1 acpi= [HW,ACPI,X86,ARM64,RISCV64]
2 Advanced Configuration and Power Interface
3 Format: { force | on | off | strict | noirq | rsdt |
5 force -- enable ACPI if default was off
6 on -- enable ACPI but allow fallback to DT [arm64,riscv64]
7 off -- disable ACPI if default was on
8 noirq -- do not use ACPI for IRQ routing
9 strict -- Be less tolerant of platforms that are not
10 strictly ACPI specification compliant.
11 rsdt -- prefer RSDT over (default) XSDT
12 copy_dsdt -- copy DSDT to memory
13 For ARM64 and RISCV64, ONLY "acpi=off", "acpi=on" or
14 "acpi=force" are available
16 See also Documentation/power/runtime_pm.rst, pci=noacpi
18 acpi_apic_instance= [ACPI, IOAPIC]
20 2: use 2nd APIC table, if available
21 1,0: use 1st APIC table
24 acpi_backlight= [HW,ACPI]
25 { vendor | video | native | none }
26 If set to vendor, prefer vendor-specific driver
27 (e.g. thinkpad_acpi, sony_acpi, etc.) instead
28 of the ACPI video.ko driver.
29 If set to video, use the ACPI video.ko driver.
30 If set to native, use the device's native backlight mode.
31 If set to none, disable the ACPI backlight interface.
33 acpi_force_32bit_fadt_addr
34 force FADT to use 32 bit addresses rather than the
35 64 bit X_* addresses. Some firmware have broken 64
36 bit addresses for force ACPI ignore these and use
37 the older legacy 32 bit addresses.
39 acpica_no_return_repair [HW, ACPI]
40 Disable AML predefined validation mechanism
41 This mechanism can repair the evaluation result to make
42 the return objects more ACPI specification compliant.
43 This option is useful for developers to identify the
44 root cause of an AML interpreter issue when the issue
45 has something to do with the repair mechanism.
47 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG]
48 acpi.debug_level= [HW,ACPI,ACPI_DEBUG]
50 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
51 debug output. Bits in debug_layer correspond to a
52 _COMPONENT in an ACPI source file, e.g.,
53 #define _COMPONENT ACPI_EVENTS
54 Bits in debug_level correspond to a level in
55 ACPI_DEBUG_PRINT statements, e.g.,
56 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
57 The debug_level mask defaults to "info". See
58 Documentation/firmware-guide/acpi/debug.rst for more information about
59 debug layers and levels.
61 Enable processor driver info messages:
62 acpi.debug_layer=0x20000000
63 Enable AML "Debug" output, i.e., stores to the Debug
64 object while interpreting AML:
65 acpi.debug_layer=0xffffffff acpi.debug_level=0x2
66 Enable all messages related to ACPI hardware:
67 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
69 Some values produce so much output that the system is
70 unusable. The "log_buf_len" parameter may be useful
71 if you need to capture more output.
73 acpi_enforce_resources= [ACPI]
75 Check for resource conflicts between native drivers
76 and ACPI OperationRegions (SystemIO and SystemMemory
77 only). IO ports and memory declared in ACPI might be
78 used by the ACPI subsystem in arbitrary AML code and
79 can interfere with legacy drivers.
80 strict (default): access to resources claimed by ACPI
81 is denied; legacy drivers trying to access reserved
82 resources will fail to bind to device using them.
83 lax: access to resources claimed by ACPI is allowed;
84 legacy drivers trying to access reserved resources
85 will bind successfully but a warning message is logged.
86 no: ACPI OperationRegions are not marked as reserved,
87 no further checks are performed.
89 acpi_force_table_verification [HW,ACPI]
90 Enable table checksum verification during early stage.
91 By default, this is disabled due to x86 early mapping
94 acpi_irq_balance [HW,ACPI]
95 ACPI will balance active IRQs
98 acpi_irq_nobalance [HW,ACPI]
99 ACPI will not move active IRQs (default)
102 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA
103 Format: <irq>,<irq>...
105 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for
107 Format: <irq>,<irq>...
109 acpi_mask_gpe= [HW,ACPI]
110 Due to the existence of _Lxx/_Exx, some GPEs triggered
111 by unsupported hardware/firmware features can result in
112 GPE floodings that cannot be automatically disabled by
114 This facility can be used to prevent such uncontrolled
116 Format: <byte> or <bitmap-list>
118 acpi_no_auto_serialize [HW,ACPI]
119 Disable auto-serialization of AML methods
120 AML control methods that contain the opcodes to create
121 named objects will be marked as "Serialized" by the
122 auto-serialization feature.
