2 Format: { eager | lazy }
4 By default, unaccepted memory is accepted lazily to
5 avoid prolonged boot times. The lazy option will add
6 some runtime overhead until all memory is eventually
7 accepted. In most cases the overhead is negligible.
8 For some workloads or for debugging purposes
9 accept_memory=eager can be used to accept all memory
12 acpi= [HW,ACPI,X86,ARM64,RISCV64,EARLY]
13 Advanced Configuration and Power Interface
14 Format: { force | on | off | strict | noirq | rsdt |
16 force -- enable ACPI if default was off
17 on -- enable ACPI but allow fallback to DT [arm64,riscv64]
18 off -- disable ACPI if default was on
19 noirq -- do not use ACPI for IRQ routing
20 strict -- Be less tolerant of platforms that are not
21 strictly ACPI specification compliant.
22 rsdt -- prefer RSDT over (default) XSDT
23 copy_dsdt -- copy DSDT to memory
24 For ARM64 and RISCV64, ONLY "acpi=off", "acpi=on" or
25 "acpi=force" are available
27 See also Documentation/power/runtime_pm.rst, pci=noacpi
29 acpi_apic_instance= [ACPI,IOAPIC,EARLY]
31 2: use 2nd APIC table, if available
32 1,0: use 1st APIC table
35 acpi_backlight= [HW,ACPI]
36 { vendor | video | native | none }
37 If set to vendor, prefer vendor-specific driver
38 (e.g. thinkpad_acpi, sony_acpi, etc.) instead
39 of the ACPI video.ko driver.
40 If set to video, use the ACPI video.ko driver.
41 If set to native, use the device's native backlight mode.
42 If set to none, disable the ACPI backlight interface.
44 acpi_force_32bit_fadt_addr [ACPI,EARLY]
45 force FADT to use 32 bit addresses rather than the
46 64 bit X_* addresses. Some firmware have broken 64
47 bit addresses for force ACPI ignore these and use
48 the older legacy 32 bit addresses.
50 acpica_no_return_repair [HW, ACPI]
51 Disable AML predefined validation mechanism
52 This mechanism can repair the evaluation result to make
53 the return objects more ACPI specification compliant.
54 This option is useful for developers to identify the
55 root cause of an AML interpreter issue when the issue
56 has something to do with the repair mechanism.
58 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG]
59 acpi.debug_level= [HW,ACPI,ACPI_DEBUG]
61 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
62 debug output. Bits in debug_layer correspond to a
63 _COMPONENT in an ACPI source file, e.g.,
64 #define _COMPONENT ACPI_EVENTS
65 Bits in debug_level correspond to a level in
66 ACPI_DEBUG_PRINT statements, e.g.,
67 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
68 The debug_level mask defaults to "info". See
69 Documentation/firmware-guide/acpi/debug.rst for more information about
70 debug layers and levels.
72 Enable processor driver info messages:
73 acpi.debug_layer=0x20000000
74 Enable AML "Debug" output, i.e., stores to the Debug
75 object while interpreting AML:
76 acpi.debug_layer=0xffffffff acpi.debug_level=0x2
77 Enable all messages related to ACPI hardware:
78 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
80 Some values produce so much output that the system is
81 unusable. The "log_buf_len" parameter may be useful
82 if you need to capture more output.
84 acpi_enforce_resources= [ACPI]
86 Check for resource conflicts between native drivers
87 and ACPI OperationRegions (SystemIO and SystemMemory
88 only). IO ports and memory declared in ACPI might be
89 used by the ACPI subsystem in arbitrary AML code and
90 can interfere with legacy drivers.
91 strict (default): access to resources claimed by ACPI
92 is denied; legacy drivers trying to access reserved
93 resources will fail to bind to device using them.
94 lax: access to resources claimed by ACPI is allowed;
95 legacy drivers trying to access reserved resources
96 will bind successfully but a warning message is logged.
97 no: ACPI OperationRegions are not marked as reserved,
98 no further checks are performed.
100 acpi_force_table_verification [HW,ACPI,EARLY]
101 Enable table checksum verification during early stage.
102 By default, this is disabled due to x86 early mapping
105 acpi_irq_balance [HW,ACPI]
106 ACPI will balance active IRQs
109 acpi_irq_nobalance [HW,ACPI]
110 ACPI will not move active IRQs (default)
113 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA
114 Format: <irq>,<irq>...
116 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for
118 Format: <irq>,<irq>...
120 acpi_mask_gpe= [HW,ACPI]
121 Due to the existence of _Lxx/_Exx, some GPEs triggered
122 by unsupported hardware/firmware features can result in
123 GPE floodings that cannot be automatically disabled by
125 This facility can be used to prevent such uncontrolled
127 Format: <byte> or <bitmap-list>
129 acpi_no_auto_serialize [HW,ACPI]
130 Disable auto-serialization of AML methods
131 AML control methods that contain the opcodes to create
132 named objects will be marked as "Serialized" by the
133 auto-serialization feature.
134 This feature is enabled by default.
135 This option allows to turn off the feature.
137 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump
140 acpi_no_static_ssdt [HW,ACPI,EARLY]
141 Disable installation of static SSDTs at early boot time
142 By default, SSDTs contained in the RSDT/XSDT will be
143 installed automatically and they will appear under
144 /sys/firmware/acpi/tables.
145 This option turns off this feature.
146 Note that specifying this option does not affect
147 dynamic table installation which will install SSDT
148 tables to /sys/firmware/acpi/tables/dynamic.
150 acpi_no_watchdog [HW,ACPI,WDT]
151 Ignore the ACPI-based watchdog interface (WDAT) and let
152 a native driver control the watchdog device instead.
154 acpi_rsdp= [ACPI,EFI,KEXEC,EARLY]
155 Pass the RSDP address to the kernel, mostly used
156 on machines running EFI runtime service to boot the
157 second kernel for kdump.
159 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
160 Format: To spoof as Windows 98: ="Microsoft Windows"
162 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead
163 of 2 which is mandated by ACPI 6) as the supported ACPI
164 specification revision (when using this switch, it may
165 be necessary to carry out a cold reboot _twice_ in a
166 row to make it take effect on the platform firmware).
168 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
169 acpi_osi="string1" # add string1
170 acpi_osi="!string2" # remove string2
171 acpi_osi=!* # remove all strings
172 acpi_osi=! # disable all built-in OS vendor
174 acpi_osi=!! # enable all built-in OS vendor
176 acpi_osi= # disable all strings
178 'acpi_osi=!' can be used in combination with single or
179 multiple 'acpi_osi="string1"' to support specific OS
180 vendor string(s). Note that such command can only
181 affect the default state of the OS vendor strings, thus
182 it cannot affect the default state of the feature group
183 strings and the current state of the OS vendor strings,
184 specifying it multiple times through kernel command line
185 is meaningless. This command is useful when one do not
186 care about the state of the feature group strings which
187 should be controlled by the OSPM.
189 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
190 to 'acpi_osi="Windows 2000" acpi_osi=!', they all
191 can make '_OSI("Windows 2000")' TRUE.
193 'acpi_osi=' cannot be used in combination with other
194 'acpi_osi=' command lines, the _OSI method will not
195 exist in the ACPI namespace. NOTE that such command can
196 only affect the _OSI support state, thus specifying it
197 multiple times through kernel command line is also
200 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
203 'acpi_osi=!*' can be used in combination with single or
204 multiple 'acpi_osi="string1"' to support specific
205 string(s). Note that such command can affect the
206 current state of both the OS vendor strings and the
207 feature group strings, thus specifying it multiple times
208 through kernel command line is meaningful. But it may
209 still not able to affect the final state of a string if
210 there are quirks related to this string. This command
211 is useful when one want to control the state of the
212 feature group strings to debug BIOS issues related to
215 1. 'acpi_osi="Module Device" acpi_osi=!*' can make
216 '_OSI("Module Device")' FALSE.
217 2. 'acpi_osi=!* acpi_osi="Module Device"' can make
218 '_OSI("Module Device")' TRUE.
219 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
221 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
223 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
224 they all will make '_OSI("Windows 2000")' TRUE.
227 Override the pmtimer bug detection: force the kernel
228 to assume that this machine's pmtimer latches its value
229 and always returns good values.
231 acpi_sci= [HW,ACPI,EARLY] ACPI System Control Interrupt trigger mode
232 Format: { level | edge | high | low }
234 acpi_skip_timer_override [HW,ACPI,EARLY]
235 Recognize and ignore IRQ0/pin2 Interrupt Override.
236 For broken nForce2 BIOS resulting in XT-PIC timer.
238 acpi_sleep= [HW,ACPI] Sleep options
239 Format: { s3_bios, s3_mode, s3_beep, s4_hwsig,
240 s4_nohwsig, old_ordering, nonvs,
241 sci_force_enable, nobl }
242 See Documentation/power/video.rst for information on
244 s3_beep is for debugging; it makes the PC's speaker beep
245 as soon as the kernel's real-mode entry point is called.
246 s4_hwsig causes the kernel to check the ACPI hardware
247 signature during resume from hibernation, and gracefully
248 refuse to resume if it has changed. This complies with
249 the ACPI specification but not with reality, since
250 Windows does not do this and many laptops do change it
251 on docking. So the default behaviour is to allow resume
252 and simply warn when the signature changes, unless the
253 s4_hwsig option is enabled.
254 s4_nohwsig prevents ACPI hardware signature from being
255 used (or even warned about) during resume.
256 old_ordering causes the ACPI 1.0 ordering of the _PTS
257 control method, with respect to putting devices into
258 low power states, to be enforced (the ACPI 2.0 ordering
259 of _PTS is used by default).
260 nonvs prevents the kernel from saving/restoring the
261 ACPI NVS memory during suspend/hibernation and resume.
262 sci_force_enable causes the kernel to set SCI_EN directly
263 on resume from S1/S3 (which is against the ACPI spec,
264 but some broken systems don't work without it).
265 nobl causes the internal blacklist of systems known to
266 behave incorrectly in some ways with respect to system
267 suspend and resume to be ignored (use wisely).
269 acpi_use_timer_override [HW,ACPI,EARLY]
270 Use timer override. For some broken Nvidia NF5 boards
271 that require a timer override, but don't have HPET
273 add_efi_memmap [EFI,X86,EARLY] Include EFI memory map in
274 kernel's map of available physical RAM.
277 { off | try_unsupported }
278 off: disable AGP support
279 try_unsupported: try to drive unsupported chipsets
280 (may crash computer or cause data corruption)
283 See Documentation/sound/alsa-configuration.rst
286 Allow the default userspace alignment fault handler
287 behaviour to be specified. Bit 0 enables warnings,
288 bit 1 enables fixups, and bit 2 sends a segfault.
290 align_va_addr= [X86-64]
291 Align virtual addresses by clearing slice [14:12] when
292 allocating a VMA at process creation time. This option
293 gives you up to 3% performance improvement on AMD F15h
294 machines (where it is enabled by default) for a
295 CPU-intensive style benchmark, and it can vary highly in
296 a microbenchmark depending on workload and compiler.
298 32: only for 32-bit processes
299 64: only for 64-bit processes
300 on: enable for both 32- and 64-bit processes
301 off: disable for both 32- and 64-bit processes
303 alloc_snapshot [FTRACE]
304 Allocate the ftrace snapshot buffer on boot up when the
305 main buffer is allocated. This is handy if debugging
306 and you need to use tracing_snapshot() on boot up, and
307 do not want to use tracing_snapshot_alloc() as it needs
308 to be done where GFP_KERNEL allocations are allowed.
310 allow_mismatched_32bit_el0 [ARM64,EARLY]
311 Allow execve() of 32-bit applications and setting of the
312 PER_LINUX32 personality on systems where only a strict
313 subset of the CPUs support 32-bit EL0. When this
314 parameter is present, the set of CPUs supporting 32-bit
315 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0
316 and hot-unplug operations may be restricted.
318 See Documentation/arch/arm64/asymmetric-32bit.rst for more
321 amd_iommu= [HW,X86-64]
322 Pass parameters to the AMD IOMMU driver in the system.
324 fullflush - Deprecated, equivalent to iommu.strict=1
325 off - do not initialize any AMD IOMMU found in
327 force_isolation - Force device isolation for all
328 devices. The IOMMU driver is not
329 allowed anymore to lift isolation
330 requirements as needed. This option
331 does not override iommu=pt
332 force_enable - Force enable the IOMMU on platforms known
333 to be buggy with IOMMU enabled. Use this
335 pgtbl_v1 - Use v1 page table for DMA-API (Default).
336 pgtbl_v2 - Use v2 page table for DMA-API.
337 irtcachedis - Disable Interrupt Remapping Table (IRT) caching.
339 amd_iommu_dump= [HW,X86-64]
340 Enable AMD IOMMU driver option to dump the ACPI table
341 for AMD IOMMU. With this option enabled, AMD IOMMU
342 driver will print ACPI tables for AMD IOMMU during
343 IOMMU initialization.
345 amd_iommu_intr= [HW,X86-64]
346 Specifies one of the following AMD IOMMU interrupt
348 legacy - Use legacy interrupt remapping mode.
349 vapic - Use virtual APIC mode, which allows IOMMU
350 to inject interrupts directly into guest.
351 This mode requires kvm-amd.avic=1.
352 (Default when IOMMU HW support is present.)
354 amd_pstate= [X86,EARLY]
356 Do not enable amd_pstate as the default
357 scaling driver for the supported processors
359 Use amd_pstate with passive mode as a scaling driver.
360 In this mode autonomous selection is disabled.
361 Driver requests a desired performance level and platform
362 tries to match the same performance level if it is
363 satisfied by guaranteed performance level.
365 Use amd_pstate_epp driver instance as the scaling driver,
366 driver provides a hint to the hardware if software wants
367 to bias toward performance (0x0) or energy efficiency (0xff)
368 to the CPPC firmware. then CPPC power algorithm will
369 calculate the runtime workload and adjust the realtime cores
372 Activate guided autonomous mode. Driver requests minimum and
373 maximum performance level and the platform autonomously
374 selects a performance level in this range and appropriate
375 to the current workload.
380 Disable amd-pstate preferred core.
382 amijoy.map= [HW,JOY] Amiga joystick support
383 Map of devices attached to JOY0DAT and JOY1DAT
385 See also Documentation/input/joydev/joystick.rst
387 analog.map= [HW,JOY] Analog joystick and gamepad support
388 Specifies type or capabilities of an analog joystick
389 connected to one of 16 gameports
390 Format: <type1>,<type2>,..<type16>
393 Power management functions (SPARCstation-4/5 + deriv.)
395 Disable APC CPU standby support. SPARCstation-Fox does
396 not play well with APC CPU idle - disable it if you have
397 APC and your system crashes randomly.
399 apic= [APIC,X86,EARLY] Advanced Programmable Interrupt Controller
400 Change the output verbosity while booting
401 Format: { quiet (default) | verbose | debug }
402 Change the amount of debugging information output
403 when initialising the APIC and IO-APIC components.
404 For X86-32, this can also be used to specify an APIC
406 Format: apic=driver_name
407 Examples: apic=bigsmp
409 apic_extnmi= [APIC,X86,EARLY] External NMI delivery setting
410 Format: { bsp (default) | all | none }
411 bsp: External NMI is delivered only to CPU 0
412 all: External NMIs are broadcast to all CPUs as a
414 none: External NMI is masked for all CPUs. This is
415 useful so that a dump capture kernel won't be
419 See Documentation/networking/ipv6.rst.
421 apm= [APM] Advanced Power Management
422 See header of arch/x86/kernel/apm_32.c.
424 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
425 Format: { "0" | "1" }
426 See security/apparmor/Kconfig help text
429 Default value is set via kernel config option.
431 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
432 Format: <io>,<irq>,<nodeID>
434 arm64.nobti [ARM64] Unconditionally disable Branch Target
435 Identification support
437 arm64.nomops [ARM64] Unconditionally disable Memory Copy and Memory
438 Set instructions support
440 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension
443 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication
446 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix
449 arm64.nosve [ARM64] Unconditionally disable Scalable Vector
454 atarimouse= [HW,MOUSE] Atari Mouse
456 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
457 EzKey and similar keyboards
459 atkbd.reset= [HW] Reset keyboard during initialization
461 atkbd.set= [HW] Select keyboard code set
462 Format: <int> (2 = AT (default), 3 = PS/2)
464 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
467 atkbd.softraw= [HW] Choose between synthetic and real raw mode
468 Format: <bool> (0 = real, 1 = synthetic (default))
470 atkbd.softrepeat= [HW]
471 Use software keyboard repeat
473 audit= [KNL] Enable the audit sub-system
474 Format: { "0" | "1" | "off" | "on" }
475 0 | off - kernel audit is disabled and can not be
476 enabled until the next reboot
477 unset - kernel audit is initialized but disabled and
478 will be fully enabled by the userspace auditd.
479 1 | on - kernel audit is initialized and partially
480 enabled, storing at most audit_backlog_limit
481 messages in RAM until it is fully enabled by the
485 audit_backlog_limit= [KNL] Set the audit queue size limit.
486 Format: <int> (must be >=0)
489 bau= [X86_UV] Enable the BAU on SGI UV. The default
490 behavior is to disable the BAU (i.e. bau=0).
491 Format: { "0" | "1" }
494 unset - Disable the BAU.
496 baycom_epp= [HW,AX25]
499 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
501 See header of drivers/net/hamradio/baycom_par.c.
503 baycom_ser_fdx= [HW,AX25]
504 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
505 Format: <io>,<irq>,<mode>[,<baud>]
506 See header of drivers/net/hamradio/baycom_ser_fdx.c.
508 baycom_ser_hdx= [HW,AX25]
509 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
510 Format: <io>,<irq>,<mode>
511 See header of drivers/net/hamradio/baycom_ser_hdx.c.
514 Disable BERT OS support on buggy BIOSes.
516 bgrt_disable [ACPI,X86,EARLY]
517 Disable BGRT to avoid flickering OEM logo.
519 blkdevparts= Manual partition parsing of block device(s) for
520 embedded devices based on command line input.
521 See Documentation/block/cmdline-partition.rst
523 boot_delay= [KNL,EARLY]
524 Milliseconds to delay each printk during boot.
525 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled,
526 and you may also have to specify "lpj=". Boot_delay
527 values larger than 10 seconds (10000) are assumed
528 erroneous and ignored.
531 bootconfig [KNL,EARLY]
532 Extended command line options can be added to an initrd
533 and this will cause the kernel to look for it.
535 See Documentation/admin-guide/bootconfig.rst
537 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
538 bttv.radio= Most important insmod options are available as
540 bttv.pll= See Documentation/admin-guide/media/bttv.rst
543 bulk_remove=off [PPC] This parameter disables the use of the pSeries
544 firmware feature for flushing multiple hpte entries
547 c101= [NET] Moxa C101 synchronous serial card
549 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
550 Sometimes CPU hardware bugs make them report the cache
551 size incorrectly. The kernel will attempt work arounds
552 to fix known problems, but for some CPUs it is not
553 possible to determine what the correct size should be.
554 This option provides an override for these situations.
557 [NET] Specifies amount of time (in seconds) that
558 the kernel should wait for a network carrier. By default
559 it waits 120 seconds.
561 ca_keys= [KEYS] This parameter identifies a specific key(s) on
562 the system trusted keyring to be used for certificate
564 format: { id:<keyid> | builtin }
566 cca= [MIPS,EARLY] Override the kernel pages' cache coherency
567 algorithm. Accepted values range from 0 to 7
568 inclusive. See arch/mips/include/asm/pgtable-bits.h
569 for platform specific values (SB1, Loongson3 and
572 ccw_timeout_log [S390]
573 See Documentation/arch/s390/common_io.rst for details.
575 cgroup_disable= [KNL] Disable a particular controller or optional feature
576 Format: {name of the controller(s) or feature(s) to disable}
577 The effects of cgroup_disable=foo are:
578 - foo isn't auto-mounted if you mount all cgroups in
580 - foo isn't visible as an individually mountable
582 - if foo is an optional feature then the feature is
583 disabled and corresponding cgroup files are not
585 {Currently only "memory" controller deal with this and
586 cut the overhead, others just disable the usage. So
587 only cgroup_disable=memory is actually worthy}
588 Specifying "pressure" disables per-cgroup pressure
589 stall information accounting feature
591 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1
592 Format: { { controller | "all" | "named" }
593 [,{ controller | "all" | "named" }...] }
594 Like cgroup_disable, but only applies to cgroup v1;
595 the blacklisted controllers remain available in cgroup2.
596 "all" blacklists all controllers and "named" disables
597 named mounts. Specifying both "all" and "named" disables
600 cgroup_favordynmods= [KNL] Enable or Disable favordynmods.
601 Format: { "true" | "false" }
602 Defaults to the value of CONFIG_CGROUP_FAVOR_DYNMODS.
604 cgroup.memory= [KNL] Pass options to the cgroup memory controller.
606 nosocket -- Disable socket memory accounting.
607 nokmem -- Disable kernel memory accounting.
608 nobpf -- Disable BPF memory accounting.
610 checkreqprot= [SELINUX] Set initial checkreqprot flag value.
611 Format: { "0" | "1" }
612 See security/selinux/Kconfig help text.
613 0 -- check protection applied by kernel (includes
614 any implied execute protection).
615 1 -- check protection requested by application.
616 Default value is set via a kernel config option.
617 Value can be changed at runtime via
618 /sys/fs/selinux/checkreqprot.
619 Setting checkreqprot to 1 is deprecated.
622 See Documentation/arch/s390/common_io.rst for details.
624 clearcpuid=X[,X...] [X86]
625 Disable CPUID feature X for the kernel. See
626 arch/x86/include/asm/cpufeatures.h for the valid bit
627 numbers X. Note the Linux-specific bits are not necessarily
628 stable over kernel options, but the vendor-specific
630 X can also be a string as appearing in the flags: line
631 in /proc/cpuinfo which does not have the above
632 instability issue. However, not all features have names
634 Note that using this option will taint your kernel.
635 Also note that user programs calling CPUID directly
636 or using the feature without checking anything
637 will still see it. This just prevents it from
638 being used by the kernel or shown in /proc/cpuinfo.
639 Also note the kernel might malfunction if you disable
644 Prevents the clock framework from automatically gating
645 clocks that have not been explicitly enabled by a Linux
646 device driver but are enabled in hardware at reset or
647 by the bootloader/firmware. Note that this does not
648 force such clocks to be always-on nor does it reserve
649 those clocks in any way. This parameter is useful for
650 debug and development, but should not be needed on a
651 platform with proper driver support. For more
652 information, see Documentation/driver-api/clk.rst.
654 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
656 Forces specified clocksource (if available) to be used
657 when calculating gettimeofday(). If specified
658 clocksource is not available, it defaults to PIT.
659 Format: { pit | tsc | cyclone | pmtmr }
661 clocksource= Override the default clocksource
663 Override the default clocksource and use the clocksource
664 with the name specified.
665 Some clocksource names to choose from, depending on
667 [all] jiffies (this is the base, fallback clocksource)
669 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
670 pxa_timer,timer3,32k_counter,timer0_1
671 [X86-32] pit,hpet,tsc;
672 scx200_hrt on Geode; cyclone on IBM x440
680 clocksource.arm_arch_timer.evtstrm=
683 Enable/disable the eventstream feature of the ARM
684 architected timer so that code using WFE-based polling
685 loops can be debugged more effectively on production
688 clocksource.verify_n_cpus= [KNL]
689 Limit the number of CPUs checked for clocksources
690 marked with CLOCK_SOURCE_VERIFY_PERCPU that
691 are marked unstable due to excessive skew.
692 A negative value says to check all CPUs, while
693 zero says not to check any. Values larger than
694 nr_cpu_ids are silently truncated to nr_cpu_ids.
695 The actual CPUs are chosen randomly, with
696 no replacement if the same CPU is chosen twice.
698 clocksource-wdtest.holdoff= [KNL]
699 Set the time in seconds that the clocksource
700 watchdog test waits before commencing its tests.
701 Defaults to zero when built as a module and to
702 10 seconds when built into the kernel.
704 cma=nn[MG]@[start[MG][-end[MG]]]
706 Sets the size of kernel global memory area for
707 contiguous memory allocations and optionally the
708 placement constraint by the physical address range of
709 memory allocations. A value of 0 disables CMA
710 altogether. For more information, see
711 kernel/dma/contiguous.c
715 Sets the size of kernel per-numa memory area for
716 contiguous memory allocations. A value of 0 disables
717 per-numa CMA altogether. And If this option is not
718 specified, the default value is 0.
719 With per-numa CMA enabled, DMA users on node nid will
720 first try to allocate buffer from the pernuma area
721 which is located in node nid, if the allocation fails,
722 they will fallback to the global default memory area.
724 numa_cma=<node>:nn[MG][,<node>:nn[MG]]
726 Sets the size of kernel numa memory area for
727 contiguous memory allocations. It will reserve CMA
728 area for the specified node.
730 With numa CMA enabled, DMA users on node nid will
731 first try to allocate buffer from the numa area
732 which is located in node nid, if the allocation fails,
733 they will fallback to the global default memory area.
735 cmo_free_hint= [PPC] Format: { yes | no }
736 Specify whether pages are marked as being inactive
737 when they are freed. This is used in CMO environments
738 to determine OS memory pressure for page stealing by
742 coherent_pool=nn[KMG] [ARM,KNL,EARLY]
743 Sets the size of memory pool for coherent, atomic dma
744 allocations, by default set to 256K.
746 com20020= [HW,NET] ARCnet - COM20020 chipset
748 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
750 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
754 ARCnet - COM90xx chipset (memory-mapped buffers)
755 Format: <io>[,<irq>[,<memstart>]]
757 condev= [HW,S390] console device
760 con3215_drop= [S390,EARLY] 3215 console drop mode.
762 When set to true, drop data on the 3215 console when
763 the console buffer is full. In this case the
764 operator using a 3270 terminal emulator (for example
765 x3270) does not have to enter the clear key for the
766 console output to advance and the kernel to continue.
767 This leads to a much faster boot time when a 3270
768 terminal emulator is active. If no 3270 terminal
769 emulator is used, this parameter has no effect.
771 console= [KNL] Output console device and options.
773 tty<n> Use the virtual console device <n>.
777 Use the specified serial port. The options are of
778 the form "bbbbpnf", where "bbbb" is the baud rate,
779 "p" is parity ("n", "o", or "e"), "n" is number of
780 bits, and "f" is flow control ("r" for RTS or
781 omit it). Default is "9600n8".
783 See Documentation/admin-guide/serial-console.rst for more
785 Documentation/networking/netconsole.rst for an
788 uart[8250],io,<addr>[,options]
789 uart[8250],mmio,<addr>[,options]
790 uart[8250],mmio16,<addr>[,options]
791 uart[8250],mmio32,<addr>[,options]
792 uart[8250],0x<addr>[,options]
793 Start an early, polled-mode console on the 8250/16550
794 UART at the specified I/O port or MMIO address,
795 switching to the matching ttyS device later.
796 MMIO inter-register address stride is either 8-bit
797 (mmio), 16-bit (mmio16), or 32-bit (mmio32).
798 If none of [io|mmio|mmio16|mmio32], <addr> is assumed
799 to be equivalent to 'mmio'. 'options' are specified in
800 the same format described for ttyS above; if unspecified,
801 the h/w is not re-initialized.
803 hvc<n> Use the hypervisor console device <n>. This is for
804 both Xen and PowerPC hypervisors.
807 Use to disable console output, i.e., to have kernel
808 console messages discarded.
809 This must be the only console= parameter used on the
812 If the device connected to the port is not a TTY but a braille
813 device, prepend "brl," before the device type, for instance
815 For now, only VisioBraille is supported.
818 [KNL] Change console messages format
820 By default we print messages on consoles in
821 "[time stamp] text\n" format (time stamp may not be
822 printed, depending on CONFIG_PRINTK_TIME or
823 `printk_time' param).
825 Switch to syslog format: "<%u>[time stamp] text\n"
826 IOW, each message will have a facility and loglevel
827 prefix. The format is similar to one used by syslog()
828 syscall, or to executing "dmesg -S --raw" or to reading
831 consoleblank= [KNL] The console blank (screen saver) timeout in
832 seconds. A value of 0 disables the blank timer.
836 [KNL] Change the default value for
837 /proc/<pid>/coredump_filter.
838 See also Documentation/filesystems/proc.rst.
840 coresight_cpu_debug.enable
843 Enable/disable the CPU sampling based debugging.
844 0: default value, disable debugging
845 1: enable debugging at boot time
847 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
849 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
851 cpuidle.off=1 [CPU_IDLE]
852 disable the cpuidle sub-system
855 [CPU_IDLE] Name of the cpuidle governor to use.
857 cpufreq.off=1 [CPU_FREQ]
858 disable the cpufreq sub-system
860 cpufreq.default_governor=
861 [CPU_FREQ] Name of the default cpufreq governor or
862 policy to use. This governor must be registered in the
863 kernel before the cpufreq driver probes.
866 [X86,EARLY] Delay for N microsec between assert and de-assert
867 of APIC INIT to start processors. This delay occurs
868 on every CPU online, such as boot, and resume from suspend.
872 [SMP] Enable/disable parallel bringup of secondary CPUs
874 Default is enabled if CONFIG_HOTPLUG_PARALLEL=y. Otherwise
875 the parameter has no effect.
877 crash_kexec_post_notifiers
878 Run kdump after running panic-notifiers and dumping
879 kmsg. This only for the users who doubt kdump always
880 succeeds in any situation.
881 Note that this also increases risks of kdump failure,
882 because some panic notifiers can make the crashed
883 kernel more unstable.
885 crashkernel=size[KMG][@offset[KMG]]
886 [KNL,EARLY] Using kexec, Linux can switch to a 'crash kernel'
887 upon panic. This parameter reserves the physical
888 memory region [offset, offset + size] for that kernel
889 image. If '@offset' is omitted, then a suitable offset
890 is selected automatically.
891 [KNL, X86-64, ARM64, RISCV, LoongArch] Select a region
892 under 4G first, and fall back to reserve region above
893 4G when '@offset' hasn't been specified.
894 See Documentation/admin-guide/kdump/kdump.rst for further details.
896 crashkernel=range1:size1[,range2:size2,...][@offset]
897 [KNL] Same as above, but depends on the memory
898 in the running system. The syntax of range is
899 start-[end] where start and end are both
900 a memory unit (amount[KMG]). See also
901 Documentation/admin-guide/kdump/kdump.rst for an example.
903 crashkernel=size[KMG],high
904 [KNL, X86-64, ARM64, RISCV, LoongArch] range could be
906 Allow kernel to allocate physical memory region from top,
907 so could be above 4G if system have more than 4G ram
908 installed. Otherwise memory region will be allocated
909 below 4G, if available.
910 It will be ignored if crashkernel=X is specified.
911 crashkernel=size[KMG],low
912 [KNL, X86-64, ARM64, RISCV, LoongArch] range under 4G.
913 When crashkernel=X,high is passed, kernel could allocate
914 physical memory region above 4G, that cause second kernel
915 crash on system that require some amount of low memory,
916 e.g. swiotlb requires at least 64M+32K low memory, also
917 enough extra low memory is needed to make sure DMA buffers
918 for 32-bit devices won't run out. Kernel would try to allocate
919 default size of memory below 4G automatically. The default
920 size is platform dependent.
921 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB)
924 --> loongarch: 128MiB
925 This one lets the user specify own low range under 4G
926 for second kernel instead.
927 0: to disable low allocation.
928 It will be ignored when crashkernel=X,high is not used
929 or memory reserved is below 4G.
932 [KNL] Disable crypto self-tests
937 cs89x0_media= [HW,NET]
938 Format: { rj45 | aui | bnc }
940 csdlock_debug= [KNL] Enable or disable debug add-ons of cross-CPU
941 function call handling. When switched on,
942 additional debug data is printed to the console
943 in case a hanging CPU is detected, and that
944 CPU is pinged again in order to try to resolve
945 the hang situation. The default value of this
946 option depends on the CSD_LOCK_WAIT_DEBUG_DEFAULT
950 See header of drivers/s390/block/dasd_devmap.c.
952 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
953 (one device per port)
954 Format: <port#>,<type>
955 See also Documentation/input/devices/joystick-parport.rst
957 debug [KNL,EARLY] Enable kernel debugging (events log level).
