.
2762 * [no]dma: Turn on or off DMA transfers.
2764 * atapi_dmadir: Enable ATAPI DMADIR bridge support.
2766 * atapi_mod16_dma: Enable the use of ATAPI DMA for
2767 commands that are not a multiple of 16 bytes.
2769 * [no]dmalog: Enable or disable the use of the
2770 READ LOG DMA EXT command to access logs.
2772 * [no]iddevlog: Enable or disable access to the
2773 identify device data log.
2775 * [no]logdir: Enable or disable access to the general
2776 purpose log directory.
2778 * max_sec_128: Set transfer size limit to 128 sectors.
2780 * max_sec_1024: Set or clear transfer size limit to
2783 * max_sec_lba48: Set or clear transfer size limit to
2786 * [no]lpm: Enable or disable link power management.
2788 * [no]setxfer: Indicate if transfer speed mode setting
2791 * [no]fua: Disable or enable FUA (Force Unit Access)
2792 support for devices supporting this feature.
2794 * dump_id: Dump IDENTIFY data.
2796 * disable: Disable this device.
2798 If there are multiple matching configurations changing
2799 the same attribute, the last one is used.
2801 load_ramdisk= [RAM] [Deprecated]
2803 lockd.nlm_grace_period=P [NFS] Assign grace period.
2806 lockd.nlm_tcpport=N [NFS] Assign TCP port.
2809 lockd.nlm_timeout=T [NFS] Assign timeout value.
2812 lockd.nlm_udpport=M [NFS] Assign UDP port.
2815 lockdown= [SECURITY]
2816 { integrity | confidentiality }
2817 Enable the kernel lockdown feature. If set to
2818 integrity, kernel features that allow userland to
2819 modify the running kernel are disabled. If set to
2820 confidentiality, kernel features that allow userland
2821 to extract confidential information from the kernel
2824 locktorture.nreaders_stress= [KNL]
2825 Set the number of locking read-acquisition kthreads.
2826 Defaults to being automatically set based on the
2827 number of online CPUs.
2829 locktorture.nwriters_stress= [KNL]
2830 Set the number of locking write-acquisition kthreads.
2832 locktorture.onoff_holdoff= [KNL]
2833 Set time (s) after boot for CPU-hotplug testing.
2835 locktorture.onoff_interval= [KNL]
2836 Set time (s) between CPU-hotplug operations, or
2837 zero to disable CPU-hotplug testing.
2839 locktorture.shuffle_interval= [KNL]
2840 Set task-shuffle interval (jiffies). Shuffling
2841 tasks allows some CPUs to go into dyntick-idle
2842 mode during the locktorture test.
2844 locktorture.shutdown_secs= [KNL]
2845 Set time (s) after boot system shutdown. This
2846 is useful for hands-off automated testing.
2848 locktorture.stat_interval= [KNL]
2849 Time (s) between statistics printk()s.
2851 locktorture.stutter= [KNL]
2852 Time (s) to stutter testing, for example,
2853 specifying five seconds causes the test to run for
2854 five seconds, wait for five seconds, and so on.
2855 This tests the locking primitive's ability to
2856 transition abruptly to and from idle.
2858 locktorture.torture_type= [KNL]
2859 Specify the locking implementation to test.
2861 locktorture.verbose= [KNL]
2862 Enable additional printk() statements.
2864 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
2867 loglevel= All Kernel Messages with a loglevel smaller than the
2868 console loglevel will be printed to the console. It can
2869 also be changed with klogd or other programs. The
2870 loglevels are defined as follows:
2872 0 (KERN_EMERG) system is unusable
2873 1 (KERN_ALERT) action must be taken immediately
2874 2 (KERN_CRIT) critical conditions
2875 3 (KERN_ERR) error conditions
2876 4 (KERN_WARNING) warning conditions
2877 5 (KERN_NOTICE) normal but significant condition
2878 6 (KERN_INFO) informational
2879 7 (KERN_DEBUG) debug-level messages
2881 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
2882 in bytes. n must be a power of two and greater
2883 than the minimal size. The minimal size is defined
2884 by LOG_BUF_SHIFT kernel config parameter. There is
2885 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
2886 that allows to increase the default size depending on
2887 the number of CPUs. See init/Kconfig for more details.
2889 logo.nologo [FB] Disables display of the built-in Linux logo.
2890 This may be used to provide more screen space for
2891 kernel log messages and is useful when debugging
2892 kernel boot problems.
2894 lp=0 [LP] Specify parallel ports to use, e.g,
2895 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
2896 lp=reset first parallel port). 'lp=0' disables the
2897 lp=auto printer driver. 'lp=reset' (which can be
2898 specified in addition to the ports) causes
2899 attached printers to be reset. Using
2900 lp=port1,port2,... specifies the parallel ports
2901 to associate lp devices with, starting with
2902 lp0. A port specification may be 'none' to skip
2903 that lp device, or a parport name such as
2904 'parport0'. Specifying 'lp=auto' instead of a
2905 port specification list means that device IDs
2906 from each port should be examined, to see if
2907 an IEEE 1284-compliant printer is attached; if
2908 so, the driver will manage that printer.
2909 See also header of drivers/char/lp.c.
2912 Sets loops_per_jiffy to given constant, thus avoiding
2913 time-consuming boot-time autodetection (up to 250 ms per
2914 CPU). 0 enables autodetection (default). To determine
2915 the correct value for your kernel, boot with normal
2916 autodetection and see what value is printed. Note that
2917 on SMP systems the preset will be applied to all CPUs,
2918 which is likely to cause problems if your CPUs need
2919 significantly divergent settings. An incorrect value
2920 will cause delays in the kernel to be wrong, leading to
2921 unpredictable I/O errors and other breakage. Although
2922 unlikely, in the extreme case this might damage your
2926 Format: <io>,<irq>,<dma>
2928 lsm.debug [SECURITY] Enable LSM initialization debugging output.
2931 [SECURITY] Choose order of LSM initialization. This
2932 overrides CONFIG_LSM, and the "security=" parameter.
2934 machvec= [IA-64] Force the use of a particular machine-vector
2935 (machvec) in a generic kernel.
2936 Example: machvec=hpzx1
2938 machtype= [Loongson] Share the same kernel image file between
2939 different yeeloong laptops.
2940 Example: machtype=lemote-yeeloong-2f-7inch
2942 max_addr=nn[KMG] [KNL,BOOT,IA-64] All physical memory greater
2943 than or equal to this physical address is ignored.
2945 maxcpus= [SMP] Maximum number of processors that an SMP kernel
2946 will bring up during bootup. maxcpus=n : n >= 0 limits
2947 the kernel to bring up 'n' processors. Surely after
2948 bootup you can bring up the other plugged cpu by executing
2949 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus
2950 only takes effect during system bootup.
2951 While n=0 is a special case, it is equivalent to "nosmp",
2952 which also disables the IO APIC.
2954 max_loop= [LOOP] The number of loop block devices that get
2955 (loop.max_loop) unconditionally pre-created at init time. The default
2956 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
2957 of statically allocating a predefined number, loop
2958 devices can be requested on-demand with the
2959 /dev/loop-control interface.
2961 mce [X86-32] Machine Check Exception
2963 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.rst
2965 md= [HW] RAID subsystems devices and level
2966 See Documentation/admin-guide/md.rst.
2969 Format: <first>,<last>
2970 Specifies range of consoles to be captured by the MDA.
2973 Control mitigation for the Micro-architectural Data
2974 Sampling (MDS) vulnerability.
2976 Certain CPUs are vulnerable to an exploit against CPU
2977 internal buffers which can forward information to a
2978 disclosure gadget under certain conditions.
2980 In vulnerable processors, the speculatively
2981 forwarded data can be used in a cache side channel
2982 attack, to access data to which the attacker does
2983 not have direct access.
2985 This parameter controls the MDS mitigation. The
2988 full - Enable MDS mitigation on vulnerable CPUs
2989 full,nosmt - Enable MDS mitigation and disable
2990 SMT on vulnerable CPUs
2991 off - Unconditionally disable MDS mitigation
2993 On TAA-affected machines, mds=off can be prevented by
2994 an active TAA mitigation as both vulnerabilities are
2995 mitigated with the same mechanism so in order to disable
2996 this mitigation, you need to specify tsx_async_abort=off
2999 Not specifying this option is equivalent to
3002 For details see: Documentation/admin-guide/hw-vuln/mds.rst
3004 mem=nn[KMG] [HEXAGON] Set the memory size.
3005 Must be specified, otherwise memory size will be 0.
3007 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
3008 Amount of memory to be used in cases as follows:
3011 2 when the kernel is not able to see the whole system memory;
3012 3 memory that lies after 'mem=' boundary is excluded from
3013 the hypervisor, then assigned to KVM guests.
3014 4 to limit the memory available for kdump kernel.
3016 [ARC,MICROBLAZE] - the limit applies only to low memory,
3017 high memory is not affected.
3019 [ARM64] - only limits memory covered by the linear
3020 mapping. The NOMAP regions are not affected.
3022 [X86] Work as limiting max address. Use together
3023 with memmap= to avoid physical address space collisions.
3024 Without memmap= PCI devices could be placed at addresses
3025 belonging to unused RAM.
3027 Note that this only takes effects during boot time since
3028 in above case 3, memory may need be hot added after boot
3029 if system memory of hypervisor is not sufficient.
3032 [ARM,MIPS] - override the memory layout reported by
3034 Define a memory region of size nn[KMG] starting at
3036 Multiple different regions can be specified with
3037 multiple mem= parameters on the command line.
3039 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
3042 memblock=debug [KNL] Enable memblock debug messages.
3045 [KNL,SH] Allow user to override the default size for
3046 per-device physically contiguous DMA buffers.
3048 memhp_default_state=online/offline
3049 [KNL] Set the initial state for the memory hotplug
3050 onlining policy. If not specified, the default value is
3051 set according to the
3052 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config
3054 See Documentation/admin-guide/mm/memory-hotplug.rst.
3056 memmap=exactmap [KNL,X86] Enable setting of an exact
3057 E820 memory map, as specified by the user.
3058 Such memmap=exactmap lines can be constructed based on
3059 BIOS output or other requirements. See the memmap=nn@ss
3062 memmap=nn[KMG]@ss[KMG]
3063 [KNL, X86, MIPS, XTENSA] Force usage of a specific region of memory.
3064 Region of memory to be used is from ss to ss+nn.
3065 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG],
3066 which limits max address to nn[KMG].
3067 Multiple different regions can be specified,
3070 memmap=100M@2G,100M#3G,1G!1024G
3072 memmap=nn[KMG]#ss[KMG]
3073 [KNL,ACPI] Mark specific memory as ACPI data.
3074 Region of memory to be marked is from ss to ss+nn.
3076 memmap=nn[KMG]$ss[KMG]
3077 [KNL,ACPI] Mark specific memory as reserved.
3078 Region of memory to be reserved is from ss to ss+nn.
3079 Example: Exclude memory from 0x18690000-0x1869ffff
3080 memmap=64K$0x18690000
3082 memmap=0x10000$0x18690000
3083 Some bootloaders may need an escape character before '$',
3084 like Grub2, otherwise '$' and the following number
3087 memmap=nn[KMG]!ss[KMG]
3088 [KNL,X86] Mark specific memory as protected.
3089 Region of memory to be used, from ss to ss+nn.
3090 The memory region may be marked as e820 type 12 (0xc)
3091 and is NVDIMM or ADR memory.
3093 memmap=<size>%<offset>-<oldtype>+<newtype>
3094 [KNL,ACPI] Convert memory within the specified region
3095 from <oldtype> to <newtype>. If "-<oldtype>" is left
3096 out, the whole region will be marked as <newtype>,
3097 even if previously unavailable. If "+<newtype>" is left
3098 out, matching memory will be removed. Types are
3099 specified as e820 types, e.g., 1 = RAM, 2 = reserved,
3100 3 = ACPI, 12 = PRAM.
3102 memory_corruption_check=0/1 [X86]
3103 Some BIOSes seem to corrupt the first 64k of
3104 memory when doing things like suspend/resume.
3105 Setting this option will scan the memory
3106 looking for corruption. Enabling this will
3107 both detect corruption and prevent the kernel
3108 from using the memory being corrupted.
3109 However, its intended as a diagnostic tool; if
3110 repeatable BIOS-originated corruption always
3111 affects the same memory, you can use memmap=
3112 to prevent the kernel from using that memory.
3114 memory_corruption_check_size=size [X86]
3115 By default it checks for corruption in the low
3116 64k, making this memory unavailable for normal
3117 use. Use this parameter to scan for
3118 corruption in more or less memory.
3120 memory_corruption_check_period=seconds [X86]
3121 By default it checks for corruption every 60
3122 seconds. Use this parameter to check at some
3123 other rate. 0 disables periodic checking.
3125 memory_hotplug.memmap_on_memory
3126 [KNL,X86,ARM] Boolean flag to enable this feature.
3127 Format: {on | off (default)}
3128 When enabled, runtime hotplugged memory will
3129 allocate its internal metadata (struct pages,
3130 those vmemmap pages cannot be optimized even
3131 if hugetlb_free_vmemmap is enabled) from the
3132 hotadded memory which will allow to hotadd a
3133 lot of memory without requiring additional
3135 This feature is disabled by default because it
3136 has some implication on large (e.g. GB)
3137 allocations in some configurations (e.g. small
3139 The state of the flag can be read in
3140 /sys/module/memory_hotplug/parameters/memmap_on_memory.
3141 Note that even when enabled, there are a few cases where
3142 the feature is not effective.
3144 memtest= [KNL,X86,ARM,M68K,PPC,RISCV] Enable memtest
3146 default : 0 <disable>
3147 Specifies the number of memtest passes to be
3148 performed. Each pass selects another test
3149 pattern from a given set of patterns. Memtest
3150 fills the memory with this pattern, validates
3151 memory contents and reserves bad memory
3152 regions that are detected.
3154 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control
3155 Valid arguments: on, off
3156 Default (depends on kernel configuration option):
3157 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y)
3158 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n)
3159 mem_encrypt=on: Activate SME
3160 mem_encrypt=off: Do not activate SME
3162 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst
3163 for details on when memory encryption can be activated.
3165 mem_sleep_default= [SUSPEND] Default system suspend mode:
3166 s2idle - Suspend-To-Idle
3167 shallow - Power-On Suspend or equivalent (if supported)
3168 deep - Suspend-To-RAM or equivalent (if supported)
3169 See Documentation/admin-guide/pm/sleep-states.rst.
3171 meye.*= [HW] Set MotionEye Camera parameters
3172 See Documentation/admin-guide/media/meye.rst.
3174 mfgpt_irq= [IA-32] Specify the IRQ to use for the
3175 Multi-Function General Purpose Timers on AMD Geode
3178 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
3179 the BIOS has incorrectly applied a workaround. TinyBIOS
3180 version 0.98 is known to be affected, 0.99 fixes the
3181 problem by letting the user disable the workaround.
3185 min_addr=nn[KMG] [KNL,BOOT,IA-64] All physical memory below this
3186 physical address is ignored.
3188 mini2440= [ARM,HW,KNL]
3189 Format:[0..2][b][c][t]
3191 MINI2440 configuration specification:
3192 0 - The attached screen is the 3.5" TFT
3193 1 - The attached screen is the 7" TFT
3194 2 - The VGA Shield is attached (1024x768)
3195 Leaving out the screen size parameter will not load
3196 the TFT driver, and the framebuffer will be left
3198 b - Enable backlight. The TFT backlight pin will be
3199 linked to the kernel VESA blanking code and a GPIO
3200 LED. This parameter is not necessary when using the
3202 c - Enable the s3c camera interface.
3203 t - Reserved for enabling touchscreen support. The
3204 touchscreen support is not enabled in the mainstream
3205 kernel as of 2.6.30, a preliminary port can be found
3206 in the "bleeding edge" mini2440 support kernel at
3207 https://repo.or.cz/w/linux-2.6/mini2440.git
3210 [X86,PPC,S390,ARM64] Control optional mitigations for
3211 CPU vulnerabilities. This is a set of curated,
3212 arch-independent options, each of which is an
3213 aggregation of existing arch-specific options.
3216 Disable all optional CPU mitigations. This
3217 improves system performance, but it may also
3218 expose users to several CPU vulnerabilities.
3219 Equivalent to: nopti [X86,PPC]
3220 if nokaslr then kpti=0 [ARM64]
3221 nospectre_v1 [X86,PPC]
3223 nospectre_v2 [X86,PPC,S390,ARM64]
3224 spectre_v2_user=off [X86]
3225 spec_store_bypass_disable=off [X86,PPC]
3226 ssbd=force-off [ARM64]
3227 nospectre_bhb [ARM64]
3230 tsx_async_abort=off [X86]
3231 kvm.nx_huge_pages=off [X86]
3232 srbds=off [X86,INTEL]
3233 no_entry_flush [PPC]
3234 no_uaccess_flush [PPC]
3235 mmio_stale_data=off [X86]
3239 This does not have any effect on
3240 kvm.nx_huge_pages when
3241 kvm.nx_huge_pages=force.
3244 Mitigate all CPU vulnerabilities, but leave SMT
3245 enabled, even if it's vulnerable. This is for
3246 users who don't want to be surprised by SMT
3247 getting disabled across kernel upgrades, or who
3248 have other ways of avoiding SMT-based attacks.
3249 Equivalent to: (default behavior)
3252 Mitigate all CPU vulnerabilities, disabling SMT
3253 if needed. This is for users who always want to
3254 be fully mitigated, even if it means losing SMT.
3255 Equivalent to: l1tf=flush,nosmt [X86]
3256 mds=full,nosmt [X86]
3257 tsx_async_abort=full,nosmt [X86]
3258 mmio_stale_data=full,nosmt [X86]
3259 retbleed=auto,nosmt [X86]
3262 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
3263 parameter allows control of the logging verbosity for
3264 the additional memory initialisation checks. A value
3265 of 0 disables mminit logging and a level of 4 will
3266 log everything. Information is printed at KERN_DEBUG
3267 so loglevel=8 may also need to be specified.
3270 [X86,INTEL] Control mitigation for the Processor
3271 MMIO Stale Data vulnerabilities.
3273 Processor MMIO Stale Data is a class of
3274 vulnerabilities that may expose data after an MMIO
3275 operation. Exposed data could originate or end in
3276 the same CPU buffers as affected by MDS and TAA.
3277 Therefore, similar to MDS and TAA, the mitigation
3278 is to clear the affected CPU buffers.