123 This feature is enabled by default.
124 This option allows to turn off the feature.
126 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump
129 acpi_no_static_ssdt [HW,ACPI]
130 Disable installation of static SSDTs at early boot time
131 By default, SSDTs contained in the RSDT/XSDT will be
132 installed automatically and they will appear under
133 /sys/firmware/acpi/tables.
134 This option turns off this feature.
135 Note that specifying this option does not affect
136 dynamic table installation which will install SSDT
137 tables to /sys/firmware/acpi/tables/dynamic.
139 acpi_no_watchdog [HW,ACPI,WDT]
140 Ignore the ACPI-based watchdog interface (WDAT) and let
141 a native driver control the watchdog device instead.
143 acpi_rsdp= [ACPI,EFI,KEXEC]
144 Pass the RSDP address to the kernel, mostly used
145 on machines running EFI runtime service to boot the
146 second kernel for kdump.
148 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
149 Format: To spoof as Windows 98: ="Microsoft Windows"
151 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead
152 of 2 which is mandated by ACPI 6) as the supported ACPI
153 specification revision (when using this switch, it may
154 be necessary to carry out a cold reboot _twice_ in a
155 row to make it take effect on the platform firmware).
157 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
158 acpi_osi="string1" # add string1
159 acpi_osi="!string2" # remove string2
160 acpi_osi=!* # remove all strings
161 acpi_osi=! # disable all built-in OS vendor
163 acpi_osi=!! # enable all built-in OS vendor
165 acpi_osi= # disable all strings
167 'acpi_osi=!' can be used in combination with single or
168 multiple 'acpi_osi="string1"' to support specific OS
169 vendor string(s). Note that such command can only
170 affect the default state of the OS vendor strings, thus
171 it cannot affect the default state of the feature group
172 strings and the current state of the OS vendor strings,
173 specifying it multiple times through kernel command line
174 is meaningless. This command is useful when one do not
175 care about the state of the feature group strings which
176 should be controlled by the OSPM.
178 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
179 to 'acpi_osi="Windows 2000" acpi_osi=!', they all
180 can make '_OSI("Windows 2000")' TRUE.
182 'acpi_osi=' cannot be used in combination with other
183 'acpi_osi=' command lines, the _OSI method will not
184 exist in the ACPI namespace. NOTE that such command can
185 only affect the _OSI support state, thus specifying it
186 multiple times through kernel command line is also
189 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
192 'acpi_osi=!*' can be used in combination with single or
193 multiple 'acpi_osi="string1"' to support specific
194 string(s). Note that such command can affect the
195 current state of both the OS vendor strings and the
196 feature group strings, thus specifying it multiple times
197 through kernel command line is meaningful. But it may
198 still not able to affect the final state of a string if
199 there are quirks related to this string. This command
200 is useful when one want to control the state of the
201 feature group strings to debug BIOS issues related to
204 1. 'acpi_osi="Module Device" acpi_osi=!*' can make
205 '_OSI("Module Device")' FALSE.
206 2. 'acpi_osi=!* acpi_osi="Module Device"' can make
207 '_OSI("Module Device")' TRUE.
208 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
210 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
212 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
213 they all will make '_OSI("Windows 2000")' TRUE.
216 Override the pmtimer bug detection: force the kernel
217 to assume that this machine's pmtimer latches its value
218 and always returns good values.
220 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
221 Format: { level | edge | high | low }
223 acpi_skip_timer_override [HW,ACPI]
224 Recognize and ignore IRQ0/pin2 Interrupt Override.
225 For broken nForce2 BIOS resulting in XT-PIC timer.
227 acpi_sleep= [HW,ACPI] Sleep options
228 Format: { s3_bios, s3_mode, s3_beep, s4_hwsig,
229 s4_nohwsig, old_ordering, nonvs,
230 sci_force_enable, nobl }
231 See Documentation/power/video.rst for information on
233 s3_beep is for debugging; it makes the PC's speaker beep
234 as soon as the kernel's real-mode entry point is called.