960 [KNL,EARLY] Enable printing [hashed] pointers early in the
961 boot sequence. If enabled, we use a weak hash instead
962 of siphash to hash pointers. Use this option if you are
963 seeing instances of '(___ptrval___)') and need to see a
964 value (hashed pointer) instead. Cryptographically
965 insecure, please do not use on production kernels.
968 [KNL] verbose locking self-tests
970 Print debugging info while doing the locking API
972 Bitmask for the various LOCKTYPE_ tests. Defaults to 0
973 (no extra messages), setting it to -1 (all bits set)
974 will print _a_lot_ more information - normally only
975 useful to lockdep developers.
977 debug_objects [KNL,EARLY] Enable object debugging
979 debug_guardpage_minorder=
980 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this
981 parameter allows control of the order of pages that will
982 be intentionally kept free (and hence protected) by the
983 buddy allocator. Bigger value increase the probability
984 of catching random memory corruption, but reduce the
985 amount of memory for normal system use. The maximum
986 possible value is MAX_PAGE_ORDER/2. Setting this
987 parameter to 1 or 2 should be enough to identify most
988 random memory corruption problems caused by bugs in
989 kernel or driver code when a CPU writes to (or reads
990 from) a random memory location. Note that there exists
991 a class of memory corruptions problems caused by buggy
992 H/W or F/W or by drivers badly programming DMA
993 (basically when memory is written at bus level and the
994 CPU MMU is bypassed) which are not detectable by
995 CONFIG_DEBUG_PAGEALLOC, hence this option will not
996 help tracking down these problems.
999 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this parameter
1000 enables the feature at boot time. By default, it is
1001 disabled and the system will work mostly the same as a
1002 kernel built without CONFIG_DEBUG_PAGEALLOC.
1003 Note: to get most of debug_pagealloc error reports, it's
1004 useful to also enable the page_owner functionality.
1005 on: enable the feature
1007 debugfs= [KNL,EARLY] This parameter enables what is exposed to
1008 userspace and debugfs internal clients.
1009 Format: { on, no-mount, off }
1010 on: All functions are enabled.
1012 Filesystem is not registered but kernel clients can
1013 access APIs and a crashkernel can be used to read
1014 its content. There is nothing to mount.
1015 off: Filesystem is not registered and clients
1016 get a -EPERM as result when trying to register files
1017 or directories within debugfs.
1018 This is equivalent of the runtime functionality if
1019 debugfs was not enabled in the kernel at all.
1020 Default value is set in build-time with a kernel configuration.
1022 debugpat [X86] Enable PAT debugging
1025 [HW] The size of the default HugeTLB page. This is
1026 the size represented by the legacy /proc/ hugepages
1027 APIs. In addition, this is the default hugetlb size
1028 used for shmget(), mmap() and mounting hugetlbfs
1029 filesystems. If not specified, defaults to the
1030 architecture's default huge page size. Huge page
1031 sizes are architecture dependent. See also
1032 Documentation/admin-guide/mm/hugetlbpage.rst.
1035 deferred_probe_timeout=
1036 [KNL] Debugging option to set a timeout in seconds for
1037 deferred probe to give up waiting on dependencies to
1038 probe. Only specific dependencies (subsystems or
1039 drivers) that have opted in will be ignored. A timeout
1040 of 0 will timeout at the end of initcalls. If the time
1041 out hasn't expired, it'll be restarted by each
1042 successful driver registration. This option will also
1043 dump out devices still on the deferred probe list after
1046 delayacct [KNL] Enable per-task delay accounting
1048 dell_smm_hwmon.ignore_dmi=
1049 [HW] Continue probing hardware even if DMI data
1050 indicates that the driver is running on unsupported
1053 dell_smm_hwmon.force=
1054 [HW] Activate driver even if SMM BIOS signature does
1055 not match list of supported models and enable otherwise
1056 blacklisted features.
1058 dell_smm_hwmon.power_status=
1059 [HW] Report power status in /proc/i8k
1060 (disabled by default).
1062 dell_smm_hwmon.restricted=
1063 [HW] Allow controlling fans only if SYS_ADMIN
1066 dell_smm_hwmon.fan_mult=
1067 [HW] Factor to multiply fan speed with.
1069 dell_smm_hwmon.fan_max=
1070 [HW] Maximum configurable fan speed.
1073 Format: { on | off | def_only | inf_only | always }
1074 on: s390 zlib hardware support for compression on
1075 level 1 and decompression (default)
1076 off: No s390 zlib hardware support
1077 def_only: s390 zlib hardware support for deflate
1078 only (compression on level 1)
1079 inf_only: s390 zlib hardware support for inflate
1080 only (decompression)
1081 always: Same as 'on' but ignores the selected compression
1082 level always using hardware support (used for debugging)
1084 dhash_entries= [KNL]
1085 Set number of hash buckets for dentry cache.
1087 disable_1tb_segments [PPC,EARLY]
1088 Disables the use of 1TB hash page table segments. This
1089 causes the kernel to fall back to 256MB segments which
1090 can be useful when debugging issues that require an SLB
1094 See Documentation/networking/ipv6.rst.
1096 disable_radix [PPC,EARLY]
1097 Disable RADIX MMU mode on POWER9
1100 Disable TLBIE instruction. Currently does not work
1101 with KVM, with HASH MMU, or with coherent accelerators.
1103 disable_ddw [PPC/PSERIES,EARLY]
1104 Disable Dynamic DMA Window support. Use this
1105 to workaround buggy firmware.
1107 disable_ipv6= [IPV6]
1108 See Documentation/networking/ipv6.rst.
1110 disable_mtrr_cleanup [X86,EARLY]
1111 The kernel tries to adjust MTRR layout from continuous
1112 to discrete, to make X server driver able to add WB
1113 entry later. This parameter disables that.
1115 disable_mtrr_trim [X86, Intel and AMD only,EARLY]
1116 By default the kernel will trim any uncacheable
1117 memory out of your available memory pool based on
1118 MTRR settings. This parameter disables that behavior,
1119 possibly causing your machine to run very slowly.
1121 disable_timer_pin_1 [X86,EARLY]
1122 Disable PIN 1 of APIC timer
1123 Can be useful to work around chipset bugs.
1125 dis_ucode_ldr [X86] Disable the microcode loader.
1127 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
1128 this option disables the debugging code at boot.
1130 dma_debug_entries=<number>
1131 This option allows to tune the number of preallocated
1132 entries for DMA-API debugging code. One entry is
1133 required per DMA-API allocation. Use this if the
1134 DMA-API debugging code disables itself because the
1135 architectural default is too low.
1137 dma_debug_driver=<driver_name>
1138 With this option the DMA-API debugging driver
1139 filter feature can be enabled at boot time. Just
1140 pass the driver to filter for as the parameter.
1141 The filter can be disabled or changed to another
1142 driver later using sysfs.
1144 reg_file_data_sampling=
1145 [X86] Controls mitigation for Register File Data
1146 Sampling (RFDS) vulnerability. RFDS is a CPU
1147 vulnerability which may allow userspace to infer
1148 kernel data values previously stored in floating point
1149 registers, vector registers, or integer registers.
1150 RFDS only affects Intel Atom processors.
1152 on: Turns ON the mitigation.
1153 off: Turns OFF the mitigation.
1155 This parameter overrides the compile time default set
1156 by CONFIG_MITIGATION_RFDS. Mitigation cannot be
1157 disabled when other VERW based mitigations (like MDS)
1158 are enabled. In order to disable RFDS mitigation all
1159 VERW based mitigations need to be disabled.
1162 Documentation/admin-guide/hw-vuln/reg-file-data-sampling.rst
1164 driver_async_probe= [KNL]
1165 List of driver names to be probed asynchronously. *
1166 matches with all driver names. If * is specified, the
1167 rest of the listed driver names are those that will NOT
1169 Format: <driver_name1>,<driver_name2>...
1171 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
1172 Broken monitors, graphic adapters, KVMs and EDIDless
1173 panels may send no or incorrect EDID data sets.
1174 This parameter allows to specify an EDID data sets
1175 in the /lib/firmware directory that are used instead.
1176 An EDID data set will only be used for a particular
1177 connector, if its name and a colon are prepended to
1178 the EDID name. Each connector may use a unique EDID
1179 data set by separating the files with a comma. An EDID
1180 data set with no connector name will be used for
1181 any connectors not explicitly specified.
1185 dt_cpu_ftrs= [PPC,EARLY]
1186 Format: {"off" | "known"}
1187 Control how the dt_cpu_ftrs device-tree binding is
1188 used for CPU feature discovery and setup (if it
1190 off: Do not use it, fall back to legacy cpu table.
1191 known: Do not pass through unknown features to guests
1192 or userspace, only those that the kernel is aware of.
1194 dump_apple_properties [X86]
1195 Dump name and content of EFI device properties on
1196 x86 Macs. Useful for driver authors to determine
1197 what data is available or for reverse-engineering.
1199 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
1200 <module>.dyndbg[="val"]
1201 Enable debug messages at boot time. See
1202 Documentation/admin-guide/dynamic-debug-howto.rst
1205 early_ioremap_debug [KNL,EARLY]
1206 Enable debug messages in early_ioremap support. This
1207 is useful for tracking down temporary early mappings
1208 which are not unmapped.
1210 earlycon= [KNL,EARLY] Output early console device and options.
1212 When used with no options, the early console is
1213 determined by stdout-path property in device tree's
1214 chosen node or the ACPI SPCR table if supported by
1217 cdns,<addr>[,options]
1218 Start an early, polled-mode console on a Cadence
1219 (xuartps) serial port at the specified address. Only
1220 supported option is baud rate. If baud rate is not
1221 specified, the serial port must already be setup and
1224 uart[8250],io,<addr>[,options[,uartclk]]
1225 uart[8250],mmio,<addr>[,options[,uartclk]]
1226 uart[8250],mmio32,<addr>[,options[,uartclk]]
1227 uart[8250],mmio32be,<addr>[,options[,uartclk]]
1228 uart[8250],0x<addr>[,options]
1229 Start an early, polled-mode console on the 8250/16550
1230 UART at the specified I/O port or MMIO address.
1231 MMIO inter-register address stride is either 8-bit
1232 (mmio) or 32-bit (mmio32 or mmio32be).
1233 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
1234 to be equivalent to 'mmio'. 'options' are specified
1235 in the same format described for "console=ttyS<n>"; if
1236 unspecified, the h/w is not initialized. 'uartclk' is
1237 the uart clock frequency; if unspecified, it is set
1238 to 'BASE_BAUD' * 16.
1242 Start an early, polled-mode console on a pl011 serial
1243 port at the specified address. The pl011 serial port
1244 must already be setup and configured. Options are not
1245 yet supported. If 'mmio32' is specified, then only
1246 the driver will use only 32-bit accessors to read/write
1247 the device registers.
1250 Start an early console on a litex serial port at the
1251 specified address. The serial port must already be
1252 setup and configured. Options are not yet supported.
1255 Start an early, polled-mode console on a meson serial
1256 port at the specified address. The serial port must
1257 already be setup and configured. Options are not yet
1261 Start an early, polled-mode console on an msm serial
1262 port at the specified address. The serial port
1263 must already be setup and configured. Options are not
1266 msm_serial_dm,<addr>
1267 Start an early, polled-mode console on an msm serial
1268 dm port at the specified address. The serial port
1269 must already be setup and configured. Options are not
1273 Start an early, polled-mode console on a serial port
1274 of an Actions Semi SoC, such as S500 or S900, at the
1275 specified address. The serial port must already be
1276 setup and configured. Options are not yet supported.
1279 Start an early, polled-mode console on a serial port
1280 of an RDA Micro SoC, such as RDA8810PL, at the
1281 specified address. The serial port must already be
1282 setup and configured. Options are not yet supported.
1285 Use RISC-V SBI (Supervisor Binary Interface) for early
1288 smh Use ARM semihosting calls for early console.
1296 Use early console provided by serial driver available
1297 on Samsung SoCs, requires selecting proper type and
1298 a correct base address of the selected UART port. The
1299 serial port must already be setup and configured.
1300 Options are not yet supported.
1303 Start an early, polled-mode console on a lantiq serial
1304 (lqasc) port at the specified address. The serial port
1305 must already be setup and configured. Options are not
1310 Use early console provided by Freescale LP UART driver
1311 found on Freescale Vybrid and QorIQ LS1021A processors.
1312 A valid base address must be provided, and the serial
1313 port must already be setup and configured.
1317 Start an early, polled-mode, output-only console on the
1318 Freescale i.MX UART at the specified address. The UART
1319 must already be setup and configured.
1322 Start an early, polled-mode console on the
1323 Armada 3700 serial port at the specified
1324 address. The serial port must already be setup
1325 and configured. Options are not yet supported.
1328 Start an early, polled-mode console on a Qualcomm
1329 Generic Interface (GENI) based serial port at the
1330 specified address. The serial port must already be
1331 setup and configured. Options are not yet supported.
1334 Start an early, unaccelerated console on the EFI
1335 memory mapped framebuffer (if available). On cache
1336 coherent non-x86 systems that use system memory for
1337 the framebuffer, pass the 'ram' option so that it is
1338 mapped with the correct attributes.
1341 Use early console provided by Freescale LINFlexD UART
1342 serial driver for NXP S32V234 SoCs. A valid base
1343 address must be provided, and the serial port must
1344 already be setup and configured.
1346 earlyprintk= [X86,SH,ARM,M68k,S390,UM,EARLY]
1350 earlyprintk=serial[,ttySn[,baudrate]]
1351 earlyprintk=serial[,0x...[,baudrate]]
1352 earlyprintk=ttySn[,baudrate]
1353 earlyprintk=dbgp[debugController#]
1354 earlyprintk=pciserial[,force],bus:device.function[,baudrate]
1355 earlyprintk=xdbc[xhciController#]
1358 earlyprintk is useful when the kernel crashes before
1359 the normal console is initialized. It is not enabled by
1360 default because it has some cosmetic problems.
1362 Append ",keep" to not disable it when the real console
1365 Only one of vga, serial, or usb debug port can
1368 Currently only ttyS0 and ttyS1 may be specified by
1369 name. Other I/O ports may be explicitly specified
1370 on some architectures (x86 and arm at least) by
1371 replacing ttySn with an I/O port address, like this:
1372 earlyprintk=serial,0x1008,115200
1373 You can find the port for a given device in
1374 /proc/tty/driver/serial:
1375 2: uart:ST16650V2 port:00001008 irq:18 ...
1377 Interaction with the standard serial driver is not
1380 The VGA output is eventually overwritten by
1383 The xen option can only be used in Xen domains.
1385 The sclp output can only be used on s390.
1387 The bios output can only be used on SuperH.
1389 The optional "force" to "pciserial" enables use of a
1390 PCI device even when its classcode is not of the
1393 edac_report= [HW,EDAC] Control how to report EDAC event
1394 Format: {"on" | "off" | "force"}
1395 on: enable EDAC to report H/W event. May be overridden
1396 by other higher priority error reporting module.
1397 off: disable H/W event reporting through EDAC.
1398 force: enforce the use of EDAC to report H/W event.
1402 Format: {"off" | "on" | "skip[mbr]"}
1405 Format: { "debug", "disable_early_pci_dma",
1406 "nochunk", "noruntime", "nosoftreserve",
1407 "novamap", "no_disable_early_pci_dma" }
1408 debug: enable misc debug output.
1409 disable_early_pci_dma: disable the busmaster bit on all
1410 PCI bridges while in the EFI boot stub.
1411 nochunk: disable reading files in "chunks" in the EFI
1412 boot stub, as chunking can cause problems with some
1413 firmware implementations.
1414 noruntime : disable EFI runtime services support
1415 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose)
1416 attribute may cause the kernel to reserve the
1417 memory range for a memory mapping driver to
1418 claim. Specify efi=nosoftreserve to disable this
1419 reservation and treat the memory by its base type
1420 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM").
1421 novamap: do not call SetVirtualAddressMap().
1422 no_disable_early_pci_dma: Leave the busmaster bit set
1423 on all PCI bridges while in the EFI boot stub
1425 efi_no_storage_paranoia [EFI,X86,EARLY]
1426 Using this parameter you can use more than 50% of
1427 your efi variable storage. Use this parameter only if
1428 you are really sure that your UEFI does sane gc and
1429 fulfills the spec otherwise your board may brick.
1431 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI,X86,EARLY]
1432 Add arbitrary attribute to specific memory range by
1433 updating original EFI memory map.
1434 Region of memory which aa attribute is added to is
1437 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1438 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1439 attribute is added to range 0x100000000-0x180000000 and
1440 0x10a0000000-0x1120000000.
1442 If efi_fake_mem=8G@9G:0x40000 is specified, the
1443 EFI_MEMORY_SP(0x40000) attribute is added to
1444 range 0x240000000-0x43fffffff.
1446 Using this parameter you can do debugging of EFI memmap
1447 related features. For example, you can do debugging of
1448 Address Range Mirroring feature even if your box
1449 doesn't support it, or mark specific memory as
1452 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT
1453 that is to be dynamically loaded by Linux. If there are
1454 multiple variables with the same name but with different
1455 vendor GUIDs, all of them will be loaded. See
1456 Documentation/admin-guide/acpi/ssdt-overlays.rst for details.
1459 eisa_irq_edge= [PARISC,HW]
1460 See header of drivers/parisc/eisa.c.
1462 ekgdboc= [X86,KGDB,EARLY] Allow early kernel console debugging
1465 This is designed to be used in conjunction with
1466 the boot argument: earlyprintk=vga
1468 This parameter works in place of the kgdboc parameter
1469 but can only be used if the backing tty is available
1470 very early in the boot process. For early debugging
1471 via a serial port see kgdboc_earlycon instead.
1474 See comment before function elanfreq_setup() in
1475 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1477 elfcorehdr=[size[KMG]@]offset[KMG] [PPC,SH,X86,S390,EARLY]
1478 Specifies physical address of start of kernel core
1479 image elf header and optionally the size. Generally
1480 kexec loader will pass this option to capture kernel.
1481 See Documentation/admin-guide/kdump/kdump.rst for details.
1483 enable_mtrr_cleanup [X86,EARLY]
1484 The kernel tries to adjust MTRR layout from continuous
1485 to discrete, to make X server driver able to add WB
1486 entry later. This parameter enables that.
1488 enable_timer_pin_1 [X86]
1489 Enable PIN 1 of APIC timer
1490 Can be useful to work around chipset bugs
1491 (in particular on some ATI chipsets).
1492 The kernel tries to set a reasonable default.
1494 enforcing= [SELINUX] Set initial enforcing status.
1496 See security/selinux/Kconfig help text.
1497 0 -- permissive (log only, no denials).
1498 1 -- enforcing (deny and log).
1500 Value can be changed at runtime via
1501 /sys/fs/selinux/enforce.
1504 Disable Error Record Serialization Table (ERST)
1507 ether= [HW,NET] Ethernet cards parameters
1508 This option is obsoleted by the "netdev=" option, which
1509 has equivalent usage. See its documentation for details.
1513 Permit 'security.evm' to be updated regardless of
1514 current integrity status.
1516 early_page_ext [KNL,EARLY] Enforces page_ext initialization to earlier
1517 stages so cover more early boot allocations.
1518 Please note that as side effect some optimizations
1519 might be disabled to achieve that (e.g. parallelized
1520 memory initialization is disabled) so the boot process
1521 might take longer, especially on systems with a lot of
1522 memory. Available with CONFIG_PAGE_EXTENSION=y.
1527 fail_make_request=[KNL]
1528 General fault injection mechanism.
1529 Format: <interval>,<probability>,<space>,<times>
1530 See also Documentation/fault-injection/.
1533 Format: { initns | none }
1534 See Documentation/admin-guide/sysctl/net.rst for
1535 fb_tunnels_only_for_init_ns
1538 See Documentation/admin-guide/blockdev/floppy.rst.
1541 Forcefully enable Physical Address Extension (PAE).
1542 Many Pentium M systems disable PAE but may have a
1543 functionally usable PAE implementation.
1544 Warning: use of this parameter will taint the kernel
1545 and may cause unknown problems.
1548 Enable/disable Flexible Return and Event Delivery.
1549 Format: { on | off }
1550 on: enable FRED when it's present.
1551 off: disable FRED, the default setting.
1554 [FTRACE] will set and start the specified tracer
1555 as early as possible in order to facilitate early
1558 ftrace_boot_snapshot
1559 [FTRACE] On boot up, a snapshot will be taken of the
1560 ftrace ring buffer that can be read at:
1561 /sys/kernel/tracing/snapshot.
1562 This is useful if you need tracing information from kernel
1563 boot up that is likely to be overridden by user space
1564 start up functionality.
1566 Optionally, the snapshot can also be defined for a tracing
1567 instance that was created by the trace_instance= command
1570 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo
1572 The above will cause the "foo" tracing instance to trigger
1573 a snapshot at the end of boot up.
1575 ftrace_dump_on_oops[=2(orig_cpu) | =<instance>][,<instance> |
1576 ,<instance>=2(orig_cpu)]
1577 [FTRACE] will dump the trace buffers on oops.
1578 If no parameter is passed, ftrace will dump global
1579 buffers of all CPUs, if you pass 2 or orig_cpu, it
1580 will dump only the buffer of the CPU that triggered
1581 the oops, or the specific instance will be dumped if
1582 its name is passed. Multiple instance dump is also
1583 supported, and instances are separated by commas. Each
1584 instance supports only dump on CPU that triggered the
1585 oops by passing 2 or orig_cpu to it.
1587 ftrace_dump_on_oops=foo=orig_cpu
1589 The above will dump only the buffer of "foo" instance
1590 on CPU that triggered the oops.
1592 ftrace_dump_on_oops,foo,bar=orig_cpu
1594 The above will dump global buffer on all CPUs, the
1595 buffer of "foo" instance on all CPUs and the buffer
1596 of "bar" instance on CPU that triggered the oops.
1598 ftrace_filter=[function-list]
1599 [FTRACE] Limit the functions traced by the function
1600 tracer at boot up. function-list is a comma-separated
1601 list of functions. This list can be changed at run
1602 time by the set_ftrace_filter file in the debugfs
1605 ftrace_notrace=[function-list]
1606 [FTRACE] Do not trace the functions specified in
1607 function-list. This list can be changed at run time
1608 by the set_ftrace_notrace file in the debugfs
1611 ftrace_graph_filter=[function-list]
1612 [FTRACE] Limit the top level callers functions traced
1613 by the function graph tracer at boot up.
1614 function-list is a comma-separated list of functions
1615 that can be changed at run time by the
1616 set_graph_function file in the debugfs tracing directory.
1618 ftrace_graph_notrace=[function-list]
1619 [FTRACE] Do not trace from the functions specified in
1620 function-list. This list is a comma-separated list of
1621 functions that can be changed at run time by the
1622 set_graph_notrace file in the debugfs tracing directory.
1624 ftrace_graph_max_depth=<uint>
1625 [FTRACE] Used with the function graph tracer. This is
1626 the max depth it will trace into a function. This value
1627 can be changed at run time by the max_graph_depth file
1628 in the tracefs tracing directory. default: 0 (no limit)
1630 fw_devlink= [KNL,EARLY] Create device links between consumer and supplier
1631 devices by scanning the firmware to infer the
1632 consumer/supplier relationships. This feature is
1633 especially useful when drivers are loaded as modules as
1634 it ensures proper ordering of tasks like device probing
1635 (suppliers first, then consumers), supplier boot state
1636 clean up (only after all consumers have probed),
1637 suspend/resume & runtime PM (consumers first, then
1639 Format: { off | permissive | on | rpm }
1640 off -- Don't create device links from firmware info.
1641 permissive -- Create device links from firmware info
1642 but use it only for ordering boot state clean
1643 up (sync_state() calls).
1644 on -- Create device links from firmware info and use it
1645 to enforce probe and suspend/resume ordering.
1646 rpm -- Like "on", but also use to order runtime PM.
1648 fw_devlink.strict=<bool>
1649 [KNL,EARLY] Treat all inferred dependencies as mandatory
1650 dependencies. This only applies for fw_devlink=on|rpm.
1653 fw_devlink.sync_state =
1654 [KNL,EARLY] When all devices that could probe have finished
1655 probing, this parameter controls what to do with
1656 devices that haven't yet received their sync_state()
1658 Format: { strict | timeout }
1659 strict -- Default. Continue waiting on consumers to
1661 timeout -- Give up waiting on consumers and call
1662 sync_state() on any devices that haven't yet
1663 received their sync_state() calls after
1664 deferred_probe_timeout has expired or by
1665 late_initcall() if !CONFIG_MODULES.
1668 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1669 support via parallel port (up to 5 devices per port)
1670 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1671 See also Documentation/input/devices/joystick-parport.rst
1675 gart_fix_e820= [X86-64,EARLY] disable the fix e820 for K8 GART
1679 gather_data_sampling=
1680 [X86,INTEL,EARLY] Control the Gather Data Sampling (GDS)
1683 Gather Data Sampling is a hardware vulnerability which
1684 allows unprivileged speculative access to data which was
1685 previously stored in vector registers.
1687 This issue is mitigated by default in updated microcode.
1688 The mitigation may have a performance impact but can be
1689 disabled. On systems without the microcode mitigation
1690 disabling AVX serves as a mitigation.
1692 force: Disable AVX to mitigate systems without
1693 microcode mitigation. No effect if the microcode
1694 mitigation is present. Known to cause crashes in
1695 userspace with buggy AVX enumeration.
1697 off: Disable GDS mitigation.
1699 gcov_persist= [GCOV] When non-zero (default), profiling data for
1700 kernel modules is saved and remains accessible via
1701 debugfs, even when the module is unloaded/reloaded.
1702 When zero, profiling data is discarded and associated
1703 debugfs files are removed at module unload time.
1705 goldfish [X86] Enable the goldfish android emulator platform.
1706 Don't use this when you are not running on the
1709 gpio-mockup.gpio_mockup_ranges
1710 [HW] Sets the ranges of gpiochip of for this device.
1711 Format: <start1>,<end1>,<start2>,<end2>...
1712 gpio-mockup.gpio_mockup_named_lines
1713 [HW] Let the driver know GPIO lines should be named.
1715 gpt [EFI] Forces disk with valid GPT signature but
1716 invalid Protective MBR to be treated as GPT. If the
1717 primary GPT is corrupted, it enables the backup/alternate
1718 GPT to be used instead.
1720 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1721 the "Enable 0" bit of the configuration register.
1724 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1725 the "Enable 0" bit of the configuration register.
1728 grcan.select= [HW] Select which physical interface to use.
1731 grcan.txsize= [HW] Sets the size of the tx buffer.
1732 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1734 grcan.rxsize= [HW] Sets the size of the rx buffer.
1735 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1739 [KNL] Under CONFIG_HARDENED_USERCOPY, whether
1740 hardening is enabled for this boot. Hardened
1741 usercopy checking is used to protect the kernel
1742 from reading or writing beyond known memory
1743 allocation boundaries as a proactive defense
1744 against bounds-checking flaws in the kernel's
1745 copy_to_user()/copy_from_user() interface.
1746 on Perform hardened usercopy checks (default).
1747 off Disable hardened usercopy checks.
1749 hardlockup_all_cpu_backtrace=
1750 [KNL] Should the hard-lockup detector generate
1751 backtraces on all cpus.
1754 hashdist= [KNL,NUMA] Large hashes allocated during boot
1755 are distributed across NUMA nodes. Defaults on
1756 for 64-bit NUMA, off otherwise.
1757 Format: 0 | 1 (for off | on)
1759 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1761 hd= [EIDE] (E)IDE hard drive subsystem geometry
1762 Format: <cyl>,<head>,<sect>
1765 Disable Hardware Error Source Table (HEST) support;
1766 corresponding firmware-first mode error processing
1767 logic will be disabled.
1769 hibernate= [HIBERNATION]
1770 noresume Don't check if there's a hibernation image
1771 present during boot.
1772 nocompress Don't compress/decompress hibernation images.
1773 no Disable hibernation and resume.
1774 protect_image Turn on image protection during restoration
1775 (that will set all pages holding image data
1776 during restoration read-only).
1778 hibernate.compressor= [HIBERNATION] Compression algorithm to be
1779 used with hibernation.
1780 Format: { lzo | lz4 }
1783 lzo: Select LZO compression algorithm to
1784 compress/decompress hibernation image.
1786 lz4: Select LZ4 compression algorithm to
1787 compress/decompress hibernation image.
1789 highmem=nn[KMG] [KNL,BOOT,EARLY] forces the highmem zone to have an exact
1790 size of <nn>. This works even on boxes that have no
1791 highmem otherwise. This also works to reduce highmem
1792 size on bigger boxes.
1794 highres= [KNL] Enable/disable high resolution timer mode.
1795 Valid parameters: "on", "off"
1800 hostname= [KNL,EARLY] Set the hostname (aka UTS nodename).
1802 This allows setting the system's hostname during early
1803 startup. This sets the name returned by gethostname.
1804 Using this parameter to set the hostname makes it
1805 possible to ensure the hostname is correctly set before
1806 any userspace processes run, avoiding the possibility
1807 that a process may call gethostname before the hostname
1808 has been explicitly set, resulting in the calling
1809 process getting an incorrect result. The string must
1810 not exceed the maximum allowed hostname length (usually
1811 64 characters) and will be truncated otherwise.
1813 hpet= [X86-32,HPET] option to control HPET usage
1814 Format: { enable (default) | disable | force |
1816 disable: disable HPET and use PIT instead
1817 force: allow force enabled of undocumented chips (ICH4,
1819 verbose: show contents of HPET registers during setup
1821 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1822 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1824 hugepages= [HW] Number of HugeTLB pages to allocate at boot.
1825 If this follows hugepagesz (below), it specifies
1826 the number of pages of hugepagesz to be allocated.
1827 If this is the first HugeTLB parameter on the command
1828 line, it specifies the number of pages to allocate for
1829 the default huge page size. If using node format, the
1830 number of pages to allocate per-node can be specified.
1831 See also Documentation/admin-guide/mm/hugetlbpage.rst.
1832 Format: <integer> or (node format)
1833 <node>:<integer>[,<node>:<integer>]
1836 [HW] The size of the HugeTLB pages. This is used in
1837 conjunction with hugepages (above) to allocate huge
1838 pages of a specific size at boot. The pair
1839 hugepagesz=X hugepages=Y can be specified once for
1840 each supported huge page size. Huge page sizes are
1841 architecture dependent. See also
1842 Documentation/admin-guide/mm/hugetlbpage.rst.
1845 hugetlb_cma= [HW,CMA,EARLY] The size of a CMA area used for allocation
1846 of gigantic hugepages. Or using node format, the size
1847 of a CMA area per node can be specified.
1848 Format: nn[KMGTPE] or (node format)
1849 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]]
1851 Reserve a CMA area of given size and allocate gigantic
1852 hugepages using the CMA allocator. If enabled, the
1853 boot-time allocation of gigantic hugepages is skipped.
1855 hugetlb_free_vmemmap=
1856 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP
1858 Control if HugeTLB Vmemmap Optimization (HVO) is enabled.
1859 Allows heavy hugetlb users to free up some more
1860 memory (7 * PAGE_SIZE for each 2MB hugetlb page).
1861 Format: { on | off (default) }
1866 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y,
1869 Note that the vmemmap pages may be allocated from the added
1870 memory block itself when memory_hotplug.memmap_on_memory is
1871 enabled, those vmemmap pages cannot be optimized even if this
1872 feature is enabled. Other vmemmap pages not allocated from
1873 the added memory block itself do not be affected.
1876 [KNL] Should the hung task detector generate panics.
1879 A value of 1 instructs the kernel to panic when a
1880 hung task is detected. The default value is controlled
1881 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time
1882 option. The value selected by this boot parameter can
1883 be changed later by the kernel.hung_task_panic sysctl.
1885 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1886 terminal devices. Valid values: 0..8
1887 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1888 If specified, z/VM IUCV HVC accepts connections
1889 from listed z/VM user IDs only.
1891 hv_nopvspin [X86,HYPER_V,EARLY]
1892 Disables the paravirt spinlock optimizations
1893 which allow the hypervisor to 'idle' the guest
1896 i2c_bus= [HW] Override the default board specific I2C bus speed
1897 or register an additional I2C bus that is not
1898 registered from board initialization code.