3280 This parameter controls the mitigation. The
3283 full - Enable mitigation on vulnerable CPUs
3285 full,nosmt - Enable mitigation and disable SMT on
3288 off - Unconditionally disable mitigation
3290 On MDS or TAA affected machines,
3291 mmio_stale_data=off can be prevented by an active
3292 MDS or TAA mitigation as these vulnerabilities are
3293 mitigated with the same mechanism so in order to
3294 disable this mitigation, you need to specify
3295 mds=off and tsx_async_abort=off too.
3297 Not specifying this option is equivalent to
3298 mmio_stale_data=full.
3301 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst
3303 <module>.async_probe[=<bool>] [KNL]
3304 If no <bool> value is specified or if the value
3305 specified is not a valid <bool>, enable asynchronous
3306 probe on this module. Otherwise, enable/disable
3307 asynchronous probe on this module as indicated by the
3308 <bool> value. See also: module.async_probe
3310 module.async_probe=<bool>
3311 [KNL] When set to true, modules will use async probing
3312 by default. To enable/disable async probing for a
3313 specific module, use the module specific control that
3314 is documented under <module>.async_probe. When both
3315 module.async_probe and <module>.async_probe are
3316 specified, <module>.async_probe takes precedence for
3317 the specific module.
3320 [KNL] When CONFIG_MODULE_SIG is set, this means that
3321 modules without (valid) signatures will fail to load.
3322 Note that if CONFIG_MODULE_SIG_FORCE is set, that
3323 is always true, so this option does nothing.
3325 module_blacklist= [KNL] Do not load a comma-separated list of
3326 modules. Useful for debugging problem modules.
3329 [MOUSE] Maximum time between finger touching and
3330 leaving touchpad surface for touch to be considered
3331 a tap and be reported as a left button click (for
3332 touchpads working in absolute mode only).
3334 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
3335 reporting absolute coordinates, such as tablets
3336 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
3337 reporting absolute coordinates, such as tablets
3339 movablecore= [KNL,X86,IA-64,PPC]
3340 Format: nn[KMGTPE] | nn%
3341 This parameter is the complement to kernelcore=, it
3342 specifies the amount of memory used for migratable
3343 allocations. If both kernelcore and movablecore is
3344 specified, then kernelcore will be at *least* the
3345 specified value but may be more. If movablecore on its
3346 own is specified, the administrator must be careful
3347 that the amount of memory usable for all allocations
3350 movable_node [KNL] Boot-time switch to make hotplugable memory
3351 NUMA nodes to be movable. This means that the memory
3352 of such nodes will be usable only for movable
3353 allocations which rules out almost all kernel
3354 allocations. Use with caution!
3356 MTD_Partition= [MTD]
3357 Format: <name>,<region-number>,<size>,<offset>
3359 MTD_Region= [MTD] Format:
3360 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
3363 See drivers/mtd/parsers/cmdlinepart.c
3366 ARM/S3C2412 JIVE boot control
3368 See arch/arm/mach-s3c/mach-jive.c
3370 mtouchusb.raw_coordinates=
3371 [HW] Make the MicroTouch USB driver use raw coordinates
3372 ('y', default) or cooked coordinates ('n')
3374 mtrr_chunk_size=nn[KMG] [X86]
3375 used for mtrr cleanup. It is largest continuous chunk
3376 that could hold holes aka. UC entries.
3378 mtrr_gran_size=nn[KMG] [X86]
3379 Used for mtrr cleanup. It is granularity of mtrr block.
3381 Large value could prevent small alignment from
3384 mtrr_spare_reg_nr=n [X86]
3386 Range: 0,7 : spare reg number
3388 Used for mtrr cleanup. It is spare mtrr entries number.
3389 Set to 2 or more if your graphical card needs more.
3391 multitce=off [PPC] This parameter disables the use of the pSeries
3392 firmware feature for updating multiple TCE entries
3395 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
3397 netdev= [NET] Network devices parameters
3398 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
3399 Note that mem_start is often overloaded to mean
3400 something different and driver-specific.
3401 This usage is only documented in each driver source
3404 netpoll.carrier_timeout=
3405 [NET] Specifies amount of time (in seconds) that
3406 netpoll should wait for a carrier. By default netpoll
3410 [NETFILTER] Enable connection tracking flow accounting
3411 0 to disable accounting
3412 1 to enable accounting
3415 nfsaddrs= [NFS] Deprecated. Use ip= instead.
3416 See Documentation/admin-guide/nfs/nfsroot.rst.
3418 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
3419 See Documentation/admin-guide/nfs/nfsroot.rst.
3421 nfsrootdebug [NFS] enable nfsroot debugging messages.
3422 See Documentation/admin-guide/nfs/nfsroot.rst.
3424 nfs.callback_nr_threads=
3425 [NFSv4] set the total number of threads that the
3426 NFS client will assign to service NFSv4 callback
3429 nfs.callback_tcpport=
3430 [NFS] set the TCP port on which the NFSv4 callback
3431 channel should listen.
3434 [NFS] sets the pathname to the program which is used
3435 to update the NFS client cache entries.
3437 nfs.cache_getent_timeout=
3438 [NFS] sets the timeout after which an attempt to
3439 update a cache entry is deemed to have failed.
3441 nfs.idmap_cache_timeout=
3442 [NFS] set the maximum lifetime for idmapper cache
3446 [NFS] enable 64-bit inode numbers.
3447 If zero, the NFS client will fake up a 32-bit inode
3448 number for the readdir() and stat() syscalls instead
3449 of returning the full 64-bit number.
3450 The default is to return 64-bit inode numbers.
3452 nfs.max_session_cb_slots=
3453 [NFSv4.1] Sets the maximum number of session
3454 slots the client will assign to the callback
3455 channel. This determines the maximum number of
3456 callbacks the client will process in parallel for
3457 a particular server.
3459 nfs.max_session_slots=
3460 [NFSv4.1] Sets the maximum number of session slots
3461 the client will attempt to negotiate with the server.
3462 This limits the number of simultaneous RPC requests
3463 that the client can send to the NFSv4.1 server.
3464 Note that there is little point in setting this
3465 value higher than the max_tcp_slot_table_limit.
3467 nfs.nfs4_disable_idmapping=
3468 [NFSv4] When set to the default of '1', this option
3469 ensures that both the RPC level authentication
3470 scheme and the NFS level operations agree to use
3471 numeric uids/gids if the mount is using the
3472 'sec=sys' security flavour. In effect it is
3473 disabling idmapping, which can make migration from
3474 legacy NFSv2/v3 systems to NFSv4 easier.
3475 Servers that do not support this mode of operation
3476 will be autodetected by the client, and it will fall
3477 back to using the idmapper.
3478 To turn off this behaviour, set the value to '0'.
3480 [NFS4] Specify an additional fixed unique ident-
3481 ification string that NFSv4 clients can insert into
3482 their nfs_client_id4 string. This is typically a
3483 UUID that is generated at system install time.
3485 nfs.send_implementation_id =
3486 [NFSv4.1] Send client implementation identification
3487 information in exchange_id requests.
3488 If zero, no implementation identification information
3490 The default is to send the implementation identification
3493 nfs.recover_lost_locks =
3494 [NFSv4] Attempt to recover locks that were lost due
3495 to a lease timeout on the server. Please note that
3496 doing this risks data corruption, since there are
3497 no guarantees that the file will remain unchanged
3498 after the locks are lost.
3499 If you want to enable the kernel legacy behaviour of
3500 attempting to recover these locks, then set this
3502 The default parameter value of '0' causes the kernel
3503 not to attempt recovery of lost locks.
3505 nfs4.layoutstats_timer =
3506 [NFSv4.2] Change the rate at which the kernel sends
3507 layoutstats to the pNFS metadata server.
3509 Setting this to value to 0 causes the kernel to use
3510 whatever value is the default set by the layout
3511 driver. A non-zero value sets the minimum interval
3512 in seconds between layoutstats transmissions.
3514 nfsd.inter_copy_offload_enable =
3515 [NFSv4.2] When set to 1, the server will support
3516 server-to-server copies for which this server is
3517 the destination of the copy.
3519 nfsd.nfsd4_ssc_umount_timeout =
3520 [NFSv4.2] When used as the destination of a
3521 server-to-server copy, knfsd temporarily mounts
3522 the source server. It caches the mount in case
3523 it will be needed again, and discards it if not
3524 used for the number of milliseconds specified by
3527 nfsd.nfs4_disable_idmapping=
3528 [NFSv4] When set to the default of '1', the NFSv4
3529 server will return only numeric uids and gids to
3530 clients using auth_sys, and will accept numeric uids
3531 and gids from such clients. This is intended to ease
3532 migration from NFSv2/v3.
3535 nmi_backtrace.backtrace_idle [KNL]
3536 Dump stacks even of idle CPUs in response to an
3537 NMI stack-backtrace request.
3539 nmi_debug= [KNL,SH] Specify one or more actions to take
3540 when a NMI is triggered.
3541 Format: [state][,regs][,debounce][,die]
3543 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
3544 Format: [panic,][nopanic,][num]
3546 0 - turn hardlockup detector in nmi_watchdog off
3547 1 - turn hardlockup detector in nmi_watchdog on
3548 When panic is specified, panic when an NMI watchdog
3549 timeout occurs (or 'nopanic' to not panic on an NMI
3550 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set)
3551 To disable both hard and soft lockup detectors,
3552 please see 'nowatchdog'.
3553 This is useful when you use a panic=... timeout and
3554 need the box quickly up again.
3556 These settings can be accessed at runtime via
3557 the nmi_watchdog and hardlockup_panic sysctls.
3559 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
3560 emulation library even if a 387 maths coprocessor
3563 no5lvl [X86-64] Disable 5-level paging mode. Forces
3564 kernel to use 4-level paging instead.
3566 nofsgsbase [X86] Disables FSGSBASE instructions.
3569 [HW] Never suspend the console
3570 Disable suspending of consoles during suspend and
3571 hibernate operations. Once disabled, debugging
3572 messages can reach various consoles while the rest
3573 of the system is being put to sleep (ie, while
3574 debugging driver suspend/resume hooks). This may
3575 not work reliably with all consoles, but is known
3576 to work with serial and VGA consoles.
3577 To facilitate more flexible debugging, we also add
3578 console_suspend, a printk module parameter to control
3579 it. Users could use console_suspend (usually
3580 /sys/module/printk/parameters/console_suspend) to
3581 turn on/off it dynamically.
3583 novmcoredd [KNL,KDUMP]
3584 Disable device dump. Device dump allows drivers to
3585 append dump data to vmcore so you can collect driver
3586 specified debug info. Drivers can append the data
3587 without any limit and this data is stored in memory,
3588 so this may cause significant memory stress. Disabling
3589 device dump can help save memory but the driver debug
3590 data will be no longer available. This parameter
3591 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP
3594 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
3595 caches in the slab allocator. Saves per-node memory,
3596 but will impact performance.
3600 noaltinstr [S390] Disables alternative instructions patching
3601 (CPU alternatives feature).
3603 noapic [SMP,APIC] Tells the kernel to not make use of any
3604 IOAPICs that may be present in the system.
3606 noautogroup Disable scheduler automatic task group creation.
3610 nodsp [SH] Disable hardware DSP at boot time.
3612 noefi Disable EFI runtime services support.
3614 no_entry_flush [PPC] Don't flush the L1-D cache when entering the kernel.
3619 Disable SMAP (Supervisor Mode Access Prevention)
3620 even if it is supported by processor.
3623 Disable SMEP (Supervisor Mode Execution Prevention)
3624 even if it is supported by processor.
3627 This affects only 32-bit executables.
3628 noexec32=on: enable non-executable mappings (default)
3629 read doesn't imply executable mappings
3630 noexec32=off: disable non-executable mappings
3631 read implies executable mappings
3633 nofpu [MIPS,SH] Disable hardware FPU at boot time.
3635 nofxsr [BUGS=X86-32] Disables x86 floating point extended
3636 register save and restore. The kernel will only save
3637 legacy floating-point registers on task switch.
3639 nohugeiomap [KNL,X86,PPC,ARM64] Disable kernel huge I/O mappings.
3641 nohugevmalloc [KNL,X86,PPC,ARM64] Disable kernel huge vmalloc mappings.
3643 nosmt [KNL,S390] Disable symmetric multithreading (SMT).
3644 Equivalent to smt=1.
3646 [KNL,X86] Disable symmetric multithreading (SMT).
3647 nosmt=force: Force disable SMT, cannot be undone
3648 via the sysfs control file.
3650 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1
3651 (bounds check bypass). With this option data leaks are
3652 possible in the system.
3654 nospectre_v2 [X86,PPC_E500,ARM64] Disable all mitigations for
3655 the Spectre variant 2 (indirect branch prediction)
3656 vulnerability. System may allow data leaks with this
3659 nospectre_bhb [ARM64] Disable all mitigations for Spectre-BHB (branch
3660 history injection) vulnerability. System may allow data leaks
3663 nospec_store_bypass_disable
3664 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability
3667 [PPC] Don't flush the L1-D cache after accessing user data.
3669 noxsave [BUGS=X86] Disables x86 extended register state save
3670 and restore using xsave. The kernel will fallback to
3671 enabling legacy floating-point and sse state.
3673 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
3674 register states. The kernel will fall back to use
3675 xsave to save the states. By using this parameter,
3676 performance of saving the states is degraded because
3677 xsave doesn't support modified optimization while
3678 xsaveopt supports it on xsaveopt enabled systems.
3680 noxsaves [X86] Disables xsaves and xrstors used in saving and
3681 restoring x86 extended register state in compacted
3682 form of xsave area. The kernel will fall back to use
3683 xsaveopt and xrstor to save and restore the states
3684 in standard form of xsave area. By using this
3685 parameter, xsave area per process might occupy more
3686 memory on xsaves enabled systems.
3688 nohlt [ARM,ARM64,MICROBLAZE,SH] Forces the kernel to busy wait
3689 in do_idle() and not use the arch_cpu_idle()
3690 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP
3691 to be effective. This is useful on platforms where the
3692 sleep(SH) or wfi(ARM,ARM64) instructions do not work
3693 correctly or when doing power measurements to evaluate
3694 the impact of the sleep instructions. This is also
3695 useful when using JTAG debugger.
3697 no_file_caps Tells the kernel not to honor file capabilities. The
3698 only way then for a file to be executed with privilege
3699 is to be setuid root or executed by root.
3701 nohalt [IA-64] Tells the kernel not to use the power saving
3702 function PAL_HALT_LIGHT when idle. This increases
3703 power-consumption. On the positive side, it reduces
3704 interrupt wake-up latency, which may improve performance
3705 in certain environments such as networked servers or
3709 Force pointers printed to the console or buffers to be
3710 unhashed. By default, when a pointer is printed via %p
3711 format string, that pointer is "hashed", i.e. obscured
3712 by hashing the pointer value. This is a security feature
3713 that hides actual kernel addresses from unprivileged
3714 users, but it also makes debugging the kernel more
3715 difficult since unequal pointers can no longer be
3716 compared. However, if this command-line option is
3717 specified, then all normal pointers will have their true
3718 value printed. This option should only be specified when
3719 debugging the kernel. Please do not use on production
3722 nohibernate [HIBERNATION] Disable hibernation and resume.
3724 nohz= [KNL] Boottime enable/disable dynamic ticks
3725 Valid arguments: on, off
3728 nohz_full= [KNL,BOOT,SMP,ISOL]
3729 The argument is a cpu list, as described above.
3730 In kernels built with CONFIG_NO_HZ_FULL=y, set
3731 the specified list of CPUs whose tick will be stopped
3732 whenever possible. The boot CPU will be forced outside
3733 the range to maintain the timekeeping. Any CPUs
3734 in this list will have their RCU callbacks offloaded,
3735 just as if they had also been called out in the
3736 rcu_nocbs= boot parameter.
3738 Note that this argument takes precedence over
3739 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option.
3741 noiotrap [SH] Disables trapped I/O port accesses.
3743 noirqdebug [X86-32] Disables the code which attempts to detect and
3744 disable unhandled interrupt sources.
3746 no_timer_check [X86,APIC] Disables the code which tests for
3747 broken timer IRQ sources.
3749 noisapnp [ISAPNP] Disables ISA PnP code.
3751 noinitrd [RAM] Tells the kernel not to load any configured
3754 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
3756 [Deprecated - use intremap=off]
3760 noinvpcid [X86] Disable the INVPCID cpu feature.
3762 nojitter [IA-64] Disables jitter checking for ITC timers.
3765 When CONFIG_RANDOMIZE_BASE is set, this disables
3766 kernel and module base offset ASLR (Address Space
3767 Layout Randomization).
3769 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
3771 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
3775 [X86,PV_OPS] Disable paravirtualized VMware scheduler
3776 clock and use the default one.
3778 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES] Disable paravirtualized
3779 steal time accounting. steal time is computed, but
3780 won't influence scheduler behaviour
3782 nolapic [X86-32,APIC] Do not enable or use the local APIC.
3784 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
3786 nomca [IA-64] Disable machine check abort handling
3788 nomce [X86-32] Disable Machine Check Exception
3790 nomfgpt [X86-32] Disable Multi-Function General Purpose
3791 Timer usage (for AMD Geode machines).
3793 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
3794 shutdown the other cpus. Instead use the REBOOT_VECTOR
3797 nomodeset Disable kernel modesetting. Most systems' firmware
3798 sets up a display mode and provides framebuffer memory
3799 for output. With nomodeset, DRM and fbdev drivers will
3800 not load if they could possibly displace the pre-
3801 initialized output. Only the system framebuffer will
3802 be available for use. The respective drivers will not
3803 perform display-mode changes or accelerated rendering.
3805 Useful as error fallback, or for testing and debugging.
3807 nomodule Disable module load
3809 nopat [X86] Disable PAT (page attribute table extension of
3810 pagetables) support.
3812 nopcid [X86-64] Disable the PCID cpu feature.
3814 nopku [X86] Disable Memory Protection Keys CPU feature found
3817 nopv= [X86,XEN,KVM,HYPER_V,VMWARE]
3818 Disables the PV optimizations forcing the guest to run
3819 as generic guest with no PV drivers. Currently support
3820 XEN HVM, KVM, HYPER_V and VMWARE guest.
3822 nopvspin [X86,XEN,KVM]
3823 Disables the qspinlock slow path using PV optimizations
3824 which allow the hypervisor to 'idle' the guest on lock
3827 norandmaps Don't use address space randomization. Equivalent to
3828 echo 0 > /proc/sys/kernel/randomize_va_space
3830 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
3831 with UP alternatives
3833 noresume [SWSUSP] Disables resume and restores original swap
3836 no-scroll [VGA] Disables scrollback.