235 s4_hwsig causes the kernel to check the ACPI hardware
236 signature during resume from hibernation, and gracefully
237 refuse to resume if it has changed. This complies with
238 the ACPI specification but not with reality, since
239 Windows does not do this and many laptops do change it
240 on docking. So the default behaviour is to allow resume
241 and simply warn when the signature changes, unless the
242 s4_hwsig option is enabled.
243 s4_nohwsig prevents ACPI hardware signature from being
244 used (or even warned about) during resume.
245 old_ordering causes the ACPI 1.0 ordering of the _PTS
246 control method, with respect to putting devices into
247 low power states, to be enforced (the ACPI 2.0 ordering
248 of _PTS is used by default).
249 nonvs prevents the kernel from saving/restoring the
250 ACPI NVS memory during suspend/hibernation and resume.
251 sci_force_enable causes the kernel to set SCI_EN directly
252 on resume from S1/S3 (which is against the ACPI spec,
253 but some broken systems don't work without it).
254 nobl causes the internal blacklist of systems known to
255 behave incorrectly in some ways with respect to system
256 suspend and resume to be ignored (use wisely).
258 acpi_use_timer_override [HW,ACPI]
259 Use timer override. For some broken Nvidia NF5 boards
260 that require a timer override, but don't have HPET
262 add_efi_memmap [EFI; X86] Include EFI memory map in
263 kernel's map of available physical RAM.
266 { off | try_unsupported }
267 off: disable AGP support
268 try_unsupported: try to drive unsupported chipsets
269 (may crash computer or cause data corruption)
272 See Documentation/sound/alsa-configuration.rst
275 Allow the default userspace alignment fault handler
276 behaviour to be specified. Bit 0 enables warnings,
277 bit 1 enables fixups, and bit 2 sends a segfault.
279 align_va_addr= [X86-64]
280 Align virtual addresses by clearing slice [14:12] when
281 allocating a VMA at process creation time. This option
282 gives you up to 3% performance improvement on AMD F15h
283 machines (where it is enabled by default) for a
284 CPU-intensive style benchmark, and it can vary highly in
285 a microbenchmark depending on workload and compiler.
287 32: only for 32-bit processes
288 64: only for 64-bit processes
289 on: enable for both 32- and 64-bit processes
290 off: disable for both 32- and 64-bit processes
292 alloc_snapshot [FTRACE]
293 Allocate the ftrace snapshot buffer on boot up when the
294 main buffer is allocated. This is handy if debugging
295 and you need to use tracing_snapshot() on boot up, and
296 do not want to use tracing_snapshot_alloc() as it needs
297 to be done where GFP_KERNEL allocations are allowed.
299 allow_mismatched_32bit_el0 [ARM64]
300 Allow execve() of 32-bit applications and setting of the
301 PER_LINUX32 personality on systems where only a strict
302 subset of the CPUs support 32-bit EL0. When this
303 parameter is present, the set of CPUs supporting 32-bit
304 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0
305 and hot-unplug operations may be restricted.
307 See Documentation/arch/arm64/asymmetric-32bit.rst for more
310 amd_iommu= [HW,X86-64]
311 Pass parameters to the AMD IOMMU driver in the system.
313 fullflush - Deprecated, equivalent to iommu.strict=1
314 off - do not initialize any AMD IOMMU found in
316 force_isolation - Force device isolation for all
317 devices. The IOMMU driver is not
318 allowed anymore to lift isolation
319 requirements as needed. This option
320 does not override iommu=pt
321 force_enable - Force enable the IOMMU on platforms known
322 to be buggy with IOMMU enabled. Use this
324 pgtbl_v1 - Use v1 page table for DMA-API (Default).
325 pgtbl_v2 - Use v2 page table for DMA-API.
326 irtcachedis - Disable Interrupt Remapping Table (IRT) caching.