1902 i8042.debug [HW] Toggle i8042 debug mode
1903 i8042.unmask_kbd_data
1904 [HW] Enable printing of interrupt data from the KBD port
1905 (disabled by default, and as a pre-condition
1906 requires that i8042.debug=1 be enabled)
1907 i8042.direct [HW] Put keyboard port into non-translated mode
1908 i8042.dumbkbd [HW] Pretend that controller can only read data from
1909 keyboard and cannot control its state
1910 (Don't attempt to blink the leds)
1911 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1912 i8042.nokbd [HW] Don't check/create keyboard port
1913 i8042.noloop [HW] Disable the AUX Loopback command while probing
1915 i8042.nomux [HW] Don't check presence of an active multiplexing
1917 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1919 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1920 i8042.reset [HW] Reset the controller during init, cleanup and
1921 suspend-to-ram transitions, only during s2r
1922 transitions, or never reset
1923 Format: { 1 | Y | y | 0 | N | n }
1924 1, Y, y: always reset controller
1925 0, N, n: don't ever reset controller
1926 Default: only on s2r transitions on x86; most other
1927 architectures force reset to be always executed
1928 i8042.unlock [HW] Unlock (ignore) the keylock
1929 i8042.kbdreset [HW] Reset device connected to KBD port
1931 [HW] Allow deferred probing upon i8042 probe errors
1935 i915.invert_brightness=
1936 [DRM] Invert the sense of the variable that is used to
1937 set the brightness of the panel backlight. Normally a
1938 brightness value of 0 indicates backlight switched off,
1939 and the maximum of the brightness value sets the backlight
1940 to maximum brightness. If this parameter is set to 0
1941 (default) and the machine requires it, or this parameter
1942 is set to 1, a brightness value of 0 sets the backlight
1943 to maximum brightness, and the maximum of the brightness
1944 value switches the backlight off.
1945 -1 -- never invert brightness
1946 0 -- machine default
1947 1 -- force brightness inversion
1949 ia32_emulation= [X86-64]
1951 When true, allows loading 32-bit programs and executing 32-bit
1952 syscalls, essentially overriding IA32_EMULATION_DEFAULT_DISABLED at
1953 boot time. When false, unconditionally disables IA32 emulation.
1956 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1960 Format: idle=poll, idle=halt, idle=nomwait
1961 Poll forces a polling idle loop that can slightly
1962 improve the performance of waking up a idle CPU, but
1963 will use a lot of power and make the system run hot.
1965 idle=halt: Halt is forced to be used for CPU idle.
1966 In such case C2/C3 won't be used again.
1967 idle=nomwait: Disable mwait for CPU C-states
1971 Allow force disabling of Shared Virtual Memory (SVA)
1972 support for the idxd driver. By default it is set to
1975 idxd.tc_override= [HW]
1977 Allow override of default traffic class configuration
1978 for the device. By default it is set to false (0).
1980 ieee754= [MIPS] Select IEEE Std 754 conformance mode
1981 Format: { strict | legacy | 2008 | relaxed }
1984 Choose which programs will be accepted for execution
1985 based on the IEEE 754 NaN encoding(s) supported by
1986 the FPU and the NaN encoding requested with the value
1987 of an ELF file header flag individually set by each
1988 binary. Hardware implementations are permitted to
1989 support either or both of the legacy and the 2008 NaN
1992 Available settings are as follows:
1993 strict accept binaries that request a NaN encoding
1994 supported by the FPU
1995 legacy only accept legacy-NaN binaries, if supported
1997 2008 only accept 2008-NaN binaries, if supported
1999 relaxed accept any binaries regardless of whether
2000 supported by the FPU
2002 The FPU emulator is always able to support both NaN
2003 encodings, so if no FPU hardware is present or it has
2004 been disabled with 'nofpu', then the settings of
2005 'legacy' and '2008' strap the emulator accordingly,
2006 'relaxed' straps the emulator for both legacy-NaN and
2007 2008-NaN, whereas 'strict' enables legacy-NaN only on
2008 legacy processors and both NaN encodings on MIPS32 or
2011 The setting for ABS.fmt/NEG.fmt instruction execution
2012 mode generally follows that for the NaN encoding,
2013 except where unsupported by hardware.
2015 ignore_loglevel [KNL,EARLY]
2016 Ignore loglevel setting - this will print /all/
2017 kernel messages to the console. Useful for debugging.
2018 We also add it as printk module parameter, so users
2019 could change it dynamically, usually by
2020 /sys/module/printk/parameters/ignore_loglevel.
2023 Ignore RLIMIT_DATA setting for data mappings,
2024 print warning at first misuse. Can be changed via
2025 /sys/module/kernel/parameters/ignore_rlimit_data.
2027 ihash_entries= [KNL]
2028 Set number of hash buckets for inode cache.
2030 ima_appraise= [IMA] appraise integrity measurements
2031 Format: { "off" | "enforce" | "fix" | "log" }
2034 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead.
2035 The builtin appraise policy appraises all files
2038 ima_canonical_fmt [IMA]
2039 Use the canonical format for the binary runtime
2040 measurements, instead of host native format.
2043 Format: { md5 | sha1 | rmd160 | sha256 | sha384
2047 The list of supported hash algorithms is defined
2048 in crypto/hash_info.h.
2051 The builtin policies to load during IMA setup.
2052 Format: "tcb | appraise_tcb | secure_boot |
2053 fail_securely | critical_data"
2055 The "tcb" policy measures all programs exec'd, files
2056 mmap'd for exec, and all files opened with the read
2057 mode bit set by either the effective uid (euid=0) or
2060 The "appraise_tcb" policy appraises the integrity of
2061 all files owned by root.
2063 The "secure_boot" policy appraises the integrity
2064 of files (eg. kexec kernel image, kernel modules,
2065 firmware, policy, etc) based on file signatures.
2067 The "fail_securely" policy forces file signature
2068 verification failure also on privileged mounted
2069 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE
2072 The "critical_data" policy measures kernel integrity
2075 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
2076 Load a policy which meets the needs of the Trusted
2077 Computing Base. This means IMA will measure all
2078 programs exec'd, files mmap'd for exec, and all files
2079 opened for read by uid=0.
2082 Select one of defined IMA measurements template formats.
2083 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" |
2088 [IMA] Define a custom template format.
2089 Format: { "field1|...|fieldN" }
2091 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
2092 Format: <min_file_size>
2093 Set the minimal file size for using asynchronous hash.
2094 If left unspecified, ahash usage is disabled.
2096 ahash performance varies for different data sizes on
2097 different crypto accelerators. This option can be used
2098 to achieve the best performance for a particular HW.
2100 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
2102 Set hashing buffer size. Default: 4k.
2104 ahash performance varies for different chunk sizes on
2105 different crypto accelerators. This option can be used
2106 to achieve best performance for particular HW.
2110 Run specified binary instead of /sbin/init as init
2113 initcall_debug [KNL] Trace initcalls as they are executed. Useful
2114 for working out where the kernel is dying during
2117 initcall_blacklist= [KNL] Do not execute a comma-separated list of
2118 initcall functions. Useful for debugging built-in
2119 modules and initcalls.
2121 initramfs_async= [KNL]
2124 This parameter controls whether the initramfs
2125 image is unpacked asynchronously, concurrently
2126 with devices being probed and
2127 initialized. This should normally just work,
2128 but as a debugging aid, one can get the
2129 historical behaviour of the initramfs
2130 unpacking being completed before device_ and
2133 initrd= [BOOT,EARLY] Specify the location of the initial ramdisk
2135 initrdmem= [KNL,EARLY] Specify a physical address and size from which to
2136 load the initrd. If an initrd is compiled in or
2137 specified in the bootparams, it takes priority over this
2139 Format: ss[KMG],nn[KMG]
2142 init_on_alloc= [MM,EARLY] Fill newly allocated pages and heap objects with
2145 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON.
2147 init_on_free= [MM,EARLY] Fill freed pages and heap objects with zeroes.
2149 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON.
2151 init_pkru= [X86] Specify the default memory protection keys rights
2152 register contents for all processes. 0x55555554 by
2153 default (disallow access to all but pkey 0). Can
2154 override in debugfs after boot.
2156 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
2159 int_pln_enable [X86] Enable power limit notification interrupt
2161 integrity_audit=[IMA]
2162 Format: { "0" | "1" }
2163 0 -- basic integrity auditing messages. (Default)
2164 1 -- additional integrity auditing messages.
2166 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
2168 Enable intel iommu driver.
2170 Disable intel iommu driver.
2171 igfx_off [Default Off]
2172 By default, gfx is mapped as normal device. If a gfx
2173 device has a dedicated DMAR unit, the DMAR unit is
2174 bypassed by not enabling DMAR with this option. In
2175 this case, gfx device will use physical address for
2177 strict [Default Off]
2178 Deprecated, equivalent to iommu.strict=1.
2179 sp_off [Default Off]
2180 By default, super page will be supported if Intel IOMMU
2181 has the capability. With this option, super page will
2184 Enable the Intel IOMMU scalable mode if the hardware
2185 advertises that it has support for the scalable mode
2188 Disallow use of the Intel IOMMU scalable mode.
2189 tboot_noforce [Default Off]
2190 Do not force the Intel IOMMU enabled under tboot.
2191 By default, tboot will force Intel IOMMU on, which
2192 could harm performance of some high-throughput
2193 devices like 40GBit network cards, even if identity
2195 Note that using this option lowers the security
2196 provided by tboot because it makes the system
2197 vulnerable to DMA attacks.
2199 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
2200 0 disables intel_idle and fall back on acpi_idle.
2201 1 to 9 specify maximum depth of C-state.
2203 intel_pstate= [X86,EARLY]
2205 Do not enable intel_pstate as the default
2206 scaling driver for the supported processors
2208 Use intel_pstate driver to bypass the scaling
2209 governors layer of cpufreq and provides it own
2210 algorithms for p-state selection. There are two
2211 P-state selection algorithms provided by
2212 intel_pstate in the active mode: powersave and
2213 performance. The way they both operate depends
2214 on whether or not the hardware managed P-states
2215 (HWP) feature has been enabled in the processor
2216 and possibly on the processor model.
2218 Use intel_pstate as a scaling driver, but configure it
2219 to work with generic cpufreq governors (instead of
2220 enabling its internal governor). This mode cannot be
2221 used along with the hardware-managed P-states (HWP)
2224 Enable intel_pstate on systems that prohibit it by default
2225 in favor of acpi-cpufreq. Forcing the intel_pstate driver
2226 instead of acpi-cpufreq may disable platform features, such
2227 as thermal controls and power capping, that rely on ACPI
2228 P-States information being indicated to OSPM and therefore
2229 should be used with caution. This option does not work with
2230 processors that aren't supported by the intel_pstate driver
2231 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
2233 Do not enable hardware P state control (HWP)
2236 Only load intel_pstate on systems which support
2237 hardware P state control (HWP) if available.
2239 Enforce ACPI _PPC performance limits. If the Fixed ACPI
2240 Description Table, specifies preferred power management
2241 profile as "Enterprise Server" or "Performance Server",
2242 then this feature is turned on by default.
2244 Allow per-logical-CPU P-State performance control limits using
2245 cpufreq sysfs interface
2247 intremap= [X86-64,Intel-IOMMU,EARLY]
2248 on enable Interrupt Remapping (default)
2249 off disable Interrupt Remapping
2250 nosid disable Source ID checking
2252 BIOS x2APIC opt-out request will be ignored
2253 nopost disable Interrupt Posting
2255 iomem= Disable strict checking of access to MMIO memory
2256 strict regions from userspace.
2271 nobypass [PPC/POWERNV]
2272 Disable IOMMU bypass, using IOMMU for PCI devices.
2274 iommu.forcedac= [ARM64,X86,EARLY] Control IOVA allocation for PCI devices.
2275 Format: { "0" | "1" }
2276 0 - Try to allocate a 32-bit DMA address first, before
2277 falling back to the full range if needed.
2278 1 - Allocate directly from the full usable range,
2279 forcing Dual Address Cycle for PCI cards supporting
2280 greater than 32-bit addressing.
2282 iommu.strict= [ARM64,X86,S390,EARLY] Configure TLB invalidation behaviour
2283 Format: { "0" | "1" }
2285 Request that DMA unmap operations use deferred
2286 invalidation of hardware TLBs, for increased
2287 throughput at the cost of reduced device isolation.
2288 Will fall back to strict mode if not supported by
2289 the relevant IOMMU driver.
2291 DMA unmap operations invalidate IOMMU hardware TLBs
2293 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}.
2294 Note: on x86, strict mode specified via one of the
2295 legacy driver-specific options takes precedence.
2298 [ARM64,X86,EARLY] Configure DMA to bypass the IOMMU by default.
2299 Format: { "0" | "1" }
2300 0 - Use IOMMU translation for DMA.
2301 1 - Bypass the IOMMU for DMA.
2302 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH.
2304 io7= [HW] IO7 for Marvel-based Alpha systems
2305 See comment before marvel_specify_io7 in
2306 arch/alpha/kernel/core_marvel.c.
2308 io_delay= [X86,EARLY] I/O delay method
2310 Standard port 0x80 based delay
2312 Alternate port 0xed based delay (needed on some systems)
2314 Simple two microseconds delay
2319 See Documentation/admin-guide/nfs/nfsroot.rst.
2321 ipcmni_extend [KNL,EARLY] Extend the maximum number of unique System V
2322 IPC identifiers from 32,768 to 16,777,216.
2324 irqaffinity= [SMP] Set the default irq affinity mask
2325 The argument is a cpu list, as described above.
2327 irqchip.gicv2_force_probe=
2330 Force the kernel to look for the second 4kB page
2331 of a GICv2 controller even if the memory range
2332 exposed by the device tree is too small.
2334 irqchip.gicv3_nolpi=
2336 Force the kernel to ignore the availability of
2337 LPIs (and by consequence ITSs). Intended for system
2338 that use the kernel as a bootloader, and thus want
2339 to let secondary kernels in charge of setting up
2342 irqchip.gicv3_pseudo_nmi= [ARM64,EARLY]
2343 Enables support for pseudo-NMIs in the kernel. This
2344 requires the kernel to be built with
2345 CONFIG_ARM64_PSEUDO_NMI.
2348 When an interrupt is not handled search all handlers
2349 for it. Intended to get systems with badly broken
2353 When an interrupt is not handled search all handlers
2354 for it. Also check all handlers each timer
2355 interrupt. Intended to get systems with badly broken
2359 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
2361 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance.
2362 [Deprecated - use cpusets instead]
2363 Format: [flag-list,]<cpu-list>
2365 Specify one or more CPUs to isolate from disturbances
2366 specified in the flag list (default: domain):
2369 Disable the tick when a single task runs.
2371 A residual 1Hz tick is offloaded to workqueues, which you
2372 need to affine to housekeeping through the global
2373 workqueue's affinity configured via the
2374 /sys/devices/virtual/workqueue/cpumask sysfs file, or
2375 by using the 'domain' flag described below.
2377 NOTE: by default the global workqueue runs on all CPUs,
2378 so to protect individual CPUs the 'cpumask' file has to
2379 be configured manually after bootup.
2382 Isolate from the general SMP balancing and scheduling
2383 algorithms. Note that performing domain isolation this way
2384 is irreversible: it's not possible to bring back a CPU to
2385 the domains once isolated through isolcpus. It's strongly
2386 advised to use cpusets instead to disable scheduler load
2387 balancing through the "cpuset.sched_load_balance" file.
2388 It offers a much more flexible interface where CPUs can
2389 move in and out of an isolated set anytime.
2391 You can move a process onto or off an "isolated" CPU via
2392 the CPU affinity syscalls or cpuset.
2393 <cpu number> begins at 0 and the maximum value is
2394 "number of CPUs in system - 1".
2398 Isolate from being targeted by managed interrupts
2399 which have an interrupt mask containing isolated
2400 CPUs. The affinity of managed interrupts is
2401 handled by the kernel and cannot be changed via
2402 the /proc/irq/* interfaces.
2404 This isolation is best effort and only effective
2405 if the automatically assigned interrupt mask of a
2406 device queue contains isolated and housekeeping
2407 CPUs. If housekeeping CPUs are online then such
2408 interrupts are directed to the housekeeping CPU
2409 so that IO submitted on the housekeeping CPU
2410 cannot disturb the isolated CPU.
2412 If a queue's affinity mask contains only isolated
2413 CPUs then this parameter has no effect on the
2414 interrupt routing decision, though interrupts are
2415 only delivered when tasks running on those
2416 isolated CPUs submit IO. IO submitted on
2417 housekeeping CPUs has no influence on those
2420 The format of <cpu-list> is described above.
2424 ivrs_ioapic [HW,X86-64]
2425 Provide an override to the IOAPIC-ID<->DEVICE-ID
2426 mapping provided in the IVRS ACPI table.
2427 By default, PCI segment is 0, and can be omitted.
2429 For example, to map IOAPIC-ID decimal 10 to
2430 PCI segment 0x1 and PCI device 00:14.0,
2431 write the parameter as:
2432 ivrs_ioapic=10@0001:00:14.0
2435 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0
2436 write the parameter as:
2437 ivrs_ioapic[10]=00:14.0
2438 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and
2439 PCI device 00:14.0 write the parameter as:
2440 ivrs_ioapic[10]=0001:00:14.0
2442 ivrs_hpet [HW,X86-64]
2443 Provide an override to the HPET-ID<->DEVICE-ID
2444 mapping provided in the IVRS ACPI table.
2445 By default, PCI segment is 0, and can be omitted.
2447 For example, to map HPET-ID decimal 10 to
2448 PCI segment 0x1 and PCI device 00:14.0,
2449 write the parameter as:
2450 ivrs_hpet=10@0001:00:14.0
2453 * To map HPET-ID decimal 0 to PCI device 00:14.0
2454 write the parameter as:
2455 ivrs_hpet[0]=00:14.0
2456 * To map HPET-ID decimal 10 to PCI segment 0x1 and
2457 PCI device 00:14.0 write the parameter as:
2458 ivrs_ioapic[10]=0001:00:14.0
2460 ivrs_acpihid [HW,X86-64]
2461 Provide an override to the ACPI-HID:UID<->DEVICE-ID
2462 mapping provided in the IVRS ACPI table.
2463 By default, PCI segment is 0, and can be omitted.
2465 For example, to map UART-HID:UID AMD0020:0 to
2466 PCI segment 0x1 and PCI device ID 00:14.5,
2467 write the parameter as:
2468 ivrs_acpihid=AMD0020:0@0001:00:14.5
2471 * To map UART-HID:UID AMD0020:0 to PCI segment is 0,
2472 PCI device ID 00:14.5, write the parameter as:
2473 ivrs_acpihid[00:14.5]=AMD0020:0
2474 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and
2475 PCI device ID 00:14.5, write the parameter as:
2476 ivrs_acpihid[0001:00:14.5]=AMD0020:0
2478 js= [HW,JOY] Analog joystick
2479 See Documentation/input/joydev/joystick.rst.
2482 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print
2483 report on every invalid memory access. Without this
2484 parameter KASAN will print report only for the first
2487 keep_bootcon [KNL,EARLY]
2488 Do not unregister boot console at start. This is only
2489 useful for debugging when something happens in the window
2490 between unregistering the boot console and initializing
2493 keepinitrd [HW,ARM] See retain_initrd.
2495 kernelcore= [KNL,X86,IA-64,PPC,EARLY]
2496 Format: nn[KMGTPE] | nn% | "mirror"
2497 This parameter specifies the amount of memory usable by
2498 the kernel for non-movable allocations. The requested
2499 amount is spread evenly throughout all nodes in the
2500 system as ZONE_NORMAL. The remaining memory is used for
2501 movable memory in its own zone, ZONE_MOVABLE. In the
2502 event, a node is too small to have both ZONE_NORMAL and
2503 ZONE_MOVABLE, kernelcore memory will take priority and
2504 other nodes will have a larger ZONE_MOVABLE.
2506 ZONE_MOVABLE is used for the allocation of pages that
2507 may be reclaimed or moved by the page migration
2508 subsystem. Note that allocations like PTEs-from-HighMem
2509 still use the HighMem zone if it exists, and the Normal
2510 zone if it does not.
2512 It is possible to specify the exact amount of memory in
2513 the form of "nn[KMGTPE]", a percentage of total system
2514 memory in the form of "nn%", or "mirror". If "mirror"
2515 option is specified, mirrored (reliable) memory is used
2516 for non-movable allocations and remaining memory is used
2517 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror"
2518 are exclusive, so you cannot specify multiple forms.
2520 kgdbdbgp= [KGDB,HW,EARLY] kgdb over EHCI usb debug port.
2521 Format: <Controller#>[,poll interval]
2522 The controller # is the number of the ehci usb debug
2523 port as it is probed via PCI. The poll interval is
2524 optional and is the number seconds in between
2525 each poll cycle to the debug port in case you need
2526 the functionality for interrupting the kernel with
2527 gdb or control-c on the dbgp connection. When
2528 not using this parameter you use sysrq-g to break into
2529 the kernel debugger.
2531 kgdboc= [KGDB,HW] kgdb over consoles.
2532 Requires a tty driver that supports console polling,
2533 or a supported polling keyboard driver (non-usb).
2534 Serial only format: <serial_device>[,baud]
2535 keyboard only format: kbd
2536 keyboard and serial format: kbd,<serial_device>[,baud]
2537 Optional Kernel mode setting:
2538 kms, kbd format: kms,kbd
2539 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
2541 kgdboc_earlycon= [KGDB,HW,EARLY]
2542 If the boot console provides the ability to read
2543 characters and can work in polling mode, you can use
2544 this parameter to tell kgdb to use it as a backend
2545 until the normal console is registered. Intended to
2546 be used together with the kgdboc parameter which
2547 specifies the normal console to transition to.
2549 The name of the early console should be specified
2550 as the value of this parameter. Note that the name of
2551 the early console might be different than the tty
2552 name passed to kgdboc. It's OK to leave the value
2553 blank and the first boot console that implements
2554 read() will be picked.
2556 kgdbwait [KGDB,EARLY] Stop kernel execution and enter the
2557 kernel debugger at the earliest opportunity.
2559 kmac= [MIPS] Korina ethernet MAC address.
2560 Configure the RouterBoard 532 series on-chip
2561 Ethernet adapter MAC address.
2563 kmemleak= [KNL,EARLY] Boot-time kmemleak enable/disable
2564 Valid arguments: on, off
2566 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
2569 kprobe_event=[probe-list]
2570 [FTRACE] Add kprobe events and enable at boot time.
2571 The probe-list is a semicolon delimited list of probe
2572 definitions. Each definition is same as kprobe_events
2573 interface, but the parameters are comma delimited.
2574 For example, to add a kprobe event on vfs_read with
2575 arg1 and arg2, add to the command line;
2577 kprobe_event=p,vfs_read,$arg1,$arg2
2579 See also Documentation/trace/kprobetrace.rst "Kernel
2580 Boot Parameter" section.
2582 kpti= [ARM64,EARLY] Control page table isolation of
2583 user and kernel address spaces.
2584 Default: enabled on cores which need mitigation.
2588 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires
2589 CONFIG_KUNIT to be set to be fully enabled. The
2590 default value can be overridden via
2591 KUNIT_DEFAULT_ENABLED.
2592 Default is 1 (enabled)
2594 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
2595 Default is 0 (don't ignore, but inject #GP)
2597 kvm.eager_page_split=
2598 [KVM,X86] Controls whether or not KVM will try to
2599 proactively split all huge pages during dirty logging.
2600 Eager page splitting reduces interruptions to vCPU
2601 execution by eliminating the write-protection faults
2602 and MMU lock contention that would otherwise be
2603 required to split huge pages lazily.
2605 VM workloads that rarely perform writes or that write
2606 only to a small region of VM memory may benefit from
2607 disabling eager page splitting to allow huge pages to
2608 still be used for reads.
2610 The behavior of eager page splitting depends on whether
2611 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If
2612 disabled, all huge pages in a memslot will be eagerly
2613 split when dirty logging is enabled on that memslot. If
2614 enabled, eager page splitting will be performed during
2615 the KVM_CLEAR_DIRTY ioctl, and only for the pages being
2618 Eager page splitting is only supported when kvm.tdp_mmu=Y.
2622 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface.
2623 Default is false (don't support).
2626 [KVM] Controls the software workaround for the
2627 X86_BUG_ITLB_MULTIHIT bug.
2628 force : Always deploy workaround.
2629 off : Never deploy workaround.
2630 auto : Deploy workaround based on the presence of
2631 X86_BUG_ITLB_MULTIHIT.
2635 If the software workaround is enabled for the host,
2636 guests do need not to enable it for nested guests.
2638 kvm.nx_huge_pages_recovery_ratio=
2639 [KVM] Controls how many 4KiB pages are periodically zapped
2640 back to huge pages. 0 disables the recovery, otherwise if
2641 the value is N KVM will zap 1/Nth of the 4KiB pages every
2642 period (see below). The default is 60.
2644 kvm.nx_huge_pages_recovery_period_ms=
2645 [KVM] Controls the time period at which KVM zaps 4KiB pages
2646 back to huge pages. If the value is a non-zero N, KVM will
2647 zap a portion (see ratio above) of the pages every N msecs.
2648 If the value is 0 (the default), KVM will pick a period based
2649 on the ratio, such that a page is zapped after 1 hour on average.
2651 kvm-amd.nested= [KVM,AMD] Control nested virtualization feature in
2652 KVM/SVM. Default is 1 (enabled).
2654 kvm-amd.npt= [KVM,AMD] Control KVM's use of Nested Page Tables,
2655 a.k.a. Two-Dimensional Page Tables. Default is 1
2656 (enabled). Disable by KVM if hardware lacks support
2660 [KVM,ARM,EARLY] Select one of KVM/arm64's modes of
2663 none: Forcefully disable KVM.
2665 nvhe: Standard nVHE-based mode, without support for
2668 protected: nVHE-based mode with support for guests whose
2669 state is kept private from the host.
2671 nested: VHE-based mode with support for nested
2672 virtualization. Requires at least ARMv8.3
2675 Defaults to VHE/nVHE based on hardware support. Setting
2676 mode to "protected" will disable kexec and hibernation
2677 for the host. "nested" is experimental and should be
2678 used with extreme caution.
2680 kvm-arm.vgic_v3_group0_trap=
2681 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-0
2684 kvm-arm.vgic_v3_group1_trap=
2685 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-1
2688 kvm-arm.vgic_v3_common_trap=
2689 [KVM,ARM,EARLY] Trap guest accesses to GICv3 common
2692 kvm-arm.vgic_v4_enable=
2693 [KVM,ARM,EARLY] Allow use of GICv4 for direct
2696 kvm_cma_resv_ratio=n [PPC,EARLY]
2697 Reserves given percentage from system memory area for
2698 contiguous memory allocation for KVM hash pagetable
2700 By default it reserves 5% of total system memory.
2704 kvm-intel.ept= [KVM,Intel] Control KVM's use of Extended Page Tables,
2705 a.k.a. Two-Dimensional Page Tables. Default is 1
2706 (enabled). Disable by KVM if hardware lacks support
2709 kvm-intel.emulate_invalid_guest_state=
2710 [KVM,Intel] Control whether to emulate invalid guest
2711 state. Ignored if kvm-intel.enable_unrestricted_guest=1,
2712 as guest state is never invalid for unrestricted
2713 guests. This param doesn't apply to nested guests (L2),
2714 as KVM never emulates invalid L2 guest state.
2715 Default is 1 (enabled).
2717 kvm-intel.flexpriority=
2718 [KVM,Intel] Control KVM's use of FlexPriority feature
2719 (TPR shadow). Default is 1 (enabled). Disable by KVM if
2720 hardware lacks support for it.
2723 [KVM,Intel] Control nested virtualization feature in
2724 KVM/VMX. Default is 1 (enabled).
2726 kvm-intel.unrestricted_guest=
2727 [KVM,Intel] Control KVM's use of unrestricted guest
2728 feature (virtualized real and unpaged mode). Default
2729 is 1 (enabled). Disable by KVM if EPT is disabled or
2730 hardware lacks support for it.
2732 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault
2735 Valid arguments: never, cond, always
2737 always: L1D cache flush on every VMENTER.
2738 cond: Flush L1D on VMENTER only when the code between
2739 VMEXIT and VMENTER can leak host memory.
2740 never: Disables the mitigation
2742 Default is cond (do L1 cache flush in specific instances)
2744 kvm-intel.vpid= [KVM,Intel] Control KVM's use of Virtual Processor
2745 Identification feature (tagged TLBs). Default is 1
2746 (enabled). Disable by KVM if hardware lacks support
2749 l1d_flush= [X86,INTEL,EARLY]
2750 Control mitigation for L1D based snooping vulnerability.
2752 Certain CPUs are vulnerable to an exploit against CPU
2753 internal buffers which can forward information to a
2754 disclosure gadget under certain conditions.
2756 In vulnerable processors, the speculatively
2757 forwarded data can be used in a cache side channel
2758 attack, to access data to which the attacker does
2759 not have direct access.
2761 This parameter controls the mitigation. The
2764 on - enable the interface for the mitigation
2766 l1tf= [X86,EARLY] Control mitigation of the L1TF vulnerability on
2769 The kernel PTE inversion protection is unconditionally
2770 enabled and cannot be disabled.
2773 Provides all available mitigations for the
2774 L1TF vulnerability. Disables SMT and
2775 enables all mitigations in the
2776 hypervisors, i.e. unconditional L1D flush.
2778 SMT control and L1D flush control via the
2779 sysfs interface is still possible after
2780 boot. Hypervisors will issue a warning
2781 when the first VM is started in a
2782 potentially insecure configuration,
2783 i.e. SMT enabled or L1D flush disabled.
2786 Same as 'full', but disables SMT and L1D
2787 flush runtime control. Implies the
2788 'nosmt=force' command line option.
2789 (i.e. sysfs control of SMT is disabled.)
2792 Leaves SMT enabled and enables the default
2793 hypervisor mitigation, i.e. conditional
2796 SMT control and L1D flush control via the
2797 sysfs interface is still possible after
2798 boot. Hypervisors will issue a warning
2799 when the first VM is started in a
2800 potentially insecure configuration,
2801 i.e. SMT enabled or L1D flush disabled.
2805 Disables SMT and enables the default
2806 hypervisor mitigation.
2808 SMT control and L1D flush control via the
2809 sysfs interface is still possible after
2810 boot. Hypervisors will issue a warning
2811 when the first VM is started in a
2812 potentially insecure configuration,
2813 i.e. SMT enabled or L1D flush disabled.
2816 Same as 'flush', but hypervisors will not
2817 warn when a VM is started in a potentially
2818 insecure configuration.
2821 Disables hypervisor mitigations and doesn't
2823 It also drops the swap size and available
2824 RAM limit restriction on both hypervisor and
2829 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst
2835 lapic [X86-32,APIC,EARLY] Enable the local APIC even if BIOS
2838 lapic= [X86,APIC] Do not use TSC deadline
2839 value for LAPIC timer one-shot implementation. Default
2840 back to the programmable timer unit in the LAPIC.
2841 Format: notscdeadline
2843 lapic_timer_c2_ok [X86,APIC,EARLY] trust the local apic timer
2846 libata.dma= [LIBATA] DMA control
2847 libata.dma=0 Disable all PATA and SATA DMA
2848 libata.dma=1 PATA and SATA Disk DMA only
2849 libata.dma=2 ATAPI (CDROM) DMA only
2850 libata.dma=4 Compact Flash DMA only
2851 Combinations also work, so libata.dma=3 enables DMA
2852 for disks and CDROMs, but not CFs.
2854 libata.ignore_hpa= [LIBATA] Ignore HPA limit
2855 libata.ignore_hpa=0 keep BIOS limits (default)
2856 libata.ignore_hpa=1 ignore limits, using full disk
2858 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
2862 libata.force= [LIBATA] Force configurations. The format is a comma-
2863 separated list of "[ID:]VAL" where ID is PORT[.DEVICE].
2864 PORT and DEVICE are decimal numbers matching port, link
2865 or device. Basically, it matches the ATA ID string
2866 printed on console by libata. If the whole ID part is
2867 omitted, the last PORT and DEVICE values are used. If
2868 ID hasn't been specified yet, the configuration applies
2869 to all ports, links and devices.
2871 If only DEVICE is omitted, the parameter applies to
2872 the port and all links and devices behind it. DEVICE
2873 number of 0 either selects the first device or the
2874 first fan-out link behind PMP device. It does not
2875 select the host link. DEVICE number of 15 selects the
2876 host link and device attached to it.
2878 The VAL specifies the configuration to force. As long
2879 as there is no ambiguity, shortcut notation is allowed.