3837 This is required for the Braillex ib80-piezo Braille
3838 reader made by F.H. Papenmeier (Germany).
3842 nosgx [X86-64,SGX] Disables Intel SGX kernel support.
3844 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
3845 and disable the IO APIC. legacy for "maxcpus=0".
3847 nosoftlockup [KNL] Disable the soft-lockup detector.
3849 nosync [HW,M68K] Disables sync negotiation for all devices.
3851 nowatchdog [KNL] Disable both lockup detectors, i.e.
3852 soft-lockup and NMI watchdog (hard-lockup).
3856 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
3858 NOTE: this parameter will be ignored on systems with the
3859 LEGACY_XAPIC_DISABLED bit set in the
3860 IA32_XAPIC_DISABLE_STATUS MSR.
3862 nps_mtm_hs_ctr= [KNL,ARC]
3863 This parameter sets the maximum duration, in
3864 cycles, each HW thread of the CTOP can run
3865 without interruptions, before HW switches it.
3866 The actual maximum duration is 16 times this
3868 Format: integer between 1 and 255
3871 nptcg= [IA-64] Override max number of concurrent global TLB
3872 purges which is reported from either PAL_VM_SUMMARY or
3875 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
3876 could support. nr_cpus=n : n >= 1 limits the kernel to
3877 support 'n' processors. It could be larger than the
3878 number of already plugged CPU during bootup, later in
3879 runtime you can physically add extra cpu until it reaches
3880 n. So during boot up some boot time memory for per-cpu
3881 variables need be pre-allocated for later physical cpu
3884 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
3886 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86] Disable NUMA, Only
3887 set up a single NUMA node spanning all memory.
3889 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic
3891 Allowed values are enable and disable
3893 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
3894 'node', 'default' can be specified
3895 This can be set from sysctl after boot.
3896 See Documentation/admin-guide/sysctl/vm.rst for details.
3898 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
3899 See Documentation/core-api/debugging-via-ohci1394.rst for more
3902 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
3903 Rather than timing out after 20 ms if an EC
3904 command is not properly ACKed, override the length
3905 of the timeout. We have interrupts disabled while
3906 waiting for the ACK, so if this is set too high
3907 interrupts *may* be lost!
3909 omap_mux= [OMAP] Override bootloader pin multiplexing.
3910 Format: <mux_mode0.mode_name=value>...
3911 For example, to override I2C bus2:
3912 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
3914 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
3916 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
3918 boundary - index of last SLC block on Flex-OneNAND.
3919 The remaining blocks are configured as MLC blocks.
3920 lock - Configure if Flex-OneNAND boundary should be locked.
3921 Once locked, the boundary cannot be changed.
3922 1 indicates lock status, 0 indicates unlock status.
3924 oops=panic Always panic on oopses. Default is to just kill the
3925 process, but there is a small probability of
3926 deadlocking the machine.
3927 This will also cause panics on machine check exceptions.
3928 Useful together with panic=30 to trigger a reboot.
3931 [KNL] Boolean flag to control whether the page allocator
3932 should randomize its free lists. The randomization may
3933 be automatically enabled if the kernel detects it is
3934 running on a platform with a direct-mapped memory-side
3935 cache, and this parameter can be used to
3936 override/disable that behavior. The state of the flag
3937 can be read from sysfs at:
3938 /sys/module/page_alloc/parameters/shuffle.
3940 page_owner= [KNL] Boot-time page_owner enabling option.
3941 Storage of the information about who allocated
3942 each page is disabled in default. With this switch,
3944 on: enable the feature
3946 page_poison= [KNL] Boot-time parameter changing the state of
3947 poisoning on the buddy allocator, available with
3948 CONFIG_PAGE_POISONING=y.
3949 off: turn off poisoning (default)
3950 on: turn on poisoning
3952 page_reporting.page_reporting_order=
3953 [KNL] Minimal page reporting order
3955 Adjust the minimal page reporting order. The page
3956 reporting is disabled when it exceeds (MAX_ORDER-1).
3958 panic= [KNL] Kernel behaviour on panic: delay <timeout>
3959 timeout > 0: seconds before rebooting
3960 timeout = 0: wait forever
3961 timeout < 0: reboot immediately
3964 panic_print= Bitmask for printing system info when panic happens.
3965 User can chose combination of the following bits:
3966 bit 0: print all tasks info
3967 bit 1: print system memory info
3968 bit 2: print timer info
3969 bit 3: print locks info if CONFIG_LOCKDEP is on
3970 bit 4: print ftrace buffer
3971 bit 5: print all printk messages in buffer
3972 bit 6: print all CPUs backtrace (if available in the arch)
3973 *Be aware* that this option may print a _lot_ of lines,
3974 so there are risks of losing older messages in the log.
3975 Use this option carefully, maybe worth to setup a
3976 bigger log buffer with "log_buf_len" along with this.
3978 panic_on_taint= Bitmask for conditionally calling panic() in add_taint()
3979 Format: <hex>[,nousertaint]
3980 Hexadecimal bitmask representing the set of TAINT flags
3981 that will cause the kernel to panic when add_taint() is
3982 called with any of the flags in this set.
3983 The optional switch "nousertaint" can be utilized to
3984 prevent userspace forced crashes by writing to sysctl
3985 /proc/sys/kernel/tainted any flagset matching with the
3986 bitmask set on panic_on_taint.
3987 See Documentation/admin-guide/tainted-kernels.rst for
3988 extra details on the taint flags that users can pick
3989 to compose the bitmask to assign to panic_on_taint.
3991 panic_on_warn panic() instead of WARN(). Useful to cause kdump
3994 parkbd.port= [HW] Parallel port number the keyboard adapter is
3995 connected to, default is 0.
3997 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
3998 0 for XT, 1 for AT (default is AT).
4001 parport= [HW,PPT] Specify parallel ports. 0 disables.
4002 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
4003 Use 'auto' to force the driver to use any
4004 IRQ/DMA settings detected (the default is to
4005 ignore detected IRQ/DMA settings because of
4006 possible conflicts). You can specify the base
4007 address, IRQ, and DMA settings; IRQ and DMA
4008 should be numbers, or 'auto' (for using detected
4009 settings on that particular port), or 'nofifo'
4010 (to avoid using a FIFO even if it is detected).
4011 Parallel ports are assigned in the order they
4012 are specified on the command line, starting
4015 parport_init_mode= [HW,PPT]
4016 Configure VIA parallel port to operate in
4017 a specific mode. This is necessary on Pegasos
4018 computer where firmware has no options for setting
4019 up parallel port mode and sets it to spp.
4020 Currently this function knows 686a and 8231 chips.
4021 Format: [spp|ps2|epp|ecp|ecpepp]
4023 pata_legacy.all= [HW,LIBATA]
4025 Set to non-zero to probe primary and secondary ISA
4026 port ranges on PCI systems where no PCI PATA device
4027 has been found at either range. Disabled by default.
4029 pata_legacy.autospeed= [HW,LIBATA]
4031 Set to non-zero if a chip is present that snoops speed
4032 changes. Disabled by default.
4034 pata_legacy.ht6560a= [HW,LIBATA]
4036 Set to 1, 2, or 3 for HT 6560A on the primary channel,
4037 the secondary channel, or both channels respectively.
4038 Disabled by default.
4040 pata_legacy.ht6560b= [HW,LIBATA]
4042 Set to 1, 2, or 3 for HT 6560B on the primary channel,
4043 the secondary channel, or both channels respectively.
4044 Disabled by default.
4046 pata_legacy.iordy_mask= [HW,LIBATA]
4048 IORDY enable mask. Set individual bits to allow IORDY
4049 for the respective channel. Bit 0 is for the first
4050 legacy channel handled by this driver, bit 1 is for
4051 the second channel, and so on. The sequence will often
4052 correspond to the primary legacy channel, the secondary
4053 legacy channel, and so on, but the handling of a PCI
4054 bus and the use of other driver options may interfere
4055 with the sequence. By default IORDY is allowed across
4058 pata_legacy.opti82c46x= [HW,LIBATA]
4060 Set to 1, 2, or 3 for Opti 82c611A on the primary
4061 channel, the secondary channel, or both channels
4062 respectively. Disabled by default.
4064 pata_legacy.opti82c611a= [HW,LIBATA]
4066 Set to 1, 2, or 3 for Opti 82c465MV on the primary
4067 channel, the secondary channel, or both channels
4068 respectively. Disabled by default.
4070 pata_legacy.pio_mask= [HW,LIBATA]
4072 PIO mode mask for autospeed devices. Set individual
4073 bits to allow the use of the respective PIO modes.
4074 Bit 0 is for mode 0, bit 1 is for mode 1, and so on.
4075 All modes allowed by default.
4077 pata_legacy.probe_all= [HW,LIBATA]
4079 Set to non-zero to probe tertiary and further ISA
4080 port ranges on PCI systems. Disabled by default.
4082 pata_legacy.probe_mask= [HW,LIBATA]
4084 Probe mask for legacy ISA PATA ports. Depending on
4085 platform configuration and the use of other driver
4086 options up to 6 legacy ports are supported: 0x1f0,
4087 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing
4088 of individual ports can be disabled by setting the
4089 corresponding bits in the mask to 1. Bit 0 is for
4090 the first port in the list above (0x1f0), and so on.
4091 By default all supported ports are probed.
4093 pata_legacy.qdi= [HW,LIBATA]
4095 Set to non-zero to probe QDI controllers. By default
4096 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise.
4098 pata_legacy.winbond= [HW,LIBATA]
4100 Set to non-zero to probe Winbond controllers. Use
4101 the standard I/O port (0x130) if 1, otherwise the
4102 value given is the I/O port to use (typically 0x1b0).
4103 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE,
4106 pata_platform.pio_mask= [HW,LIBATA]
4108 Supported PIO mode mask. Set individual bits to allow
4109 the use of the respective PIO modes. Bit 0 is for
4110 mode 0, bit 1 is for mode 1, and so on. Mode 0 only
4114 Halt all CPUs after the first oops has been printed for
4115 the specified number of seconds. This is to be used if
4116 your oopses keep scrolling off the screen.
4120 pci=option[,option...] [PCI] various PCI subsystem options.
4122 Some options herein operate on a specific device
4123 or a set of devices (<pci_dev>). These are
4124 specified in one of the following formats:
4126 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]*
4127 pci:<vendor>:<device>[:<subvendor>:<subdevice>]
4129 Note: the first format specifies a PCI
4130 bus/device/function address which may change
4131 if new hardware is inserted, if motherboard
4132 firmware changes, or due to changes caused
4133 by other kernel parameters. If the
4134 domain is left unspecified, it is
4135 taken to be zero. Optionally, a path
4136 to a device through multiple device/function
4137 addresses can be specified after the base
4138 address (this is more robust against
4139 renumbering issues). The second format
4140 selects devices using IDs from the
4141 configuration space which may match multiple
4142 devices in the system.
4144 earlydump dump PCI config space before the kernel
4146 off [X86] don't probe for the PCI bus
4147 bios [X86-32] force use of PCI BIOS, don't access
4148 the hardware directly. Use this if your machine
4149 has a non-standard PCI host bridge.
4150 nobios [X86-32] disallow use of PCI BIOS, only direct
4151 hardware access methods are allowed. Use this
4152 if you experience crashes upon bootup and you
4153 suspect they are caused by the BIOS.
4154 conf1 [X86] Force use of PCI Configuration Access
4155 Mechanism 1 (config address in IO port 0xCF8,
4156 data in IO port 0xCFC, both 32-bit).
4157 conf2 [X86] Force use of PCI Configuration Access
4158 Mechanism 2 (IO port 0xCF8 is an 8-bit port for
4159 the function, IO port 0xCFA, also 8-bit, sets
4160 bus number. The config space is then accessed
4161 through ports 0xC000-0xCFFF).
4162 See http://wiki.osdev.org/PCI for more info
4163 on the configuration access mechanisms.
4164 noaer [PCIE] If the PCIEAER kernel config parameter is
4165 enabled, this kernel boot option can be used to
4166 disable the use of PCIE advanced error reporting.
4167 nodomains [PCI] Disable support for multiple PCI
4168 root domains (aka PCI segments, in ACPI-speak).
4169 nommconf [X86] Disable use of MMCONFIG for PCI
4171 check_enable_amd_mmconf [X86] check for and enable
4172 properly configured MMIO access to PCI
4173 config space on AMD family 10h CPU
4174 nomsi [MSI] If the PCI_MSI kernel config parameter is
4175 enabled, this kernel boot option can be used to
4176 disable the use of MSI interrupts system-wide.
4177 noioapicquirk [APIC] Disable all boot interrupt quirks.
4178 Safety option to keep boot IRQs enabled. This
4179 should never be necessary.
4180 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
4181 primary IO-APIC for bridges that cannot disable
4182 boot IRQs. This fixes a source of spurious IRQs
4183 when the system masks IRQs.
4184 noioapicreroute [APIC] Disable workaround that uses the
4185 boot IRQ equivalent of an IRQ that connects to
4186 a chipset where boot IRQs cannot be disabled.
4187 The opposite of ioapicreroute.
4188 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
4189 routing table. These calls are known to be buggy
4190 on several machines and they hang the machine
4191 when used, but on other computers it's the only
4192 way to get the interrupt routing table. Try
4193 this option if the kernel is unable to allocate
4194 IRQs or discover secondary PCI buses on your
4196 rom [X86] Assign address space to expansion ROMs.
4197 Use with caution as certain devices share
4198 address decoders between ROMs and other
4200 norom [X86] Do not assign address space to
4201 expansion ROMs that do not already have
4202 BIOS assigned address ranges.
4203 nobar [X86] Do not assign address space to the
4204 BARs that weren't assigned by the BIOS.
4205 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
4206 assigned automatically to PCI devices. You can
4207 make the kernel exclude IRQs of your ISA cards
4209 pirqaddr=0xAAAAA [X86] Specify the physical address
4210 of the PIRQ table (normally generated
4211 by the BIOS) if it is outside the
4212 F0000h-100000h range.
4213 lastbus=N [X86] Scan all buses thru bus #N. Can be
4214 useful if the kernel is unable to find your
4215 secondary buses and you want to tell it
4216 explicitly which ones they are.
4217 assign-busses [X86] Always assign all PCI bus
4218 numbers ourselves, overriding
4219 whatever the firmware may have done.
4220 usepirqmask [X86] Honor the possible IRQ mask stored
4221 in the BIOS $PIR table. This is needed on
4222 some systems with broken BIOSes, notably
4223 some HP Pavilion N5400 and Omnibook XE3
4224 notebooks. This will have no effect if ACPI
4225 IRQ routing is enabled.
4226 noacpi [X86] Do not use ACPI for IRQ routing
4227 or for PCI scanning.
4228 use_crs [X86] Use PCI host bridge window information
4229 from ACPI. On BIOSes from 2008 or later, this
4230 is enabled by default. If you need to use this,
4231 please report a bug.
4232 nocrs [X86] Ignore PCI host bridge windows from ACPI.
4233 If you need to use this, please report a bug.
4234 use_e820 [X86] Use E820 reservations to exclude parts of
4235 PCI host bridge windows. This is a workaround
4236 for BIOS defects in host bridge _CRS methods.
4237 If you need to use this, please report a bug to
4238 <linux-pci@vger.kernel.org>.
4239 no_e820 [X86] Ignore E820 reservations for PCI host
4240 bridge windows. This is the default on modern
4241 hardware. If you need to use this, please report
4242 a bug to <linux-pci@vger.kernel.org>.
4243 routeirq Do IRQ routing for all PCI devices.
4244 This is normally done in pci_enable_device(),
4245 so this option is a temporary workaround
4246 for broken drivers that don't call it.
4247 skip_isa_align [X86] do not align io start addr, so can
4248 handle more pci cards
4249 noearly [X86] Don't do any early type 1 scanning.
4250 This might help on some broken boards which
4251 machine check when some devices' config space
4252 is read. But various workarounds are disabled
4253 and some IOMMU drivers will not work.
4254 bfsort Sort PCI devices into breadth-first order.
4255 This sorting is done to get a device
4256 order compatible with older (<= 2.4) kernels.
4257 nobfsort Don't sort PCI devices into breadth-first order.
4258 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
4259 tuning and use the BIOS-configured MPS defaults.
4260 pcie_bus_safe Set every device's MPS to the largest value
4261 supported by all devices below the root complex.
4262 pcie_bus_perf Set device MPS to the largest allowable MPS
4263 based on its parent bus. Also set MRRS (Max
4264 Read Request Size) to the largest supported
4265 value (no larger than the MPS that the device
4266 or bus can support) for best performance.
4267 pcie_bus_peer2peer Set every device's MPS to 128B, which
4268 every device is guaranteed to support. This
4269 configuration allows peer-to-peer DMA between
4270 any pair of devices, possibly at the cost of
4271 reduced performance. This also guarantees
4272 that hot-added devices will work.
4273 cbiosize=nn[KMG] The fixed amount of bus space which is
4274 reserved for the CardBus bridge's IO window.
4275 The default value is 256 bytes.
4276 cbmemsize=nn[KMG] The fixed amount of bus space which is
4277 reserved for the CardBus bridge's memory
4278 window. The default value is 64 megabytes.
4281 [<order of align>@]<pci_dev>[; ...]
4282 Specifies alignment and device to reassign
4283 aligned memory resources. How to
4284 specify the device is described above.
4285 If <order of align> is not specified,
4286 PAGE_SIZE is used as alignment.
4287 A PCI-PCI bridge can be specified if resource
4288 windows need to be expanded.
4289 To specify the alignment for several
4290 instances of a device, the PCI vendor,
4291 device, subvendor, and subdevice may be
4292 specified, e.g., 12@pci:8086:9c22:103c:198f
4293 for 4096-byte alignment.
4294 ecrc= Enable/disable PCIe ECRC (transaction layer
4295 end-to-end CRC checking).
4296 bios: Use BIOS/firmware settings. This is the
4300 hpiosize=nn[KMG] The fixed amount of bus space which is
4301 reserved for hotplug bridge's IO window.
4302 Default size is 256 bytes.
4303 hpmmiosize=nn[KMG] The fixed amount of bus space which is
4304 reserved for hotplug bridge's MMIO window.
4305 Default size is 2 megabytes.
4306 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is
4307 reserved for hotplug bridge's MMIO_PREF window.
4308 Default size is 2 megabytes.
4309 hpmemsize=nn[KMG] The fixed amount of bus space which is
4310 reserved for hotplug bridge's MMIO and
4312 Default size is 2 megabytes.
4313 hpbussize=nn The minimum amount of additional bus numbers
4314 reserved for buses below a hotplug bridge.