328 amd_iommu_dump= [HW,X86-64]
329 Enable AMD IOMMU driver option to dump the ACPI table
330 for AMD IOMMU. With this option enabled, AMD IOMMU
331 driver will print ACPI tables for AMD IOMMU during
332 IOMMU initialization.
334 amd_iommu_intr= [HW,X86-64]
335 Specifies one of the following AMD IOMMU interrupt
337 legacy - Use legacy interrupt remapping mode.
338 vapic - Use virtual APIC mode, which allows IOMMU
339 to inject interrupts directly into guest.
340 This mode requires kvm-amd.avic=1.
341 (Default when IOMMU HW support is present.)
345 Do not enable amd_pstate as the default
346 scaling driver for the supported processors
348 Use amd_pstate with passive mode as a scaling driver.
349 In this mode autonomous selection is disabled.
350 Driver requests a desired performance level and platform
351 tries to match the same performance level if it is
352 satisfied by guaranteed performance level.
354 Use amd_pstate_epp driver instance as the scaling driver,
355 driver provides a hint to the hardware if software wants
356 to bias toward performance (0x0) or energy efficiency (0xff)
357 to the CPPC firmware. then CPPC power algorithm will
358 calculate the runtime workload and adjust the realtime cores
361 Activate guided autonomous mode. Driver requests minimum and
362 maximum performance level and the platform autonomously
363 selects a performance level in this range and appropriate
364 to the current workload.
366 amijoy.map= [HW,JOY] Amiga joystick support
367 Map of devices attached to JOY0DAT and JOY1DAT
369 See also Documentation/input/joydev/joystick.rst
371 analog.map= [HW,JOY] Analog joystick and gamepad support
372 Specifies type or capabilities of an analog joystick
373 connected to one of 16 gameports
374 Format: <type1>,<type2>,..<type16>
377 Power management functions (SPARCstation-4/5 + deriv.)
379 Disable APC CPU standby support. SPARCstation-Fox does
380 not play well with APC CPU idle - disable it if you have
381 APC and your system crashes randomly.
383 apic= [APIC,X86] Advanced Programmable Interrupt Controller
384 Change the output verbosity while booting
385 Format: { quiet (default) | verbose | debug }
386 Change the amount of debugging information output
387 when initialising the APIC and IO-APIC components.
388 For X86-32, this can also be used to specify an APIC
390 Format: apic=driver_name
391 Examples: apic=bigsmp
393 apic_extnmi= [APIC,X86] External NMI delivery setting
394 Format: { bsp (default) | all | none }
395 bsp: External NMI is delivered only to CPU 0
396 all: External NMIs are broadcast to all CPUs as a
398 none: External NMI is masked for all CPUs. This is
399 useful so that a dump capture kernel won't be
403 See Documentation/networking/ipv6.rst.
405 apm= [APM] Advanced Power Management
406 See header of arch/x86/kernel/apm_32.c.
408 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
409 Format: { "0" | "1" }
410 See security/apparmor/Kconfig help text
413 Default value is set via kernel config option.
415 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
416 Format: <io>,<irq>,<nodeID>
418 arm64.nobti [ARM64] Unconditionally disable Branch Target
419 Identification support
421 arm64.nomops [ARM64] Unconditionally disable Memory Copy and Memory
422 Set instructions support
424 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension
427 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication
430 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix
433 arm64.nosve [ARM64] Unconditionally disable Scalable Vector
438 atarimouse= [HW,MOUSE] Atari Mouse
440 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
441 EzKey and similar keyboards
443 atkbd.reset= [HW] Reset keyboard during initialization
445 atkbd.set= [HW] Select keyboard code set
446 Format: <int> (2 = AT (default), 3 = PS/2)
448 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
451 atkbd.softraw= [HW] Choose between synthetic and real raw mode
452 Format: <bool> (0 = real, 1 = synthetic (default))
454 atkbd.softrepeat= [HW]
455 Use software keyboard repeat
457 audit= [KNL] Enable the audit sub-system
458 Format: { "0" | "1" | "off" | "on" }
459 0 | off - kernel audit is disabled and can not be
460 enabled until the next reboot
461 unset - kernel audit is initialized but disabled and
462 will be fully enabled by the userspace auditd.