2880 For example, both 1.5 and 1.5G would work for 1.5Gbps.
2881 The following configurations can be forced.
2883 * Cable type: 40c, 80c, short40c, unk, ign or sata.
2884 Any ID with matching PORT is used.
2886 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
2888 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
2889 udma[/][16,25,33,44,66,100,133] notation is also
2892 * nohrst, nosrst, norst: suppress hard, soft and both
2895 * rstonce: only attempt one reset during hot-unplug
2898 * [no]dbdelay: Enable or disable the extra 200ms delay
2899 before debouncing a link PHY and device presence
2902 * [no]ncq: Turn on or off NCQ.
2904 * [no]ncqtrim: Enable or disable queued DSM TRIM.
2906 * [no]ncqati: Enable or disable NCQ trim on ATI chipset.
2908 * [no]trim: Enable or disable (unqueued) TRIM.
2910 * trim_zero: Indicate that TRIM command zeroes data.
2912 * max_trim_128m: Set 128M maximum trim size limit.
2914 * [no]dma: Turn on or off DMA transfers.
2916 * atapi_dmadir: Enable ATAPI DMADIR bridge support.
2918 * atapi_mod16_dma: Enable the use of ATAPI DMA for
2919 commands that are not a multiple of 16 bytes.
2921 * [no]dmalog: Enable or disable the use of the
2922 READ LOG DMA EXT command to access logs.
2924 * [no]iddevlog: Enable or disable access to the
2925 identify device data log.
2927 * [no]logdir: Enable or disable access to the general
2928 purpose log directory.
2930 * max_sec_128: Set transfer size limit to 128 sectors.
2932 * max_sec_1024: Set or clear transfer size limit to
2935 * max_sec_lba48: Set or clear transfer size limit to
2938 * [no]lpm: Enable or disable link power management.
2940 * [no]setxfer: Indicate if transfer speed mode setting
2943 * [no]fua: Disable or enable FUA (Force Unit Access)
2944 support for devices supporting this feature.
2946 * dump_id: Dump IDENTIFY data.
2948 * disable: Disable this device.
2950 If there are multiple matching configurations changing
2951 the same attribute, the last one is used.
2953 load_ramdisk= [RAM] [Deprecated]
2955 lockd.nlm_grace_period=P [NFS] Assign grace period.
2958 lockd.nlm_tcpport=N [NFS] Assign TCP port.
2961 lockd.nlm_timeout=T [NFS] Assign timeout value.
2964 lockd.nlm_udpport=M [NFS] Assign UDP port.
2967 lockdown= [SECURITY,EARLY]
2968 { integrity | confidentiality }
2969 Enable the kernel lockdown feature. If set to
2970 integrity, kernel features that allow userland to
2971 modify the running kernel are disabled. If set to
2972 confidentiality, kernel features that allow userland
2973 to extract confidential information from the kernel
2976 locktorture.acq_writer_lim= [KNL]
2977 Set the time limit in jiffies for a lock
2978 acquisition. Acquisitions exceeding this limit
2979 will result in a splat once they do complete.
2981 locktorture.bind_readers= [KNL]
2982 Specify the list of CPUs to which the readers are
2985 locktorture.bind_writers= [KNL]
2986 Specify the list of CPUs to which the writers are
2989 locktorture.call_rcu_chains= [KNL]
2990 Specify the number of self-propagating call_rcu()
2991 chains to set up. These are used to ensure that
2992 there is a high probability of an RCU grace period
2993 in progress at any given time. Defaults to 0,
2994 which disables these call_rcu() chains.
2996 locktorture.long_hold= [KNL]
2997 Specify the duration in milliseconds for the
2998 occasional long-duration lock hold time. Defaults
2999 to 100 milliseconds. Select 0 to disable.
3001 locktorture.nested_locks= [KNL]
3002 Specify the maximum lock nesting depth that
3003 locktorture is to exercise, up to a limit of 8
3004 (MAX_NESTED_LOCKS). Specify zero to disable.
3005 Note that this parameter is ineffective on types
3006 of locks that do not support nested acquisition.
3008 locktorture.nreaders_stress= [KNL]
3009 Set the number of locking read-acquisition kthreads.
3010 Defaults to being automatically set based on the
3011 number of online CPUs.
3013 locktorture.nwriters_stress= [KNL]
3014 Set the number of locking write-acquisition kthreads.
3016 locktorture.onoff_holdoff= [KNL]
3017 Set time (s) after boot for CPU-hotplug testing.
3019 locktorture.onoff_interval= [KNL]
3020 Set time (s) between CPU-hotplug operations, or
3021 zero to disable CPU-hotplug testing.
3023 locktorture.rt_boost= [KNL]
3024 Do periodic testing of real-time lock priority
3025 boosting. Select 0 to disable, 1 to boost
3026 only rt_mutex, and 2 to boost unconditionally.
3027 Defaults to 2, which might seem to be an
3028 odd choice, but which should be harmless for
3029 non-real-time spinlocks, due to their disabling
3030 of preemption. Note that non-realtime mutexes
3033 locktorture.rt_boost_factor= [KNL]
3034 Number that determines how often and for how
3035 long priority boosting is exercised. This is
3036 scaled down by the number of writers, so that the
3037 number of boosts per unit time remains roughly
3038 constant as the number of writers increases.
3039 On the other hand, the duration of each boost
3040 increases with the number of writers.
3042 locktorture.shuffle_interval= [KNL]
3043 Set task-shuffle interval (jiffies). Shuffling
3044 tasks allows some CPUs to go into dyntick-idle
3045 mode during the locktorture test.
3047 locktorture.shutdown_secs= [KNL]
3048 Set time (s) after boot system shutdown. This
3049 is useful for hands-off automated testing.
3051 locktorture.stat_interval= [KNL]
3052 Time (s) between statistics printk()s.
3054 locktorture.stutter= [KNL]
3055 Time (s) to stutter testing, for example,
3056 specifying five seconds causes the test to run for
3057 five seconds, wait for five seconds, and so on.
3058 This tests the locking primitive's ability to
3059 transition abruptly to and from idle.
3061 locktorture.torture_type= [KNL]
3062 Specify the locking implementation to test.
3064 locktorture.verbose= [KNL]
3065 Enable additional printk() statements.
3067 locktorture.writer_fifo= [KNL]
3068 Run the write-side locktorture kthreads at
3069 sched_set_fifo() real-time priority.
3071 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
3074 loglevel= [KNL,EARLY]
3075 All Kernel Messages with a loglevel smaller than the
3076 console loglevel will be printed to the console. It can
3077 also be changed with klogd or other programs. The
3078 loglevels are defined as follows:
3080 0 (KERN_EMERG) system is unusable
3081 1 (KERN_ALERT) action must be taken immediately
3082 2 (KERN_CRIT) critical conditions
3083 3 (KERN_ERR) error conditions
3084 4 (KERN_WARNING) warning conditions
3085 5 (KERN_NOTICE) normal but significant condition
3086 6 (KERN_INFO) informational
3087 7 (KERN_DEBUG) debug-level messages
3089 log_buf_len=n[KMG] [KNL,EARLY]
3090 Sets the size of the printk ring buffer, in bytes.
3091 n must be a power of two and greater than the
3092 minimal size. The minimal size is defined by
3093 LOG_BUF_SHIFT kernel config parameter. There
3094 is also CONFIG_LOG_CPU_MAX_BUF_SHIFT config
3095 parameter that allows to increase the default size
3096 depending on the number of CPUs. See init/Kconfig
3099 logo.nologo [FB] Disables display of the built-in Linux logo.
3100 This may be used to provide more screen space for
3101 kernel log messages and is useful when debugging
3102 kernel boot problems.
3104 lp=0 [LP] Specify parallel ports to use, e.g,
3105 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
3106 lp=reset first parallel port). 'lp=0' disables the
3107 lp=auto printer driver. 'lp=reset' (which can be
3108 specified in addition to the ports) causes
3109 attached printers to be reset. Using
3110 lp=port1,port2,... specifies the parallel ports
3111 to associate lp devices with, starting with
3112 lp0. A port specification may be 'none' to skip
3113 that lp device, or a parport name such as
3114 'parport0'. Specifying 'lp=auto' instead of a
3115 port specification list means that device IDs
3116 from each port should be examined, to see if
3117 an IEEE 1284-compliant printer is attached; if
3118 so, the driver will manage that printer.
3119 See also header of drivers/char/lp.c.
3122 Sets loops_per_jiffy to given constant, thus avoiding
3123 time-consuming boot-time autodetection (up to 250 ms per
3124 CPU). 0 enables autodetection (default). To determine
3125 the correct value for your kernel, boot with normal
3126 autodetection and see what value is printed. Note that
3127 on SMP systems the preset will be applied to all CPUs,
3128 which is likely to cause problems if your CPUs need
3129 significantly divergent settings. An incorrect value
3130 will cause delays in the kernel to be wrong, leading to
3131 unpredictable I/O errors and other breakage. Although
3132 unlikely, in the extreme case this might damage your
3136 Format: <io>,<irq>,<dma>
3138 lsm.debug [SECURITY] Enable LSM initialization debugging output.
3141 [SECURITY] Choose order of LSM initialization. This
3142 overrides CONFIG_LSM, and the "security=" parameter.
3144 machvec= [IA-64] Force the use of a particular machine-vector
3145 (machvec) in a generic kernel.
3146 Example: machvec=hpzx1
3148 machtype= [Loongson] Share the same kernel image file between
3149 different yeeloong laptops.
3150 Example: machtype=lemote-yeeloong-2f-7inch
3152 max_addr=nn[KMG] [KNL,BOOT,IA-64] All physical memory greater
3153 than or equal to this physical address is ignored.
3155 maxcpus= [SMP,EARLY] Maximum number of processors that an SMP kernel
3156 will bring up during bootup. maxcpus=n : n >= 0 limits
3157 the kernel to bring up 'n' processors. Surely after
3158 bootup you can bring up the other plugged cpu by executing
3159 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus
3160 only takes effect during system bootup.
3161 While n=0 is a special case, it is equivalent to "nosmp",
3162 which also disables the IO APIC.
3164 max_loop= [LOOP] The number of loop block devices that get
3165 (loop.max_loop) unconditionally pre-created at init time. The default
3166 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
3167 of statically allocating a predefined number, loop
3168 devices can be requested on-demand with the
3169 /dev/loop-control interface.
3171 mce [X86-32] Machine Check Exception
3173 mce=option [X86-64] See Documentation/arch/x86/x86_64/boot-options.rst
3175 md= [HW] RAID subsystems devices and level
3176 See Documentation/admin-guide/md.rst.
3179 Format: <first>,<last>
3180 Specifies range of consoles to be captured by the MDA.
3182 mds= [X86,INTEL,EARLY]
3183 Control mitigation for the Micro-architectural Data
3184 Sampling (MDS) vulnerability.
3186 Certain CPUs are vulnerable to an exploit against CPU
3187 internal buffers which can forward information to a
3188 disclosure gadget under certain conditions.
3190 In vulnerable processors, the speculatively
3191 forwarded data can be used in a cache side channel
3192 attack, to access data to which the attacker does
3193 not have direct access.
3195 This parameter controls the MDS mitigation. The
3198 full - Enable MDS mitigation on vulnerable CPUs
3199 full,nosmt - Enable MDS mitigation and disable
3200 SMT on vulnerable CPUs
3201 off - Unconditionally disable MDS mitigation
3203 On TAA-affected machines, mds=off can be prevented by
3204 an active TAA mitigation as both vulnerabilities are
3205 mitigated with the same mechanism so in order to disable
3206 this mitigation, you need to specify tsx_async_abort=off
3209 Not specifying this option is equivalent to
3212 For details see: Documentation/admin-guide/hw-vuln/mds.rst
3214 mem=nn[KMG] [HEXAGON,EARLY] Set the memory size.
3215 Must be specified, otherwise memory size will be 0.
3217 mem=nn[KMG] [KNL,BOOT,EARLY] Force usage of a specific amount
3218 of memory Amount of memory to be used in cases
3222 2 when the kernel is not able to see the whole system memory;
3223 3 memory that lies after 'mem=' boundary is excluded from
3224 the hypervisor, then assigned to KVM guests.
3225 4 to limit the memory available for kdump kernel.
3227 [ARC,MICROBLAZE] - the limit applies only to low memory,
3228 high memory is not affected.
3230 [ARM64] - only limits memory covered by the linear
3231 mapping. The NOMAP regions are not affected.
3233 [X86] Work as limiting max address. Use together
3234 with memmap= to avoid physical address space collisions.
3235 Without memmap= PCI devices could be placed at addresses
3236 belonging to unused RAM.
3238 Note that this only takes effects during boot time since
3239 in above case 3, memory may need be hot added after boot
3240 if system memory of hypervisor is not sufficient.
3243 [ARM,MIPS,EARLY] - override the memory layout
3244 reported by firmware.
3245 Define a memory region of size nn[KMG] starting at
3247 Multiple different regions can be specified with
3248 multiple mem= parameters on the command line.
3250 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
3253 memblock=debug [KNL,EARLY] Enable memblock debug messages.
3256 [KNL,SH] Allow user to override the default size for
3257 per-device physically contiguous DMA buffers.
3259 memhp_default_state=online/offline/online_kernel/online_movable
3260 [KNL] Set the initial state for the memory hotplug
3261 onlining policy. If not specified, the default value is
3262 set according to the
3263 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config
3265 See Documentation/admin-guide/mm/memory-hotplug.rst.
3267 memmap=exactmap [KNL,X86,EARLY] Enable setting of an exact
3268 E820 memory map, as specified by the user.
3269 Such memmap=exactmap lines can be constructed based on
3270 BIOS output or other requirements. See the memmap=nn@ss
3273 memmap=nn[KMG]@ss[KMG]
3274 [KNL, X86,MIPS,XTENSA,EARLY] Force usage of a specific region of memory.
3275 Region of memory to be used is from ss to ss+nn.
3276 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG],
3277 which limits max address to nn[KMG].
3278 Multiple different regions can be specified,
3281 memmap=100M@2G,100M#3G,1G!1024G
3283 memmap=nn[KMG]#ss[KMG]
3284 [KNL,ACPI,EARLY] Mark specific memory as ACPI data.
3285 Region of memory to be marked is from ss to ss+nn.
3287 memmap=nn[KMG]$ss[KMG]
3288 [KNL,ACPI,EARLY] Mark specific memory as reserved.
3289 Region of memory to be reserved is from ss to ss+nn.
3290 Example: Exclude memory from 0x18690000-0x1869ffff
3291 memmap=64K$0x18690000
3293 memmap=0x10000$0x18690000
3294 Some bootloaders may need an escape character before '$',
3295 like Grub2, otherwise '$' and the following number
3298 memmap=nn[KMG]!ss[KMG,EARLY]
3299 [KNL,X86] Mark specific memory as protected.
3300 Region of memory to be used, from ss to ss+nn.
3301 The memory region may be marked as e820 type 12 (0xc)
3302 and is NVDIMM or ADR memory.
3304 memmap=<size>%<offset>-<oldtype>+<newtype>
3305 [KNL,ACPI,EARLY] Convert memory within the specified region
3306 from <oldtype> to <newtype>. If "-<oldtype>" is left
3307 out, the whole region will be marked as <newtype>,
3308 even if previously unavailable. If "+<newtype>" is left
3309 out, matching memory will be removed. Types are
3310 specified as e820 types, e.g., 1 = RAM, 2 = reserved,
3311 3 = ACPI, 12 = PRAM.
3313 memory_corruption_check=0/1 [X86,EARLY]
3314 Some BIOSes seem to corrupt the first 64k of
3315 memory when doing things like suspend/resume.
3316 Setting this option will scan the memory
3317 looking for corruption. Enabling this will
3318 both detect corruption and prevent the kernel
3319 from using the memory being corrupted.
3320 However, its intended as a diagnostic tool; if
3321 repeatable BIOS-originated corruption always
3322 affects the same memory, you can use memmap=
3323 to prevent the kernel from using that memory.
3325 memory_corruption_check_size=size [X86,EARLY]
3326 By default it checks for corruption in the low
3327 64k, making this memory unavailable for normal
3328 use. Use this parameter to scan for
3329 corruption in more or less memory.
3331 memory_corruption_check_period=seconds [X86,EARLY]
3332 By default it checks for corruption every 60
3333 seconds. Use this parameter to check at some
3334 other rate. 0 disables periodic checking.
3336 memory_hotplug.memmap_on_memory
3337 [KNL,X86,ARM] Boolean flag to enable this feature.
3338 Format: {on | off (default)}
3339 When enabled, runtime hotplugged memory will
3340 allocate its internal metadata (struct pages,
3341 those vmemmap pages cannot be optimized even
3342 if hugetlb_free_vmemmap is enabled) from the
3343 hotadded memory which will allow to hotadd a
3344 lot of memory without requiring additional
3346 This feature is disabled by default because it
3347 has some implication on large (e.g. GB)
3348 allocations in some configurations (e.g. small
3350 The state of the flag can be read in
3351 /sys/module/memory_hotplug/parameters/memmap_on_memory.
3352 Note that even when enabled, there are a few cases where
3353 the feature is not effective.
3355 memtest= [KNL,X86,ARM,M68K,PPC,RISCV,EARLY] Enable memtest
3357 default : 0 <disable>
3358 Specifies the number of memtest passes to be
3359 performed. Each pass selects another test
3360 pattern from a given set of patterns. Memtest
3361 fills the memory with this pattern, validates
3362 memory contents and reserves bad memory
3363 regions that are detected.
3365 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control
3366 Valid arguments: on, off
3368 mem_encrypt=on: Activate SME
3369 mem_encrypt=off: Do not activate SME
3371 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst
3372 for details on when memory encryption can be activated.
3374 mem_sleep_default= [SUSPEND] Default system suspend mode:
3375 s2idle - Suspend-To-Idle
3376 shallow - Power-On Suspend or equivalent (if supported)
3377 deep - Suspend-To-RAM or equivalent (if supported)
3378 See Documentation/admin-guide/pm/sleep-states.rst.
3380 mfgpt_irq= [IA-32] Specify the IRQ to use for the
3381 Multi-Function General Purpose Timers on AMD Geode
3384 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
3385 the BIOS has incorrectly applied a workaround. TinyBIOS
3386 version 0.98 is known to be affected, 0.99 fixes the
3387 problem by letting the user disable the workaround.
3391 microcode.force_minrev= [X86]
3393 Enable or disable the microcode minimal revision
3394 enforcement for the runtime microcode loader.
3396 min_addr=nn[KMG] [KNL,BOOT,IA-64] All physical memory below this
3397 physical address is ignored.
3399 mini2440= [ARM,HW,KNL]
3400 Format:[0..2][b][c][t]
3402 MINI2440 configuration specification:
3403 0 - The attached screen is the 3.5" TFT
3404 1 - The attached screen is the 7" TFT
3405 2 - The VGA Shield is attached (1024x768)
3406 Leaving out the screen size parameter will not load
3407 the TFT driver, and the framebuffer will be left
3409 b - Enable backlight. The TFT backlight pin will be
3410 linked to the kernel VESA blanking code and a GPIO
3411 LED. This parameter is not necessary when using the
3413 c - Enable the s3c camera interface.
3414 t - Reserved for enabling touchscreen support. The
3415 touchscreen support is not enabled in the mainstream
3416 kernel as of 2.6.30, a preliminary port can be found
3417 in the "bleeding edge" mini2440 support kernel at
3418 https://repo.or.cz/w/linux-2.6/mini2440.git
3421 [X86,PPC,S390,ARM64,EARLY] Control optional mitigations for
3422 CPU vulnerabilities. This is a set of curated,
3423 arch-independent options, each of which is an
3424 aggregation of existing arch-specific options.
3427 Disable all optional CPU mitigations. This
3428 improves system performance, but it may also
3429 expose users to several CPU vulnerabilities.
3430 Equivalent to: if nokaslr then kpti=0 [ARM64]
3431 gather_data_sampling=off [X86]
3432 kvm.nx_huge_pages=off [X86]
3435 mmio_stale_data=off [X86]
3436 no_entry_flush [PPC]
3437 no_uaccess_flush [PPC]
3440 nospectre_bhb [ARM64]
3441 nospectre_v1 [X86,PPC]
3442 nospectre_v2 [X86,PPC,S390,ARM64]
3443 reg_file_data_sampling=off [X86]
3445 spec_rstack_overflow=off [X86]
3446 spec_store_bypass_disable=off [X86,PPC]
3447 spectre_v2_user=off [X86]
3448 srbds=off [X86,INTEL]
3449 ssbd=force-off [ARM64]
3450 tsx_async_abort=off [X86]
3453 This does not have any effect on
3454 kvm.nx_huge_pages when
3455 kvm.nx_huge_pages=force.
3458 Mitigate all CPU vulnerabilities, but leave SMT
3459 enabled, even if it's vulnerable. This is for
3460 users who don't want to be surprised by SMT
3461 getting disabled across kernel upgrades, or who
3462 have other ways of avoiding SMT-based attacks.
3463 Equivalent to: (default behavior)
3466 Mitigate all CPU vulnerabilities, disabling SMT
3467 if needed. This is for users who always want to
3468 be fully mitigated, even if it means losing SMT.
3469 Equivalent to: l1tf=flush,nosmt [X86]
3470 mds=full,nosmt [X86]
3471 tsx_async_abort=full,nosmt [X86]
3472 mmio_stale_data=full,nosmt [X86]
3473 retbleed=auto,nosmt [X86]
3476 [KNL,EARLY] When CONFIG_DEBUG_MEMORY_INIT is set, this
3477 parameter allows control of the logging verbosity for
3478 the additional memory initialisation checks. A value
3479 of 0 disables mminit logging and a level of 4 will
3480 log everything. Information is printed at KERN_DEBUG
3481 so loglevel=8 may also need to be specified.
3484 [X86,INTEL,EARLY] Control mitigation for the Processor
3485 MMIO Stale Data vulnerabilities.
3487 Processor MMIO Stale Data is a class of
3488 vulnerabilities that may expose data after an MMIO
3489 operation. Exposed data could originate or end in
3490 the same CPU buffers as affected by MDS and TAA.
3491 Therefore, similar to MDS and TAA, the mitigation
3492 is to clear the affected CPU buffers.
3494 This parameter controls the mitigation. The
3497 full - Enable mitigation on vulnerable CPUs
3499 full,nosmt - Enable mitigation and disable SMT on
3502 off - Unconditionally disable mitigation
3504 On MDS or TAA affected machines,
3505 mmio_stale_data=off can be prevented by an active
3506 MDS or TAA mitigation as these vulnerabilities are
3507 mitigated with the same mechanism so in order to
3508 disable this mitigation, you need to specify
3509 mds=off and tsx_async_abort=off too.
3511 Not specifying this option is equivalent to
3512 mmio_stale_data=full.
3515 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst
3517 <module>.async_probe[=<bool>] [KNL]
3518 If no <bool> value is specified or if the value
3519 specified is not a valid <bool>, enable asynchronous
3520 probe on this module. Otherwise, enable/disable
3521 asynchronous probe on this module as indicated by the
3522 <bool> value. See also: module.async_probe
3524 module.async_probe=<bool>
3525 [KNL] When set to true, modules will use async probing
3526 by default. To enable/disable async probing for a
3527 specific module, use the module specific control that
3528 is documented under <module>.async_probe. When both
3529 module.async_probe and <module>.async_probe are
3530 specified, <module>.async_probe takes precedence for
3531 the specific module.
3533 module.enable_dups_trace
3534 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set,
3535 this means that duplicate request_module() calls will
3536 trigger a WARN_ON() instead of a pr_warn(). Note that
3537 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s
3538 will always be issued and this option does nothing.
3540 [KNL] When CONFIG_MODULE_SIG is set, this means that
3541 modules without (valid) signatures will fail to load.
3542 Note that if CONFIG_MODULE_SIG_FORCE is set, that
3543 is always true, so this option does nothing.
3545 module_blacklist= [KNL] Do not load a comma-separated list of
3546 modules. Useful for debugging problem modules.
3549 [MOUSE] Maximum time between finger touching and
3550 leaving touchpad surface for touch to be considered
3551 a tap and be reported as a left button click (for
3552 touchpads working in absolute mode only).
3554 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
3555 reporting absolute coordinates, such as tablets
3556 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
3557 reporting absolute coordinates, such as tablets
3559 movablecore= [KNL,X86,IA-64,PPC,EARLY]
3560 Format: nn[KMGTPE] | nn%
3561 This parameter is the complement to kernelcore=, it
3562 specifies the amount of memory used for migratable
3563 allocations. If both kernelcore and movablecore is
3564 specified, then kernelcore will be at *least* the
3565 specified value but may be more. If movablecore on its
3566 own is specified, the administrator must be careful
3567 that the amount of memory usable for all allocations
3570 movable_node [KNL,EARLY] Boot-time switch to make hotplugable memory
3571 NUMA nodes to be movable. This means that the memory
3572 of such nodes will be usable only for movable
3573 allocations which rules out almost all kernel
3574 allocations. Use with caution!
3576 MTD_Partition= [MTD]
3577 Format: <name>,<region-number>,<size>,<offset>
3579 MTD_Region= [MTD] Format:
3580 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
3583 See drivers/mtd/parsers/cmdlinepart.c
3586 ARM/S3C2412 JIVE boot control
3588 See arch/arm/mach-s3c/mach-jive.c
3590 mtouchusb.raw_coordinates=
3591 [HW] Make the MicroTouch USB driver use raw coordinates
3592 ('y', default) or cooked coordinates ('n')
3594 mtrr=debug [X86,EARLY]
3595 Enable printing debug information related to MTRR
3596 registers at boot time.
3598 mtrr_chunk_size=nn[KMG,X86,EARLY]
3599 used for mtrr cleanup. It is largest continuous chunk
3600 that could hold holes aka. UC entries.
3602 mtrr_gran_size=nn[KMG,X86,EARLY]
3603 Used for mtrr cleanup. It is granularity of mtrr block.
3605 Large value could prevent small alignment from
3608 mtrr_spare_reg_nr=n [X86,EARLY]
3610 Range: 0,7 : spare reg number
3612 Used for mtrr cleanup. It is spare mtrr entries number.
3613 Set to 2 or more if your graphical card needs more.
3615 multitce=off [PPC] This parameter disables the use of the pSeries
3616 firmware feature for updating multiple TCE entries
3619 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
3621 netdev= [NET] Network devices parameters
3622 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
3623 Note that mem_start is often overloaded to mean
3624 something different and driver-specific.
3625 This usage is only documented in each driver source
3628 netpoll.carrier_timeout=
3629 [NET] Specifies amount of time (in seconds) that
3630 netpoll should wait for a carrier. By default netpoll
3634 [NETFILTER] Enable connection tracking flow accounting
3635 0 to disable accounting
3636 1 to enable accounting
3640 [NFS] sets the pathname to the program which is used
3641 to update the NFS client cache entries.
3643 nfs.cache_getent_timeout=
3644 [NFS] sets the timeout after which an attempt to
3645 update a cache entry is deemed to have failed.
3647 nfs.callback_nr_threads=
3648 [NFSv4] set the total number of threads that the
3649 NFS client will assign to service NFSv4 callback
3652 nfs.callback_tcpport=
3653 [NFS] set the TCP port on which the NFSv4 callback
3654 channel should listen.
3657 [NFS] specifies the number of times the NFSv4 client
3658 retries the request before returning an EAGAIN error,
3659 after a reply of NFS4ERR_DELAY from the server.
3660 Only applies if the softerr mount option is enabled,
3661 and the specified value is >= 0.
3664 [NFS] enable 64-bit inode numbers.
3665 If zero, the NFS client will fake up a 32-bit inode
3666 number for the readdir() and stat() syscalls instead
3667 of returning the full 64-bit number.
3668 The default is to return 64-bit inode numbers.
3670 nfs.idmap_cache_timeout=
3671 [NFS] set the maximum lifetime for idmapper cache
3674 nfs.max_session_cb_slots=
3675 [NFSv4.1] Sets the maximum number of session
3676 slots the client will assign to the callback
3677 channel. This determines the maximum number of
3678 callbacks the client will process in parallel for
3679 a particular server.
3681 nfs.max_session_slots=
3682 [NFSv4.1] Sets the maximum number of session slots
3683 the client will attempt to negotiate with the server.
3684 This limits the number of simultaneous RPC requests
3685 that the client can send to the NFSv4.1 server.
3686 Note that there is little point in setting this
3687 value higher than the max_tcp_slot_table_limit.
3689 nfs.nfs4_disable_idmapping=
3690 [NFSv4] When set to the default of '1', this option
3691 ensures that both the RPC level authentication
3692 scheme and the NFS level operations agree to use
3693 numeric uids/gids if the mount is using the
3694 'sec=sys' security flavour. In effect it is
3695 disabling idmapping, which can make migration from
3696 legacy NFSv2/v3 systems to NFSv4 easier.
3697 Servers that do not support this mode of operation
3698 will be autodetected by the client, and it will fall
3699 back to using the idmapper.
3700 To turn off this behaviour, set the value to '0'.
3703 [NFS4] Specify an additional fixed unique ident-
3704 ification string that NFSv4 clients can insert into
3705 their nfs_client_id4 string. This is typically a
3706 UUID that is generated at system install time.
3708 nfs.recover_lost_locks=
3709 [NFSv4] Attempt to recover locks that were lost due
3710 to a lease timeout on the server. Please note that
3711 doing this risks data corruption, since there are
3712 no guarantees that the file will remain unchanged
3713 after the locks are lost.
3714 If you want to enable the kernel legacy behaviour of
3715 attempting to recover these locks, then set this
3717 The default parameter value of '0' causes the kernel
3718 not to attempt recovery of lost locks.
3720 nfs.send_implementation_id=
3721 [NFSv4.1] Send client implementation identification
3722 information in exchange_id requests.
3723 If zero, no implementation identification information
3725 The default is to send the implementation identification
3728 nfs4.layoutstats_timer=
3729 [NFSv4.2] Change the rate at which the kernel sends
3730 layoutstats to the pNFS metadata server.
3732 Setting this to value to 0 causes the kernel to use
3733 whatever value is the default set by the layout
3734 driver. A non-zero value sets the minimum interval
3735 in seconds between layoutstats transmissions.
3737 nfsd.inter_copy_offload_enable=
3738 [NFSv4.2] When set to 1, the server will support
3739 server-to-server copies for which this server is
3740 the destination of the copy.
3742 nfsd.nfs4_disable_idmapping=
3743 [NFSv4] When set to the default of '1', the NFSv4
3744 server will return only numeric uids and gids to
3745 clients using auth_sys, and will accept numeric uids
3746 and gids from such clients. This is intended to ease
3747 migration from NFSv2/v3.
3749 nfsd.nfsd4_ssc_umount_timeout=
3750 [NFSv4.2] When used as the destination of a
3751 server-to-server copy, knfsd temporarily mounts
3752 the source server. It caches the mount in case
3753 it will be needed again, and discards it if not
3754 used for the number of milliseconds specified by
3757 nfsaddrs= [NFS] Deprecated. Use ip= instead.
3758 See Documentation/admin-guide/nfs/nfsroot.rst.
3760 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
3761 See Documentation/admin-guide/nfs/nfsroot.rst.
3763 nfsrootdebug [NFS] enable nfsroot debugging messages.
3764 See Documentation/admin-guide/nfs/nfsroot.rst.
3766 nmi_backtrace.backtrace_idle [KNL]
3767 Dump stacks even of idle CPUs in response to an
3768 NMI stack-backtrace request.
3770 nmi_debug= [KNL,SH] Specify one or more actions to take
3771 when a NMI is triggered.
3772 Format: [state][,regs][,debounce][,die]
3774 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
3775 Format: [panic,][nopanic,][num]
3777 0 - turn hardlockup detector in nmi_watchdog off
3778 1 - turn hardlockup detector in nmi_watchdog on
3779 When panic is specified, panic when an NMI watchdog
3780 timeout occurs (or 'nopanic' to not panic on an NMI
3781 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set)
3782 To disable both hard and soft lockup detectors,
3783 please see 'nowatchdog'.
3784 This is useful when you use a panic=... timeout and
3785 need the box quickly up again.
3787 These settings can be accessed at runtime via
3788 the nmi_watchdog and hardlockup_panic sysctls.
3790 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
3791 emulation library even if a 387 maths coprocessor
3794 no4lvl [RISCV,EARLY] Disable 4-level and 5-level paging modes.