4316 realloc= Enable/disable reallocating PCI bridge resources
4317 if allocations done by BIOS are too small to
4318 accommodate resources required by all child
4320 off: Turn realloc off
4322 realloc same as realloc=on
4323 noari do not use PCIe ARI.
4324 noats [PCIE, Intel-IOMMU, AMD-IOMMU]
4325 do not use PCIe ATS (and IOMMU device IOTLB).
4326 pcie_scan_all Scan all possible PCIe devices. Otherwise we
4327 only look for one device below a PCIe downstream
4329 big_root_window Try to add a big 64bit memory window to the PCIe
4330 root complex on AMD CPUs. Some GFX hardware
4331 can resize a BAR to allow access to all VRAM.
4332 Adding the window is slightly risky (it may
4333 conflict with unreported devices), so this
4335 disable_acs_redir=<pci_dev>[; ...]
4336 Specify one or more PCI devices (in the format
4337 specified above) separated by semicolons.
4338 Each device specified will have the PCI ACS
4339 redirect capabilities forced off which will
4340 allow P2P traffic between devices through
4341 bridges without forcing it upstream. Note:
4342 this removes isolation between devices and
4343 may put more devices in an IOMMU group.
4344 force_floating [S390] Force usage of floating interrupts.
4345 nomio [S390] Do not use MIO instructions.
4346 norid [S390] ignore the RID field and force use of
4347 one PCI domain per PCI function
4349 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
4352 force Enable ASPM even on devices that claim not to support it.
4353 WARNING: Forcing ASPM on may cause system lockups.
4355 pcie_ports= [PCIE] PCIe port services handling:
4356 native Use native PCIe services (PME, AER, DPC, PCIe hotplug)
4357 even if the platform doesn't give the OS permission to
4358 use them. This may cause conflicts if the platform
4359 also tries to use these services.
4360 dpc-native Use native PCIe service for DPC only. May
4361 cause conflicts if firmware uses AER or DPC.
4362 compat Disable native PCIe services (PME, AER, DPC, PCIe
4365 pcie_port_pm= [PCIE] PCIe port power management handling:
4366 off Disable power management of all PCIe ports
4367 force Forcibly enable power management of all PCIe ports
4369 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
4370 nomsi Do not use MSI for native PCIe PME signaling (this makes
4371 all PCIe root ports use INTx for all services).
4373 pcmv= [HW,PCMCIA] BadgePAD 4
4377 Keep all power-domains already enabled by bootloader on,
4378 even if no driver has claimed them. This is useful
4379 for debug and development, but should not be
4380 needed on a platform with proper driver support.
4382 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
4385 See arch/parisc/kernel/pdc_chassis.c
4387 percpu_alloc= Select which percpu first chunk allocator to use.
4388 Currently supported values are "embed" and "page".
4389 Archs may support subset or none of the selections.
4390 See comments in mm/percpu.c for details on each
4391 allocator. This parameter is primarily for debugging
4392 and performance comparison.
4394 pirq= [SMP,APIC] Manual mp-table setup
4395 See Documentation/x86/i386/IO-APIC.rst.
4397 plip= [PPT,NET] Parallel port network link
4398 Format: { parport<nr> | timid | 0 }
4399 See also Documentation/admin-guide/parport.rst.
4401 pmtmr= [X86] Manual setup of pmtmr I/O Port.
4402 Override pmtimer IOPort with a hex value.
4405 pmu_override= [PPC] Override the PMU.
4406 This option takes over the PMU facility, so it is no
4407 longer usable by perf. Setting this option starts the
4408 PMU counters by setting MMCR0 to 0 (the FC bit is
4409 cleared). If a number is given, then MMCR1 is set to
4410 that number, otherwise (e.g., 'pmu_override=on'), MMCR1
4413 pm_debug_messages [SUSPEND,KNL]
4414 Enable suspend/resume debug messages during boot up.
4417 Enable PNP debug messages (depends on the
4418 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
4419 via /sys/module/pnp/parameters/debug. We always show
4420 current resource usage; turning this on also shows
4421 possible settings and some assignment information.
4427 { on | off | curr | res | no-curr | no-res }
4430 [ISAPNP] Exclude IRQs for the autoconfiguration
4433 [ISAPNP] Exclude DMAs for the autoconfiguration
4435 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
4436 Ranges are in pairs (I/O port base and size).
4439 [ISAPNP] Exclude memory regions for the
4441 Ranges are in pairs (memory base and size).
4443 ports= [IP_VS_FTP] IPVS ftp helper module
4445 Up to 8 (IP_VS_APP_MAX_PORTS) ports
4447 Format: <port>,<port>....
4449 powersave=off [PPC] This option disables power saving features.
4450 It specifically disables cpuidle and sets the
4451 platform machine description specific power_save
4452 function to NULL. On Idle the CPU just reduces
4455 ppc_strict_facility_enable
4456 [PPC] This option catches any kernel floating point,
4457 Altivec, VSX and SPE outside of regions specifically
4458 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
4459 There is some performance impact when enabling this.
4463 Disable Hardware Transactional Memory
4466 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC
4467 none - Limited to cond_resched() calls
4468 voluntary - Limited to cond_resched() and might_sleep() calls
4469 full - Any section that isn't explicitly preempt disabled
4470 can be preempted anytime.
4472 print-fatal-signals=
4473 [KNL] debug: print fatal signals
4475 If enabled, warn about various signal handling
4476 related application anomalies: too many signals,
4477 too many POSIX.1 timers, fatal signals causing a
4480 If you hit the warning due to signal overflow,
4481 you might want to try "ulimit -i unlimited".
4485 printk.always_kmsg_dump=
4486 Trigger kmsg_dump for cases other than kernel oops or
4488 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
4491 printk.console_no_auto_verbose=
4492 Disable console loglevel raise on oops, panic
4493 or lockdep-detected issues (only if lock debug is on).
4494 With an exception to setups with low baudrate on
4495 serial console, keeping this 0 is a good choice
4496 in order to provide more debug information.
4498 default: 0 (auto_verbose is enabled)
4500 printk.devkmsg={on,off,ratelimit}
4501 Control writing to /dev/kmsg.
4502 on - unlimited logging to /dev/kmsg from userspace
4503 off - logging to /dev/kmsg disabled
4504 ratelimit - ratelimit the logging
4507 printk.time= Show timing data prefixed to each printk message line
4508 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
4510 processor.max_cstate= [HW,ACPI]
4511 Limit processor to maximum C-state
4512 max_cstate=9 overrides any DMI blacklist limit.
4514 processor.nocst [HW,ACPI]
4515 Ignore the _CST method to determine C-states,
4516 instead using the legacy FADT method
4518 profile= [KNL] Enable kernel profiling via /proc/profile
4519 Format: [<profiletype>,]<number>
4520 Param: <profiletype>: "schedule", "sleep", or "kvm"
4521 [defaults to kernel profiling]
4522 Param: "schedule" - profile schedule points.
4523 Param: "sleep" - profile D-state sleeping (millisecs).
4524 Requires CONFIG_SCHEDSTATS
4525 Param: "kvm" - profile VM exits.
4526 Param: <number> - step/bucket size as a power of 2 for
4527 statistical time based profiling.
4529 prompt_ramdisk= [RAM] [Deprecated]
4531 prot_virt= [S390] enable hosting protected virtual machines
4532 isolated from the hypervisor (if hardware supports
4536 psi= [KNL] Enable or disable pressure stall information
4540 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
4541 probe for; one of (bare|imps|exps|lifebook|any).
4542 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
4544 psmouse.resetafter= [HW,MOUSE]
4545 Try to reset the device after so many bad packets
4548 [HW,MOUSE] Set desired mouse resolution, in dpi.
4549 psmouse.smartscroll=
4550 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
4551 0 = disabled, 1 = enabled (default).
4553 pstore.backend= Specify the name of the pstore backend to use
4555 pti= [X86-64] Control Page Table Isolation of user and
4556 kernel address spaces. Disabling this feature
4557 removes hardening, but improves performance of
4558 system calls and interrupts.
4560 on - unconditionally enable
4561 off - unconditionally disable
4562 auto - kernel detects whether your CPU model is
4563 vulnerable to issues that PTI mitigates
4565 Not specifying this option is equivalent to pti=auto.
4568 Equivalent to pti=off
4571 [KNL] Number of legacy pty's. Overwrites compiled-in
4574 quiet [KNL] Disable most log messages
4578 radix_hcall_invalidate=on [PPC/PSERIES]
4579 Disable RADIX GTSE feature and use hcall for TLB
4583 See Documentation/admin-guide/md.rst.
4585 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
4586 See Documentation/admin-guide/blockdev/ramdisk.rst.
4588 ramdisk_start= [RAM] RAM disk image start address
4590 random.trust_cpu=off
4591 [KNL] Disable trusting the use of the CPU's
4592 random number generator (if available) to
4593 initialize the kernel's RNG.
4595 random.trust_bootloader=off
4596 [KNL] Disable trusting the use of the a seed
4597 passed by the bootloader (if available) to
4598 initialize the kernel's RNG.
4600 randomize_kstack_offset=
4601 [KNL] Enable or disable kernel stack offset
4602 randomization, which provides roughly 5 bits of
4603 entropy, frustrating memory corruption attacks
4604 that depend on stack address determinism or
4605 cross-syscall address exposures. This is only
4606 available on architectures that have defined
4607 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET.
4608 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
4609 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT.
4611 ras=option[,option,...] [KNL] RAS-specific options
4614 Disable the Correctable Errors Collector,
4615 see CONFIG_RAS_CEC help text.
4617 rcu_nocbs[=cpu-list]
4618 [KNL] The optional argument is a cpu list,
4621 In kernels built with CONFIG_RCU_NOCB_CPU=y,
4622 enable the no-callback CPU mode, which prevents
4623 such CPUs' callbacks from being invoked in
4624 softirq context. Invocation of such CPUs' RCU
4625 callbacks will instead be offloaded to "rcuox/N"
4626 kthreads created for that purpose, where "x" is
4627 "p" for RCU-preempt, "s" for RCU-sched, and "g"
4628 for the kthreads that mediate grace periods; and
4629 "N" is the CPU number. This reduces OS jitter on
4630 the offloaded CPUs, which can be useful for HPC
4631 and real-time workloads. It can also improve
4632 energy efficiency for asymmetric multiprocessors.
4634 If a cpulist is passed as an argument, the specified
4635 list of CPUs is set to no-callback mode from boot.
4637 Otherwise, if the '=' sign and the cpulist
4638 arguments are omitted, no CPU will be set to
4639 no-callback mode from boot but the mode may be
4640 toggled at runtime via cpusets.
4642 Note that this argument takes precedence over
4643 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option.
4646 Rather than requiring that offloaded CPUs
4647 (specified by rcu_nocbs= above) explicitly
4648 awaken the corresponding "rcuoN" kthreads,
4649 make these kthreads poll for callbacks.
4650 This improves the real-time response for the
4651 offloaded CPUs by relieving them of the need to
4652 wake up the corresponding kthread, but degrades
4653 energy efficiency by requiring that the kthreads
4654 periodically wake up to do the polling.
4656 rcutree.blimit= [KNL]
4657 Set maximum number of finished RCU callbacks to
4658 process in one batch.
4660 rcutree.dump_tree= [KNL]
4661 Dump the structure of the rcu_node combining tree
4662 out at early boot. This is used for diagnostic
4663 purposes, to verify correct tree setup.
4665 rcutree.gp_cleanup_delay= [KNL]
4666 Set the number of jiffies to delay each step of
4667 RCU grace-period cleanup.
4669 rcutree.gp_init_delay= [KNL]
4670 Set the number of jiffies to delay each step of
4671 RCU grace-period initialization.
4673 rcutree.gp_preinit_delay= [KNL]
4674 Set the number of jiffies to delay each step of
4675 RCU grace-period pre-initialization, that is,
4676 the propagation of recent CPU-hotplug changes up
4677 the rcu_node combining tree.
4679 rcutree.use_softirq= [KNL]
4680 If set to zero, move all RCU_SOFTIRQ processing to
4681 per-CPU rcuc kthreads. Defaults to a non-zero
4682 value, meaning that RCU_SOFTIRQ is used by default.
4683 Specify rcutree.use_softirq=0 to use rcuc kthreads.
4685 But note that CONFIG_PREEMPT_RT=y kernels disable
4686 this kernel boot parameter, forcibly setting it
4689 rcutree.rcu_fanout_exact= [KNL]
4690 Disable autobalancing of the rcu_node combining
4691 tree. This is used by rcutorture, and might
4692 possibly be useful for architectures having high
4693 cache-to-cache transfer latencies.
4695 rcutree.rcu_fanout_leaf= [KNL]
4696 Change the number of CPUs assigned to each
4697 leaf rcu_node structure. Useful for very
4698 large systems, which will choose the value 64,
4699 and for NUMA systems with large remote-access
4700 latencies, which will choose a value aligned
4701 with the appropriate hardware boundaries.
4703 rcutree.rcu_min_cached_objs= [KNL]
4704 Minimum number of objects which are cached and
4705 maintained per one CPU. Object size is equal
4706 to PAGE_SIZE. The cache allows to reduce the
4707 pressure to page allocator, also it makes the
4708 whole algorithm to behave better in low memory
4711 rcutree.rcu_delay_page_cache_fill_msec= [KNL]
4712 Set the page-cache refill delay (in milliseconds)
4713 in response to low-memory conditions. The range
4714 of permitted values is in the range 0:100000.
4716 rcutree.jiffies_till_first_fqs= [KNL]
4717 Set delay from grace-period initialization to
4718 first attempt to force quiescent states.
4719 Units are jiffies, minimum value is zero,
4720 and maximum value is HZ.
4722 rcutree.jiffies_till_next_fqs= [KNL]
4723 Set delay between subsequent attempts to force
4724 quiescent states. Units are jiffies, minimum
4725 value is one, and maximum value is HZ.
4727 rcutree.jiffies_till_sched_qs= [KNL]
4728 Set required age in jiffies for a
4729 given grace period before RCU starts
4730 soliciting quiescent-state help from
4731 rcu_note_context_switch() and cond_resched().
4732 If not specified, the kernel will calculate
4733 a value based on the most recent settings
4734 of rcutree.jiffies_till_first_fqs
4735 and rcutree.jiffies_till_next_fqs.
4736 This calculated value may be viewed in
4737 rcutree.jiffies_to_sched_qs. Any attempt to set
4738 rcutree.jiffies_to_sched_qs will be cheerfully
4741 rcutree.kthread_prio= [KNL,BOOT]
4742 Set the SCHED_FIFO priority of the RCU per-CPU
4743 kthreads (rcuc/N). This value is also used for
4744 the priority of the RCU boost threads (rcub/N)
4745 and for the RCU grace-period kthreads (rcu_bh,
4746 rcu_preempt, and rcu_sched). If RCU_BOOST is
4747 set, valid values are 1-99 and the default is 1
4748 (the least-favored priority). Otherwise, when
4749 RCU_BOOST is not set, valid values are 0-99 and
4750 the default is zero (non-realtime operation).
4751 When RCU_NOCB_CPU is set, also adjust the
4752 priority of NOCB callback kthreads.
4754 rcutree.rcu_divisor= [KNL]
4755 Set the shift-right count to use to compute
4756 the callback-invocation batch limit bl from
4757 the number of callbacks queued on this CPU.
4758 The result will be bounded below by the value of
4759 the rcutree.blimit kernel parameter. Every bl
4760 callbacks, the softirq handler will exit in
4761 order to allow the CPU to do other work.
4763 Please note that this callback-invocation batch
4764 limit applies only to non-offloaded callback
4765 invocation. Offloaded callbacks are instead
4766 invoked in the context of an rcuoc kthread, which
4767 scheduler will preempt as it does any other task.
4769 rcutree.nocb_nobypass_lim_per_jiffy= [KNL]
4770 On callback-offloaded (rcu_nocbs) CPUs,
4771 RCU reduces the lock contention that would
4772 otherwise be caused by callback floods through
4773 use of the ->nocb_bypass list. However, in the
4774 common non-flooded case, RCU queues directly to
4775 the main ->cblist in order to avoid the extra
4776 overhead of the ->nocb_bypass list and its lock.
4777 But if there are too many callbacks queued during
4778 a single jiffy, RCU pre-queues the callbacks into
4779 the ->nocb_bypass queue. The definition of "too
4780 many" is supplied by this kernel boot parameter.
4782 rcutree.rcu_nocb_gp_stride= [KNL]
4783 Set the number of NOCB callback kthreads in
4784 each group, which defaults to the square root
4785 of the number of CPUs. Larger numbers reduce
4786 the wakeup overhead on the global grace-period
4787 kthread, but increases that same overhead on
4788 each group's NOCB grace-period kthread.
4790 rcutree.qhimark= [KNL]
4791 Set threshold of queued RCU callbacks beyond which
4792 batch limiting is disabled.
4794 rcutree.qlowmark= [KNL]
4795 Set threshold of queued RCU callbacks below which
4796 batch limiting is re-enabled.
4798 rcutree.qovld= [KNL]
4799 Set threshold of queued RCU callbacks beyond which
4800 RCU's force-quiescent-state scan will aggressively
4801 enlist help from cond_resched() and sched IPIs to
4802 help CPUs more quickly reach quiescent states.
4803 Set to less than zero to make this be set based
4804 on rcutree.qhimark at boot time and to zero to
4805 disable more aggressive help enlistment.
4807 rcutree.rcu_kick_kthreads= [KNL]
4808 Cause the grace-period kthread to get an extra
4809 wake_up() if it sleeps three times longer than
4810 it should at force-quiescent-state time.
4811 This wake_up() will be accompanied by a
4812 WARN_ONCE() splat and an ftrace_dump().
4814 rcutree.rcu_unlock_delay= [KNL]
4815 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels,
4816 this specifies an rcu_read_unlock()-time delay
4817 in microseconds. This defaults to zero.
4818 Larger delays increase the probability of
4819 catching RCU pointer leaks, that is, buggy use
4820 of RCU-protected pointers after the relevant
4821 rcu_read_unlock() has completed.
4823 rcutree.sysrq_rcu= [KNL]
4824 Commandeer a sysrq key to dump out Tree RCU's
4825 rcu_node tree with an eye towards determining
4826 why a new grace period has not yet started.
4828 rcuscale.gp_async= [KNL]
4829 Measure performance of asynchronous
4830 grace-period primitives such as call_rcu().
4832 rcuscale.gp_async_max= [KNL]
4833 Specify the maximum number of outstanding
4834 callbacks per writer thread. When a writer
4835 thread exceeds this limit, it invokes the
4836 corresponding flavor of rcu_barrier() to allow
4837 previously posted callbacks to drain.