463 1 | on - kernel audit is initialized and partially
464 enabled, storing at most audit_backlog_limit
465 messages in RAM until it is fully enabled by the
469 audit_backlog_limit= [KNL] Set the audit queue size limit.
470 Format: <int> (must be >=0)
473 bau= [X86_UV] Enable the BAU on SGI UV. The default
474 behavior is to disable the BAU (i.e. bau=0).
475 Format: { "0" | "1" }
478 unset - Disable the BAU.
480 baycom_epp= [HW,AX25]
483 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
485 See header of drivers/net/hamradio/baycom_par.c.
487 baycom_ser_fdx= [HW,AX25]
488 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
489 Format: <io>,<irq>,<mode>[,<baud>]
490 See header of drivers/net/hamradio/baycom_ser_fdx.c.
492 baycom_ser_hdx= [HW,AX25]
493 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
494 Format: <io>,<irq>,<mode>
495 See header of drivers/net/hamradio/baycom_ser_hdx.c.
498 Disable BERT OS support on buggy BIOSes.
500 bgrt_disable [ACPI][X86]
501 Disable BGRT to avoid flickering OEM logo.
503 blkdevparts= Manual partition parsing of block device(s) for
504 embedded devices based on command line input.
505 See Documentation/block/cmdline-partition.rst
507 boot_delay= Milliseconds to delay each printk during boot.
508 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled,
509 and you may also have to specify "lpj=". Boot_delay
510 values larger than 10 seconds (10000) are assumed
511 erroneous and ignored.
515 Extended command line options can be added to an initrd
516 and this will cause the kernel to look for it.
518 See Documentation/admin-guide/bootconfig.rst
520 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
521 bttv.radio= Most important insmod options are available as
523 bttv.pll= See Documentation/admin-guide/media/bttv.rst
526 bulk_remove=off [PPC] This parameter disables the use of the pSeries
527 firmware feature for flushing multiple hpte entries
530 c101= [NET] Moxa C101 synchronous serial card
532 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
533 Sometimes CPU hardware bugs make them report the cache
534 size incorrectly. The kernel will attempt work arounds
535 to fix known problems, but for some CPUs it is not
536 possible to determine what the correct size should be.
537 This option provides an override for these situations.
540 [NET] Specifies amount of time (in seconds) that
541 the kernel should wait for a network carrier. By default
542 it waits 120 seconds.
544 ca_keys= [KEYS] This parameter identifies a specific key(s) on
545 the system trusted keyring to be used for certificate
547 format: { id:<keyid> | builtin }
549 cca= [MIPS] Override the kernel pages' cache coherency
550 algorithm. Accepted values range from 0 to 7
551 inclusive. See arch/mips/include/asm/pgtable-bits.h
552 for platform specific values (SB1, Loongson3 and
555 ccw_timeout_log [S390]
556 See Documentation/arch/s390/common_io.rst for details.
558 cgroup_disable= [KNL] Disable a particular controller or optional feature
559 Format: {name of the controller(s) or feature(s) to disable}
560 The effects of cgroup_disable=foo are:
561 - foo isn't auto-mounted if you mount all cgroups in
563 - foo isn't visible as an individually mountable
565 - if foo is an optional feature then the feature is
566 disabled and corresponding cgroup files are not
568 {Currently only "memory" controller deal with this and
569 cut the overhead, others just disable the usage. So
570 only cgroup_disable=memory is actually worthy}
571 Specifying "pressure" disables per-cgroup pressure
572 stall information accounting feature
574 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1
575 Format: { { controller | "all" | "named" }
576 [,{ controller | "all" | "named" }...] }
577 Like cgroup_disable, but only applies to cgroup v1;
578 the blacklisted controllers remain available in cgroup2.
579 "all" blacklists all controllers and "named" disables
580 named mounts. Specifying both "all" and "named" disables
583 cgroup.memory= [KNL] Pass options to the cgroup memory controller.
585 nosocket -- Disable socket memory accounting.
586 nokmem -- Disable kernel memory accounting.
587 nobpf -- Disable BPF memory accounting.