3795 Forces kernel to use 3-level paging instead.
3797 no5lvl [X86-64,RISCV,EARLY] Disable 5-level paging mode. Forces
3798 kernel to use 4-level paging instead.
3802 noaltinstr [S390,EARLY] Disables alternative instructions
3803 patching (CPU alternatives feature).
3805 noapic [SMP,APIC,EARLY] Tells the kernel to not make use of any
3806 IOAPICs that may be present in the system.
3808 noautogroup Disable scheduler automatic task group creation.
3813 [HW] Never suspend the console
3814 Disable suspending of consoles during suspend and
3815 hibernate operations. Once disabled, debugging
3816 messages can reach various consoles while the rest
3817 of the system is being put to sleep (ie, while
3818 debugging driver suspend/resume hooks). This may
3819 not work reliably with all consoles, but is known
3820 to work with serial and VGA consoles.
3821 To facilitate more flexible debugging, we also add
3822 console_suspend, a printk module parameter to control
3823 it. Users could use console_suspend (usually
3824 /sys/module/printk/parameters/console_suspend) to
3825 turn on/off it dynamically.
3828 [KNL,EARLY] Disable object debugging
3830 nodsp [SH] Disable hardware DSP at boot time.
3832 noefi [EFI,EARLY] Disable EFI runtime services support.
3834 no_entry_flush [PPC,EARLY] Don't flush the L1-D cache when entering the kernel.
3839 This affects only 32-bit executables.
3840 noexec32=on: enable non-executable mappings (default)
3841 read doesn't imply executable mappings
3842 noexec32=off: disable non-executable mappings
3843 read implies executable mappings
3845 no_file_caps Tells the kernel not to honor file capabilities. The
3846 only way then for a file to be executed with privilege
3847 is to be setuid root or executed by root.
3849 nofpu [MIPS,SH] Disable hardware FPU at boot time.
3851 nofsgsbase [X86] Disables FSGSBASE instructions.
3853 nofxsr [BUGS=X86-32] Disables x86 floating point extended
3854 register save and restore. The kernel will only save
3855 legacy floating-point registers on task switch.
3857 nohalt [IA-64] Tells the kernel not to use the power saving
3858 function PAL_HALT_LIGHT when idle. This increases
3859 power-consumption. On the positive side, it reduces
3860 interrupt wake-up latency, which may improve performance
3861 in certain environments such as networked servers or
3866 Force pointers printed to the console or buffers to be
3867 unhashed. By default, when a pointer is printed via %p
3868 format string, that pointer is "hashed", i.e. obscured
3869 by hashing the pointer value. This is a security feature
3870 that hides actual kernel addresses from unprivileged
3871 users, but it also makes debugging the kernel more
3872 difficult since unequal pointers can no longer be
3873 compared. However, if this command-line option is
3874 specified, then all normal pointers will have their true
3875 value printed. This option should only be specified when
3876 debugging the kernel. Please do not use on production
3879 nohibernate [HIBERNATION] Disable hibernation and resume.
3881 nohlt [ARM,ARM64,MICROBLAZE,MIPS,PPC,SH] Forces the kernel to
3882 busy wait in do_idle() and not use the arch_cpu_idle()
3883 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP
3884 to be effective. This is useful on platforms where the
3885 sleep(SH) or wfi(ARM,ARM64) instructions do not work
3886 correctly or when doing power measurements to evaluate
3887 the impact of the sleep instructions. This is also
3888 useful when using JTAG debugger.
3890 nohugeiomap [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge I/O mappings.
3892 nohugevmalloc [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge vmalloc mappings.
3894 nohz= [KNL] Boottime enable/disable dynamic ticks
3895 Valid arguments: on, off
3898 nohz_full= [KNL,BOOT,SMP,ISOL]
3899 The argument is a cpu list, as described above.
3900 In kernels built with CONFIG_NO_HZ_FULL=y, set
3901 the specified list of CPUs whose tick will be stopped
3902 whenever possible. The boot CPU will be forced outside
3903 the range to maintain the timekeeping. Any CPUs
3904 in this list will have their RCU callbacks offloaded,
3905 just as if they had also been called out in the
3906 rcu_nocbs= boot parameter.
3908 Note that this argument takes precedence over
3909 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option.
3911 noinitrd [RAM] Tells the kernel not to load any configured
3914 nointremap [X86-64,Intel-IOMMU,EARLY] Do not enable interrupt
3916 [Deprecated - use intremap=off]
3920 noinvpcid [X86,EARLY] Disable the INVPCID cpu feature.
3922 noiotrap [SH] Disables trapped I/O port accesses.
3924 noirqdebug [X86-32] Disables the code which attempts to detect and
3925 disable unhandled interrupt sources.
3927 noisapnp [ISAPNP] Disables ISA PnP code.
3929 nojitter [IA-64] Disables jitter checking for ITC timers.
3932 When CONFIG_RANDOMIZE_BASE is set, this disables
3933 kernel and module base offset ASLR (Address Space
3934 Layout Randomization).
3936 no-kvmapf [X86,KVM,EARLY] Disable paravirtualized asynchronous page
3939 no-kvmclock [X86,KVM,EARLY] Disable paravirtualized KVM clock driver
3941 nolapic [X86-32,APIC,EARLY] Do not enable or use the local APIC.
3943 nolapic_timer [X86-32,APIC,EARLY] Do not use the local APIC timer.
3945 nomca [IA-64] Disable machine check abort handling
3947 nomce [X86-32] Disable Machine Check Exception
3949 nomfgpt [X86-32] Disable Multi-Function General Purpose
3950 Timer usage (for AMD Geode machines).
3952 nomodeset Disable kernel modesetting. Most systems' firmware
3953 sets up a display mode and provides framebuffer memory
3954 for output. With nomodeset, DRM and fbdev drivers will
3955 not load if they could possibly displace the pre-
3956 initialized output. Only the system framebuffer will
3957 be available for use. The respective drivers will not
3958 perform display-mode changes or accelerated rendering.
3960 Useful as error fallback, or for testing and debugging.
3962 nomodule Disable module load
3964 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
3965 shutdown the other cpus. Instead use the REBOOT_VECTOR
3968 nopat [X86,EARLY] Disable PAT (page attribute table extension of
3969 pagetables) support.
3971 nopcid [X86-64,EARLY] Disable the PCID cpu feature.
3973 nopku [X86] Disable Memory Protection Keys CPU feature found
3976 nopti [X86-64,EARLY]
3977 Equivalent to pti=off
3979 nopv= [X86,XEN,KVM,HYPER_V,VMWARE,EARLY]
3980 Disables the PV optimizations forcing the guest to run
3981 as generic guest with no PV drivers. Currently support
3982 XEN HVM, KVM, HYPER_V and VMWARE guest.
3984 nopvspin [X86,XEN,KVM,EARLY]
3985 Disables the qspinlock slow path using PV optimizations
3986 which allow the hypervisor to 'idle' the guest on lock
3989 norandmaps Don't use address space randomization. Equivalent to
3990 echo 0 > /proc/sys/kernel/randomize_va_space
3992 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
3993 with UP alternatives
3995 noresume [SWSUSP] Disables resume and restores original swap
4000 no-scroll [VGA] Disables scrollback.
4001 This is required for the Braillex ib80-piezo Braille
4002 reader made by F.H. Papenmeier (Germany).
4004 nosgx [X86-64,SGX,EARLY] Disables Intel SGX kernel support.
4007 Disable SMAP (Supervisor Mode Access Prevention)
4008 even if it is supported by processor.
4010 nosmep [PPC64s,EARLY]
4011 Disable SMEP (Supervisor Mode Execution Prevention)
4012 even if it is supported by processor.
4014 nosmp [SMP,EARLY] Tells an SMP kernel to act as a UP kernel,
4015 and disable the IO APIC. legacy for "maxcpus=0".
4017 nosmt [KNL,MIPS,PPC,S390,EARLY] Disable symmetric multithreading (SMT).
4018 Equivalent to smt=1.
4020 [KNL,X86,PPC] Disable symmetric multithreading (SMT).
4021 nosmt=force: Force disable SMT, cannot be undone
4022 via the sysfs control file.
4024 nosoftlockup [KNL] Disable the soft-lockup detector.
4026 nospec_store_bypass_disable
4027 [HW,EARLY] Disable all mitigations for the Speculative
4028 Store Bypass vulnerability
4030 nospectre_bhb [ARM64,EARLY] Disable all mitigations for Spectre-BHB (branch
4031 history injection) vulnerability. System may allow data leaks
4034 nospectre_v1 [X86,PPC,EARLY] Disable mitigations for Spectre Variant 1
4035 (bounds check bypass). With this option data leaks are
4036 possible in the system.
4038 nospectre_v2 [X86,PPC_E500,ARM64,EARLY] Disable all mitigations
4039 for the Spectre variant 2 (indirect branch
4040 prediction) vulnerability. System may allow data
4041 leaks with this option.
4043 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES,RISCV,EARLY] Disable
4044 paravirtualized steal time accounting. steal time is
4045 computed, but won't influence scheduler behaviour
4047 nosync [HW,M68K] Disables sync negotiation for all devices.
4049 no_timer_check [X86,APIC] Disables the code which tests for
4050 broken timer IRQ sources.
4053 [PPC,EARLY] Don't flush the L1-D cache after accessing user data.
4055 novmcoredd [KNL,KDUMP]
4056 Disable device dump. Device dump allows drivers to
4057 append dump data to vmcore so you can collect driver
4058 specified debug info. Drivers can append the data
4059 without any limit and this data is stored in memory,
4060 so this may cause significant memory stress. Disabling
4061 device dump can help save memory but the driver debug
4062 data will be no longer available. This parameter
4063 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP
4067 [X86,PV_OPS,EARLY] Disable paravirtualized VMware
4068 scheduler clock and use the default one.
4070 nowatchdog [KNL] Disable both lockup detectors, i.e.
4071 soft-lockup and NMI watchdog (hard-lockup).
4075 nox2apic [X86-64,APIC,EARLY] Do not enable x2APIC mode.
4077 NOTE: this parameter will be ignored on systems with the
4078 LEGACY_XAPIC_DISABLED bit set in the
4079 IA32_XAPIC_DISABLE_STATUS MSR.
4081 noxsave [BUGS=X86] Disables x86 extended register state save
4082 and restore using xsave. The kernel will fallback to
4083 enabling legacy floating-point and sse state.
4085 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
4086 register states. The kernel will fall back to use
4087 xsave to save the states. By using this parameter,
4088 performance of saving the states is degraded because
4089 xsave doesn't support modified optimization while
4090 xsaveopt supports it on xsaveopt enabled systems.
4092 noxsaves [X86] Disables xsaves and xrstors used in saving and
4093 restoring x86 extended register state in compacted
4094 form of xsave area. The kernel will fall back to use
4095 xsaveopt and xrstor to save and restore the states
4096 in standard form of xsave area. By using this
4097 parameter, xsave area per process might occupy more
4098 memory on xsaves enabled systems.
4100 nps_mtm_hs_ctr= [KNL,ARC]
4101 This parameter sets the maximum duration, in
4102 cycles, each HW thread of the CTOP can run
4103 without interruptions, before HW switches it.
4104 The actual maximum duration is 16 times this
4106 Format: integer between 1 and 255
4109 nptcg= [IA-64] Override max number of concurrent global TLB
4110 purges which is reported from either PAL_VM_SUMMARY or
4113 nr_cpus= [SMP,EARLY] Maximum number of processors that an SMP kernel
4114 could support. nr_cpus=n : n >= 1 limits the kernel to
4115 support 'n' processors. It could be larger than the
4116 number of already plugged CPU during bootup, later in
4117 runtime you can physically add extra cpu until it reaches
4118 n. So during boot up some boot time memory for per-cpu
4119 variables need be pre-allocated for later physical cpu
4122 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
4124 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86, EARLY]
4125 Disable NUMA, Only set up a single NUMA node
4126 spanning all memory.
4128 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic
4130 Allowed values are enable and disable
4132 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
4133 'node', 'default' can be specified
4134 This can be set from sysctl after boot.
4135 See Documentation/admin-guide/sysctl/vm.rst for details.
4137 ohci1394_dma=early [HW,EARLY] enable debugging via the ohci1394 driver.
4138 See Documentation/core-api/debugging-via-ohci1394.rst for more
4141 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
4142 Rather than timing out after 20 ms if an EC
4143 command is not properly ACKed, override the length
4144 of the timeout. We have interrupts disabled while
4145 waiting for the ACK, so if this is set too high
4146 interrupts *may* be lost!
4148 omap_mux= [OMAP] Override bootloader pin multiplexing.
4149 Format: <mux_mode0.mode_name=value>...
4150 For example, to override I2C bus2:
4151 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
4153 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
4155 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
4157 boundary - index of last SLC block on Flex-OneNAND.
4158 The remaining blocks are configured as MLC blocks.
4159 lock - Configure if Flex-OneNAND boundary should be locked.
4160 Once locked, the boundary cannot be changed.
4161 1 indicates lock status, 0 indicates unlock status.
4163 oops=panic [KNL,EARLY]
4164 Always panic on oopses. Default is to just kill the
4165 process, but there is a small probability of
4166 deadlocking the machine.
4167 This will also cause panics on machine check exceptions.
4168 Useful together with panic=30 to trigger a reboot.
4171 [KNL] Boolean flag to control whether the page allocator
4172 should randomize its free lists. The randomization may
4173 be automatically enabled if the kernel detects it is
4174 running on a platform with a direct-mapped memory-side
4175 cache, and this parameter can be used to
4176 override/disable that behavior. The state of the flag
4177 can be read from sysfs at:
4178 /sys/module/page_alloc/parameters/shuffle.
4180 page_owner= [KNL,EARLY] Boot-time page_owner enabling option.
4181 Storage of the information about who allocated
4182 each page is disabled in default. With this switch,
4184 on: enable the feature
4186 page_poison= [KNL,EARLY] Boot-time parameter changing the state of
4187 poisoning on the buddy allocator, available with
4188 CONFIG_PAGE_POISONING=y.
4189 off: turn off poisoning (default)
4190 on: turn on poisoning
4192 page_reporting.page_reporting_order=
4193 [KNL] Minimal page reporting order
4195 Adjust the minimal page reporting order. The page
4196 reporting is disabled when it exceeds MAX_PAGE_ORDER.
4198 panic= [KNL] Kernel behaviour on panic: delay <timeout>
4199 timeout > 0: seconds before rebooting
4200 timeout = 0: wait forever
4201 timeout < 0: reboot immediately
4204 panic_on_taint= [KNL,EARLY]
4205 Bitmask for conditionally calling panic() in add_taint()
4206 Format: <hex>[,nousertaint]
4207 Hexadecimal bitmask representing the set of TAINT flags
4208 that will cause the kernel to panic when add_taint() is
4209 called with any of the flags in this set.
4210 The optional switch "nousertaint" can be utilized to
4211 prevent userspace forced crashes by writing to sysctl
4212 /proc/sys/kernel/tainted any flagset matching with the
4213 bitmask set on panic_on_taint.
4214 See Documentation/admin-guide/tainted-kernels.rst for
4215 extra details on the taint flags that users can pick
4216 to compose the bitmask to assign to panic_on_taint.
4218 panic_on_warn=1 panic() instead of WARN(). Useful to cause kdump
4221 panic_print= Bitmask for printing system info when panic happens.
4222 User can chose combination of the following bits:
4223 bit 0: print all tasks info
4224 bit 1: print system memory info
4225 bit 2: print timer info
4226 bit 3: print locks info if CONFIG_LOCKDEP is on
4227 bit 4: print ftrace buffer
4228 bit 5: print all printk messages in buffer
4229 bit 6: print all CPUs backtrace (if available in the arch)
4230 bit 7: print only tasks in uninterruptible (blocked) state
4231 *Be aware* that this option may print a _lot_ of lines,
4232 so there are risks of losing older messages in the log.
4233 Use this option carefully, maybe worth to setup a
4234 bigger log buffer with "log_buf_len" along with this.
4236 parkbd.port= [HW] Parallel port number the keyboard adapter is
4237 connected to, default is 0.
4239 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
4240 0 for XT, 1 for AT (default is AT).
4243 parport= [HW,PPT] Specify parallel ports. 0 disables.
4244 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
4245 Use 'auto' to force the driver to use any
4246 IRQ/DMA settings detected (the default is to
4247 ignore detected IRQ/DMA settings because of
4248 possible conflicts). You can specify the base
4249 address, IRQ, and DMA settings; IRQ and DMA
4250 should be numbers, or 'auto' (for using detected
4251 settings on that particular port), or 'nofifo'
4252 (to avoid using a FIFO even if it is detected).
4253 Parallel ports are assigned in the order they
4254 are specified on the command line, starting
4257 parport_init_mode= [HW,PPT]
4258 Configure VIA parallel port to operate in
4259 a specific mode. This is necessary on Pegasos
4260 computer where firmware has no options for setting
4261 up parallel port mode and sets it to spp.
4262 Currently this function knows 686a and 8231 chips.
4263 Format: [spp|ps2|epp|ecp|ecpepp]
4265 pata_legacy.all= [HW,LIBATA]
4267 Set to non-zero to probe primary and secondary ISA
4268 port ranges on PCI systems where no PCI PATA device
4269 has been found at either range. Disabled by default.
4271 pata_legacy.autospeed= [HW,LIBATA]
4273 Set to non-zero if a chip is present that snoops speed
4274 changes. Disabled by default.
4276 pata_legacy.ht6560a= [HW,LIBATA]
4278 Set to 1, 2, or 3 for HT 6560A on the primary channel,
4279 the secondary channel, or both channels respectively.
4280 Disabled by default.
4282 pata_legacy.ht6560b= [HW,LIBATA]
4284 Set to 1, 2, or 3 for HT 6560B on the primary channel,
4285 the secondary channel, or both channels respectively.
4286 Disabled by default.
4288 pata_legacy.iordy_mask= [HW,LIBATA]
4290 IORDY enable mask. Set individual bits to allow IORDY
4291 for the respective channel. Bit 0 is for the first
4292 legacy channel handled by this driver, bit 1 is for
4293 the second channel, and so on. The sequence will often
4294 correspond to the primary legacy channel, the secondary
4295 legacy channel, and so on, but the handling of a PCI
4296 bus and the use of other driver options may interfere
4297 with the sequence. By default IORDY is allowed across
4300 pata_legacy.opti82c46x= [HW,LIBATA]
4302 Set to 1, 2, or 3 for Opti 82c611A on the primary
4303 channel, the secondary channel, or both channels
4304 respectively. Disabled by default.
4306 pata_legacy.opti82c611a= [HW,LIBATA]
4308 Set to 1, 2, or 3 for Opti 82c465MV on the primary
4309 channel, the secondary channel, or both channels
4310 respectively. Disabled by default.
4312 pata_legacy.pio_mask= [HW,LIBATA]
4314 PIO mode mask for autospeed devices. Set individual
4315 bits to allow the use of the respective PIO modes.
4316 Bit 0 is for mode 0, bit 1 is for mode 1, and so on.
4317 All modes allowed by default.
4319 pata_legacy.probe_all= [HW,LIBATA]
4321 Set to non-zero to probe tertiary and further ISA
4322 port ranges on PCI systems. Disabled by default.
4324 pata_legacy.probe_mask= [HW,LIBATA]
4326 Probe mask for legacy ISA PATA ports. Depending on
4327 platform configuration and the use of other driver
4328 options up to 6 legacy ports are supported: 0x1f0,
4329 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing
4330 of individual ports can be disabled by setting the
4331 corresponding bits in the mask to 1. Bit 0 is for
4332 the first port in the list above (0x1f0), and so on.
4333 By default all supported ports are probed.
4335 pata_legacy.qdi= [HW,LIBATA]
4337 Set to non-zero to probe QDI controllers. By default
4338 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise.
4340 pata_legacy.winbond= [HW,LIBATA]
4342 Set to non-zero to probe Winbond controllers. Use
4343 the standard I/O port (0x130) if 1, otherwise the
4344 value given is the I/O port to use (typically 0x1b0).
4345 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE,
4348 pata_platform.pio_mask= [HW,LIBATA]
4350 Supported PIO mode mask. Set individual bits to allow
4351 the use of the respective PIO modes. Bit 0 is for
4352 mode 0, bit 1 is for mode 1, and so on. Mode 0 only
4356 Halt all CPUs after the first oops has been printed for
4357 the specified number of seconds. This is to be used if
4358 your oopses keep scrolling off the screen.
4362 pci=option[,option...] [PCI,EARLY] various PCI subsystem options.
4364 Some options herein operate on a specific device
4365 or a set of devices (<pci_dev>). These are
4366 specified in one of the following formats:
4368 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]*
4369 pci:<vendor>:<device>[:<subvendor>:<subdevice>]
4371 Note: the first format specifies a PCI
4372 bus/device/function address which may change
4373 if new hardware is inserted, if motherboard
4374 firmware changes, or due to changes caused
4375 by other kernel parameters. If the
4376 domain is left unspecified, it is
4377 taken to be zero. Optionally, a path
4378 to a device through multiple device/function
4379 addresses can be specified after the base
4380 address (this is more robust against
4381 renumbering issues). The second format
4382 selects devices using IDs from the
4383 configuration space which may match multiple
4384 devices in the system.
4386 earlydump dump PCI config space before the kernel
4388 off [X86] don't probe for the PCI bus
4389 bios [X86-32] force use of PCI BIOS, don't access
4390 the hardware directly. Use this if your machine
4391 has a non-standard PCI host bridge.
4392 nobios [X86-32] disallow use of PCI BIOS, only direct
4393 hardware access methods are allowed. Use this
4394 if you experience crashes upon bootup and you
4395 suspect they are caused by the BIOS.
4396 conf1 [X86] Force use of PCI Configuration Access
4397 Mechanism 1 (config address in IO port 0xCF8,
4398 data in IO port 0xCFC, both 32-bit).
4399 conf2 [X86] Force use of PCI Configuration Access
4400 Mechanism 2 (IO port 0xCF8 is an 8-bit port for
4401 the function, IO port 0xCFA, also 8-bit, sets
4402 bus number. The config space is then accessed
4403 through ports 0xC000-0xCFFF).
4404 See http://wiki.osdev.org/PCI for more info
4405 on the configuration access mechanisms.
4406 noaer [PCIE] If the PCIEAER kernel config parameter is
4407 enabled, this kernel boot option can be used to
4408 disable the use of PCIE advanced error reporting.
4409 nodomains [PCI] Disable support for multiple PCI
4410 root domains (aka PCI segments, in ACPI-speak).
4411 nommconf [X86] Disable use of MMCONFIG for PCI
4413 check_enable_amd_mmconf [X86] check for and enable
4414 properly configured MMIO access to PCI
4415 config space on AMD family 10h CPU
4416 nomsi [MSI] If the PCI_MSI kernel config parameter is
4417 enabled, this kernel boot option can be used to
4418 disable the use of MSI interrupts system-wide.
4419 noioapicquirk [APIC] Disable all boot interrupt quirks.
4420 Safety option to keep boot IRQs enabled. This
4421 should never be necessary.
4422 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
4423 primary IO-APIC for bridges that cannot disable
4424 boot IRQs. This fixes a source of spurious IRQs
4425 when the system masks IRQs.
4426 noioapicreroute [APIC] Disable workaround that uses the
4427 boot IRQ equivalent of an IRQ that connects to
4428 a chipset where boot IRQs cannot be disabled.
4429 The opposite of ioapicreroute.
4430 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
4431 routing table. These calls are known to be buggy
4432 on several machines and they hang the machine
4433 when used, but on other computers it's the only
4434 way to get the interrupt routing table. Try
4435 this option if the kernel is unable to allocate
4436 IRQs or discover secondary PCI buses on your
4438 rom [X86] Assign address space to expansion ROMs.
4439 Use with caution as certain devices share
4440 address decoders between ROMs and other
4442 norom [X86] Do not assign address space to
4443 expansion ROMs that do not already have
4444 BIOS assigned address ranges.
4445 nobar [X86] Do not assign address space to the
4446 BARs that weren't assigned by the BIOS.
4447 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
4448 assigned automatically to PCI devices. You can
4449 make the kernel exclude IRQs of your ISA cards
4451 pirqaddr=0xAAAAA [X86] Specify the physical address
4452 of the PIRQ table (normally generated
4453 by the BIOS) if it is outside the
4454 F0000h-100000h range.
4455 lastbus=N [X86] Scan all buses thru bus #N. Can be
4456 useful if the kernel is unable to find your
4457 secondary buses and you want to tell it
4458 explicitly which ones they are.
4459 assign-busses [X86] Always assign all PCI bus
4460 numbers ourselves, overriding
4461 whatever the firmware may have done.
4462 usepirqmask [X86] Honor the possible IRQ mask stored
4463 in the BIOS $PIR table. This is needed on
4464 some systems with broken BIOSes, notably
4465 some HP Pavilion N5400 and Omnibook XE3
4466 notebooks. This will have no effect if ACPI
4467 IRQ routing is enabled.
4468 noacpi [X86] Do not use ACPI for IRQ routing
4469 or for PCI scanning.
4470 use_crs [X86] Use PCI host bridge window information
4471 from ACPI. On BIOSes from 2008 or later, this
4472 is enabled by default. If you need to use this,
4473 please report a bug.
4474 nocrs [X86] Ignore PCI host bridge windows from ACPI.
4475 If you need to use this, please report a bug.
4476 use_e820 [X86] Use E820 reservations to exclude parts of
4477 PCI host bridge windows. This is a workaround
4478 for BIOS defects in host bridge _CRS methods.
4479 If you need to use this, please report a bug to
4480 <linux-pci@vger.kernel.org>.
4481 no_e820 [X86] Ignore E820 reservations for PCI host
4482 bridge windows. This is the default on modern
4483 hardware. If you need to use this, please report
4484 a bug to <linux-pci@vger.kernel.org>.
4485 routeirq Do IRQ routing for all PCI devices.
4486 This is normally done in pci_enable_device(),
4487 so this option is a temporary workaround
4488 for broken drivers that don't call it.
4489 skip_isa_align [X86] do not align io start addr, so can
4490 handle more pci cards
4491 noearly [X86] Don't do any early type 1 scanning.
4492 This might help on some broken boards which
4493 machine check when some devices' config space
4494 is read. But various workarounds are disabled
4495 and some IOMMU drivers will not work.
4496 bfsort Sort PCI devices into breadth-first order.
4497 This sorting is done to get a device
4498 order compatible with older (<= 2.4) kernels.
4499 nobfsort Don't sort PCI devices into breadth-first order.
4500 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
4501 tuning and use the BIOS-configured MPS defaults.
4502 pcie_bus_safe Set every device's MPS to the largest value
4503 supported by all devices below the root complex.
4504 pcie_bus_perf Set device MPS to the largest allowable MPS
4505 based on its parent bus. Also set MRRS (Max
4506 Read Request Size) to the largest supported
4507 value (no larger than the MPS that the device
4508 or bus can support) for best performance.
4509 pcie_bus_peer2peer Set every device's MPS to 128B, which
4510 every device is guaranteed to support. This
4511 configuration allows peer-to-peer DMA between
4512 any pair of devices, possibly at the cost of
4513 reduced performance. This also guarantees
4514 that hot-added devices will work.
4515 cbiosize=nn[KMG] The fixed amount of bus space which is
4516 reserved for the CardBus bridge's IO window.
4517 The default value is 256 bytes.
4518 cbmemsize=nn[KMG] The fixed amount of bus space which is
4519 reserved for the CardBus bridge's memory
4520 window. The default value is 64 megabytes.
4523 [<order of align>@]<pci_dev>[; ...]
4524 Specifies alignment and device to reassign
4525 aligned memory resources. How to
4526 specify the device is described above.
4527 If <order of align> is not specified,
4528 PAGE_SIZE is used as alignment.
4529 A PCI-PCI bridge can be specified if resource
4530 windows need to be expanded.
4531 To specify the alignment for several
4532 instances of a device, the PCI vendor,
4533 device, subvendor, and subdevice may be
4534 specified, e.g., 12@pci:8086:9c22:103c:198f
4535 for 4096-byte alignment.
4536 ecrc= Enable/disable PCIe ECRC (transaction layer
4537 end-to-end CRC checking). Only effective if
4538 OS has native AER control (either granted by
4539 ACPI _OSC or forced via "pcie_ports=native")
4540 bios: Use BIOS/firmware settings. This is the
4544 hpiosize=nn[KMG] The fixed amount of bus space which is
4545 reserved for hotplug bridge's IO window.
4546 Default size is 256 bytes.
4547 hpmmiosize=nn[KMG] The fixed amount of bus space which is
4548 reserved for hotplug bridge's MMIO window.
4549 Default size is 2 megabytes.
4550 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is
4551 reserved for hotplug bridge's MMIO_PREF window.
4552 Default size is 2 megabytes.
4553 hpmemsize=nn[KMG] The fixed amount of bus space which is
4554 reserved for hotplug bridge's MMIO and
4556 Default size is 2 megabytes.
4557 hpbussize=nn The minimum amount of additional bus numbers
4558 reserved for buses below a hotplug bridge.
4560 realloc= Enable/disable reallocating PCI bridge resources
4561 if allocations done by BIOS are too small to
4562 accommodate resources required by all child
4564 off: Turn realloc off
4566 realloc same as realloc=on
4567 noari do not use PCIe ARI.
4568 noats [PCIE, Intel-IOMMU, AMD-IOMMU]
4569 do not use PCIe ATS (and IOMMU device IOTLB).
4570 pcie_scan_all Scan all possible PCIe devices. Otherwise we
4571 only look for one device below a PCIe downstream
4573 big_root_window Try to add a big 64bit memory window to the PCIe
4574 root complex on AMD CPUs. Some GFX hardware
4575 can resize a BAR to allow access to all VRAM.
4576 Adding the window is slightly risky (it may
4577 conflict with unreported devices), so this
4579 disable_acs_redir=<pci_dev>[; ...]
4580 Specify one or more PCI devices (in the format
4581 specified above) separated by semicolons.
4582 Each device specified will have the PCI ACS
4583 redirect capabilities forced off which will
4584 allow P2P traffic between devices through
4585 bridges without forcing it upstream. Note:
4586 this removes isolation between devices and
4587 may put more devices in an IOMMU group.
4588 force_floating [S390] Force usage of floating interrupts.
4589 nomio [S390] Do not use MIO instructions.
4590 norid [S390] ignore the RID field and force use of
4591 one PCI domain per PCI function
4593 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
4596 force Enable ASPM even on devices that claim not to support it.
4597 WARNING: Forcing ASPM on may cause system lockups.
4599 pcie_ports= [PCIE] PCIe port services handling:
4600 native Use native PCIe services (PME, AER, DPC, PCIe hotplug)
4601 even if the platform doesn't give the OS permission to
4602 use them. This may cause conflicts if the platform
4603 also tries to use these services.
4604 dpc-native Use native PCIe service for DPC only. May
4605 cause conflicts if firmware uses AER or DPC.
4606 compat Disable native PCIe services (PME, AER, DPC, PCIe
4609 pcie_port_pm= [PCIE] PCIe port power management handling:
4610 off Disable power management of all PCIe ports
4611 force Forcibly enable power management of all PCIe ports
4613 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
4614 nomsi Do not use MSI for native PCIe PME signaling (this makes
4615 all PCIe root ports use INTx for all services).
4617 pcmv= [HW,PCMCIA] BadgePAD 4
4621 Keep all power-domains already enabled by bootloader on,
4622 even if no driver has claimed them. This is useful
4623 for debug and development, but should not be
4624 needed on a platform with proper driver support.
4626 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
4629 See arch/parisc/kernel/pdc_chassis.c
4631 percpu_alloc= [MM,EARLY]
4632 Select which percpu first chunk allocator to use.
4633 Currently supported values are "embed" and "page".
4634 Archs may support subset or none of the selections.
4635 See comments in mm/percpu.c for details on each
4636 allocator. This parameter is primarily for debugging
4637 and performance comparison.
4639 pirq= [SMP,APIC] Manual mp-table setup
4640 See Documentation/arch/x86/i386/IO-APIC.rst.
4642 plip= [PPT,NET] Parallel port network link
4643 Format: { parport<nr> | timid | 0 }
4644 See also Documentation/admin-guide/parport.rst.
4646 pmtmr= [X86] Manual setup of pmtmr I/O Port.
4647 Override pmtimer IOPort with a hex value.
4650 pmu_override= [PPC] Override the PMU.