4839 rcuscale.gp_exp= [KNL]
4840 Measure performance of expedited synchronous
4841 grace-period primitives.
4843 rcuscale.holdoff= [KNL]
4844 Set test-start holdoff period. The purpose of
4845 this parameter is to delay the start of the
4846 test until boot completes in order to avoid
4849 rcuscale.kfree_rcu_test= [KNL]
4850 Set to measure performance of kfree_rcu() flooding.
4852 rcuscale.kfree_rcu_test_double= [KNL]
4853 Test the double-argument variant of kfree_rcu().
4854 If this parameter has the same value as
4855 rcuscale.kfree_rcu_test_single, both the single-
4856 and double-argument variants are tested.
4858 rcuscale.kfree_rcu_test_single= [KNL]
4859 Test the single-argument variant of kfree_rcu().
4860 If this parameter has the same value as
4861 rcuscale.kfree_rcu_test_double, both the single-
4862 and double-argument variants are tested.
4864 rcuscale.kfree_nthreads= [KNL]
4865 The number of threads running loops of kfree_rcu().
4867 rcuscale.kfree_alloc_num= [KNL]
4868 Number of allocations and frees done in an iteration.
4870 rcuscale.kfree_loops= [KNL]
4871 Number of loops doing rcuscale.kfree_alloc_num number
4872 of allocations and frees.
4874 rcuscale.nreaders= [KNL]
4875 Set number of RCU readers. The value -1 selects
4876 N, where N is the number of CPUs. A value
4877 "n" less than -1 selects N-n+1, where N is again
4878 the number of CPUs. For example, -2 selects N
4879 (the number of CPUs), -3 selects N+1, and so on.
4880 A value of "n" less than or equal to -N selects
4883 rcuscale.nwriters= [KNL]
4884 Set number of RCU writers. The values operate
4885 the same as for rcuscale.nreaders.
4886 N, where N is the number of CPUs
4888 rcuscale.perf_type= [KNL]
4889 Specify the RCU implementation to test.
4891 rcuscale.shutdown= [KNL]
4892 Shut the system down after performance tests
4893 complete. This is useful for hands-off automated
4896 rcuscale.verbose= [KNL]
4897 Enable additional printk() statements.
4899 rcuscale.writer_holdoff= [KNL]
4900 Write-side holdoff between grace periods,
4901 in microseconds. The default of zero says
4904 rcutorture.fqs_duration= [KNL]
4905 Set duration of force_quiescent_state bursts
4908 rcutorture.fqs_holdoff= [KNL]
4909 Set holdoff time within force_quiescent_state bursts
4912 rcutorture.fqs_stutter= [KNL]
4913 Set wait time between force_quiescent_state bursts
4916 rcutorture.fwd_progress= [KNL]
4917 Specifies the number of kthreads to be used
4918 for RCU grace-period forward-progress testing
4919 for the types of RCU supporting this notion.
4920 Defaults to 1 kthread, values less than zero or
4921 greater than the number of CPUs cause the number
4924 rcutorture.fwd_progress_div= [KNL]
4925 Specify the fraction of a CPU-stall-warning
4926 period to do tight-loop forward-progress testing.
4928 rcutorture.fwd_progress_holdoff= [KNL]
4929 Number of seconds to wait between successive
4930 forward-progress tests.
4932 rcutorture.fwd_progress_need_resched= [KNL]
4933 Enclose cond_resched() calls within checks for
4934 need_resched() during tight-loop forward-progress
4937 rcutorture.gp_cond= [KNL]
4938 Use conditional/asynchronous update-side
4939 primitives, if available.
4941 rcutorture.gp_exp= [KNL]
4942 Use expedited update-side primitives, if available.
4944 rcutorture.gp_normal= [KNL]
4945 Use normal (non-expedited) asynchronous
4946 update-side primitives, if available.
4948 rcutorture.gp_sync= [KNL]
4949 Use normal (non-expedited) synchronous
4950 update-side primitives, if available. If all
4951 of rcutorture.gp_cond=, rcutorture.gp_exp=,
4952 rcutorture.gp_normal=, and rcutorture.gp_sync=
4953 are zero, rcutorture acts as if is interpreted
4954 they are all non-zero.
4956 rcutorture.irqreader= [KNL]
4957 Run RCU readers from irq handlers, or, more
4958 accurately, from a timer handler. Not all RCU
4959 flavors take kindly to this sort of thing.
4961 rcutorture.leakpointer= [KNL]
4962 Leak an RCU-protected pointer out of the reader.
4963 This can of course result in splats, and is
4964 intended to test the ability of things like
4965 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect
4968 rcutorture.n_barrier_cbs= [KNL]
4969 Set callbacks/threads for rcu_barrier() testing.
4971 rcutorture.nfakewriters= [KNL]
4972 Set number of concurrent RCU writers. These just
4973 stress RCU, they don't participate in the actual
4974 test, hence the "fake".
4976 rcutorture.nocbs_nthreads= [KNL]
4977 Set number of RCU callback-offload togglers.
4978 Zero (the default) disables toggling.
4980 rcutorture.nocbs_toggle= [KNL]
4981 Set the delay in milliseconds between successive
4982 callback-offload toggling attempts.
4984 rcutorture.nreaders= [KNL]
4985 Set number of RCU readers. The value -1 selects
4986 N-1, where N is the number of CPUs. A value
4987 "n" less than -1 selects N-n-2, where N is again
4988 the number of CPUs. For example, -2 selects N
4989 (the number of CPUs), -3 selects N+1, and so on.
4991 rcutorture.object_debug= [KNL]
4992 Enable debug-object double-call_rcu() testing.
4994 rcutorture.onoff_holdoff= [KNL]
4995 Set time (s) after boot for CPU-hotplug testing.
4997 rcutorture.onoff_interval= [KNL]
4998 Set time (jiffies) between CPU-hotplug operations,
4999 or zero to disable CPU-hotplug testing.
5001 rcutorture.read_exit= [KNL]
5002 Set the number of read-then-exit kthreads used
5003 to test the interaction of RCU updaters and
5004 task-exit processing.
5006 rcutorture.read_exit_burst= [KNL]
5007 The number of times in a given read-then-exit
5008 episode that a set of read-then-exit kthreads
5011 rcutorture.read_exit_delay= [KNL]
5012 The delay, in seconds, between successive
5013 read-then-exit testing episodes.
5015 rcutorture.shuffle_interval= [KNL]
5016 Set task-shuffle interval (s). Shuffling tasks
5017 allows some CPUs to go into dyntick-idle mode
5018 during the rcutorture test.
5020 rcutorture.shutdown_secs= [KNL]
5021 Set time (s) after boot system shutdown. This
5022 is useful for hands-off automated testing.
5024 rcutorture.stall_cpu= [KNL]
5025 Duration of CPU stall (s) to test RCU CPU stall
5026 warnings, zero to disable.
5028 rcutorture.stall_cpu_block= [KNL]
5029 Sleep while stalling if set. This will result
5030 in warnings from preemptible RCU in addition
5031 to any other stall-related activity.
5033 rcutorture.stall_cpu_holdoff= [KNL]
5034 Time to wait (s) after boot before inducing stall.
5036 rcutorture.stall_cpu_irqsoff= [KNL]
5037 Disable interrupts while stalling if set.
5039 rcutorture.stall_gp_kthread= [KNL]
5040 Duration (s) of forced sleep within RCU
5041 grace-period kthread to test RCU CPU stall
5042 warnings, zero to disable. If both stall_cpu
5043 and stall_gp_kthread are specified, the
5044 kthread is starved first, then the CPU.
5046 rcutorture.stat_interval= [KNL]
5047 Time (s) between statistics printk()s.
5049 rcutorture.stutter= [KNL]
5050 Time (s) to stutter testing, for example, specifying
5051 five seconds causes the test to run for five seconds,
5052 wait for five seconds, and so on. This tests RCU's
5053 ability to transition abruptly to and from idle.
5055 rcutorture.test_boost= [KNL]
5056 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
5057 "Maybe" means test if the RCU implementation
5058 under test support RCU priority boosting.
5060 rcutorture.test_boost_duration= [KNL]
5061 Duration (s) of each individual boost test.
5063 rcutorture.test_boost_interval= [KNL]
5064 Interval (s) between each boost test.
5066 rcutorture.test_no_idle_hz= [KNL]
5067 Test RCU's dyntick-idle handling. See also the
5068 rcutorture.shuffle_interval parameter.
5070 rcutorture.torture_type= [KNL]
5071 Specify the RCU implementation to test.
5073 rcutorture.verbose= [KNL]
5074 Enable additional printk() statements.
5076 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL]
5077 Dump ftrace buffer after reporting RCU CPU
5080 rcupdate.rcu_cpu_stall_suppress= [KNL]
5081 Suppress RCU CPU stall warning messages.
5083 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL]
5084 Suppress RCU CPU stall warning messages and
5085 rcutorture writer stall warnings that occur
5086 during early boot, that is, during the time
5087 before the init task is spawned.
5089 rcupdate.rcu_cpu_stall_timeout= [KNL]
5090 Set timeout for RCU CPU stall warning messages.
5091 The value is in seconds and the maximum allowed
5092 value is 300 seconds.
5094 rcupdate.rcu_exp_cpu_stall_timeout= [KNL]
5095 Set timeout for expedited RCU CPU stall warning
5096 messages. The value is in milliseconds
5097 and the maximum allowed value is 21000
5098 milliseconds. Please note that this value is
5099 adjusted to an arch timer tick resolution.
5100 Setting this to zero causes the value from
5101 rcupdate.rcu_cpu_stall_timeout to be used (after
5102 conversion from seconds to milliseconds).
5104 rcupdate.rcu_cpu_stall_cputime= [KNL]
5105 Provide statistics on the cputime and count of
5106 interrupts and tasks during the sampling period. For
5107 multiple continuous RCU stalls, all sampling periods
5108 begin at half of the first RCU stall timeout.
5110 rcupdate.rcu_exp_stall_task_details= [KNL]
5111 Print stack dumps of any tasks blocking the
5112 current expedited RCU grace period during an
5113 expedited RCU CPU stall warning.
5115 rcupdate.rcu_expedited= [KNL]
5116 Use expedited grace-period primitives, for
5117 example, synchronize_rcu_expedited() instead
5118 of synchronize_rcu(). This reduces latency,
5119 but can increase CPU utilization, degrade
5120 real-time latency, and degrade energy efficiency.
5121 No effect on CONFIG_TINY_RCU kernels.
5123 rcupdate.rcu_normal= [KNL]
5124 Use only normal grace-period primitives,
5125 for example, synchronize_rcu() instead of
5126 synchronize_rcu_expedited(). This improves
5127 real-time latency, CPU utilization, and
5128 energy efficiency, but can expose users to
5129 increased grace-period latency. This parameter
5130 overrides rcupdate.rcu_expedited. No effect on
5131 CONFIG_TINY_RCU kernels.
5133 rcupdate.rcu_normal_after_boot= [KNL]
5134 Once boot has completed (that is, after
5135 rcu_end_inkernel_boot() has been invoked), use
5136 only normal grace-period primitives. No effect
5137 on CONFIG_TINY_RCU kernels.
5139 But note that CONFIG_PREEMPT_RT=y kernels enables
5140 this kernel boot parameter, forcibly setting
5141 it to the value one, that is, converting any
5142 post-boot attempt at an expedited RCU grace
5143 period to instead use normal non-expedited
5144 grace-period processing.
5146 rcupdate.rcu_task_collapse_lim= [KNL]
5147 Set the maximum number of callbacks present
5148 at the beginning of a grace period that allows
5149 the RCU Tasks flavors to collapse back to using
5150 a single callback queue. This switching only
5151 occurs when rcupdate.rcu_task_enqueue_lim is
5152 set to the default value of -1.
5154 rcupdate.rcu_task_contend_lim= [KNL]
5155 Set the minimum number of callback-queuing-time
5156 lock-contention events per jiffy required to
5157 cause the RCU Tasks flavors to switch to per-CPU
5158 callback queuing. This switching only occurs
5159 when rcupdate.rcu_task_enqueue_lim is set to
5160 the default value of -1.
5162 rcupdate.rcu_task_enqueue_lim= [KNL]
5163 Set the number of callback queues to use for the
5164 RCU Tasks family of RCU flavors. The default
5165 of -1 allows this to be automatically (and
5166 dynamically) adjusted. This parameter is intended
5169 rcupdate.rcu_task_ipi_delay= [KNL]
5170 Set time in jiffies during which RCU tasks will
5171 avoid sending IPIs, starting with the beginning
5172 of a given grace period. Setting a large
5173 number avoids disturbing real-time workloads,
5174 but lengthens grace periods.
5176 rcupdate.rcu_task_stall_info= [KNL]
5177 Set initial timeout in jiffies for RCU task stall
5178 informational messages, which give some indication
5179 of the problem for those not patient enough to
5180 wait for ten minutes. Informational messages are
5181 only printed prior to the stall-warning message
5182 for a given grace period. Disable with a value
5183 less than or equal to zero. Defaults to ten
5184 seconds. A change in value does not take effect
5185 until the beginning of the next grace period.
5187 rcupdate.rcu_task_stall_info_mult= [KNL]
5188 Multiplier for time interval between successive
5189 RCU task stall informational messages for a given
5190 RCU tasks grace period. This value is clamped
5191 to one through ten, inclusive. It defaults to
5192 the value three, so that the first informational
5193 message is printed 10 seconds into the grace
5194 period, the second at 40 seconds, the third at
5195 160 seconds, and then the stall warning at 600
5196 seconds would prevent a fourth at 640 seconds.
5198 rcupdate.rcu_task_stall_timeout= [KNL]
5199 Set timeout in jiffies for RCU task stall
5200 warning messages. Disable with a value less
5201 than or equal to zero. Defaults to ten minutes.
5202 A change in value does not take effect until
5203 the beginning of the next grace period.
5205 rcupdate.rcu_self_test= [KNL]
5206 Run the RCU early boot self tests
5210 Run specified binary instead of /init from the ramdisk,
5211 used for early userspace startup. See initrd.
5214 force - Override the decision by the kernel to hide the
5215 advertisement of RDRAND support (this affects
5216 certain AMD processors because of buggy BIOS
5217 support, specifically around the suspend/resume
5221 Turn on/off individual RDT features. List is:
5222 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp,
5224 E.g. to turn on cmt and turn off mba use:
5228 Format (x86 or x86_64):
5229 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \
5231 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
5233 Where reboot_mode is one of warm (soft) or cold (hard) or gpio
5234 (prefix with 'panic_' to set mode for panic
5236 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
5237 reboot_force is either force or not specified,
5238 reboot_cpu is s[mp]#### with #### being the processor
5239 to be used for rebooting.
5241 refscale.holdoff= [KNL]
5242 Set test-start holdoff period. The purpose of
5243 this parameter is to delay the start of the
5244 test until boot completes in order to avoid
5247 refscale.loops= [KNL]
5248 Set the number of loops over the synchronization
5249 primitive under test. Increasing this number
5250 reduces noise due to loop start/end overhead,
5251 but the default has already reduced the per-pass
5252 noise to a handful of picoseconds on ca. 2020
5255 refscale.nreaders= [KNL]
5256 Set number of readers. The default value of -1
5257 selects N, where N is roughly 75% of the number
5258 of CPUs. A value of zero is an interesting choice.
5260 refscale.nruns= [KNL]
5261 Set number of runs, each of which is dumped onto
5264 refscale.readdelay= [KNL]
5265 Set the read-side critical-section duration,
5266 measured in microseconds.
5268 refscale.scale_type= [KNL]
5269 Specify the read-protection implementation to test.
5271 refscale.shutdown= [KNL]
5272 Shut down the system at the end of the performance
5273 test. This defaults to 1 (shut it down) when
5274 refscale is built into the kernel and to 0 (leave
5275 it running) when refscale is built as a module.
5277 refscale.verbose= [KNL]
5278 Enable additional printk() statements.
5280 refscale.verbose_batched= [KNL]
5281 Batch the additional printk() statements. If zero
5282 (the default) or negative, print everything. Otherwise,
5283 print every Nth verbose statement, where N is the value
5287 [KNL, SMP] Set scheduler's default relax_domain_level.
5288 See Documentation/admin-guide/cgroup-v1/cpusets.rst.
5290 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory
5291 Format: <base1>,<size1>[,<base2>,<size2>,...]
5292 Reserve I/O ports or memory so the kernel won't use
5293 them. If <base> is less than 0x10000, the region
5294 is assumed to be I/O ports; otherwise it is memory.
5296 reservetop= [X86-32]
5298 Reserves a hole at the top of the kernel virtual
5301 reset_devices [KNL] Force drivers to reset the underlying device
5302 during initialization.
5305 Specify the partition device for software suspend
5307 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
5309 resume_offset= [SWSUSP]
5310 Specify the offset from the beginning of the partition
5311 given by "resume=" at which the swap header is located,
5312 in <PAGE_SIZE> units (needed only for swap files).
5313 See Documentation/power/swsusp-and-swap-files.rst
5315 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
5316 read the resume files
5318 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
5319 Useful for devices that are detected asynchronously
5320 (e.g. USB and MMC devices).
5322 retain_initrd [RAM] Keep initrd memory after extraction
5324 retbleed= [X86] Control mitigation of RETBleed (Arbitrary
5325 Speculative Code Execution with Return Instructions)
5328 AMD-based UNRET and IBPB mitigations alone do not stop
5329 sibling threads from influencing the predictions of other
5330 sibling threads. For that reason, STIBP is used on pro-
5331 cessors that support it, and mitigate SMT on processors
5335 auto - automatically select a migitation
5336 auto,nosmt - automatically select a mitigation,
5337 disabling SMT if necessary for
5338 the full mitigation (only on Zen1
5339 and older without STIBP).
5340 ibpb - On AMD, mitigate short speculation
5341 windows on basic block boundaries too.
5342 Safe, highest perf impact. It also
5343 enables STIBP if present. Not suitable
5345 ibpb,nosmt - Like "ibpb" above but will disable SMT
5346 when STIBP is not available. This is
5347 the alternative for systems which do not
5349 unret - Force enable untrained return thunks,
5350 only effective on AMD f15h-f17h based
5352 unret,nosmt - Like unret, but will disable SMT when STIBP
5353 is not available. This is the alternative for
5354 systems which do not have STIBP.
5356 Selecting 'auto' will choose a mitigation method at run
5357 time according to the CPU.