589 checkreqprot= [SELINUX] Set initial checkreqprot flag value.
590 Format: { "0" | "1" }
591 See security/selinux/Kconfig help text.
592 0 -- check protection applied by kernel (includes
593 any implied execute protection).
594 1 -- check protection requested by application.
595 Default value is set via a kernel config option.
596 Value can be changed at runtime via
597 /sys/fs/selinux/checkreqprot.
598 Setting checkreqprot to 1 is deprecated.
601 See Documentation/arch/s390/common_io.rst for details.
603 clearcpuid=X[,X...] [X86]
604 Disable CPUID feature X for the kernel. See
605 arch/x86/include/asm/cpufeatures.h for the valid bit
606 numbers X. Note the Linux-specific bits are not necessarily
607 stable over kernel options, but the vendor-specific
609 X can also be a string as appearing in the flags: line
610 in /proc/cpuinfo which does not have the above
611 instability issue. However, not all features have names
613 Note that using this option will taint your kernel.
614 Also note that user programs calling CPUID directly
615 or using the feature without checking anything
616 will still see it. This just prevents it from
617 being used by the kernel or shown in /proc/cpuinfo.
618 Also note the kernel might malfunction if you disable
623 Prevents the clock framework from automatically gating
624 clocks that have not been explicitly enabled by a Linux
625 device driver but are enabled in hardware at reset or
626 by the bootloader/firmware. Note that this does not
627 force such clocks to be always-on nor does it reserve
628 those clocks in any way. This parameter is useful for
629 debug and development, but should not be needed on a
630 platform with proper driver support. For more
631 information, see Documentation/driver-api/clk.rst.
633 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
635 Forces specified clocksource (if available) to be used
636 when calculating gettimeofday(). If specified
637 clocksource is not available, it defaults to PIT.
638 Format: { pit | tsc | cyclone | pmtmr }
640 clocksource= Override the default clocksource
642 Override the default clocksource and use the clocksource
643 with the name specified.
644 Some clocksource names to choose from, depending on
646 [all] jiffies (this is the base, fallback clocksource)
648 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
649 pxa_timer,timer3,32k_counter,timer0_1
650 [X86-32] pit,hpet,tsc;
651 scx200_hrt on Geode; cyclone on IBM x440
659 clocksource.arm_arch_timer.evtstrm=
662 Enable/disable the eventstream feature of the ARM
663 architected timer so that code using WFE-based polling
664 loops can be debugged more effectively on production
667 clocksource.max_cswd_read_retries= [KNL]
668 Number of clocksource_watchdog() retries due to
669 external delays before the clock will be marked
670 unstable. Defaults to two retries, that is,
671 three attempts to read the clock under test.
673 clocksource.verify_n_cpus= [KNL]
674 Limit the number of CPUs checked for clocksources
675 marked with CLOCK_SOURCE_VERIFY_PERCPU that
676 are marked unstable due to excessive skew.
677 A negative value says to check all CPUs, while
678 zero says not to check any. Values larger than
679 nr_cpu_ids are silently truncated to nr_cpu_ids.
680 The actual CPUs are chosen randomly, with
681 no replacement if the same CPU is chosen twice.
683 clocksource-wdtest.holdoff= [KNL]
684 Set the time in seconds that the clocksource
685 watchdog test waits before commencing its tests.
686 Defaults to zero when built as a module and to
687 10 seconds when built into the kernel.
689 cma=nn[MG]@[start[MG][-end[MG]]]
691 Sets the size of kernel global memory area for
692 contiguous memory allocations and optionally the
693 placement constraint by the physical address range of
694 memory allocations. A value of 0 disables CMA
695 altogether. For more information, see
696 kernel/dma/contiguous.c
700 Sets the size of kernel per-numa memory area for
701 contiguous memory allocations. A value of 0 disables
702 per-numa CMA altogether. And If this option is not
703 specified, the default value is 0.
704 With per-numa CMA enabled, DMA users on node nid will
705 first try to allocate buffer from the pernuma area
706 which is located in node nid, if the allocation fails,
707 they will fallback to the global default memory area.