4651 This option takes over the PMU facility, so it is no
4652 longer usable by perf. Setting this option starts the
4653 PMU counters by setting MMCR0 to 0 (the FC bit is
4654 cleared). If a number is given, then MMCR1 is set to
4655 that number, otherwise (e.g., 'pmu_override=on'), MMCR1
4658 pm_debug_messages [SUSPEND,KNL]
4659 Enable suspend/resume debug messages during boot up.
4662 Enable PNP debug messages (depends on the
4663 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
4664 via /sys/module/pnp/parameters/debug. We always show
4665 current resource usage; turning this on also shows
4666 possible settings and some assignment information.
4672 { on | off | curr | res | no-curr | no-res }
4675 [ISAPNP] Exclude IRQs for the autoconfiguration
4678 [ISAPNP] Exclude DMAs for the autoconfiguration
4680 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
4681 Ranges are in pairs (I/O port base and size).
4684 [ISAPNP] Exclude memory regions for the
4686 Ranges are in pairs (memory base and size).
4688 ports= [IP_VS_FTP] IPVS ftp helper module
4690 Up to 8 (IP_VS_APP_MAX_PORTS) ports
4692 Format: <port>,<port>....
4694 possible_cpus= [SMP,S390,X86]
4695 Format: <unsigned int>
4696 Set the number of possible CPUs, overriding the
4697 regular discovery mechanisms (such as ACPI/FW, etc).
4699 powersave=off [PPC] This option disables power saving features.
4700 It specifically disables cpuidle and sets the
4701 platform machine description specific power_save
4702 function to NULL. On Idle the CPU just reduces
4705 ppc_strict_facility_enable
4706 [PPC,ENABLE] This option catches any kernel floating point,
4707 Altivec, VSX and SPE outside of regions specifically
4708 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
4709 There is some performance impact when enabling this.
4713 Disable Hardware Transactional Memory
4716 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC
4717 none - Limited to cond_resched() calls
4718 voluntary - Limited to cond_resched() and might_sleep() calls
4719 full - Any section that isn't explicitly preempt disabled
4720 can be preempted anytime.
4722 print-fatal-signals=
4723 [KNL] debug: print fatal signals
4725 If enabled, warn about various signal handling
4726 related application anomalies: too many signals,
4727 too many POSIX.1 timers, fatal signals causing a
4730 If you hit the warning due to signal overflow,
4731 you might want to try "ulimit -i unlimited".
4735 printk.always_kmsg_dump=
4736 Trigger kmsg_dump for cases other than kernel oops or
4738 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
4741 printk.console_no_auto_verbose=
4742 Disable console loglevel raise on oops, panic
4743 or lockdep-detected issues (only if lock debug is on).
4744 With an exception to setups with low baudrate on
4745 serial console, keeping this 0 is a good choice
4746 in order to provide more debug information.
4748 default: 0 (auto_verbose is enabled)
4750 printk.devkmsg={on,off,ratelimit}
4751 Control writing to /dev/kmsg.
4752 on - unlimited logging to /dev/kmsg from userspace
4753 off - logging to /dev/kmsg disabled
4754 ratelimit - ratelimit the logging
4757 printk.time= Show timing data prefixed to each printk message line
4758 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
4760 processor.max_cstate= [HW,ACPI]
4761 Limit processor to maximum C-state
4762 max_cstate=9 overrides any DMI blacklist limit.
4764 processor.nocst [HW,ACPI]
4765 Ignore the _CST method to determine C-states,
4766 instead using the legacy FADT method
4768 profile= [KNL] Enable kernel profiling via /proc/profile
4769 Format: [<profiletype>,]<number>
4770 Param: <profiletype>: "schedule", "sleep", or "kvm"
4771 [defaults to kernel profiling]
4772 Param: "schedule" - profile schedule points.
4773 Param: "sleep" - profile D-state sleeping (millisecs).
4774 Requires CONFIG_SCHEDSTATS
4775 Param: "kvm" - profile VM exits.
4776 Param: <number> - step/bucket size as a power of 2 for
4777 statistical time based profiling.
4779 prompt_ramdisk= [RAM] [Deprecated]
4781 prot_virt= [S390] enable hosting protected virtual machines
4782 isolated from the hypervisor (if hardware supports
4786 psi= [KNL] Enable or disable pressure stall information
4790 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
4791 probe for; one of (bare|imps|exps|lifebook|any).
4792 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
4794 psmouse.resetafter= [HW,MOUSE]
4795 Try to reset the device after so many bad packets
4798 [HW,MOUSE] Set desired mouse resolution, in dpi.
4799 psmouse.smartscroll=
4800 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
4801 0 = disabled, 1 = enabled (default).
4803 pstore.backend= Specify the name of the pstore backend to use
4805 pti= [X86-64] Control Page Table Isolation of user and
4806 kernel address spaces. Disabling this feature
4807 removes hardening, but improves performance of
4808 system calls and interrupts.
4810 on - unconditionally enable
4811 off - unconditionally disable
4812 auto - kernel detects whether your CPU model is
4813 vulnerable to issues that PTI mitigates
4815 Not specifying this option is equivalent to pti=auto.
4818 [KNL] Number of legacy pty's. Overwrites compiled-in
4821 quiet [KNL,EARLY] Disable most log messages
4825 radix_hcall_invalidate=on [PPC/PSERIES]
4826 Disable RADIX GTSE feature and use hcall for TLB
4830 See Documentation/admin-guide/md.rst.
4832 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
4833 See Documentation/admin-guide/blockdev/ramdisk.rst.
4835 ramdisk_start= [RAM] RAM disk image start address
4837 random.trust_cpu=off
4838 [KNL,EARLY] Disable trusting the use of the CPU's
4839 random number generator (if available) to
4840 initialize the kernel's RNG.
4842 random.trust_bootloader=off
4843 [KNL,EARLY] Disable trusting the use of the a seed
4844 passed by the bootloader (if available) to
4845 initialize the kernel's RNG.
4847 randomize_kstack_offset=
4848 [KNL,EARLY] Enable or disable kernel stack offset
4849 randomization, which provides roughly 5 bits of
4850 entropy, frustrating memory corruption attacks
4851 that depend on stack address determinism or
4852 cross-syscall address exposures. This is only
4853 available on architectures that have defined
4854 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET.
4855 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
4856 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT.
4858 ras=option[,option,...] [KNL] RAS-specific options
4861 Disable the Correctable Errors Collector,
4862 see CONFIG_RAS_CEC help text.
4864 rcu_nocbs[=cpu-list]
4865 [KNL] The optional argument is a cpu list,
4868 In kernels built with CONFIG_RCU_NOCB_CPU=y,
4869 enable the no-callback CPU mode, which prevents
4870 such CPUs' callbacks from being invoked in
4871 softirq context. Invocation of such CPUs' RCU
4872 callbacks will instead be offloaded to "rcuox/N"
4873 kthreads created for that purpose, where "x" is
4874 "p" for RCU-preempt, "s" for RCU-sched, and "g"
4875 for the kthreads that mediate grace periods; and
4876 "N" is the CPU number. This reduces OS jitter on
4877 the offloaded CPUs, which can be useful for HPC
4878 and real-time workloads. It can also improve
4879 energy efficiency for asymmetric multiprocessors.
4881 If a cpulist is passed as an argument, the specified
4882 list of CPUs is set to no-callback mode from boot.
4884 Otherwise, if the '=' sign and the cpulist
4885 arguments are omitted, no CPU will be set to
4886 no-callback mode from boot but the mode may be
4887 toggled at runtime via cpusets.
4889 Note that this argument takes precedence over
4890 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option.
4893 Rather than requiring that offloaded CPUs
4894 (specified by rcu_nocbs= above) explicitly
4895 awaken the corresponding "rcuoN" kthreads,
4896 make these kthreads poll for callbacks.
4897 This improves the real-time response for the
4898 offloaded CPUs by relieving them of the need to
4899 wake up the corresponding kthread, but degrades
4900 energy efficiency by requiring that the kthreads
4901 periodically wake up to do the polling.
4903 rcutree.blimit= [KNL]
4904 Set maximum number of finished RCU callbacks to
4905 process in one batch.
4907 rcutree.do_rcu_barrier= [KNL]
4908 Request a call to rcu_barrier(). This is
4909 throttled so that userspace tests can safely
4910 hammer on the sysfs variable if they so choose.
4911 If triggered before the RCU grace-period machinery
4912 is fully active, this will error out with EAGAIN.
4914 rcutree.dump_tree= [KNL]
4915 Dump the structure of the rcu_node combining tree
4916 out at early boot. This is used for diagnostic
4917 purposes, to verify correct tree setup.
4919 rcutree.gp_cleanup_delay= [KNL]
4920 Set the number of jiffies to delay each step of
4921 RCU grace-period cleanup.
4923 rcutree.gp_init_delay= [KNL]
4924 Set the number of jiffies to delay each step of
4925 RCU grace-period initialization.
4927 rcutree.gp_preinit_delay= [KNL]
4928 Set the number of jiffies to delay each step of
4929 RCU grace-period pre-initialization, that is,
4930 the propagation of recent CPU-hotplug changes up
4931 the rcu_node combining tree.
4933 rcutree.jiffies_till_first_fqs= [KNL]
4934 Set delay from grace-period initialization to
4935 first attempt to force quiescent states.
4936 Units are jiffies, minimum value is zero,
4937 and maximum value is HZ.
4939 rcutree.jiffies_till_next_fqs= [KNL]
4940 Set delay between subsequent attempts to force
4941 quiescent states. Units are jiffies, minimum
4942 value is one, and maximum value is HZ.
4944 rcutree.jiffies_till_sched_qs= [KNL]
4945 Set required age in jiffies for a
4946 given grace period before RCU starts
4947 soliciting quiescent-state help from
4948 rcu_note_context_switch() and cond_resched().
4949 If not specified, the kernel will calculate
4950 a value based on the most recent settings
4951 of rcutree.jiffies_till_first_fqs
4952 and rcutree.jiffies_till_next_fqs.
4953 This calculated value may be viewed in
4954 rcutree.jiffies_to_sched_qs. Any attempt to set
4955 rcutree.jiffies_to_sched_qs will be cheerfully
4958 rcutree.kthread_prio= [KNL,BOOT]
4959 Set the SCHED_FIFO priority of the RCU per-CPU
4960 kthreads (rcuc/N). This value is also used for
4961 the priority of the RCU boost threads (rcub/N)
4962 and for the RCU grace-period kthreads (rcu_bh,
4963 rcu_preempt, and rcu_sched). If RCU_BOOST is
4964 set, valid values are 1-99 and the default is 1
4965 (the least-favored priority). Otherwise, when
4966 RCU_BOOST is not set, valid values are 0-99 and
4967 the default is zero (non-realtime operation).
4968 When RCU_NOCB_CPU is set, also adjust the
4969 priority of NOCB callback kthreads.
4971 rcutree.nocb_nobypass_lim_per_jiffy= [KNL]
4972 On callback-offloaded (rcu_nocbs) CPUs,
4973 RCU reduces the lock contention that would
4974 otherwise be caused by callback floods through
4975 use of the ->nocb_bypass list. However, in the
4976 common non-flooded case, RCU queues directly to
4977 the main ->cblist in order to avoid the extra
4978 overhead of the ->nocb_bypass list and its lock.
4979 But if there are too many callbacks queued during
4980 a single jiffy, RCU pre-queues the callbacks into
4981 the ->nocb_bypass queue. The definition of "too
4982 many" is supplied by this kernel boot parameter.
4984 rcutree.qhimark= [KNL]
4985 Set threshold of queued RCU callbacks beyond which
4986 batch limiting is disabled.
4988 rcutree.qlowmark= [KNL]
4989 Set threshold of queued RCU callbacks below which
4990 batch limiting is re-enabled.
4992 rcutree.qovld= [KNL]
4993 Set threshold of queued RCU callbacks beyond which
4994 RCU's force-quiescent-state scan will aggressively
4995 enlist help from cond_resched() and sched IPIs to
4996 help CPUs more quickly reach quiescent states.
4997 Set to less than zero to make this be set based
4998 on rcutree.qhimark at boot time and to zero to
4999 disable more aggressive help enlistment.
5001 rcutree.rcu_delay_page_cache_fill_msec= [KNL]
5002 Set the page-cache refill delay (in milliseconds)
5003 in response to low-memory conditions. The range
5004 of permitted values is in the range 0:100000.
5006 rcutree.rcu_divisor= [KNL]
5007 Set the shift-right count to use to compute
5008 the callback-invocation batch limit bl from
5009 the number of callbacks queued on this CPU.
5010 The result will be bounded below by the value of
5011 the rcutree.blimit kernel parameter. Every bl
5012 callbacks, the softirq handler will exit in
5013 order to allow the CPU to do other work.
5015 Please note that this callback-invocation batch
5016 limit applies only to non-offloaded callback
5017 invocation. Offloaded callbacks are instead
5018 invoked in the context of an rcuoc kthread, which
5019 scheduler will preempt as it does any other task.
5021 rcutree.rcu_fanout_exact= [KNL]
5022 Disable autobalancing of the rcu_node combining
5023 tree. This is used by rcutorture, and might
5024 possibly be useful for architectures having high
5025 cache-to-cache transfer latencies.
5027 rcutree.rcu_fanout_leaf= [KNL]
5028 Change the number of CPUs assigned to each
5029 leaf rcu_node structure. Useful for very
5030 large systems, which will choose the value 64,
5031 and for NUMA systems with large remote-access
5032 latencies, which will choose a value aligned
5033 with the appropriate hardware boundaries.
5035 rcutree.rcu_min_cached_objs= [KNL]
5036 Minimum number of objects which are cached and
5037 maintained per one CPU. Object size is equal
5038 to PAGE_SIZE. The cache allows to reduce the
5039 pressure to page allocator, also it makes the
5040 whole algorithm to behave better in low memory
5043 rcutree.rcu_nocb_gp_stride= [KNL]
5044 Set the number of NOCB callback kthreads in
5045 each group, which defaults to the square root
5046 of the number of CPUs. Larger numbers reduce
5047 the wakeup overhead on the global grace-period
5048 kthread, but increases that same overhead on
5049 each group's NOCB grace-period kthread.
5051 rcutree.rcu_kick_kthreads= [KNL]
5052 Cause the grace-period kthread to get an extra
5053 wake_up() if it sleeps three times longer than
5054 it should at force-quiescent-state time.
5055 This wake_up() will be accompanied by a
5056 WARN_ONCE() splat and an ftrace_dump().
5058 rcutree.rcu_resched_ns= [KNL]
5059 Limit the time spend invoking a batch of RCU
5060 callbacks to the specified number of nanoseconds.
5061 By default, this limit is checked only once
5062 every 32 callbacks in order to limit the pain
5063 inflicted by local_clock() overhead.
5065 rcutree.rcu_unlock_delay= [KNL]
5066 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels,
5067 this specifies an rcu_read_unlock()-time delay
5068 in microseconds. This defaults to zero.
5069 Larger delays increase the probability of
5070 catching RCU pointer leaks, that is, buggy use
5071 of RCU-protected pointers after the relevant
5072 rcu_read_unlock() has completed.
5074 rcutree.sysrq_rcu= [KNL]
5075 Commandeer a sysrq key to dump out Tree RCU's
5076 rcu_node tree with an eye towards determining
5077 why a new grace period has not yet started.
5079 rcutree.use_softirq= [KNL]
5080 If set to zero, move all RCU_SOFTIRQ processing to
5081 per-CPU rcuc kthreads. Defaults to a non-zero
5082 value, meaning that RCU_SOFTIRQ is used by default.
5083 Specify rcutree.use_softirq=0 to use rcuc kthreads.
5085 But note that CONFIG_PREEMPT_RT=y kernels disable
5086 this kernel boot parameter, forcibly setting it
5089 rcutree.enable_rcu_lazy= [KNL]
5090 To save power, batch RCU callbacks and flush after
5091 delay, memory pressure or callback list growing too
5094 rcuscale.gp_async= [KNL]
5095 Measure performance of asynchronous
5096 grace-period primitives such as call_rcu().
5098 rcuscale.gp_async_max= [KNL]
5099 Specify the maximum number of outstanding
5100 callbacks per writer thread. When a writer
5101 thread exceeds this limit, it invokes the
5102 corresponding flavor of rcu_barrier() to allow
5103 previously posted callbacks to drain.
5105 rcuscale.gp_exp= [KNL]
5106 Measure performance of expedited synchronous
5107 grace-period primitives.
5109 rcuscale.holdoff= [KNL]
5110 Set test-start holdoff period. The purpose of
5111 this parameter is to delay the start of the
5112 test until boot completes in order to avoid
5115 rcuscale.kfree_by_call_rcu= [KNL]
5116 In kernels built with CONFIG_RCU_LAZY=y, test
5117 call_rcu() instead of kfree_rcu().
5119 rcuscale.kfree_mult= [KNL]
5120 Instead of allocating an object of size kfree_obj,
5121 allocate one of kfree_mult * sizeof(kfree_obj).
5124 rcuscale.kfree_rcu_test= [KNL]
5125 Set to measure performance of kfree_rcu() flooding.
5127 rcuscale.kfree_rcu_test_double= [KNL]
5128 Test the double-argument variant of kfree_rcu().
5129 If this parameter has the same value as
5130 rcuscale.kfree_rcu_test_single, both the single-
5131 and double-argument variants are tested.
5133 rcuscale.kfree_rcu_test_single= [KNL]
5134 Test the single-argument variant of kfree_rcu().
5135 If this parameter has the same value as
5136 rcuscale.kfree_rcu_test_double, both the single-
5137 and double-argument variants are tested.
5139 rcuscale.kfree_nthreads= [KNL]
5140 The number of threads running loops of kfree_rcu().
5142 rcuscale.kfree_alloc_num= [KNL]
5143 Number of allocations and frees done in an iteration.
5145 rcuscale.kfree_loops= [KNL]
5146 Number of loops doing rcuscale.kfree_alloc_num number
5147 of allocations and frees.
5149 rcuscale.minruntime= [KNL]
5150 Set the minimum test run time in seconds. This
5151 does not affect the data-collection interval,
5152 but instead allows better measurement of things
5153 like CPU consumption.
5155 rcuscale.nreaders= [KNL]
5156 Set number of RCU readers. The value -1 selects
5157 N, where N is the number of CPUs. A value
5158 "n" less than -1 selects N-n+1, where N is again
5159 the number of CPUs. For example, -2 selects N
5160 (the number of CPUs), -3 selects N+1, and so on.
5161 A value of "n" less than or equal to -N selects
5164 rcuscale.nwriters= [KNL]
5165 Set number of RCU writers. The values operate
5166 the same as for rcuscale.nreaders.
5167 N, where N is the number of CPUs
5169 rcuscale.scale_type= [KNL]
5170 Specify the RCU implementation to test.
5172 rcuscale.shutdown= [KNL]
5173 Shut the system down after performance tests
5174 complete. This is useful for hands-off automated
5177 rcuscale.verbose= [KNL]
5178 Enable additional printk() statements.
5180 rcuscale.writer_holdoff= [KNL]
5181 Write-side holdoff between grace periods,
5182 in microseconds. The default of zero says
5185 rcuscale.writer_holdoff_jiffies= [KNL]
5186 Additional write-side holdoff between grace
5187 periods, but in jiffies. The default of zero
5190 rcutorture.fqs_duration= [KNL]
5191 Set duration of force_quiescent_state bursts
5194 rcutorture.fqs_holdoff= [KNL]
5195 Set holdoff time within force_quiescent_state bursts
5198 rcutorture.fqs_stutter= [KNL]
5199 Set wait time between force_quiescent_state bursts
5202 rcutorture.fwd_progress= [KNL]
5203 Specifies the number of kthreads to be used
5204 for RCU grace-period forward-progress testing
5205 for the types of RCU supporting this notion.
5206 Defaults to 1 kthread, values less than zero or
5207 greater than the number of CPUs cause the number
5210 rcutorture.fwd_progress_div= [KNL]
5211 Specify the fraction of a CPU-stall-warning
5212 period to do tight-loop forward-progress testing.
5214 rcutorture.fwd_progress_holdoff= [KNL]
5215 Number of seconds to wait between successive
5216 forward-progress tests.
5218 rcutorture.fwd_progress_need_resched= [KNL]
5219 Enclose cond_resched() calls within checks for
5220 need_resched() during tight-loop forward-progress
5223 rcutorture.gp_cond= [KNL]
5224 Use conditional/asynchronous update-side
5225 primitives, if available.
5227 rcutorture.gp_exp= [KNL]
5228 Use expedited update-side primitives, if available.
5230 rcutorture.gp_normal= [KNL]
5231 Use normal (non-expedited) asynchronous
5232 update-side primitives, if available.
5234 rcutorture.gp_sync= [KNL]
5235 Use normal (non-expedited) synchronous
5236 update-side primitives, if available. If all
5237 of rcutorture.gp_cond=, rcutorture.gp_exp=,
5238 rcutorture.gp_normal=, and rcutorture.gp_sync=
5239 are zero, rcutorture acts as if is interpreted
5240 they are all non-zero.
5242 rcutorture.irqreader= [KNL]
5243 Run RCU readers from irq handlers, or, more
5244 accurately, from a timer handler. Not all RCU
5245 flavors take kindly to this sort of thing.
5247 rcutorture.leakpointer= [KNL]
5248 Leak an RCU-protected pointer out of the reader.
5249 This can of course result in splats, and is
5250 intended to test the ability of things like
5251 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect
5254 rcutorture.n_barrier_cbs= [KNL]
5255 Set callbacks/threads for rcu_barrier() testing.
5257 rcutorture.nfakewriters= [KNL]
5258 Set number of concurrent RCU writers. These just
5259 stress RCU, they don't participate in the actual
5260 test, hence the "fake".
5262 rcutorture.nocbs_nthreads= [KNL]
5263 Set number of RCU callback-offload togglers.
5264 Zero (the default) disables toggling.
5266 rcutorture.nocbs_toggle= [KNL]
5267 Set the delay in milliseconds between successive
5268 callback-offload toggling attempts.
5270 rcutorture.nreaders= [KNL]
5271 Set number of RCU readers. The value -1 selects
5272 N-1, where N is the number of CPUs. A value
5273 "n" less than -1 selects N-n-2, where N is again
5274 the number of CPUs. For example, -2 selects N
5275 (the number of CPUs), -3 selects N+1, and so on.
5277 rcutorture.object_debug= [KNL]
5278 Enable debug-object double-call_rcu() testing.
5280 rcutorture.onoff_holdoff= [KNL]
5281 Set time (s) after boot for CPU-hotplug testing.
5283 rcutorture.onoff_interval= [KNL]
5284 Set time (jiffies) between CPU-hotplug operations,
5285 or zero to disable CPU-hotplug testing.
5287 rcutorture.read_exit= [KNL]
5288 Set the number of read-then-exit kthreads used
5289 to test the interaction of RCU updaters and
5290 task-exit processing.
5292 rcutorture.read_exit_burst= [KNL]
5293 The number of times in a given read-then-exit
5294 episode that a set of read-then-exit kthreads
5297 rcutorture.read_exit_delay= [KNL]
5298 The delay, in seconds, between successive
5299 read-then-exit testing episodes.
5301 rcutorture.shuffle_interval= [KNL]
5302 Set task-shuffle interval (s). Shuffling tasks
5303 allows some CPUs to go into dyntick-idle mode
5304 during the rcutorture test.
5306 rcutorture.shutdown_secs= [KNL]
5307 Set time (s) after boot system shutdown. This
5308 is useful for hands-off automated testing.
5310 rcutorture.stall_cpu= [KNL]
5311 Duration of CPU stall (s) to test RCU CPU stall
5312 warnings, zero to disable.
5314 rcutorture.stall_cpu_block= [KNL]
5315 Sleep while stalling if set. This will result
5316 in warnings from preemptible RCU in addition to
5317 any other stall-related activity. Note that
5318 in kernels built with CONFIG_PREEMPTION=n and
5319 CONFIG_PREEMPT_COUNT=y, this parameter will
5320 cause the CPU to pass through a quiescent state.
5321 Given CONFIG_PREEMPTION=n, this will suppress
5322 RCU CPU stall warnings, but will instead result
5323 in scheduling-while-atomic splats.
5325 Use of this module parameter results in splats.
5328 rcutorture.stall_cpu_holdoff= [KNL]
5329 Time to wait (s) after boot before inducing stall.
5331 rcutorture.stall_cpu_irqsoff= [KNL]
5332 Disable interrupts while stalling if set.
5334 rcutorture.stall_gp_kthread= [KNL]
5335 Duration (s) of forced sleep within RCU
5336 grace-period kthread to test RCU CPU stall
5337 warnings, zero to disable. If both stall_cpu
5338 and stall_gp_kthread are specified, the
5339 kthread is starved first, then the CPU.
5341 rcutorture.stat_interval= [KNL]
5342 Time (s) between statistics printk()s.
5344 rcutorture.stutter= [KNL]
5345 Time (s) to stutter testing, for example, specifying
5346 five seconds causes the test to run for five seconds,
5347 wait for five seconds, and so on. This tests RCU's
5348 ability to transition abruptly to and from idle.
5350 rcutorture.test_boost= [KNL]
5351 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
5352 "Maybe" means test if the RCU implementation
5353 under test support RCU priority boosting.
5355 rcutorture.test_boost_duration= [KNL]
5356 Duration (s) of each individual boost test.
5358 rcutorture.test_boost_interval= [KNL]
5359 Interval (s) between each boost test.
5361 rcutorture.test_no_idle_hz= [KNL]
5362 Test RCU's dyntick-idle handling. See also the
5363 rcutorture.shuffle_interval parameter.
5365 rcutorture.torture_type= [KNL]
5366 Specify the RCU implementation to test.
5368 rcutorture.verbose= [KNL]
5369 Enable additional printk() statements.
5371 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL]
5372 Dump ftrace buffer after reporting RCU CPU
5375 rcupdate.rcu_cpu_stall_notifiers= [KNL]
5376 Provide RCU CPU stall notifiers, but see the
5377 warnings in the RCU_CPU_STALL_NOTIFIER Kconfig
5378 option's help text. TL;DR: You almost certainly
5379 do not want rcupdate.rcu_cpu_stall_notifiers.
5381 rcupdate.rcu_cpu_stall_suppress= [KNL]
5382 Suppress RCU CPU stall warning messages.
5384 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL]
5385 Suppress RCU CPU stall warning messages and
5386 rcutorture writer stall warnings that occur
5387 during early boot, that is, during the time
5388 before the init task is spawned.
5390 rcupdate.rcu_cpu_stall_timeout= [KNL]
5391 Set timeout for RCU CPU stall warning messages.
5392 The value is in seconds and the maximum allowed
5393 value is 300 seconds.
5395 rcupdate.rcu_exp_cpu_stall_timeout= [KNL]
5396 Set timeout for expedited RCU CPU stall warning
5397 messages. The value is in milliseconds
5398 and the maximum allowed value is 21000
5399 milliseconds. Please note that this value is
5400 adjusted to an arch timer tick resolution.
5401 Setting this to zero causes the value from
5402 rcupdate.rcu_cpu_stall_timeout to be used (after
5403 conversion from seconds to milliseconds).
5405 rcupdate.rcu_cpu_stall_cputime= [KNL]
5406 Provide statistics on the cputime and count of
5407 interrupts and tasks during the sampling period. For
5408 multiple continuous RCU stalls, all sampling periods
5409 begin at half of the first RCU stall timeout.
5411 rcupdate.rcu_exp_stall_task_details= [KNL]
5412 Print stack dumps of any tasks blocking the
5413 current expedited RCU grace period during an
5414 expedited RCU CPU stall warning.
5416 rcupdate.rcu_expedited= [KNL]
5417 Use expedited grace-period primitives, for
5418 example, synchronize_rcu_expedited() instead
5419 of synchronize_rcu(). This reduces latency,
5420 but can increase CPU utilization, degrade
5421 real-time latency, and degrade energy efficiency.
5422 No effect on CONFIG_TINY_RCU kernels.
5424 rcupdate.rcu_normal= [KNL]
5425 Use only normal grace-period primitives,
5426 for example, synchronize_rcu() instead of
5427 synchronize_rcu_expedited(). This improves
5428 real-time latency, CPU utilization, and
5429 energy efficiency, but can expose users to
5430 increased grace-period latency. This parameter
5431 overrides rcupdate.rcu_expedited. No effect on
5432 CONFIG_TINY_RCU kernels.
5434 rcupdate.rcu_normal_after_boot= [KNL]
5435 Once boot has completed (that is, after
5436 rcu_end_inkernel_boot() has been invoked), use
5437 only normal grace-period primitives. No effect
5438 on CONFIG_TINY_RCU kernels.
5440 But note that CONFIG_PREEMPT_RT=y kernels enables
5441 this kernel boot parameter, forcibly setting
5442 it to the value one, that is, converting any
5443 post-boot attempt at an expedited RCU grace
5444 period to instead use normal non-expedited
5445 grace-period processing.
5447 rcupdate.rcu_task_collapse_lim= [KNL]
5448 Set the maximum number of callbacks present
5449 at the beginning of a grace period that allows
5450 the RCU Tasks flavors to collapse back to using
5451 a single callback queue. This switching only
5452 occurs when rcupdate.rcu_task_enqueue_lim is
5453 set to the default value of -1.
5455 rcupdate.rcu_task_contend_lim= [KNL]
5456 Set the minimum number of callback-queuing-time
5457 lock-contention events per jiffy required to
5458 cause the RCU Tasks flavors to switch to per-CPU
5459 callback queuing. This switching only occurs
5460 when rcupdate.rcu_task_enqueue_lim is set to
5461 the default value of -1.
5463 rcupdate.rcu_task_enqueue_lim= [KNL]
5464 Set the number of callback queues to use for the
5465 RCU Tasks family of RCU flavors. The default
5466 of -1 allows this to be automatically (and
5467 dynamically) adjusted. This parameter is intended
5470 rcupdate.rcu_task_ipi_delay= [KNL]
5471 Set time in jiffies during which RCU tasks will
5472 avoid sending IPIs, starting with the beginning
5473 of a given grace period. Setting a large
5474 number avoids disturbing real-time workloads,
5475 but lengthens grace periods.
5477 rcupdate.rcu_task_lazy_lim= [KNL]
5478 Number of callbacks on a given CPU that will
5479 cancel laziness on that CPU. Use -1 to disable
5480 cancellation of laziness, but be advised that
5481 doing so increases the danger of OOM due to
5484 rcupdate.rcu_task_stall_info= [KNL]
5485 Set initial timeout in jiffies for RCU task stall
5486 informational messages, which give some indication
5487 of the problem for those not patient enough to
5488 wait for ten minutes. Informational messages are
5489 only printed prior to the stall-warning message
5490 for a given grace period. Disable with a value
5491 less than or equal to zero. Defaults to ten
5492 seconds. A change in value does not take effect
5493 until the beginning of the next grace period.
5495 rcupdate.rcu_task_stall_info_mult= [KNL]
5496 Multiplier for time interval between successive
5497 RCU task stall informational messages for a given
5498 RCU tasks grace period. This value is clamped
5499 to one through ten, inclusive. It defaults to
5500 the value three, so that the first informational
5501 message is printed 10 seconds into the grace
5502 period, the second at 40 seconds, the third at
5503 160 seconds, and then the stall warning at 600
5504 seconds would prevent a fourth at 640 seconds.
5506 rcupdate.rcu_task_stall_timeout= [KNL]
5507 Set timeout in jiffies for RCU task stall
5508 warning messages. Disable with a value less
5509 than or equal to zero. Defaults to ten minutes.
5510 A change in value does not take effect until
5511 the beginning of the next grace period.
5513 rcupdate.rcu_tasks_lazy_ms= [KNL]
5514 Set timeout in milliseconds RCU Tasks asynchronous
5515 callback batching for call_rcu_tasks().
5516 A negative value will take the default. A value
5517 of zero will disable batching. Batching is
5518 always disabled for synchronize_rcu_tasks().
5520 rcupdate.rcu_tasks_rude_lazy_ms= [KNL]
5521 Set timeout in milliseconds RCU Tasks
5522 Rude asynchronous callback batching for
5523 call_rcu_tasks_rude(). A negative value
5524 will take the default. A value of zero will
5525 disable batching. Batching is always disabled
5526 for synchronize_rcu_tasks_rude().