5359 Not specifying this option is equivalent to retbleed=auto.
5361 rfkill.default_state=
5362 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
5363 etc. communication is blocked by default.
5366 rfkill.master_switch_mode=
5367 0 The "airplane mode" button does nothing.
5368 1 The "airplane mode" button toggles between everything
5369 blocked and the previous configuration.
5370 2 The "airplane mode" button toggles between everything
5371 blocked and everything unblocked.
5373 rhash_entries= [KNL,NET]
5374 Set number of hash buckets for route cache
5377 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported
5380 ro [KNL] Mount root device read-only on boot
5383 on Mark read-only kernel memory as read-only (default).
5384 off Leave read-only kernel memory writable for debugging.
5385 full Mark read-only kernel memory and aliases as read-only
5389 Enable the uart passthrough on the designated usb port
5390 on Rockchip SoCs. When active, the signals of the
5391 debug-uart get routed to the D+ and D- pins of the usb
5392 port and the regular usb controller gets disabled.
5394 root= [KNL] Root filesystem
5395 See name_to_dev_t comment in init/do_mounts.c.
5397 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
5398 mount the root filesystem
5400 rootflags= [KNL] Set root filesystem mount option string
5402 rootfstype= [KNL] Set root filesystem type
5404 rootwait [KNL] Wait (indefinitely) for root device to show up.
5405 Useful for devices that are detected asynchronously
5406 (e.g. USB and MMC devices).
5408 rproc_mem=nn[KMG][@address]
5409 [KNL,ARM,CMA] Remoteproc physical memory block.
5410 Memory area to be used by remote processor image,
5413 rw [KNL] Mount root device read-write on boot
5415 S [KNL] Run init in single mode
5417 s390_iommu= [HW,S390]
5418 Set s390 IOTLB flushing mode
5420 With strict flushing every unmap operation will result in
5421 an IOTLB flush. Default is lazy flushing before reuse,
5424 s390_iommu_aperture= [KNL,S390]
5425 Specifies the size of the per device DMA address space
5426 accessible through the DMA and IOMMU APIs as a decimal
5427 factor of the size of main memory.
5428 The default is 1 meaning that one can concurrently use
5429 as many DMA addresses as physical memory is installed,
5430 if supported by hardware, and thus map all of memory
5431 once. With a value of 2 one can map all of memory twice
5432 and so on. As a special case a factor of 0 imposes no
5433 restrictions other than those given by hardware at the
5434 cost of significant additional memory use for tables.
5437 See drivers/net/irda/sa1100_ir.c.
5439 sched_verbose [KNL] Enables verbose scheduler debug messages.
5441 schedstats= [KNL,X86] Enable or disable scheduled statistics.
5442 Allowed values are enable and disable. This feature
5443 incurs a small amount of overhead in the scheduler
5444 but is useful for debugging and performance tuning.
5446 sched_thermal_decay_shift=
5447 [KNL, SMP] Set a decay shift for scheduler thermal
5448 pressure signal. Thermal pressure signal follows the
5449 default decay period of other scheduler pelt
5450 signals(usually 32 ms but configurable). Setting
5451 sched_thermal_decay_shift will left shift the decay
5452 period for the thermal pressure signal by the shift
5454 i.e. with the default pelt decay period of 32 ms
5455 sched_thermal_decay_shift thermal pressure decay pr
5459 Format: integer between 0 and 10
5462 scftorture.holdoff= [KNL]
5463 Number of seconds to hold off before starting
5464 test. Defaults to zero for module insertion and
5465 to 10 seconds for built-in smp_call_function()
5468 scftorture.longwait= [KNL]
5469 Request ridiculously long waits randomly selected
5470 up to the chosen limit in seconds. Zero (the
5471 default) disables this feature. Please note
5472 that requesting even small non-zero numbers of
5473 seconds can result in RCU CPU stall warnings,
5474 softlockup complaints, and so on.
5476 scftorture.nthreads= [KNL]
5477 Number of kthreads to spawn to invoke the
5478 smp_call_function() family of functions.
5479 The default of -1 specifies a number of kthreads
5480 equal to the number of CPUs.
5482 scftorture.onoff_holdoff= [KNL]
5483 Number seconds to wait after the start of the
5484 test before initiating CPU-hotplug operations.
5486 scftorture.onoff_interval= [KNL]
5487 Number seconds to wait between successive
5488 CPU-hotplug operations. Specifying zero (which
5489 is the default) disables CPU-hotplug operations.
5491 scftorture.shutdown_secs= [KNL]
5492 The number of seconds following the start of the
5493 test after which to shut down the system. The
5494 default of zero avoids shutting down the system.
5495 Non-zero values are useful for automated tests.
5497 scftorture.stat_interval= [KNL]
5498 The number of seconds between outputting the
5499 current test statistics to the console. A value
5500 of zero disables statistics output.
5502 scftorture.stutter_cpus= [KNL]
5503 The number of jiffies to wait between each change
5504 to the set of CPUs under test.
5506 scftorture.use_cpus_read_lock= [KNL]
5507 Use use_cpus_read_lock() instead of the default
5508 preempt_disable() to disable CPU hotplug
5509 while invoking one of the smp_call_function*()
5512 scftorture.verbose= [KNL]
5513 Enable additional printk() statements.
5515 scftorture.weight_single= [KNL]
5516 The probability weighting to use for the
5517 smp_call_function_single() function with a zero
5518 "wait" parameter. A value of -1 selects the
5519 default if all other weights are -1. However,
5520 if at least one weight has some other value, a
5521 value of -1 will instead select a weight of zero.
5523 scftorture.weight_single_wait= [KNL]
5524 The probability weighting to use for the
5525 smp_call_function_single() function with a
5526 non-zero "wait" parameter. See weight_single.
5528 scftorture.weight_many= [KNL]
5529 The probability weighting to use for the
5530 smp_call_function_many() function with a zero
5531 "wait" parameter. See weight_single.
5532 Note well that setting a high probability for
5533 this weighting can place serious IPI load
5536 scftorture.weight_many_wait= [KNL]
5537 The probability weighting to use for the
5538 smp_call_function_many() function with a
5539 non-zero "wait" parameter. See weight_single
5542 scftorture.weight_all= [KNL]
5543 The probability weighting to use for the
5544 smp_call_function_all() function with a zero
5545 "wait" parameter. See weight_single and
5548 scftorture.weight_all_wait= [KNL]
5549 The probability weighting to use for the
5550 smp_call_function_all() function with a
5551 non-zero "wait" parameter. See weight_single
5554 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
5555 xtime_lock contention on larger systems, and/or RCU lock
5556 contention on all systems with CONFIG_MAXSMP set.
5557 Format: { "0" | "1" }
5558 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
5560 Note: increases power consumption, thus should only be
5561 enabled if running jitter sensitive (HPC/RT) workloads.
5563 security= [SECURITY] Choose a legacy "major" security module to
5564 enable at boot. This has been deprecated by the
5567 selinux= [SELINUX] Disable or enable SELinux at boot time.
5568 Format: { "0" | "1" }
5569 See security/selinux/Kconfig help text.
5574 serialnumber [BUGS=X86-32]
5576 sev=option[,option...] [X86-64] See Documentation/x86/x86_64/boot-options.rst
5579 Maximal number of shapers.
5581 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
5582 Limit apic dumping. The parameter defines the maximal
5583 number of local apics being dumped. Also it is possible
5584 to set it to "all" by meaning -- no limit here.
5585 Format: { 1 (default) | 2 | ... | all }.
5586 The parameter valid if only apic=debug or
5587 apic=verbose is specified.
5588 Example: apic=debug show_lapic=all
5596 Enable merging of slabs with similar size when the
5597 kernel is built without CONFIG_SLAB_MERGE_DEFAULT.
5600 Disable merging of slabs with similar size. May be
5601 necessary if there is some reason to distinguish
5602 allocs to different slabs, especially in hardened
5603 environments where the risk of heap overflows and
5604 layout control by attackers can usually be
5605 frustrated by disabling merging. This will reduce
5606 most of the exposure of a heap attack to a single
5607 cache (risks via metadata attacks are mostly
5608 unchanged). Debug options disable merging on their
5610 For more information see Documentation/mm/slub.rst.
5612 slab_max_order= [MM, SLAB]
5613 Determines the maximum allowed order for slabs.
5614 A high setting may cause OOMs due to memory
5615 fragmentation. Defaults to 1 for systems with
5616 more than 32MB of RAM, 0 otherwise.
5618 slub_debug[=options[,slabs][;[options[,slabs]]...] [MM, SLUB]
5619 Enabling slub_debug allows one to determine the
5620 culprit if slab objects become corrupted. Enabling
5621 slub_debug can create guard zones around objects and
5622 may poison objects when not in use. Also tracks the
5623 last alloc / free. For more information see
5624 Documentation/mm/slub.rst.
5626 slub_max_order= [MM, SLUB]
5627 Determines the maximum allowed order for slabs.
5628 A high setting may cause OOMs due to memory
5629 fragmentation. For more information see
5630 Documentation/mm/slub.rst.
5632 slub_min_objects= [MM, SLUB]
5633 The minimum number of objects per slab. SLUB will
5634 increase the slab order up to slub_max_order to
5635 generate a sufficiently large slab able to contain
5636 the number of objects indicated. The higher the number
5637 of objects the smaller the overhead of tracking slabs
5638 and the less frequently locks need to be acquired.
5639 For more information see Documentation/mm/slub.rst.
5641 slub_min_order= [MM, SLUB]
5642 Determines the minimum page order for slabs. Must be
5643 lower than slub_max_order.
5644 For more information see Documentation/mm/slub.rst.
5646 slub_merge [MM, SLUB]
5647 Same with slab_merge.
5649 slub_nomerge [MM, SLUB]
5650 Same with slab_nomerge. This is supported for legacy.
5651 See slab_nomerge for more information.
5654 Format: <io1>[,<io2>[,...,<io8>]]
5656 smp.csd_lock_timeout= [KNL]
5657 Specify the period of time in milliseconds
5658 that smp_call_function() and friends will wait
5659 for a CPU to release the CSD lock. This is
5660 useful when diagnosing bugs involving CPUs
5661 disabling interrupts for extended periods
5662 of time. Defaults to 5,000 milliseconds, and
5663 setting a value of zero disables this feature.
5664 This feature may be more efficiently disabled
5665 using the csdlock_debug- kernel parameter.
5667 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
5668 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
5669 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
5670 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
5671 smsc-ircc2.ircc_irq= [HW] IRQ line
5672 smsc-ircc2.ircc_dma= [HW] DMA channel
5673 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
5674 0: Toshiba Satellite 1800 (GP data pin select)
5675 1: Fast pin select (default)
5678 smt= [KNL,S390] Set the maximum number of threads (logical
5679 CPUs) to use per physical CPU on systems capable of
5680 symmetric multithreading (SMT). Will be capped to the
5681 actual hardware limit.
5683 Default: -1 (no limit)
5686 [KNL] Should the soft-lockup detector generate panics.
5689 A value of 1 instructs the soft-lockup detector
5690 to panic the machine when a soft-lockup occurs. It is
5691 also controlled by the kernel.softlockup_panic sysctl
5692 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the
5693 respective build-time switch to that functionality.
5695 softlockup_all_cpu_backtrace=
5696 [KNL] Should the soft-lockup detector generate
5697 backtraces on all cpus.
5700 sonypi.*= [HW] Sony Programmable I/O Control Device driver
5701 See Documentation/admin-guide/laptops/sonypi.rst
5703 spectre_v2= [X86] Control mitigation of Spectre variant 2
5704 (indirect branch speculation) vulnerability.
5705 The default operation protects the kernel from
5708 on - unconditionally enable, implies
5710 off - unconditionally disable, implies
5712 auto - kernel detects whether your CPU model is
5715 Selecting 'on' will, and 'auto' may, choose a
5716 mitigation method at run time according to the
5717 CPU, the available microcode, the setting of the
5718 CONFIG_RETPOLINE configuration option, and the
5719 compiler with which the kernel was built.
5721 Selecting 'on' will also enable the mitigation
5722 against user space to user space task attacks.
5724 Selecting 'off' will disable both the kernel and
5725 the user space protections.
5727 Specific mitigations can also be selected manually:
5729 retpoline - replace indirect branches
5730 retpoline,generic - Retpolines
5731 retpoline,lfence - LFENCE; indirect branch
5732 retpoline,amd - alias for retpoline,lfence
5733 eibrs - Enhanced/Auto IBRS
5734 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines
5735 eibrs,lfence - Enhanced/Auto IBRS + LFENCE
5736 ibrs - use IBRS to protect kernel
5738 Not specifying this option is equivalent to
5742 [X86] Control mitigation of Spectre variant 2
5743 (indirect branch speculation) vulnerability between
5746 on - Unconditionally enable mitigations. Is
5747 enforced by spectre_v2=on
5749 off - Unconditionally disable mitigations. Is
5750 enforced by spectre_v2=off
5752 prctl - Indirect branch speculation is enabled,
5753 but mitigation can be enabled via prctl
5754 per thread. The mitigation control state
5755 is inherited on fork.
5758 - Like "prctl" above, but only STIBP is
5759 controlled per thread. IBPB is issued
5760 always when switching between different user
5764 - Same as "prctl" above, but all seccomp
5765 threads will enable the mitigation unless
5766 they explicitly opt out.
5769 - Like "seccomp" above, but only STIBP is
5770 controlled per thread. IBPB is issued
5771 always when switching between different
5772 user space processes.
5774 auto - Kernel selects the mitigation depending on
5775 the available CPU features and vulnerability.
5777 Default mitigation: "prctl"
5779 Not specifying this option is equivalent to
5780 spectre_v2_user=auto.
5782 spec_store_bypass_disable=
5783 [HW] Control Speculative Store Bypass (SSB) Disable mitigation
5784 (Speculative Store Bypass vulnerability)
5786 Certain CPUs are vulnerable to an exploit against a
5787 a common industry wide performance optimization known
5788 as "Speculative Store Bypass" in which recent stores
5789 to the same memory location may not be observed by
5790 later loads during speculative execution. The idea
5791 is that such stores are unlikely and that they can
5792 be detected prior to instruction retirement at the
5793 end of a particular speculation execution window.
5795 In vulnerable processors, the speculatively forwarded
5796 store can be used in a cache side channel attack, for
5797 example to read memory to which the attacker does not
5798 directly have access (e.g. inside sandboxed code).
5800 This parameter controls whether the Speculative Store
5801 Bypass optimization is used.
5803 On x86 the options are:
5805 on - Unconditionally disable Speculative Store Bypass
5806 off - Unconditionally enable Speculative Store Bypass
5807 auto - Kernel detects whether the CPU model contains an
5808 implementation of Speculative Store Bypass and
5809 picks the most appropriate mitigation. If the
5810 CPU is not vulnerable, "off" is selected. If the
5811 CPU is vulnerable the default mitigation is
5812 architecture and Kconfig dependent. See below.
5813 prctl - Control Speculative Store Bypass per thread
5814 via prctl. Speculative Store Bypass is enabled
5815 for a process by default. The state of the control
5816 is inherited on fork.
5817 seccomp - Same as "prctl" above, but all seccomp threads
5818 will disable SSB unless they explicitly opt out.
5820 Default mitigations:
5823 On powerpc the options are:
5825 on,auto - On Power8 and Power9 insert a store-forwarding
5826 barrier on kernel entry and exit. On Power7
5827 perform a software flush on kernel entry and
5831 Not specifying this option is equivalent to
5832 spec_store_bypass_disable=auto.
5834 spia_io_base= [HW,MTD]
5840 [X86] Enable split lock detection or bus lock detection
5842 When enabled (and if hardware support is present), atomic
5843 instructions that access data across cache line
5844 boundaries will result in an alignment check exception
5845 for split lock detection or a debug exception for
5850 warn - the kernel will emit rate-limited warnings
5851 about applications triggering the #AC
5852 exception or the #DB exception. This mode is
5853 the default on CPUs that support split lock
5854 detection or bus lock detection. Default
5855 behavior is by #AC if both features are
5856 enabled in hardware.
5858 fatal - the kernel will send SIGBUS to applications
5859 that trigger the #AC exception or the #DB
5860 exception. Default behavior is by #AC if
5861 both features are enabled in hardware.
5864 Set system wide rate limit to N bus locks
5865 per second for bus lock detection.
5868 N/A for split lock detection.
5871 If an #AC exception is hit in the kernel or in
5872 firmware (i.e. not while executing in user mode)
5873 the kernel will oops in either "warn" or "fatal"
5876 #DB exception for bus lock is triggered only when
5880 Control the Special Register Buffer Data Sampling
5883 Certain CPUs are vulnerable to an MDS-like
5884 exploit which can leak bits from the random
5887 By default, this issue is mitigated by
5888 microcode. However, the microcode fix can cause
5889 the RDRAND and RDSEED instructions to become
5890 much slower. Among other effects, this will
5891 result in reduced throughput from /dev/urandom.
5893 The microcode mitigation can be disabled with
5894 the following option:
5896 off: Disable mitigation and remove
5897 performance impact to RDRAND and RDSEED
5899 srcutree.big_cpu_lim [KNL]
5900 Specifies the number of CPUs constituting a
5901 large system, such that srcu_struct structures
5902 should immediately allocate an srcu_node array.
5903 This kernel-boot parameter defaults to 128,
5904 but takes effect only when the low-order four
5905 bits of srcutree.convert_to_big is equal to 3
5908 srcutree.convert_to_big [KNL]
5909 Specifies under what conditions an SRCU tree
5910 srcu_struct structure will be converted to big
5911 form, that is, with an rcu_node tree:
5914 1: At init_srcu_struct() time.
5915 2: When rcutorture decides to.
5916 3: Decide at boot time (default).
5917 0x1X: Above plus if high contention.
5919 Either way, the srcu_node tree will be sized based
5920 on the actual runtime number of CPUs (nr_cpu_ids)
5921 instead of the compile-time CONFIG_NR_CPUS.
5923 srcutree.counter_wrap_check [KNL]
5924 Specifies how frequently to check for
5925 grace-period sequence counter wrap for the
5926 srcu_data structure's ->srcu_gp_seq_needed field.
5927 The greater the number of bits set in this kernel
5928 parameter, the less frequently counter wrap will
5929 be checked for. Note that the bottom two bits
5932 srcutree.exp_holdoff [KNL]
5933 Specifies how many nanoseconds must elapse
5934 since the end of the last SRCU grace period for
5935 a given srcu_struct until the next normal SRCU
5936 grace period will be considered for automatic
5937 expediting. Set to zero to disable automatic
5940 srcutree.srcu_max_nodelay [KNL]
5941 Specifies the number of no-delay instances
5942 per jiffy for which the SRCU grace period
5943 worker thread will be rescheduled with zero
5944 delay. Beyond this limit, worker thread will
5945 be rescheduled with a sleep delay of one jiffy.