5528 rcupdate.rcu_tasks_trace_lazy_ms= [KNL]
5529 Set timeout in milliseconds RCU Tasks
5530 Trace asynchronous callback batching for
5531 call_rcu_tasks_trace(). A negative value
5532 will take the default. A value of zero will
5533 disable batching. Batching is always disabled
5534 for synchronize_rcu_tasks_trace().
5536 rcupdate.rcu_self_test= [KNL]
5537 Run the RCU early boot self tests
5541 Run specified binary instead of /init from the ramdisk,
5542 used for early userspace startup. See initrd.
5545 force - Override the decision by the kernel to hide the
5546 advertisement of RDRAND support (this affects
5547 certain AMD processors because of buggy BIOS
5548 support, specifically around the suspend/resume
5552 Turn on/off individual RDT features. List is:
5553 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp,
5555 E.g. to turn on cmt and turn off mba use:
5559 Format (x86 or x86_64):
5560 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \
5562 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
5564 Where reboot_mode is one of warm (soft) or cold (hard) or gpio
5565 (prefix with 'panic_' to set mode for panic
5567 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
5568 reboot_force is either force or not specified,
5569 reboot_cpu is s[mp]#### with #### being the processor
5570 to be used for rebooting.
5572 refscale.holdoff= [KNL]
5573 Set test-start holdoff period. The purpose of
5574 this parameter is to delay the start of the
5575 test until boot completes in order to avoid
5578 refscale.lookup_instances= [KNL]
5579 Number of data elements to use for the forms of
5580 SLAB_TYPESAFE_BY_RCU testing. A negative number
5581 is negated and multiplied by nr_cpu_ids, while
5582 zero specifies nr_cpu_ids.
5584 refscale.loops= [KNL]
5585 Set the number of loops over the synchronization
5586 primitive under test. Increasing this number
5587 reduces noise due to loop start/end overhead,
5588 but the default has already reduced the per-pass
5589 noise to a handful of picoseconds on ca. 2020
5592 refscale.nreaders= [KNL]
5593 Set number of readers. The default value of -1
5594 selects N, where N is roughly 75% of the number
5595 of CPUs. A value of zero is an interesting choice.
5597 refscale.nruns= [KNL]
5598 Set number of runs, each of which is dumped onto
5601 refscale.readdelay= [KNL]
5602 Set the read-side critical-section duration,
5603 measured in microseconds.
5605 refscale.scale_type= [KNL]
5606 Specify the read-protection implementation to test.
5608 refscale.shutdown= [KNL]
5609 Shut down the system at the end of the performance
5610 test. This defaults to 1 (shut it down) when
5611 refscale is built into the kernel and to 0 (leave
5612 it running) when refscale is built as a module.
5614 refscale.verbose= [KNL]
5615 Enable additional printk() statements.
5617 refscale.verbose_batched= [KNL]
5618 Batch the additional printk() statements. If zero
5619 (the default) or negative, print everything. Otherwise,
5620 print every Nth verbose statement, where N is the value
5623 regulator_ignore_unused
5625 Prevents regulator framework from disabling regulators
5626 that are unused, due no driver claiming them. This may
5627 be useful for debug and development, but should not be
5628 needed on a platform with proper driver support.
5631 [KNL, SMP] Set scheduler's default relax_domain_level.
5632 See Documentation/admin-guide/cgroup-v1/cpusets.rst.
5634 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory
5635 Format: <base1>,<size1>[,<base2>,<size2>,...]
5636 Reserve I/O ports or memory so the kernel won't use
5637 them. If <base> is less than 0x10000, the region
5638 is assumed to be I/O ports; otherwise it is memory.
5640 reservetop= [X86-32,EARLY]
5642 Reserves a hole at the top of the kernel virtual
5645 reset_devices [KNL] Force drivers to reset the underlying device
5646 during initialization.
5649 Specify the partition device for software suspend
5651 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
5653 resume_offset= [SWSUSP]
5654 Specify the offset from the beginning of the partition
5655 given by "resume=" at which the swap header is located,
5656 in <PAGE_SIZE> units (needed only for swap files).
5657 See Documentation/power/swsusp-and-swap-files.rst
5659 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
5660 read the resume files
5662 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
5663 Useful for devices that are detected asynchronously
5664 (e.g. USB and MMC devices).
5666 retain_initrd [RAM] Keep initrd memory after extraction. After boot, it will
5667 be accessible via /sys/firmware/initrd.
5669 retbleed= [X86] Control mitigation of RETBleed (Arbitrary
5670 Speculative Code Execution with Return Instructions)
5673 AMD-based UNRET and IBPB mitigations alone do not stop
5674 sibling threads from influencing the predictions of other
5675 sibling threads. For that reason, STIBP is used on pro-
5676 cessors that support it, and mitigate SMT on processors
5680 auto - automatically select a migitation
5681 auto,nosmt - automatically select a mitigation,
5682 disabling SMT if necessary for
5683 the full mitigation (only on Zen1
5684 and older without STIBP).
5685 ibpb - On AMD, mitigate short speculation
5686 windows on basic block boundaries too.
5687 Safe, highest perf impact. It also
5688 enables STIBP if present. Not suitable
5690 ibpb,nosmt - Like "ibpb" above but will disable SMT
5691 when STIBP is not available. This is
5692 the alternative for systems which do not
5694 unret - Force enable untrained return thunks,
5695 only effective on AMD f15h-f17h based
5697 unret,nosmt - Like unret, but will disable SMT when STIBP
5698 is not available. This is the alternative for
5699 systems which do not have STIBP.
5701 Selecting 'auto' will choose a mitigation method at run
5702 time according to the CPU.
5704 Not specifying this option is equivalent to retbleed=auto.
5706 rfkill.default_state=
5707 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
5708 etc. communication is blocked by default.
5711 rfkill.master_switch_mode=
5712 0 The "airplane mode" button does nothing.
5713 1 The "airplane mode" button toggles between everything
5714 blocked and the previous configuration.
5715 2 The "airplane mode" button toggles between everything
5716 blocked and everything unblocked.
5718 rhash_entries= [KNL,NET]
5719 Set number of hash buckets for route cache
5722 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported
5725 riscv_isa_fallback [RISCV,EARLY]
5726 When CONFIG_RISCV_ISA_FALLBACK is not enabled, permit
5727 falling back to detecting extension support by parsing
5728 "riscv,isa" property on devicetree systems when the
5729 replacement properties are not found. See the Kconfig
5730 entry for RISCV_ISA_FALLBACK.
5732 ro [KNL] Mount root device read-only on boot
5735 on Mark read-only kernel memory as read-only (default).
5736 off Leave read-only kernel memory writable for debugging.
5737 full Mark read-only kernel memory and aliases as read-only
5742 Enable the uart passthrough on the designated usb port
5743 on Rockchip SoCs. When active, the signals of the
5744 debug-uart get routed to the D+ and D- pins of the usb
5745 port and the regular usb controller gets disabled.
5747 root= [KNL] Root filesystem
5748 Usually this a a block device specifier of some kind,
5749 see the early_lookup_bdev comment in
5750 block/early-lookup.c for details.
5751 Alternatively this can be "ram" for the legacy initial
5752 ramdisk, "nfs" and "cifs" for root on a network file
5753 system, or "mtd" and "ubi" for mounting from raw flash.
5755 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
5756 mount the root filesystem
5758 rootflags= [KNL] Set root filesystem mount option string
5760 rootfstype= [KNL] Set root filesystem type
5762 rootwait [KNL] Wait (indefinitely) for root device to show up.
5763 Useful for devices that are detected asynchronously
5764 (e.g. USB and MMC devices).
5766 rootwait= [KNL] Maximum time (in seconds) to wait for root device
5767 to show up before attempting to mount the root
5770 rproc_mem=nn[KMG][@address]
5771 [KNL,ARM,CMA] Remoteproc physical memory block.
5772 Memory area to be used by remote processor image,
5775 rw [KNL] Mount root device read-write on boot
5777 S [KNL] Run init in single mode
5779 s390_iommu= [HW,S390]
5780 Set s390 IOTLB flushing mode
5782 With strict flushing every unmap operation will result
5783 in an IOTLB flush. Default is lazy flushing before
5784 reuse, which is faster. Deprecated, equivalent to
5787 s390_iommu_aperture= [KNL,S390]
5788 Specifies the size of the per device DMA address space
5789 accessible through the DMA and IOMMU APIs as a decimal
5790 factor of the size of main memory.
5791 The default is 1 meaning that one can concurrently use
5792 as many DMA addresses as physical memory is installed,
5793 if supported by hardware, and thus map all of memory
5794 once. With a value of 2 one can map all of memory twice
5795 and so on. As a special case a factor of 0 imposes no
5796 restrictions other than those given by hardware at the
5797 cost of significant additional memory use for tables.
5800 See drivers/net/irda/sa1100_ir.c.
5802 sched_verbose [KNL,EARLY] Enables verbose scheduler debug messages.
5804 schedstats= [KNL,X86] Enable or disable scheduled statistics.
5805 Allowed values are enable and disable. This feature
5806 incurs a small amount of overhead in the scheduler
5807 but is useful for debugging and performance tuning.
5809 sched_thermal_decay_shift=
5810 [KNL, SMP] Set a decay shift for scheduler thermal
5811 pressure signal. Thermal pressure signal follows the
5812 default decay period of other scheduler pelt
5813 signals(usually 32 ms but configurable). Setting
5814 sched_thermal_decay_shift will left shift the decay
5815 period for the thermal pressure signal by the shift
5817 i.e. with the default pelt decay period of 32 ms
5818 sched_thermal_decay_shift thermal pressure decay pr
5822 Format: integer between 0 and 10
5825 scftorture.holdoff= [KNL]
5826 Number of seconds to hold off before starting
5827 test. Defaults to zero for module insertion and
5828 to 10 seconds for built-in smp_call_function()
5831 scftorture.longwait= [KNL]
5832 Request ridiculously long waits randomly selected
5833 up to the chosen limit in seconds. Zero (the
5834 default) disables this feature. Please note
5835 that requesting even small non-zero numbers of
5836 seconds can result in RCU CPU stall warnings,
5837 softlockup complaints, and so on.
5839 scftorture.nthreads= [KNL]
5840 Number of kthreads to spawn to invoke the
5841 smp_call_function() family of functions.
5842 The default of -1 specifies a number of kthreads
5843 equal to the number of CPUs.
5845 scftorture.onoff_holdoff= [KNL]
5846 Number seconds to wait after the start of the
5847 test before initiating CPU-hotplug operations.
5849 scftorture.onoff_interval= [KNL]
5850 Number seconds to wait between successive
5851 CPU-hotplug operations. Specifying zero (which
5852 is the default) disables CPU-hotplug operations.
5854 scftorture.shutdown_secs= [KNL]
5855 The number of seconds following the start of the
5856 test after which to shut down the system. The
5857 default of zero avoids shutting down the system.
5858 Non-zero values are useful for automated tests.
5860 scftorture.stat_interval= [KNL]
5861 The number of seconds between outputting the
5862 current test statistics to the console. A value
5863 of zero disables statistics output.
5865 scftorture.stutter_cpus= [KNL]
5866 The number of jiffies to wait between each change
5867 to the set of CPUs under test.
5869 scftorture.use_cpus_read_lock= [KNL]
5870 Use use_cpus_read_lock() instead of the default
5871 preempt_disable() to disable CPU hotplug
5872 while invoking one of the smp_call_function*()
5875 scftorture.verbose= [KNL]
5876 Enable additional printk() statements.
5878 scftorture.weight_single= [KNL]
5879 The probability weighting to use for the
5880 smp_call_function_single() function with a zero
5881 "wait" parameter. A value of -1 selects the
5882 default if all other weights are -1. However,
5883 if at least one weight has some other value, a
5884 value of -1 will instead select a weight of zero.
5886 scftorture.weight_single_wait= [KNL]
5887 The probability weighting to use for the
5888 smp_call_function_single() function with a
5889 non-zero "wait" parameter. See weight_single.
5891 scftorture.weight_many= [KNL]
5892 The probability weighting to use for the
5893 smp_call_function_many() function with a zero
5894 "wait" parameter. See weight_single.
5895 Note well that setting a high probability for
5896 this weighting can place serious IPI load
5899 scftorture.weight_many_wait= [KNL]
5900 The probability weighting to use for the
5901 smp_call_function_many() function with a
5902 non-zero "wait" parameter. See weight_single
5905 scftorture.weight_all= [KNL]
5906 The probability weighting to use for the
5907 smp_call_function_all() function with a zero
5908 "wait" parameter. See weight_single and
5911 scftorture.weight_all_wait= [KNL]
5912 The probability weighting to use for the
5913 smp_call_function_all() function with a
5914 non-zero "wait" parameter. See weight_single
5917 skew_tick= [KNL,EARLY] Offset the periodic timer tick per cpu to mitigate
5918 xtime_lock contention on larger systems, and/or RCU lock
5919 contention on all systems with CONFIG_MAXSMP set.
5920 Format: { "0" | "1" }
5921 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
5923 Note: increases power consumption, thus should only be
5924 enabled if running jitter sensitive (HPC/RT) workloads.
5926 security= [SECURITY] Choose a legacy "major" security module to
5927 enable at boot. This has been deprecated by the
5930 selinux= [SELINUX] Disable or enable SELinux at boot time.
5931 Format: { "0" | "1" }
5932 See security/selinux/Kconfig help text.
5937 serialnumber [BUGS=X86-32]
5939 sev=option[,option...] [X86-64] See Documentation/arch/x86/x86_64/boot-options.rst
5942 Maximal number of shapers.
5944 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
5945 Limit apic dumping. The parameter defines the maximal
5946 number of local apics being dumped. Also it is possible
5947 to set it to "all" by meaning -- no limit here.
5948 Format: { 1 (default) | 2 | ... | all }.
5949 The parameter valid if only apic=debug or
5950 apic=verbose is specified.
5951 Example: apic=debug show_lapic=all
5956 slab_debug[=options[,slabs][;[options[,slabs]]...] [MM]
5957 Enabling slab_debug allows one to determine the
5958 culprit if slab objects become corrupted. Enabling
5959 slab_debug can create guard zones around objects and
5960 may poison objects when not in use. Also tracks the
5961 last alloc / free. For more information see
5962 Documentation/mm/slub.rst.
5963 (slub_debug legacy name also accepted for now)
5965 slab_max_order= [MM]
5966 Determines the maximum allowed order for slabs.
5967 A high setting may cause OOMs due to memory
5968 fragmentation. For more information see
5969 Documentation/mm/slub.rst.
5970 (slub_max_order legacy name also accepted for now)
5973 Enable merging of slabs with similar size when the
5974 kernel is built without CONFIG_SLAB_MERGE_DEFAULT.
5975 (slub_merge legacy name also accepted for now)
5977 slab_min_objects= [MM]
5978 The minimum number of objects per slab. SLUB will
5979 increase the slab order up to slab_max_order to
5980 generate a sufficiently large slab able to contain
5981 the number of objects indicated. The higher the number
5982 of objects the smaller the overhead of tracking slabs
5983 and the less frequently locks need to be acquired.
5984 For more information see Documentation/mm/slub.rst.
5985 (slub_min_objects legacy name also accepted for now)
5987 slab_min_order= [MM]
5988 Determines the minimum page order for slabs. Must be
5989 lower or equal to slab_max_order. For more information see
5990 Documentation/mm/slub.rst.
5991 (slub_min_order legacy name also accepted for now)
5994 Disable merging of slabs with similar size. May be
5995 necessary if there is some reason to distinguish
5996 allocs to different slabs, especially in hardened
5997 environments where the risk of heap overflows and
5998 layout control by attackers can usually be
5999 frustrated by disabling merging. This will reduce
6000 most of the exposure of a heap attack to a single
6001 cache (risks via metadata attacks are mostly
6002 unchanged). Debug options disable merging on their
6004 For more information see Documentation/mm/slub.rst.
6005 (slub_nomerge legacy name also accepted for now)
6010 Format: <io1>[,<io2>[,...,<io8>]]
6012 smp.csd_lock_timeout= [KNL]
6013 Specify the period of time in milliseconds
6014 that smp_call_function() and friends will wait
6015 for a CPU to release the CSD lock. This is
6016 useful when diagnosing bugs involving CPUs
6017 disabling interrupts for extended periods
6018 of time. Defaults to 5,000 milliseconds, and
6019 setting a value of zero disables this feature.
6020 This feature may be more efficiently disabled
6021 using the csdlock_debug- kernel parameter.
6023 smp.panic_on_ipistall= [KNL]
6024 If a csd_lock_timeout extends for more than
6025 the specified number of milliseconds, panic the
6026 system. By default, let CSD-lock acquisition
6027 take as long as they take. Specifying 300,000
6028 for this value provides a 5-minute timeout.
6030 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
6031 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
6032 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
6033 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
6034 smsc-ircc2.ircc_irq= [HW] IRQ line
6035 smsc-ircc2.ircc_dma= [HW] DMA channel
6036 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
6037 0: Toshiba Satellite 1800 (GP data pin select)
6038 1: Fast pin select (default)
6041 smt= [KNL,MIPS,S390,EARLY] Set the maximum number of threads
6042 (logical CPUs) to use per physical CPU on systems
6043 capable of symmetric multithreading (SMT). Will
6044 be capped to the actual hardware limit.
6046 Default: -1 (no limit)
6049 [KNL] Should the soft-lockup detector generate panics.
6052 A value of 1 instructs the soft-lockup detector
6053 to panic the machine when a soft-lockup occurs. It is
6054 also controlled by the kernel.softlockup_panic sysctl
6055 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the
6056 respective build-time switch to that functionality.
6058 softlockup_all_cpu_backtrace=
6059 [KNL] Should the soft-lockup detector generate
6060 backtraces on all cpus.
6063 sonypi.*= [HW] Sony Programmable I/O Control Device driver
6064 See Documentation/admin-guide/laptops/sonypi.rst
6066 spectre_v2= [X86,EARLY] Control mitigation of Spectre variant 2
6067 (indirect branch speculation) vulnerability.
6068 The default operation protects the kernel from
6071 on - unconditionally enable, implies
6073 off - unconditionally disable, implies
6075 auto - kernel detects whether your CPU model is
6078 Selecting 'on' will, and 'auto' may, choose a
6079 mitigation method at run time according to the
6080 CPU, the available microcode, the setting of the
6081 CONFIG_MITIGATION_RETPOLINE configuration option,
6082 and the compiler with which the kernel was built.
6084 Selecting 'on' will also enable the mitigation
6085 against user space to user space task attacks.
6087 Selecting 'off' will disable both the kernel and
6088 the user space protections.
6090 Specific mitigations can also be selected manually:
6092 retpoline - replace indirect branches
6093 retpoline,generic - Retpolines
6094 retpoline,lfence - LFENCE; indirect branch
6095 retpoline,amd - alias for retpoline,lfence
6096 eibrs - Enhanced/Auto IBRS
6097 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines
6098 eibrs,lfence - Enhanced/Auto IBRS + LFENCE
6099 ibrs - use IBRS to protect kernel
6101 Not specifying this option is equivalent to
6105 [X86] Control mitigation of Spectre variant 2
6106 (indirect branch speculation) vulnerability between
6109 on - Unconditionally enable mitigations. Is
6110 enforced by spectre_v2=on
6112 off - Unconditionally disable mitigations. Is
6113 enforced by spectre_v2=off
6115 prctl - Indirect branch speculation is enabled,
6116 but mitigation can be enabled via prctl
6117 per thread. The mitigation control state
6118 is inherited on fork.
6121 - Like "prctl" above, but only STIBP is
6122 controlled per thread. IBPB is issued
6123 always when switching between different user
6127 - Same as "prctl" above, but all seccomp
6128 threads will enable the mitigation unless
6129 they explicitly opt out.
6132 - Like "seccomp" above, but only STIBP is
6133 controlled per thread. IBPB is issued
6134 always when switching between different
6135 user space processes.
6137 auto - Kernel selects the mitigation depending on
6138 the available CPU features and vulnerability.
6140 Default mitigation: "prctl"
6142 Not specifying this option is equivalent to
6143 spectre_v2_user=auto.
6145 spec_rstack_overflow=
6146 [X86,EARLY] Control RAS overflow mitigation on AMD Zen CPUs
6148 off - Disable mitigation
6149 microcode - Enable microcode mitigation only
6150 safe-ret - Enable sw-only safe RET mitigation (default)
6151 ibpb - Enable mitigation by issuing IBPB on
6153 ibpb-vmexit - Issue IBPB only on VMEXIT
6154 (cloud-specific mitigation)
6156 spec_store_bypass_disable=
6157 [HW,EARLY] Control Speculative Store Bypass (SSB) Disable mitigation
6158 (Speculative Store Bypass vulnerability)
6160 Certain CPUs are vulnerable to an exploit against a
6161 a common industry wide performance optimization known
6162 as "Speculative Store Bypass" in which recent stores
6163 to the same memory location may not be observed by
6164 later loads during speculative execution. The idea
6165 is that such stores are unlikely and that they can
6166 be detected prior to instruction retirement at the
6167 end of a particular speculation execution window.
6169 In vulnerable processors, the speculatively forwarded
6170 store can be used in a cache side channel attack, for
6171 example to read memory to which the attacker does not
6172 directly have access (e.g. inside sandboxed code).
6174 This parameter controls whether the Speculative Store
6175 Bypass optimization is used.
6177 On x86 the options are:
6179 on - Unconditionally disable Speculative Store Bypass
6180 off - Unconditionally enable Speculative Store Bypass
6181 auto - Kernel detects whether the CPU model contains an
6182 implementation of Speculative Store Bypass and
6183 picks the most appropriate mitigation. If the
6184 CPU is not vulnerable, "off" is selected. If the
6185 CPU is vulnerable the default mitigation is
6186 architecture and Kconfig dependent. See below.
6187 prctl - Control Speculative Store Bypass per thread
6188 via prctl. Speculative Store Bypass is enabled
6189 for a process by default. The state of the control
6190 is inherited on fork.
6191 seccomp - Same as "prctl" above, but all seccomp threads
6192 will disable SSB unless they explicitly opt out.
6194 Default mitigations:
6197 On powerpc the options are:
6199 on,auto - On Power8 and Power9 insert a store-forwarding
6200 barrier on kernel entry and exit. On Power7
6201 perform a software flush on kernel entry and
6205 Not specifying this option is equivalent to
6206 spec_store_bypass_disable=auto.
6208 spia_io_base= [HW,MTD]
6214 [X86] Enable split lock detection or bus lock detection
6216 When enabled (and if hardware support is present), atomic
6217 instructions that access data across cache line
6218 boundaries will result in an alignment check exception
6219 for split lock detection or a debug exception for
6224 warn - the kernel will emit rate-limited warnings
6225 about applications triggering the #AC
6226 exception or the #DB exception. This mode is
6227 the default on CPUs that support split lock
6228 detection or bus lock detection. Default
6229 behavior is by #AC if both features are
6230 enabled in hardware.
6232 fatal - the kernel will send SIGBUS to applications
6233 that trigger the #AC exception or the #DB
6234 exception. Default behavior is by #AC if
6235 both features are enabled in hardware.
6238 Set system wide rate limit to N bus locks
6239 per second for bus lock detection.
6242 N/A for split lock detection.
6245 If an #AC exception is hit in the kernel or in
6246 firmware (i.e. not while executing in user mode)
6247 the kernel will oops in either "warn" or "fatal"
6250 #DB exception for bus lock is triggered only when
6253 srbds= [X86,INTEL,EARLY]
6254 Control the Special Register Buffer Data Sampling
6257 Certain CPUs are vulnerable to an MDS-like
6258 exploit which can leak bits from the random
6261 By default, this issue is mitigated by
6262 microcode. However, the microcode fix can cause
6263 the RDRAND and RDSEED instructions to become
6264 much slower. Among other effects, this will
6265 result in reduced throughput from /dev/urandom.
6267 The microcode mitigation can be disabled with
6268 the following option:
6270 off: Disable mitigation and remove
6271 performance impact to RDRAND and RDSEED
6273 srcutree.big_cpu_lim [KNL]
6274 Specifies the number of CPUs constituting a
6275 large system, such that srcu_struct structures
6276 should immediately allocate an srcu_node array.
6277 This kernel-boot parameter defaults to 128,
6278 but takes effect only when the low-order four
6279 bits of srcutree.convert_to_big is equal to 3
6282 srcutree.convert_to_big [KNL]
6283 Specifies under what conditions an SRCU tree
6284 srcu_struct structure will be converted to big
6285 form, that is, with an rcu_node tree:
6288 1: At init_srcu_struct() time.
6289 2: When rcutorture decides to.
6290 3: Decide at boot time (default).
6291 0x1X: Above plus if high contention.
6293 Either way, the srcu_node tree will be sized based
6294 on the actual runtime number of CPUs (nr_cpu_ids)
6295 instead of the compile-time CONFIG_NR_CPUS.
6297 srcutree.counter_wrap_check [KNL]
6298 Specifies how frequently to check for
6299 grace-period sequence counter wrap for the
6300 srcu_data structure's ->srcu_gp_seq_needed field.
6301 The greater the number of bits set in this kernel
6302 parameter, the less frequently counter wrap will
6303 be checked for. Note that the bottom two bits
6306 srcutree.exp_holdoff [KNL]
6307 Specifies how many nanoseconds must elapse
6308 since the end of the last SRCU grace period for
6309 a given srcu_struct until the next normal SRCU
6310 grace period will be considered for automatic
6311 expediting. Set to zero to disable automatic
6314 srcutree.srcu_max_nodelay [KNL]
6315 Specifies the number of no-delay instances
6316 per jiffy for which the SRCU grace period
6317 worker thread will be rescheduled with zero
6318 delay. Beyond this limit, worker thread will
6319 be rescheduled with a sleep delay of one jiffy.
6321 srcutree.srcu_max_nodelay_phase [KNL]
6322 Specifies the per-grace-period phase, number of
6323 non-sleeping polls of readers. Beyond this limit,
6324 grace period worker thread will be rescheduled
6325 with a sleep delay of one jiffy, between each
6326 rescan of the readers, for a grace period phase.
6328 srcutree.srcu_retry_check_delay [KNL]
6329 Specifies number of microseconds of non-sleeping
6330 delay between each non-sleeping poll of readers.
6332 srcutree.small_contention_lim [KNL]
6333 Specifies the number of update-side contention
6334 events per jiffy will be tolerated before
6335 initiating a conversion of an srcu_struct
6336 structure to big form. Note that the value of
6337 srcutree.convert_to_big must have the 0x10 bit
6338 set for contention-based conversions to occur.
6340 ssbd= [ARM64,HW,EARLY]
6341 Speculative Store Bypass Disable control
6343 On CPUs that are vulnerable to the Speculative
6344 Store Bypass vulnerability and offer a
6345 firmware based mitigation, this parameter
6346 indicates how the mitigation should be used:
6348 force-on: Unconditionally enable mitigation for
6349 for both kernel and userspace
6350 force-off: Unconditionally disable mitigation for
6351 for both kernel and userspace
6352 kernel: Always enable mitigation in the
6353 kernel, and offer a prctl interface
6354 to allow userspace to register its
6355 interest in being mitigated too.
6357 stack_guard_gap= [MM]
6358 override the default stack gap protection. The value
6359 is in page units and it defines how many pages prior
6360 to (for stacks growing down) resp. after (for stacks
6361 growing up) the main stack are reserved for no other
6362 mapping. Default value is 256 pages.
6364 stack_depot_disable= [KNL,EARLY]
6365 Setting this to true through kernel command line will
6366 disable the stack depot thereby saving the static memory
6367 consumed by the stack hash table. By default this is set
6371 Enabled the stack tracer on boot up.
6373 stacktrace_filter=[function-list]
6374 [FTRACE] Limit the functions that the stack tracer
6375 will trace at boot up. function-list is a comma-separated
6376 list of functions. This list can be changed at run
6377 time by the stack_trace_filter file in the debugfs
6378 tracing directory. Note, this enables stack tracing
6379 and the stacktrace above is not needed.
6383 Set the STI (builtin display/keyboard on the HP-PARISC
6384 machines) console (graphic card) which should be used
6385 as the initial boot-console.
6386 See also comment in drivers/video/console/sticore.c.
6389 See comment in drivers/video/console/sticore.c.
6392 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
6397 Enable or disable strict sigaltstack size checks
6398 against the required signal frame size which
6399 depends on the supported FPU features. This can
6400 be used to filter out binaries which have
6401 not yet been made aware of AT_MINSIGSTKSZ.
6403 stress_hpt [PPC,EARLY]
6404 Limits the number of kernel HPT entries in the hash
6405 page table to increase the rate of hash page table
6406 faults on kernel addresses.
6408 stress_slb [PPC,EARLY]
6409 Limits the number of kernel SLB entries, and flushes
6410 them frequently to increase the rate of SLB faults
6411 on kernel addresses.
6413 sunrpc.min_resvport=
6414 sunrpc.max_resvport=
6416 SunRPC servers often require that client requests
6417 originate from a privileged port (i.e. a port in the
6418 range 0 < portnr < 1024).
6419 An administrator who wishes to reserve some of these
6420 ports for other uses may adjust the range that the
6421 kernel's sunrpc client considers to be privileged
6422 using these two parameters to set the minimum and
6423 maximum port values.
6425 sunrpc.svc_rpc_per_connection_limit=
6427 Limit the number of requests that the server will
6428 process in parallel from a single connection.
6429 The default value is 0 (no limit).
6433 Control how the NFS server code allocates CPUs to
6434 service thread pools. Depending on how many NICs
6435 you have and where their interrupts are bound, this
6436 option will affect which CPUs will do NFS serving.
6437 Note: this parameter cannot be changed while the
6438 NFS server is running.
6440 auto the server chooses an appropriate mode
6441 automatically using heuristics
6442 global a single global pool contains all CPUs
6443 percpu one pool for each CPU
6444 pernode one pool for each NUMA node (equivalent
6445 to global on non-NUMA machines)
6447 sunrpc.tcp_slot_table_entries=
6448 sunrpc.udp_slot_table_entries=
6450 Sets the upper limit on the number of simultaneous
6451 RPC calls that can be sent from the client to a
6452 server. Increasing these values may allow you to
6453 improve throughput, but will also increase the
6454 amount of memory reserved for use by the client.
6456 suspend.pm_test_delay=
6458 Sets the number of seconds to remain in a suspend test
6459 mode before resuming the system (see
6460 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
6461 is set. Default value is 5.
6464 Format: { on | off | y | n | 1 | 0 }
6465 This parameter controls use of the Protected
6466 Execution Facility on pSeries.
6468 swiotlb= [ARM,IA-64,PPC,MIPS,X86,EARLY]
6469 Format: { <int> [,<int>] | force | noforce }
6470 <int> -- Number of I/O TLB slabs
6471 <int> -- Second integer after comma. Number of swiotlb
6472 areas with their own lock. Will be rounded up
6474 force -- force using of bounce buffers even if they
6475 wouldn't be automatically used by the kernel
6476 noforce -- Never use bounce buffers (for debugging)
6478 switches= [HW,M68k,EARLY]
6481 Set a sysctl parameter, right before loading the init
6482 process, as if the value was written to the respective
6483 /proc/sys/... file. Both '.' and '/' are recognized as
6484 separators. Unrecognized parameters and invalid values
6485 are reported in the kernel log. Sysctls registered
6486 later by a loaded module cannot be set this way.
6487 Example: sysctl.vm.swappiness=40
6489 sysrq_always_enabled
6491 Ignore sysrq setting - this boot parameter will
6492 neutralize any effect of /proc/sys/kernel/sysrq.
6493 Useful for debugging.
6495 tcpmhash_entries= [KNL,NET]
6496 Set the number of tcp_metrics_hash slots.
6497 Default value is 8192 or 16384 depending on total
6498 ram pages. This is used to specify the TCP metrics
6499 cache size. See Documentation/networking/ip-sysctl.rst
6500 "tcp_no_metrics_save" section for more details.
6504 test_suspend= [SUSPEND]
6505 Format: { "mem" | "standby" | "freeze" }[,N]
6506 Specify "mem" (for Suspend-to-RAM) or "standby" (for
6507 standby suspend) or "freeze" (for suspend type freeze)
6508 as the system sleep state during system startup with
6509 the optional capability to repeat N number of times.
6510 The system is woken from this state using a
6511 wakeup-capable RTC alarm.
6513 thash_entries= [KNL,NET]
6514 Set number of hash buckets for TCP connection
6516 thermal.act= [HW,ACPI]
6517 -1: disable all active trip points in all thermal zones
6518 <degrees C>: override all lowest active trip points
6520 thermal.crt= [HW,ACPI]
6521 -1: disable all critical trip points in all thermal zones
6522 <degrees C>: override all critical trip points
6524 thermal.off= [HW,ACPI]
6525 1: disable ACPI thermal control
6527 thermal.psv= [HW,ACPI]
6528 -1: disable all passive trip points
6529 <degrees C>: override all passive trip points to this
6532 thermal.tzp= [HW,ACPI]
6533 Specify global default ACPI thermal zone polling rate
6534 <deci-seconds>: poll all this frequency
6535 0: no polling (default)
6537 threadirqs [KNL,EARLY]
6538 Force threading of all interrupt handlers except those
6539 marked explicitly IRQF_NO_THREAD.
6541 topology= [S390,EARLY]
6543 Specify if the kernel should make use of the cpu
6544 topology information if the hardware supports this.
6545 The scheduler will make use of this information and
6546 e.g. base its process migration decisions on it.
6549 topology_updates= [KNL, PPC, NUMA]
6551 Specify if the kernel should ignore (off)
6552 topology updates sent by the hypervisor to this
6555 torture.disable_onoff_at_boot= [KNL]
6556 Prevent the CPU-hotplug component of torturing
6557 until after init has spawned.
6559 torture.ftrace_dump_at_shutdown= [KNL]
6560 Dump the ftrace buffer at torture-test shutdown,
6561 even if there were no errors. This can be a
6562 very costly operation when many torture tests
6563 are running concurrently, especially on systems
6564 with rotating-rust storage.
6566 torture.verbose_sleep_frequency= [KNL]
6567 Specifies how many verbose printk()s should be
6568 emitted between each sleep. The default of zero
6569 disables verbose-printk() sleeping.
6571 torture.verbose_sleep_duration= [KNL]
6572 Duration of each verbose-printk() sleep in jiffies.
6576 tpm_suspend_pcr=[HW,TPM]
6577 Format: integer pcr id
6578 Specify that at suspend time, the tpm driver
6579 should extend the specified pcr with zeros,
6580 as a workaround for some chips which fail to
6581 flush the last written pcr on TPM_SaveState.
6582 This will guarantee that all the other pcrs
6585 tpm_tis.interrupts= [HW,TPM]
6586 Enable interrupts for the MMIO based physical layer
6587 for the FIFO interface. By default it is set to false
6588 (0). For more information about TPM hardware interfaces
6589 defined by Trusted Computing Group (TCG) see
6590 https://trustedcomputinggroup.org/resource/pc-client-platform-tpm-profile-ptp-specification/
6593 Have the tracepoints sent to printk as well as the
6594 tracing ring buffer. This is useful for early boot up
6595 where the system hangs or reboots and does not give the
6596 option for reading the tracing buffer or performing a
6597 ftrace_dump_on_oops.
6599 To turn off having tracepoints sent to printk,
6600 echo 0 > /proc/sys/kernel/tracepoint_printk
6601 Note, echoing 1 into this file without the
6602 tracepoint_printk kernel cmdline option has no effect.
6604 The tp_printk_stop_on_boot (see below) can also be used
6605 to stop the printing of events to console at
6610 Having tracepoints sent to printk() and activating high
6611 frequency tracepoints such as irq or sched, can cause
6612 the system to live lock.
6614 tp_printk_stop_on_boot [FTRACE]
6615 When tp_printk (above) is set, it can cause a lot of noise
6616 on the console. It may be useful to only include the
6617 printing of events during boot up, as user space may
6618 make the system inoperable.
6620 This command line option will stop the printing of events
6621 to console at the late_initcall_sync() time frame.
6623 trace_buf_size=nn[KMG]
6624 [FTRACE] will set tracing buffer size on each cpu.
6626 trace_clock= [FTRACE] Set the clock used for tracing events
6628 local - Use the per CPU time stamp counter
6629 (converted into nanoseconds). Fast, but
6630 depending on the architecture, may not be
6631 in sync between CPUs.
6632 global - Event time stamps are synchronize across
6633 CPUs. May be slower than the local clock,
6634 but better for some race conditions.
6635 counter - Simple counting of events (1, 2, ..)
6636 note, some counts may be skipped due to the
6637 infrastructure grabbing the clock more than
6639 uptime - Use jiffies as the time stamp.
6640 perf - Use the same clock that perf uses.
6641 mono - Use ktime_get_mono_fast_ns() for time stamps.
6642 mono_raw - Use ktime_get_raw_fast_ns() for time
6644 boot - Use ktime_get_boot_fast_ns() for time stamps.
6645 Architectures may add more clocks. See
6646 Documentation/trace/ftrace.rst for more details.
6648 trace_event=[event-list]
6649 [FTRACE] Set and start specified trace events in order
6650 to facilitate early boot debugging. The event-list is a
6651 comma-separated list of trace events to enable. See
6652 also Documentation/trace/events.rst
6654 trace_instance=[instance-info]
6655 [FTRACE] Create a ring buffer instance early in boot up.
6656 This will be listed in:
6658 /sys/kernel/tracing/instances
6660 Events can be enabled at the time the instance is created
6663 trace_instance=<name>,<system1>:<event1>,<system2>:<event2>
6665 Note, the "<system*>:" portion is optional if the event is
6668 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall
6670 will enable the "sched_switch" event (note, the "sched:" is optional, and
6671 the same thing would happen if it was left off). The irq_handler_entry
6672 event, and all events under the "initcall" system.
6674 trace_options=[option-list]
6675 [FTRACE] Enable or disable tracer options at boot.
6676 The option-list is a comma delimited list of options
6677 that can be enabled or disabled just as if you were
6678 to echo the option name into
6680 /sys/kernel/tracing/trace_options
6682 For example, to enable stacktrace option (to dump the
6683 stack trace of each event), add to the command line:
6685 trace_options=stacktrace
6687 See also Documentation/trace/ftrace.rst "trace options"
6690 trace_trigger=[trigger-list]
6691 [FTRACE] Add a event trigger on specific events.
6692 Set a trigger on top of a specific event, with an optional
6695 The format is is "trace_trigger=<event>.<trigger>[ if <filter>],..."
6696 Where more than one trigger may be specified that are comma deliminated.
6700 trace_trigger="sched_switch.stacktrace if prev_state == 2"
6702 The above will enable the "stacktrace" trigger on the "sched_switch"
6703 event but only trigger it if the "prev_state" of the "sched_switch"
6704 event is "2" (TASK_UNINTERUPTIBLE).
6706 See also "Event triggers" in Documentation/trace/events.rst
6710 [FTRACE] enable this option to disable tracing when a
6711 warning is hit. This turns off "tracing_on". Tracing can
6712 be enabled again by echoing '1' into the "tracing_on"
6713 file located in /sys/kernel/tracing/
6715 This option is useful, as it disables the trace before
6716 the WARNING dump is called, which prevents the trace to
6717 be filled with content caused by the warning output.
6719 This option can also be set at run time via the sysctl
6720 option: kernel/traceoff_on_warning
6722 transparent_hugepage=
6724 Format: [always|madvise|never]
6725 Can be used to control the default behavior of the system
6726 with respect to transparent hugepages.
6727 See Documentation/admin-guide/mm/transhuge.rst
6730 trusted.source= [KEYS]
6732 This parameter identifies the trust source as a backend
6733 for trusted keys implementation. Supported trust
6738 If not specified then it defaults to iterating through
6739 the trust source list starting with TPM and assigns the
6740 first trust source as a backend which is initialized
6741 successfully during iteration.
6745 The RNG used to generate key material for trusted keys.
6748 - the same value as trusted.source: "tpm" or "tee"
6750 If not specified, "default" is used. In this case,
6751 the RNG's choice is left to each individual trust source.
6753 tsc= Disable clocksource stability checks for TSC.
6755 [x86] reliable: mark tsc clocksource as reliable, this
6756 disables clocksource verification at runtime, as well
6757 as the stability checks done at bootup. Used to enable
6758 high-resolution timer mode on older hardware, and in
6759 virtualized environment.
6760 [x86] noirqtime: Do not use TSC to do irq accounting.
6761 Used to run time disable IRQ_TIME_ACCOUNTING on any
6762 platforms where RDTSC is slow and this accounting
6764 [x86] unstable: mark the TSC clocksource as unstable, this
6765 marks the TSC unconditionally unstable at bootup and
6766 avoids any further wobbles once the TSC watchdog notices.
6767 [x86] nowatchdog: disable clocksource watchdog. Used
6768 in situations with strict latency requirements (where
6769 interruptions from clocksource watchdog are not
6771 [x86] recalibrate: force recalibration against a HW timer
6772 (HPET or PM timer) on systems whose TSC frequency was
6773 obtained from HW or FW using either an MSR or CPUID(0x15).
6774 Warn if the difference is more than 500 ppm.
6775 [x86] watchdog: Use TSC as the watchdog clocksource with
6776 which to check other HW timers (HPET or PM timer), but
6777 only on systems where TSC has been deemed trustworthy.
6778 This will be suppressed by an earlier tsc=nowatchdog and
6779 can be overridden by a later tsc=nowatchdog. A console
6780 message will flag any such suppression or overriding.
6782 tsc_early_khz= [X86,EARLY] Skip early TSC calibration and use the given
6783 value instead. Useful when the early TSC frequency discovery
6784 procedure is not reliable, such as on overclocked systems
6785 with CPUID.16h support and partial CPUID.15h support.
6786 Format: <unsigned int>
6788 tsx= [X86] Control Transactional Synchronization
6789 Extensions (TSX) feature in Intel processors that
6790 support TSX control.
6792 This parameter controls the TSX feature. The options are:
6794 on - Enable TSX on the system. Although there are
6795 mitigations for all known security vulnerabilities,
6796 TSX has been known to be an accelerator for
6797 several previous speculation-related CVEs, and
6798 so there may be unknown security risks associated
6799 with leaving it enabled.
6801 off - Disable TSX on the system. (Note that this
6802 option takes effect only on newer CPUs which are
6803 not vulnerable to MDS, i.e., have
6804 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get
6805 the new IA32_TSX_CTRL MSR through a microcode
6806 update. This new MSR allows for the reliable
6807 deactivation of the TSX functionality.)
6809 auto - Disable TSX if X86_BUG_TAA is present,
6810 otherwise enable TSX on the system.
6812 Not specifying this option is equivalent to tsx=off.
6814 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
6817 tsx_async_abort= [X86,INTEL,EARLY] Control mitigation for the TSX Async
6818 Abort (TAA) vulnerability.
6820 Similar to Micro-architectural Data Sampling (MDS)
6821 certain CPUs that support Transactional
6822 Synchronization Extensions (TSX) are vulnerable to an
6823 exploit against CPU internal buffers which can forward
6824 information to a disclosure gadget under certain
6827 In vulnerable processors, the speculatively forwarded
6828 data can be used in a cache side channel attack, to
6829 access data to which the attacker does not have direct
6832 This parameter controls the TAA mitigation. The
6835 full - Enable TAA mitigation on vulnerable CPUs
6838 full,nosmt - Enable TAA mitigation and disable SMT on
6839 vulnerable CPUs. If TSX is disabled, SMT
6840 is not disabled because CPU is not
6841 vulnerable to cross-thread TAA attacks.
6842 off - Unconditionally disable TAA mitigation
6844 On MDS-affected machines, tsx_async_abort=off can be
6845 prevented by an active MDS mitigation as both vulnerabilities
6846 are mitigated with the same mechanism so in order to disable
6847 this mitigation, you need to specify mds=off too.
6849 Not specifying this option is equivalent to
6850 tsx_async_abort=full. On CPUs which are MDS affected
6851 and deploy MDS mitigation, TAA mitigation is not
6852 required and doesn't provide any additional
6856 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
6858 turbografx.map[2|3]= [HW,JOY]
6859 TurboGraFX parallel port interface
6861 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
6862 See also Documentation/input/devices/joystick-parport.rst
6864 udbg-immortal [PPC] When debugging early kernel crashes that
6865 happen after console_init() and before a proper
6866 console driver takes over, this boot options might
6867 help "seeing" what's going on.
6869 uhash_entries= [KNL,NET]
6870 Set number of hash buckets for UDP/UDP-Lite connections
6873 [USB] Ignore overcurrent events (default N).
6874 Some badly-designed motherboards generate lots of
6875 bogus events, for ports that aren't wired to
6876 anything. Set this parameter to avoid log spamming.
6877 Note that genuine overcurrent events won't be
6881 [X86] Cause panic on unknown NMI.
6883 unwind_debug [X86-64,EARLY]
6884 Enable unwinder debug output. This can be
6885 useful for debugging certain unwinder error
6886 conditions, including corrupt stacks and
6887 bad/missing unwinder metadata.
6889 usbcore.authorized_default=
6890 [USB] Default USB device authorization:
6891 (default -1 = authorized (same as 1),
6892 0 = not authorized, 1 = authorized, 2 = authorized
6893 if device connected to internal port)
6895 usbcore.autosuspend=
6896 [USB] The autosuspend time delay (in seconds) used
6897 for newly-detected USB devices (default 2). This
6898 is the time required before an idle device will be
6899 autosuspended. Devices for which the delay is set
6900 to a negative value won't be autosuspended at all.
6902 usbcore.usbfs_snoop=
6903 [USB] Set to log all usbfs traffic (default 0 = off).
6905 usbcore.usbfs_snoop_max=
6906 [USB] Maximum number of bytes to snoop in each URB
6909 usbcore.blinkenlights=
6910 [USB] Set to cycle leds on hubs (default 0 = off).
6912 usbcore.old_scheme_first=
6913 [USB] Start with the old device initialization
6914 scheme (default 0 = off).
6916 usbcore.usbfs_memory_mb=
6917 [USB] Memory limit (in MB) for buffers allocated by
6918 usbfs (default = 16, 0 = max = 2047).
6920 usbcore.use_both_schemes=
6921 [USB] Try the other device initialization scheme
6922 if the first one fails (default 1 = enabled).
6924 usbcore.initial_descriptor_timeout=
6925 [USB] Specifies timeout for the initial 64-byte
6926 USB_REQ_GET_DESCRIPTOR request in milliseconds
6927 (default 5000 = 5.0 seconds).
6929 usbcore.nousb [USB] Disable the USB subsystem
6932 [USB] A list of quirk entries to augment the built-in
6933 usb core quirk list. List entries are separated by
6934 commas. Each entry has the form
6935 VendorID:ProductID:Flags. The IDs are 4-digit hex
6936 numbers and Flags is a set of letters. Each letter
6937 will change the built-in quirk; setting it if it is
6938 clear and clearing it if it is set. The letters have
6939 the following meanings:
6940 a = USB_QUIRK_STRING_FETCH_255 (string
6941 descriptors must not be fetched using
6943 b = USB_QUIRK_RESET_RESUME (device can't resume
6944 correctly so reset it instead);
6945 c = USB_QUIRK_NO_SET_INTF (device can't handle
6946 Set-Interface requests);
6947 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't
6948 handle its Configuration or Interface
6950 e = USB_QUIRK_RESET (device can't be reset
6951 (e.g morph devices), don't use reset);
6952 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has
6953 more interface descriptions than the
6954 bNumInterfaces count, and can't handle
6955 talking to these interfaces);
6956 g = USB_QUIRK_DELAY_INIT (device needs a pause
6957 during initialization, after we read
6958 the device descriptor);
6959 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For
6960 high speed and super speed interrupt
6961 endpoints, the USB 2.0 and USB 3.0 spec
6962 require the interval in microframes (1
6963 microframe = 125 microseconds) to be
6964 calculated as interval = 2 ^
6966 Devices with this quirk report their
6967 bInterval as the result of this
6968 calculation instead of the exponent
6969 variable used in the calculation);
6970 i = USB_QUIRK_DEVICE_QUALIFIER (device can't
6971 handle device_qualifier descriptor
6973 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device
6974 generates spurious wakeup, ignore
6975 remote wakeup capability);
6976 k = USB_QUIRK_NO_LPM (device can't handle Link
6978 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL
6979 (Device reports its bInterval as linear
6980 frames instead of the USB 2.0
6982 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs
6983 to be disconnected before suspend to
6984 prevent spurious wakeup);
6985 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a
6986 pause after every control message);
6987 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra
6988 delay after resetting its port);
6989 p = USB_QUIRK_SHORT_SET_ADDRESS_REQ_TIMEOUT
6990 (Reduce timeout of the SET_ADDRESS
6991 request from 5000 ms to 500 ms);
6992 Example: quirks=0781:5580:bk,0a5c:5834:gij
6995 [USBHID] The interval which mice are to be polled at.
6998 [USBHID] The interval which joysticks are to be polled at.
7001 [USBHID] The interval which keyboards are to be polled at.
7003 usb-storage.delay_use=
7004 [UMS] The delay in seconds before a new device is
7005 scanned for Logical Units (default 1).
7008 [UMS] A list of quirks entries to supplement or
7009 override the built-in unusual_devs list. List
7010 entries are separated by commas. Each entry has
7011 the form VID:PID:Flags where VID and PID are Vendor
7012 and Product ID values (4-digit hex numbers) and
7013 Flags is a set of characters, each corresponding
7014 to a common usb-storage quirk flag as follows:
7015 a = SANE_SENSE (collect more than 18 bytes
7016 of sense data, not on uas);
7017 b = BAD_SENSE (don't collect more than 18
7018 bytes of sense data, not on uas);
7019 c = FIX_CAPACITY (decrease the reported
7020 device capacity by one sector);
7021 d = NO_READ_DISC_INFO (don't use
7022 READ_DISC_INFO command, not on uas);
7023 e = NO_READ_CAPACITY_16 (don't use
7024 READ_CAPACITY_16 command);
7025 f = NO_REPORT_OPCODES (don't use report opcodes
7027 g = MAX_SECTORS_240 (don't transfer more than
7028 240 sectors at a time, uas only);
7029 h = CAPACITY_HEURISTICS (decrease the
7030 reported device capacity by one
7031 sector if the number is odd);
7032 i = IGNORE_DEVICE (don't bind to this
7034 j = NO_REPORT_LUNS (don't use report luns
7036 k = NO_SAME (do not use WRITE_SAME, uas only)
7037 l = NOT_LOCKABLE (don't try to lock and
7038 unlock ejectable media, not on uas);
7039 m = MAX_SECTORS_64 (don't transfer more
7040 than 64 sectors = 32 KB at a time,
7042 n = INITIAL_READ10 (force a retry of the
7043 initial READ(10) command, not on uas);
7044 o = CAPACITY_OK (accept the capacity
7045 reported by the device, not on uas);
7046 p = WRITE_CACHE (the device cache is ON
7047 by default, not on uas);
7048 r = IGNORE_RESIDUE (the device reports
7049 bogus residue values, not on uas);
7050 s = SINGLE_LUN (the device has only one
7052 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
7053 commands, uas only);
7054 u = IGNORE_UAS (don't bind to the uas driver);
7055 w = NO_WP_DETECT (don't test whether the
7056 medium is write-protected).
7057 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE
7058 even if the device claims no cache,
7060 Example: quirks=0419:aaf5:rl,0421:0433:rc
7062 user_debug= [KNL,ARM]
7064 See arch/arm/Kconfig.debug help text.
7065 1 - undefined instruction events
7067 4 - invalid data aborts
7070 Example: user_debug=31
7073 [X86,EARLY] Flags controlling user PTE allocations.
7075 nohigh = do not allocate PTE pages in
7076 HIGHMEM regardless of setting
7079 vdso= [X86,SH,SPARC]
7080 On X86_32, this is an alias for vdso32=. Otherwise:
7082 vdso=1: enable VDSO (the default)
7083 vdso=0: disable VDSO mapping
7085 vdso32= [X86] Control the 32-bit vDSO
7086 vdso32=1: enable 32-bit VDSO
7087 vdso32=0 or vdso32=2: disable 32-bit VDSO
7089 See the help text for CONFIG_COMPAT_VDSO for more
7090 details. If CONFIG_COMPAT_VDSO is set, the default is
7091 vdso32=0; otherwise, the default is vdso32=1.
7093 For compatibility with older kernels, vdso32=2 is an
7096 Try vdso32=0 if you encounter an error that says:
7097 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
7100 vector=percpu: enable percpu vector domain
7102 video= [FB,EARLY] Frame buffer configuration
7103 See Documentation/fb/modedb.rst.
7105 video.brightness_switch_enabled= [ACPI]
7107 If set to 1, on receiving an ACPI notify event
7108 generated by hotkey, video driver will adjust brightness
7109 level and then send out the event to user space through
7110 the allocated input device. If set to 0, video driver
7111 will only send out the event without touching backlight
7116 [VMMIO] Memory mapped virtio (platform) device.
7118 <size>@<baseaddr>:<irq>[:<id>]
7120 <size> := size (can use standard suffixes
7122 <baseaddr> := physical base address
7123 <irq> := interrupt number (as passed to
7125 <id> := (optional) platform device id
7127 virtio_mmio.device=1K@0x100b0000:48:7
7129 Can be used multiple times for multiple devices.
7131 vga= [BOOT,X86-32] Select a particular video mode
7132 See Documentation/arch/x86/boot.rst and
7133 Documentation/admin-guide/svga.rst.
7134 Use vga=ask for menu.
7135 This is actually a boot loader parameter; the value is
7136 passed to the kernel using a special protocol.
7138 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y.
7139 May slow down system boot speed, especially when
7140 enabled on systems with a large amount of memory.
7141 All options are enabled by default, and this
7142 interface is meant to allow for selectively
7143 enabling or disabling specific virtual memory
7146 Available options are:
7147 P Enable page structure init time poisoning
7148 - Disable all of the above options
7150 vmalloc=nn[KMG] [KNL,BOOT,EARLY] Forces the vmalloc area to have an
7151 exact size of <nn>. This can be used to increase
7152 the minimum size (128MB on x86). It can also be
7153 used to decrease the size and leave more room
7154 for directly mapped kernel RAM.
7156 vmcp_cma=nn[MG] [KNL,S390,EARLY]
7157 Sets the memory size reserved for contiguous memory
7158 allocations for the vmcp device driver.
7160 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
7163 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
7166 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
7169 vsyscall= [X86-64,EARLY]
7170 Controls the behavior of vsyscalls (i.e. calls to
7171 fixed addresses of 0xffffffffff600x00 from legacy
7172 code). Most statically-linked binaries and older
7173 versions of glibc use these calls. Because these
7174 functions are at fixed addresses, they make nice
7175 targets for exploits that can control RIP.
7177 emulate Vsyscalls turn into traps and are emulated
7178 reasonably safely. The vsyscall page is
7181 xonly [default] Vsyscalls turn into traps and are
7182 emulated reasonably safely. The vsyscall
7183 page is not readable.
7185 none Vsyscalls don't work at all. This makes
7186 them quite hard to use for exploits but
7187 might break your system.
7189 vt.color= [VT] Default text color.
7190 Format: 0xYX, X = foreground, Y = background.
7191 Default: 0x07 = light gray on black.
7193 vt.cur_default= [VT] Default cursor shape.
7194 Format: 0xCCBBAA, where AA, BB, and CC are the same as
7195 the parameters of the <Esc>[?A;B;Cc escape sequence;
7196 see VGA-softcursor.txt. Default: 2 = underline.
7198 vt.default_blu= [VT]
7199 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
7200 Change the default blue palette of the console.
7201 This is a 16-member array composed of values
7204 vt.default_grn= [VT]
7205 Format: <green0>,<green1>,<green2>,...,<green15>
7206 Change the default green palette of the console.
7207 This is a 16-member array composed of values
7210 vt.default_red= [VT]
7211 Format: <red0>,<red1>,<red2>,...,<red15>
7212 Change the default red palette of the console.
7213 This is a 16-member array composed of values
7219 Set system-wide default UTF-8 mode for all tty's.
7220 Default is 1, i.e. UTF-8 mode is enabled for all
7221 newly opened terminals.
7223 vt.global_cursor_default=
7226 Set system-wide default for whether a cursor
7227 is shown on new VTs. Default is -1,
7228 i.e. cursors will be created by default unless
7229 overridden by individual drivers. 0 will hide
7230 cursors, 1 will display them.
7232 vt.italic= [VT] Default color for italic text; 0-15.
7235 vt.underline= [VT] Default color for underlined text; 0-15.
7238 watchdog timers [HW,WDT] For information on watchdog timers,
7239 see Documentation/watchdog/watchdog-parameters.rst
7240 or other driver-specific files in the
7241 Documentation/watchdog/ directory.
7245 Set the hard lockup detector stall duration
7246 threshold in seconds. The soft lockup detector
7247 threshold is set to twice the value. A value of 0
7248 disables both lockup detectors. Default is 10
7251 workqueue.unbound_cpus=
7252 [KNL,SMP] Specify to constrain one or some CPUs
7253 to use in unbound workqueues.
7255 By default, all online CPUs are available for
7258 workqueue.watchdog_thresh=
7259 If CONFIG_WQ_WATCHDOG is configured, workqueue can
7260 warn stall conditions and dump internal state to
7261 help debugging. 0 disables workqueue stall
7262 detection; otherwise, it's the stall threshold
7263 duration in seconds. The default value is 30 and
7264 it can be updated at runtime by writing to the
7265 corresponding sysfs file.
7267 workqueue.cpu_intensive_thresh_us=
7268 Per-cpu work items which run for longer than this
7269 threshold are automatically considered CPU intensive
7270 and excluded from concurrency management to prevent
7271 them from noticeably delaying other per-cpu work
7272 items. Default is 10000 (10ms).
7274 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel
7275 will report the work functions which violate this
7276 threshold repeatedly. They are likely good
7277 candidates for using WQ_UNBOUND workqueues instead.
7279 workqueue.cpu_intensive_warning_thresh=<uint>
7280 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel
7281 will report the work functions which violate the
7282 intensive_threshold_us repeatedly. In order to prevent
7283 spurious warnings, start printing only after a work
7284 function has violated this threshold number of times.
7286 The default is 4 times. 0 disables the warning.
7288 workqueue.power_efficient
7289 Per-cpu workqueues are generally preferred because
7290 they show better performance thanks to cache
7291 locality; unfortunately, per-cpu workqueues tend to
7292 be more power hungry than unbound workqueues.
7294 Enabling this makes the per-cpu workqueues which
7295 were observed to contribute significantly to power
7296 consumption unbound, leading to measurably lower
7297 power usage at the cost of small performance
7300 The default value of this parameter is determined by
7301 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
7303 workqueue.default_affinity_scope=
7304 Select the default affinity scope to use for unbound
7305 workqueues. Can be one of "cpu", "smt", "cache",
7306 "numa" and "system". Default is "cache". For more
7307 information, see the Affinity Scopes section in
7308 Documentation/core-api/workqueue.rst.
7310 This can be changed after boot by writing to the
7311 matching /sys/module/workqueue/parameters file. All
7312 workqueues with the "default" affinity scope will be
7313 updated accordignly.
7315 workqueue.debug_force_rr_cpu
7316 Workqueue used to implicitly guarantee that work
7317 items queued without explicit CPU specified are put
7318 on the local CPU. This guarantee is no longer true
7319 and while local CPU is still preferred work items
7320 may be put on foreign CPUs. This debug option
7321 forces round-robin CPU selection to flush out
7322 usages which depend on the now broken guarantee.
7323 When enabled, memory and cache locality will be
7326 writecombine= [LOONGARCH,EARLY] Control the MAT (Memory Access
7327 Type) of ioremap_wc().
7329 on - Enable writecombine, use WUC for ioremap_wc()
7330 off - Disable writecombine, use SUC for ioremap_wc()
7332 x2apic_phys [X86-64,APIC,EARLY] Use x2apic physical mode instead of
7333 default x2apic cluster mode on platforms
7336 xen_512gb_limit [KNL,X86-64,XEN]
7337 Restricts the kernel running paravirtualized under Xen
7338 to use only up to 512 GB of RAM. The reason to do so is
7339 crash analysis tools and Xen tools for doing domain
7340 save/restore/migration must be enabled to handle larger
7343 xen_emul_unplug= [HW,X86,XEN,EARLY]
7344 Unplug Xen emulated devices
7345 Format: [unplug0,][unplug1]
7346 ide-disks -- unplug primary master IDE devices
7347 aux-ide-disks -- unplug non-primary-master IDE devices
7348 nics -- unplug network devices
7349 all -- unplug all emulated devices (NICs and IDE disks)
7350 unnecessary -- unplugging emulated devices is
7351 unnecessary even if the host did not respond to
7353 never -- do not unplug even if version check succeeds
7355 xen_legacy_crash [X86,XEN,EARLY]
7356 Crash from Xen panic notifier, without executing late
7357 panic() code such as dumping handler.
7359 xen_msr_safe= [X86,XEN,EARLY]
7361 Select whether to always use non-faulting (safe) MSR
7362 access functions when running as Xen PV guest. The
7363 default value is controlled by CONFIG_XEN_PV_MSR_SAFE.
7365 xen_nopvspin [X86,XEN,EARLY]
7366 Disables the qspinlock slowpath using Xen PV optimizations.
7367 This parameter is obsoleted by "nopvspin" parameter, which
7368 has equivalent effect for XEN platform.
7371 Disables the PV optimizations forcing the HVM guest to
7372 run as generic HVM guest with no PV drivers.
7373 This option is obsoleted by the "nopv" option, which
7374 has equivalent effect for XEN platform.
7376 xen_no_vector_callback
7377 [KNL,X86,XEN,EARLY] Disable the vector callback for Xen
7378 event channel interrupts.
7380 xen_scrub_pages= [XEN]
7381 Boolean option to control scrubbing pages before giving them back
7382 to Xen, for use by other domains. Can be also changed at runtime
7383 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages.
7384 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT.
7386 xen_timer_slop= [X86-64,XEN,EARLY]
7387 Set the timer slop (in nanoseconds) for the virtual Xen
7388 timers (default is 100000). This adjusts the minimum
7389 delta of virtualized Xen timers, where lower values
7390 improve timer resolution at the expense of processing
7391 more timer interrupts.
7393 xen.balloon_boot_timeout= [XEN]
7394 The time (in seconds) to wait before giving up to boot
7395 in case initial ballooning fails to free enough memory.
7396 Applies only when running as HVM or PVH guest and
7397 started with less memory configured than allowed at
7398 max. Default is 180.
7400 xen.event_eoi_delay= [XEN]
7401 How long to delay EOI handling in case of event
7402 storms (jiffies). Default is 10.
7404 xen.event_loop_timeout= [XEN]
7405 After which time (jiffies) the event handling loop
7406 should start to delay EOI handling. Default is 2.
7408 xen.fifo_events= [XEN]
7409 Boolean parameter to disable using fifo event handling
7410 even if available. Normally fifo event handling is
7411 preferred over the 2-level event handling, as it is
7412 fairer and the number of possible event channels is
7413 much higher. Default is on (use fifo events).
7415 xirc2ps_cs= [NET,PCMCIA]
7417 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
7420 By default on POWER9 and above, the kernel will
7421 natively use the XIVE interrupt controller. This option
7422 allows the fallback firmware mode to be used:
7424 off Fallback to firmware control of XIVE interrupt
7425 controller on both pseries and powernv
7426 platforms. Only useful on POWER9 and above.
7428 xive.store-eoi=off [PPC]
7429 By default on POWER10 and above, the kernel will use
7430 stores for EOI handling when the XIVE interrupt mode
7431 is active. This option allows the XIVE driver to use
7432 loads instead, as on POWER9.
7434 xhci-hcd.quirks [USB,KNL]
7435 A hex value specifying bitmask with supplemental xhci
7436 host controller quirks. Meaning of each bit can be
7437 consulted in header drivers/usb/host/xhci.h.
7440 Format: { early | on | rw | ro | off }
7441 Controls if xmon debugger is enabled. Default is off.
7442 Passing only "xmon" is equivalent to "xmon=early".
7443 early Call xmon as early as possible on boot; xmon
7444 debugger is called from setup_arch().
7445 on xmon debugger hooks will be installed so xmon
7446 is only called on a kernel crash. Default mode,
7447 i.e. either "ro" or "rw" mode, is controlled
7448 with CONFIG_XMON_DEFAULT_RO_MODE.
7449 rw xmon debugger hooks will be installed so xmon
7450 is called only on a kernel crash, mode is write,
7451 meaning SPR registers, memory and, other data
7452 can be written using xmon commands.
7453 ro same as "rw" option above but SPR registers,
7454 memory, and other data can't be written using
7456 off xmon is disabled.