5947 srcutree.srcu_max_nodelay_phase [KNL]
5948 Specifies the per-grace-period phase, number of
5949 non-sleeping polls of readers. Beyond this limit,
5950 grace period worker thread will be rescheduled
5951 with a sleep delay of one jiffy, between each
5952 rescan of the readers, for a grace period phase.
5954 srcutree.srcu_retry_check_delay [KNL]
5955 Specifies number of microseconds of non-sleeping
5956 delay between each non-sleeping poll of readers.
5958 srcutree.small_contention_lim [KNL]
5959 Specifies the number of update-side contention
5960 events per jiffy will be tolerated before
5961 initiating a conversion of an srcu_struct
5962 structure to big form. Note that the value of
5963 srcutree.convert_to_big must have the 0x10 bit
5964 set for contention-based conversions to occur.
5967 Speculative Store Bypass Disable control
5969 On CPUs that are vulnerable to the Speculative
5970 Store Bypass vulnerability and offer a
5971 firmware based mitigation, this parameter
5972 indicates how the mitigation should be used:
5974 force-on: Unconditionally enable mitigation for
5975 for both kernel and userspace
5976 force-off: Unconditionally disable mitigation for
5977 for both kernel and userspace
5978 kernel: Always enable mitigation in the
5979 kernel, and offer a prctl interface
5980 to allow userspace to register its
5981 interest in being mitigated too.
5983 stack_guard_gap= [MM]
5984 override the default stack gap protection. The value
5985 is in page units and it defines how many pages prior
5986 to (for stacks growing down) resp. after (for stacks
5987 growing up) the main stack are reserved for no other
5988 mapping. Default value is 256 pages.
5990 stack_depot_disable= [KNL]
5991 Setting this to true through kernel command line will
5992 disable the stack depot thereby saving the static memory
5993 consumed by the stack hash table. By default this is set
5997 Enabled the stack tracer on boot up.
5999 stacktrace_filter=[function-list]
6000 [FTRACE] Limit the functions that the stack tracer
6001 will trace at boot up. function-list is a comma-separated
6002 list of functions. This list can be changed at run
6003 time by the stack_trace_filter file in the debugfs
6004 tracing directory. Note, this enables stack tracing
6005 and the stacktrace above is not needed.
6009 Set the STI (builtin display/keyboard on the HP-PARISC
6010 machines) console (graphic card) which should be used
6011 as the initial boot-console.
6012 See also comment in drivers/video/console/sticore.c.
6015 See comment in drivers/video/console/sticore.c.
6018 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
6023 Enable or disable strict sigaltstack size checks
6024 against the required signal frame size which
6025 depends on the supported FPU features. This can
6026 be used to filter out binaries which have
6027 not yet been made aware of AT_MINSIGSTKSZ.
6030 Limits the number of kernel HPT entries in the hash
6031 page table to increase the rate of hash page table
6032 faults on kernel addresses.
6035 Limits the number of kernel SLB entries, and flushes
6036 them frequently to increase the rate of SLB faults
6037 on kernel addresses.
6039 sunrpc.min_resvport=
6040 sunrpc.max_resvport=
6042 SunRPC servers often require that client requests
6043 originate from a privileged port (i.e. a port in the
6044 range 0 < portnr < 1024).
6045 An administrator who wishes to reserve some of these
6046 ports for other uses may adjust the range that the
6047 kernel's sunrpc client considers to be privileged
6048 using these two parameters to set the minimum and
6049 maximum port values.
6051 sunrpc.svc_rpc_per_connection_limit=
6053 Limit the number of requests that the server will
6054 process in parallel from a single connection.
6055 The default value is 0 (no limit).
6059 Control how the NFS server code allocates CPUs to
6060 service thread pools. Depending on how many NICs
6061 you have and where their interrupts are bound, this
6062 option will affect which CPUs will do NFS serving.
6063 Note: this parameter cannot be changed while the
6064 NFS server is running.
6066 auto the server chooses an appropriate mode
6067 automatically using heuristics
6068 global a single global pool contains all CPUs
6069 percpu one pool for each CPU
6070 pernode one pool for each NUMA node (equivalent
6071 to global on non-NUMA machines)
6073 sunrpc.tcp_slot_table_entries=
6074 sunrpc.udp_slot_table_entries=
6076 Sets the upper limit on the number of simultaneous
6077 RPC calls that can be sent from the client to a
6078 server. Increasing these values may allow you to
6079 improve throughput, but will also increase the
6080 amount of memory reserved for use by the client.
6082 suspend.pm_test_delay=
6084 Sets the number of seconds to remain in a suspend test
6085 mode before resuming the system (see
6086 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
6087 is set. Default value is 5.
6090 Format: { on | off | y | n | 1 | 0 }
6091 This parameter controls use of the Protected
6092 Execution Facility on pSeries.
6094 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
6095 Format: { <int> [,<int>] | force | noforce }
6096 <int> -- Number of I/O TLB slabs
6097 <int> -- Second integer after comma. Number of swiotlb
6098 areas with their own lock. Will be rounded up
6100 force -- force using of bounce buffers even if they
6101 wouldn't be automatically used by the kernel
6102 noforce -- Never use bounce buffers (for debugging)
6107 Set a sysctl parameter, right before loading the init
6108 process, as if the value was written to the respective
6109 /proc/sys/... file. Both '.' and '/' are recognized as
6110 separators. Unrecognized parameters and invalid values
6111 are reported in the kernel log. Sysctls registered
6112 later by a loaded module cannot be set this way.
6113 Example: sysctl.vm.swappiness=40
6115 sysfs.deprecated=0|1 [KNL]
6116 Enable/disable old style sysfs layout for old udev
6117 on older distributions. When this option is enabled
6118 very new udev will not work anymore. When this option
6119 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
6120 in older udev will not work anymore.
6121 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
6122 the kernel configuration.
6124 sysrq_always_enabled
6126 Ignore sysrq setting - this boot parameter will
6127 neutralize any effect of /proc/sys/kernel/sysrq.
6128 Useful for debugging.
6130 tcpmhash_entries= [KNL,NET]
6131 Set the number of tcp_metrics_hash slots.
6132 Default value is 8192 or 16384 depending on total
6133 ram pages. This is used to specify the TCP metrics
6134 cache size. See Documentation/networking/ip-sysctl.rst
6135 "tcp_no_metrics_save" section for more details.
6139 test_suspend= [SUSPEND]
6140 Format: { "mem" | "standby" | "freeze" }[,N]
6141 Specify "mem" (for Suspend-to-RAM) or "standby" (for
6142 standby suspend) or "freeze" (for suspend type freeze)
6143 as the system sleep state during system startup with
6144 the optional capability to repeat N number of times.
6145 The system is woken from this state using a
6146 wakeup-capable RTC alarm.
6148 thash_entries= [KNL,NET]
6149 Set number of hash buckets for TCP connection
6151 thermal.act= [HW,ACPI]
6152 -1: disable all active trip points in all thermal zones
6153 <degrees C>: override all lowest active trip points
6155 thermal.crt= [HW,ACPI]
6156 -1: disable all critical trip points in all thermal zones
6157 <degrees C>: override all critical trip points
6159 thermal.nocrt= [HW,ACPI]
6160 Set to disable actions on ACPI thermal zone
6161 critical and hot trip points.
6163 thermal.off= [HW,ACPI]
6164 1: disable ACPI thermal control
6166 thermal.psv= [HW,ACPI]
6167 -1: disable all passive trip points
6168 <degrees C>: override all passive trip points to this
6171 thermal.tzp= [HW,ACPI]
6172 Specify global default ACPI thermal zone polling rate
6173 <deci-seconds>: poll all this frequency
6174 0: no polling (default)
6177 Force threading of all interrupt handlers except those
6178 marked explicitly IRQF_NO_THREAD.
6182 Specify if the kernel should make use of the cpu
6183 topology information if the hardware supports this.
6184 The scheduler will make use of this information and
6185 e.g. base its process migration decisions on it.
6188 topology_updates= [KNL, PPC, NUMA]
6190 Specify if the kernel should ignore (off)
6191 topology updates sent by the hypervisor to this
6194 torture.disable_onoff_at_boot= [KNL]
6195 Prevent the CPU-hotplug component of torturing
6196 until after init has spawned.
6198 torture.ftrace_dump_at_shutdown= [KNL]
6199 Dump the ftrace buffer at torture-test shutdown,
6200 even if there were no errors. This can be a
6201 very costly operation when many torture tests
6202 are running concurrently, especially on systems
6203 with rotating-rust storage.
6205 torture.verbose_sleep_frequency= [KNL]
6206 Specifies how many verbose printk()s should be
6207 emitted between each sleep. The default of zero
6208 disables verbose-printk() sleeping.
6210 torture.verbose_sleep_duration= [KNL]
6211 Duration of each verbose-printk() sleep in jiffies.
6215 tpm_suspend_pcr=[HW,TPM]
6216 Format: integer pcr id
6217 Specify that at suspend time, the tpm driver
6218 should extend the specified pcr with zeros,
6219 as a workaround for some chips which fail to
6220 flush the last written pcr on TPM_SaveState.
6221 This will guarantee that all the other pcrs
6225 Have the tracepoints sent to printk as well as the
6226 tracing ring buffer. This is useful for early boot up
6227 where the system hangs or reboots and does not give the
6228 option for reading the tracing buffer or performing a
6229 ftrace_dump_on_oops.
6231 To turn off having tracepoints sent to printk,
6232 echo 0 > /proc/sys/kernel/tracepoint_printk
6233 Note, echoing 1 into this file without the
6234 tracepoint_printk kernel cmdline option has no effect.
6236 The tp_printk_stop_on_boot (see below) can also be used
6237 to stop the printing of events to console at
6242 Having tracepoints sent to printk() and activating high
6243 frequency tracepoints such as irq or sched, can cause
6244 the system to live lock.
6246 tp_printk_stop_on_boot [FTRACE]
6247 When tp_printk (above) is set, it can cause a lot of noise
6248 on the console. It may be useful to only include the
6249 printing of events during boot up, as user space may
6250 make the system inoperable.
6252 This command line option will stop the printing of events
6253 to console at the late_initcall_sync() time frame.
6255 trace_buf_size=nn[KMG]
6256 [FTRACE] will set tracing buffer size on each cpu.
6258 trace_clock= [FTRACE] Set the clock used for tracing events
6260 local - Use the per CPU time stamp counter
6261 (converted into nanoseconds). Fast, but
6262 depending on the architecture, may not be
6263 in sync between CPUs.
6264 global - Event time stamps are synchronize across
6265 CPUs. May be slower than the local clock,
6266 but better for some race conditions.
6267 counter - Simple counting of events (1, 2, ..)
6268 note, some counts may be skipped due to the
6269 infrastructure grabbing the clock more than
6271 uptime - Use jiffies as the time stamp.
6272 perf - Use the same clock that perf uses.
6273 mono - Use ktime_get_mono_fast_ns() for time stamps.
6274 mono_raw - Use ktime_get_raw_fast_ns() for time
6276 boot - Use ktime_get_boot_fast_ns() for time stamps.
6277 Architectures may add more clocks. See
6278 Documentation/trace/ftrace.rst for more details.
6280 trace_event=[event-list]
6281 [FTRACE] Set and start specified trace events in order
6282 to facilitate early boot debugging. The event-list is a
6283 comma-separated list of trace events to enable. See
6284 also Documentation/trace/events.rst
6286 trace_options=[option-list]
6287 [FTRACE] Enable or disable tracer options at boot.
6288 The option-list is a comma delimited list of options
6289 that can be enabled or disabled just as if you were
6290 to echo the option name into
6292 /sys/kernel/tracing/trace_options
6294 For example, to enable stacktrace option (to dump the
6295 stack trace of each event), add to the command line:
6297 trace_options=stacktrace
6299 See also Documentation/trace/ftrace.rst "trace options"
6302 trace_trigger=[trigger-list]
6303 [FTRACE] Add a event trigger on specific events.
6304 Set a trigger on top of a specific event, with an optional
6307 The format is is "trace_trigger=<event>.<trigger>[ if <filter>],..."
6308 Where more than one trigger may be specified that are comma deliminated.
6312 trace_trigger="sched_switch.stacktrace if prev_state == 2"
6314 The above will enable the "stacktrace" trigger on the "sched_switch"
6315 event but only trigger it if the "prev_state" of the "sched_switch"
6316 event is "2" (TASK_UNINTERUPTIBLE).
6318 See also "Event triggers" in Documentation/trace/events.rst
6322 [FTRACE] enable this option to disable tracing when a
6323 warning is hit. This turns off "tracing_on". Tracing can
6324 be enabled again by echoing '1' into the "tracing_on"
6325 file located in /sys/kernel/tracing/
6327 This option is useful, as it disables the trace before
6328 the WARNING dump is called, which prevents the trace to
6329 be filled with content caused by the warning output.
6331 This option can also be set at run time via the sysctl
6332 option: kernel/traceoff_on_warning
6334 transparent_hugepage=
6336 Format: [always|madvise|never]
6337 Can be used to control the default behavior of the system
6338 with respect to transparent hugepages.
6339 See Documentation/admin-guide/mm/transhuge.rst
6342 trusted.source= [KEYS]
6344 This parameter identifies the trust source as a backend
6345 for trusted keys implementation. Supported trust
6350 If not specified then it defaults to iterating through
6351 the trust source list starting with TPM and assigns the
6352 first trust source as a backend which is initialized
6353 successfully during iteration.
6357 The RNG used to generate key material for trusted keys.
6360 - the same value as trusted.source: "tpm" or "tee"
6362 If not specified, "default" is used. In this case,
6363 the RNG's choice is left to each individual trust source.
6365 tsc= Disable clocksource stability checks for TSC.
6367 [x86] reliable: mark tsc clocksource as reliable, this
6368 disables clocksource verification at runtime, as well
6369 as the stability checks done at bootup. Used to enable
6370 high-resolution timer mode on older hardware, and in
6371 virtualized environment.
6372 [x86] noirqtime: Do not use TSC to do irq accounting.
6373 Used to run time disable IRQ_TIME_ACCOUNTING on any
6374 platforms where RDTSC is slow and this accounting
6376 [x86] unstable: mark the TSC clocksource as unstable, this
6377 marks the TSC unconditionally unstable at bootup and
6378 avoids any further wobbles once the TSC watchdog notices.
6379 [x86] nowatchdog: disable clocksource watchdog. Used
6380 in situations with strict latency requirements (where
6381 interruptions from clocksource watchdog are not
6383 [x86] recalibrate: force recalibration against a HW timer
6384 (HPET or PM timer) on systems whose TSC frequency was
6385 obtained from HW or FW using either an MSR or CPUID(0x15).
6386 Warn if the difference is more than 500 ppm.
6387 [x86] watchdog: Use TSC as the watchdog clocksource with
6388 which to check other HW timers (HPET or PM timer), but
6389 only on systems where TSC has been deemed trustworthy.
6390 This will be suppressed by an earlier tsc=nowatchdog and
6391 can be overridden by a later tsc=nowatchdog. A console
6392 message will flag any such suppression or overriding.
6394 tsc_early_khz= [X86] Skip early TSC calibration and use the given
6395 value instead. Useful when the early TSC frequency discovery
6396 procedure is not reliable, such as on overclocked systems
6397 with CPUID.16h support and partial CPUID.15h support.
6398 Format: <unsigned int>
6400 tsx= [X86] Control Transactional Synchronization
6401 Extensions (TSX) feature in Intel processors that
6402 support TSX control.
6404 This parameter controls the TSX feature. The options are:
6406 on - Enable TSX on the system. Although there are
6407 mitigations for all known security vulnerabilities,
6408 TSX has been known to be an accelerator for
6409 several previous speculation-related CVEs, and
6410 so there may be unknown security risks associated
6411 with leaving it enabled.
6413 off - Disable TSX on the system. (Note that this
6414 option takes effect only on newer CPUs which are
6415 not vulnerable to MDS, i.e., have
6416 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get
6417 the new IA32_TSX_CTRL MSR through a microcode
6418 update. This new MSR allows for the reliable
6419 deactivation of the TSX functionality.)
6421 auto - Disable TSX if X86_BUG_TAA is present,
6422 otherwise enable TSX on the system.
6424 Not specifying this option is equivalent to tsx=off.
6426 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
6429 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async
6430 Abort (TAA) vulnerability.
6432 Similar to Micro-architectural Data Sampling (MDS)
6433 certain CPUs that support Transactional
6434 Synchronization Extensions (TSX) are vulnerable to an
6435 exploit against CPU internal buffers which can forward
6436 information to a disclosure gadget under certain
6439 In vulnerable processors, the speculatively forwarded
6440 data can be used in a cache side channel attack, to
6441 access data to which the attacker does not have direct
6444 This parameter controls the TAA mitigation. The
6447 full - Enable TAA mitigation on vulnerable CPUs
6450 full,nosmt - Enable TAA mitigation and disable SMT on
6451 vulnerable CPUs. If TSX is disabled, SMT
6452 is not disabled because CPU is not
6453 vulnerable to cross-thread TAA attacks.
6454 off - Unconditionally disable TAA mitigation
6456 On MDS-affected machines, tsx_async_abort=off can be
6457 prevented by an active MDS mitigation as both vulnerabilities
6458 are mitigated with the same mechanism so in order to disable
6459 this mitigation, you need to specify mds=off too.
6461 Not specifying this option is equivalent to
6462 tsx_async_abort=full. On CPUs which are MDS affected
6463 and deploy MDS mitigation, TAA mitigation is not
6464 required and doesn't provide any additional
6468 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
6470 turbografx.map[2|3]= [HW,JOY]
6471 TurboGraFX parallel port interface
6473 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
6474 See also Documentation/input/devices/joystick-parport.rst
6476 udbg-immortal [PPC] When debugging early kernel crashes that
6477 happen after console_init() and before a proper
6478 console driver takes over, this boot options might
6479 help "seeing" what's going on.
6481 uhash_entries= [KNL,NET]
6482 Set number of hash buckets for UDP/UDP-Lite connections
6485 [USB] Ignore overcurrent events (default N).
6486 Some badly-designed motherboards generate lots of
6487 bogus events, for ports that aren't wired to
6488 anything. Set this parameter to avoid log spamming.
6489 Note that genuine overcurrent events won't be
6493 [X86] Cause panic on unknown NMI.
6495 usbcore.authorized_default=
6496 [USB] Default USB device authorization:
6497 (default -1 = authorized except for wireless USB,
6498 0 = not authorized, 1 = authorized, 2 = authorized
6499 if device connected to internal port)
6501 usbcore.autosuspend=
6502 [USB] The autosuspend time delay (in seconds) used
6503 for newly-detected USB devices (default 2). This
6504 is the time required before an idle device will be
6505 autosuspended. Devices for which the delay is set
6506 to a negative value won't be autosuspended at all.
6508 usbcore.usbfs_snoop=
6509 [USB] Set to log all usbfs traffic (default 0 = off).
6511 usbcore.usbfs_snoop_max=
6512 [USB] Maximum number of bytes to snoop in each URB
6515 usbcore.blinkenlights=
6516 [USB] Set to cycle leds on hubs (default 0 = off).
6518 usbcore.old_scheme_first=
6519 [USB] Start with the old device initialization
6520 scheme (default 0 = off).
6522 usbcore.usbfs_memory_mb=
6523 [USB] Memory limit (in MB) for buffers allocated by
6524 usbfs (default = 16, 0 = max = 2047).
6526 usbcore.use_both_schemes=
6527 [USB] Try the other device initialization scheme
6528 if the first one fails (default 1 = enabled).
6530 usbcore.initial_descriptor_timeout=
6531 [USB] Specifies timeout for the initial 64-byte
6532 USB_REQ_GET_DESCRIPTOR request in milliseconds
6533 (default 5000 = 5.0 seconds).
6535 usbcore.nousb [USB] Disable the USB subsystem
6538 [USB] A list of quirk entries to augment the built-in
6539 usb core quirk list. List entries are separated by
6540 commas. Each entry has the form
6541 VendorID:ProductID:Flags. The IDs are 4-digit hex
6542 numbers and Flags is a set of letters. Each letter
6543 will change the built-in quirk; setting it if it is
6544 clear and clearing it if it is set. The letters have
6545 the following meanings:
6546 a = USB_QUIRK_STRING_FETCH_255 (string
6547 descriptors must not be fetched using
6549 b = USB_QUIRK_RESET_RESUME (device can't resume
6550 correctly so reset it instead);
6551 c = USB_QUIRK_NO_SET_INTF (device can't handle
6552 Set-Interface requests);
6553 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't
6554 handle its Configuration or Interface
6556 e = USB_QUIRK_RESET (device can't be reset
6557 (e.g morph devices), don't use reset);
6558 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has
6559 more interface descriptions than the
6560 bNumInterfaces count, and can't handle
6561 talking to these interfaces);
6562 g = USB_QUIRK_DELAY_INIT (device needs a pause
6563 during initialization, after we read
6564 the device descriptor);
6565 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For
6566 high speed and super speed interrupt
6567 endpoints, the USB 2.0 and USB 3.0 spec
6568 require the interval in microframes (1
6569 microframe = 125 microseconds) to be
6570 calculated as interval = 2 ^
6572 Devices with this quirk report their
6573 bInterval as the result of this
6574 calculation instead of the exponent
6575 variable used in the calculation);
6576 i = USB_QUIRK_DEVICE_QUALIFIER (device can't
6577 handle device_qualifier descriptor
6579 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device
6580 generates spurious wakeup, ignore
6581 remote wakeup capability);
6582 k = USB_QUIRK_NO_LPM (device can't handle Link
6584 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL
6585 (Device reports its bInterval as linear
6586 frames instead of the USB 2.0
6588 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs
6589 to be disconnected before suspend to
6590 prevent spurious wakeup);
6591 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a
6592 pause after every control message);
6593 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra
6594 delay after resetting its port);
6595 Example: quirks=0781:5580:bk,0a5c:5834:gij
6598 [USBHID] The interval which mice are to be polled at.
6601 [USBHID] The interval which joysticks are to be polled at.
6604 [USBHID] The interval which keyboards are to be polled at.
6606 usb-storage.delay_use=
6607 [UMS] The delay in seconds before a new device is
6608 scanned for Logical Units (default 1).
6611 [UMS] A list of quirks entries to supplement or
6612 override the built-in unusual_devs list. List
6613 entries are separated by commas. Each entry has
6614 the form VID:PID:Flags where VID and PID are Vendor
6615 and Product ID values (4-digit hex numbers) and
6616 Flags is a set of characters, each corresponding
6617 to a common usb-storage quirk flag as follows:
6618 a = SANE_SENSE (collect more than 18 bytes
6619 of sense data, not on uas);
6620 b = BAD_SENSE (don't collect more than 18
6621 bytes of sense data, not on uas);
6622 c = FIX_CAPACITY (decrease the reported
6623 device capacity by one sector);
6624 d = NO_READ_DISC_INFO (don't use
6625 READ_DISC_INFO command, not on uas);
6626 e = NO_READ_CAPACITY_16 (don't use
6627 READ_CAPACITY_16 command);
6628 f = NO_REPORT_OPCODES (don't use report opcodes
6630 g = MAX_SECTORS_240 (don't transfer more than
6631 240 sectors at a time, uas only);
6632 h = CAPACITY_HEURISTICS (decrease the
6633 reported device capacity by one
6634 sector if the number is odd);
6635 i = IGNORE_DEVICE (don't bind to this
6637 j = NO_REPORT_LUNS (don't use report luns
6639 k = NO_SAME (do not use WRITE_SAME, uas only)
6640 l = NOT_LOCKABLE (don't try to lock and
6641 unlock ejectable media, not on uas);
6642 m = MAX_SECTORS_64 (don't transfer more
6643 than 64 sectors = 32 KB at a time,
6645 n = INITIAL_READ10 (force a retry of the
6646 initial READ(10) command, not on uas);
6647 o = CAPACITY_OK (accept the capacity
6648 reported by the device, not on uas);
6649 p = WRITE_CACHE (the device cache is ON
6650 by default, not on uas);
6651 r = IGNORE_RESIDUE (the device reports
6652 bogus residue values, not on uas);
6653 s = SINGLE_LUN (the device has only one
6655 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
6656 commands, uas only);
6657 u = IGNORE_UAS (don't bind to the uas driver);
6658 w = NO_WP_DETECT (don't test whether the
6659 medium is write-protected).
6660 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE
6661 even if the device claims no cache,
6663 Example: quirks=0419:aaf5:rl,0421:0433:rc
6665 user_debug= [KNL,ARM]
6667 See arch/arm/Kconfig.debug help text.
6668 1 - undefined instruction events
6670 4 - invalid data aborts
6673 Example: user_debug=31
6676 [X86] Flags controlling user PTE allocations.
6678 nohigh = do not allocate PTE pages in
6679 HIGHMEM regardless of setting
6682 vdso= [X86,SH,SPARC]
6683 On X86_32, this is an alias for vdso32=. Otherwise:
6685 vdso=1: enable VDSO (the default)
6686 vdso=0: disable VDSO mapping
6688 vdso32= [X86] Control the 32-bit vDSO
6689 vdso32=1: enable 32-bit VDSO
6690 vdso32=0 or vdso32=2: disable 32-bit VDSO
6692 See the help text for CONFIG_COMPAT_VDSO for more
6693 details. If CONFIG_COMPAT_VDSO is set, the default is
6694 vdso32=0; otherwise, the default is vdso32=1.
6696 For compatibility with older kernels, vdso32=2 is an
6699 Try vdso32=0 if you encounter an error that says:
6700 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
6703 vector=percpu: enable percpu vector domain
6705 video= [FB] Frame buffer configuration
6706 See Documentation/fb/modedb.rst.
6708 video.brightness_switch_enabled= [ACPI]
6710 If set to 1, on receiving an ACPI notify event
6711 generated by hotkey, video driver will adjust brightness
6712 level and then send out the event to user space through
6713 the allocated input device. If set to 0, video driver
6714 will only send out the event without touching backlight
6719 [VMMIO] Memory mapped virtio (platform) device.
6721 <size>@<baseaddr>:<irq>[:<id>]
6723 <size> := size (can use standard suffixes
6725 <baseaddr> := physical base address
6726 <irq> := interrupt number (as passed to
6728 <id> := (optional) platform device id
6730 virtio_mmio.device=1K@0x100b0000:48:7
6732 Can be used multiple times for multiple devices.
6734 vga= [BOOT,X86-32] Select a particular video mode
6735 See Documentation/x86/boot.rst and
6736 Documentation/admin-guide/svga.rst.
6737 Use vga=ask for menu.
6738 This is actually a boot loader parameter; the value is
6739 passed to the kernel using a special protocol.
6741 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y.
6742 May slow down system boot speed, especially when
6743 enabled on systems with a large amount of memory.
6744 All options are enabled by default, and this
6745 interface is meant to allow for selectively
6746 enabling or disabling specific virtual memory
6749 Available options are:
6750 P Enable page structure init time poisoning
6751 - Disable all of the above options
6753 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
6754 size of <nn>. This can be used to increase the
6755 minimum size (128MB on x86). It can also be used to
6756 decrease the size and leave more room for directly
6759 vmcp_cma=nn[MG] [KNL,S390]
6760 Sets the memory size reserved for contiguous memory
6761 allocations for the vmcp device driver.
6763 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
6766 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
6769 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
6773 Controls the behavior of vsyscalls (i.e. calls to
6774 fixed addresses of 0xffffffffff600x00 from legacy
6775 code). Most statically-linked binaries and older
6776 versions of glibc use these calls. Because these
6777 functions are at fixed addresses, they make nice
6778 targets for exploits that can control RIP.
6780 emulate Vsyscalls turn into traps and are emulated
6781 reasonably safely. The vsyscall page is
6784 xonly [default] Vsyscalls turn into traps and are
6785 emulated reasonably safely. The vsyscall
6786 page is not readable.
6788 none Vsyscalls don't work at all. This makes
6789 them quite hard to use for exploits but
6790 might break your system.
6792 vt.color= [VT] Default text color.
6793 Format: 0xYX, X = foreground, Y = background.
6794 Default: 0x07 = light gray on black.
6796 vt.cur_default= [VT] Default cursor shape.
6797 Format: 0xCCBBAA, where AA, BB, and CC are the same as
6798 the parameters of the <Esc>[?A;B;Cc escape sequence;
6799 see VGA-softcursor.txt. Default: 2 = underline.
6801 vt.default_blu= [VT]
6802 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
6803 Change the default blue palette of the console.
6804 This is a 16-member array composed of values
6807 vt.default_grn= [VT]
6808 Format: <green0>,<green1>,<green2>,...,<green15>
6809 Change the default green palette of the console.
6810 This is a 16-member array composed of values
6813 vt.default_red= [VT]
6814 Format: <red0>,<red1>,<red2>,...,<red15>
6815 Change the default red palette of the console.
6816 This is a 16-member array composed of values
6822 Set system-wide default UTF-8 mode for all tty's.
6823 Default is 1, i.e. UTF-8 mode is enabled for all
6824 newly opened terminals.
6826 vt.global_cursor_default=
6829 Set system-wide default for whether a cursor
6830 is shown on new VTs. Default is -1,
6831 i.e. cursors will be created by default unless
6832 overridden by individual drivers. 0 will hide
6833 cursors, 1 will display them.
6835 vt.italic= [VT] Default color for italic text; 0-15.
6838 vt.underline= [VT] Default color for underlined text; 0-15.
6841 watchdog timers [HW,WDT] For information on watchdog timers,
6842 see Documentation/watchdog/watchdog-parameters.rst
6843 or other driver-specific files in the
6844 Documentation/watchdog/ directory.
6848 Set the hard lockup detector stall duration
6849 threshold in seconds. The soft lockup detector
6850 threshold is set to twice the value. A value of 0
6851 disables both lockup detectors. Default is 10
6854 workqueue.watchdog_thresh=
6855 If CONFIG_WQ_WATCHDOG is configured, workqueue can
6856 warn stall conditions and dump internal state to
6857 help debugging. 0 disables workqueue stall
6858 detection; otherwise, it's the stall threshold
6859 duration in seconds. The default value is 30 and
6860 it can be updated at runtime by writing to the
6861 corresponding sysfs file.
6863 workqueue.disable_numa
6864 By default, all work items queued to unbound
6865 workqueues are affine to the NUMA nodes they're
6866 issued on, which results in better behavior in
6867 general. If NUMA affinity needs to be disabled for
6868 whatever reason, this option can be used. Note
6869 that this also can be controlled per-workqueue for
6870 workqueues visible under /sys/bus/workqueue/.
6872 workqueue.power_efficient
6873 Per-cpu workqueues are generally preferred because
6874 they show better performance thanks to cache
6875 locality; unfortunately, per-cpu workqueues tend to
6876 be more power hungry than unbound workqueues.
6878 Enabling this makes the per-cpu workqueues which
6879 were observed to contribute significantly to power
6880 consumption unbound, leading to measurably lower
6881 power usage at the cost of small performance
6884 The default value of this parameter is determined by
6885 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
6887 workqueue.debug_force_rr_cpu
6888 Workqueue used to implicitly guarantee that work
6889 items queued without explicit CPU specified are put
6890 on the local CPU. This guarantee is no longer true
6891 and while local CPU is still preferred work items
6892 may be put on foreign CPUs. This debug option
6893 forces round-robin CPU selection to flush out
6894 usages which depend on the now broken guarantee.
6895 When enabled, memory and cache locality will be
6898 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
6899 default x2apic cluster mode on platforms
6902 xen_512gb_limit [KNL,X86-64,XEN]
6903 Restricts the kernel running paravirtualized under Xen
6904 to use only up to 512 GB of RAM. The reason to do so is
6905 crash analysis tools and Xen tools for doing domain
6906 save/restore/migration must be enabled to handle larger
6909 xen_emul_unplug= [HW,X86,XEN]
6910 Unplug Xen emulated devices
6911 Format: [unplug0,][unplug1]
6912 ide-disks -- unplug primary master IDE devices
6913 aux-ide-disks -- unplug non-primary-master IDE devices
6914 nics -- unplug network devices
6915 all -- unplug all emulated devices (NICs and IDE disks)
6916 unnecessary -- unplugging emulated devices is
6917 unnecessary even if the host did not respond to
6919 never -- do not unplug even if version check succeeds
6921 xen_legacy_crash [X86,XEN]
6922 Crash from Xen panic notifier, without executing late
6923 panic() code such as dumping handler.
6925 xen_msr_safe= [X86,XEN]
6927 Select whether to always use non-faulting (safe) MSR
6928 access functions when running as Xen PV guest. The
6929 default value is controlled by CONFIG_XEN_PV_MSR_SAFE.
6931 xen_nopvspin [X86,XEN]
6932 Disables the qspinlock slowpath using Xen PV optimizations.
6933 This parameter is obsoleted by "nopvspin" parameter, which
6934 has equivalent effect for XEN platform.
6937 Disables the PV optimizations forcing the HVM guest to
6938 run as generic HVM guest with no PV drivers.
6939 This option is obsoleted by the "nopv" option, which
6940 has equivalent effect for XEN platform.
6942 xen_no_vector_callback
6943 [KNL,X86,XEN] Disable the vector callback for Xen
6944 event channel interrupts.
6946 xen_scrub_pages= [XEN]
6947 Boolean option to control scrubbing pages before giving them back
6948 to Xen, for use by other domains. Can be also changed at runtime
6949 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages.
6950 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT.
6952 xen_timer_slop= [X86-64,XEN]
6953 Set the timer slop (in nanoseconds) for the virtual Xen
6954 timers (default is 100000). This adjusts the minimum
6955 delta of virtualized Xen timers, where lower values
6956 improve timer resolution at the expense of processing
6957 more timer interrupts.
6959 xen.balloon_boot_timeout= [XEN]
6960 The time (in seconds) to wait before giving up to boot
6961 in case initial ballooning fails to free enough memory.
6962 Applies only when running as HVM or PVH guest and
6963 started with less memory configured than allowed at
6964 max. Default is 180.
6966 xen.event_eoi_delay= [XEN]
6967 How long to delay EOI handling in case of event
6968 storms (jiffies). Default is 10.
6970 xen.event_loop_timeout= [XEN]
6971 After which time (jiffies) the event handling loop
6972 should start to delay EOI handling. Default is 2.
6974 xen.fifo_events= [XEN]
6975 Boolean parameter to disable using fifo event handling
6976 even if available. Normally fifo event handling is
6977 preferred over the 2-level event handling, as it is
6978 fairer and the number of possible event channels is
6979 much higher. Default is on (use fifo events).
6981 xirc2ps_cs= [NET,PCMCIA]
6983 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
6986 By default on POWER9 and above, the kernel will
6987 natively use the XIVE interrupt controller. This option
6988 allows the fallback firmware mode to be used:
6990 off Fallback to firmware control of XIVE interrupt
6991 controller on both pseries and powernv
6992 platforms. Only useful on POWER9 and above.
6994 xive.store-eoi=off [PPC]
6995 By default on POWER10 and above, the kernel will use
6996 stores for EOI handling when the XIVE interrupt mode
6997 is active. This option allows the XIVE driver to use
6998 loads instead, as on POWER9.
7000 xhci-hcd.quirks [USB,KNL]
7001 A hex value specifying bitmask with supplemental xhci
7002 host controller quirks. Meaning of each bit can be
7003 consulted in header drivers/usb/host/xhci.h.
7006 Format: { early | on | rw | ro | off }
7007 Controls if xmon debugger is enabled. Default is off.
7008 Passing only "xmon" is equivalent to "xmon=early".
7009 early Call xmon as early as possible on boot; xmon
7010 debugger is called from setup_arch().
7011 on xmon debugger hooks will be installed so xmon
7012 is only called on a kernel crash. Default mode,
7013 i.e. either "ro" or "rw" mode, is controlled
7014 with CONFIG_XMON_DEFAULT_RO_MODE.
7015 rw xmon debugger hooks will be installed so xmon
7016 is called only on a kernel crash, mode is write,
7017 meaning SPR registers, memory and, other data
7018 can be written using xmon commands.
7019 ro same as "rw" option above but SPR registers,
7020 memory, and other data can't be written using
7022 off xmon is disabled.
7026 Do not enable amd_pstate as the default
7027 scaling driver for the supported processors
7029 Use amd_pstate as a scaling driver, driver requests a
7030 desired performance on this abstract scale and the power
7031 management firmware translates the requests into actual
7032 hardware states (core frequency, data fabric and memory
7035 Use amd_pstate_epp driver instance as the scaling driver,
7036 driver provides a hint to the hardware if software wants
7037 to bias toward performance (0x0) or energy efficiency (0xff)
7038 to the CPPC firmware. then CPPC power algorithm will
7039 calculate the runtime workload and adjust the realtime cores