709 numa_cma=<node>:nn[MG][,<node>:nn[MG]]
711 Sets the size of kernel numa memory area for
712 contiguous memory allocations. It will reserve CMA
713 area for the specified node.
715 With numa CMA enabled, DMA users on node nid will
716 first try to allocate buffer from the numa area
717 which is located in node nid, if the allocation fails,
718 they will fallback to the global default memory area.
720 cmo_free_hint= [PPC] Format: { yes | no }
721 Specify whether pages are marked as being inactive
722 when they are freed. This is used in CMO environments
723 to determine OS memory pressure for page stealing by
727 coherent_pool=nn[KMG] [ARM,KNL]
728 Sets the size of memory pool for coherent, atomic dma
729 allocations, by default set to 256K.
731 com20020= [HW,NET] ARCnet - COM20020 chipset
733 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
735 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
739 ARCnet - COM90xx chipset (memory-mapped buffers)
740 Format: <io>[,<irq>[,<memstart>]]
742 condev= [HW,S390] console device
745 con3215_drop= [S390] 3215 console drop mode.
747 When set to true, drop data on the 3215 console when
748 the console buffer is full. In this case the
749 operator using a 3270 terminal emulator (for example
750 x3270) does not have to enter the clear key for the
751 console output to advance and the kernel to continue.
752 This leads to a much faster boot time when a 3270
753 terminal emulator is active. If no 3270 terminal
754 emulator is used, this parameter has no effect.
756 console= [KNL] Output console device and options.
758 tty<n> Use the virtual console device <n>.
762 Use the specified serial port. The options are of
763 the form "bbbbpnf", where "bbbb" is the baud rate,
764 "p" is parity ("n", "o", or "e"), "n" is number of
765 bits, and "f" is flow control ("r" for RTS or
766 omit it). Default is "9600n8".
768 See Documentation/admin-guide/serial-console.rst for more
770 Documentation/networking/netconsole.rst for an
773 uart[8250],io,<addr>[,options]
774 uart[8250],mmio,<addr>[,options]
775 uart[8250],mmio16,<addr>[,options]
776 uart[8250],mmio32,<addr>[,options]
777 uart[8250],0x<addr>[,options]
778 Start an early, polled-mode console on the 8250/16550
779 UART at the specified I/O port or MMIO address,
780 switching to the matching ttyS device later.
781 MMIO inter-register address stride is either 8-bit
782 (mmio), 16-bit (mmio16), or 32-bit (mmio32).
783 If none of [io|mmio|mmio16|mmio32], <addr> is assumed
784 to be equivalent to 'mmio'. 'options' are specified in
785 the same format described for ttyS above; if unspecified,
786 the h/w is not re-initialized.
788 hvc<n> Use the hypervisor console device <n>. This is for
789 both Xen and PowerPC hypervisors.
792 Use to disable console output, i.e., to have kernel
793 console messages discarded.
794 This must be the only console= parameter used on the
797 If the device connected to the port is not a TTY but a braille
798 device, prepend "brl," before the device type, for instance
800 For now, only VisioBraille is supported.
803 [KNL] Change console messages format
805 By default we print messages on consoles in
806 "[time stamp] text\n" format (time stamp may not be
807 printed, depending on CONFIG_PRINTK_TIME or
808 `printk_time' param).
810 Switch to syslog format: "<%u>[time stamp] text\n"
811 IOW, each message will have a facility and loglevel
812 prefix. The format is similar to one used by syslog()
813 syscall, or to executing "dmesg -S --raw" or to reading
816 consoleblank= [KNL] The console blank (screen saver) timeout in
817 seconds. A value of 0 disables the blank timer.
821 [KNL] Change the default value for
822 /proc/<pid>/coredump_filter.
823 See also Documentation/filesystems/proc.rst.
825 coresight_cpu_debug.enable
828 Enable/disable the CPU sampling based debugging.
829 0: default value, disable debugging
830 1: enable debugging at boot time
832 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
834 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
836 cpuidle.off=1 [CPU_IDLE]
837 disable the cpuidle sub-system
840 [CPU_IDLE] Name of the cpuidle governor to use.
842 cpufreq.off=1 [CPU_FREQ]
843 disable the cpufreq sub-system
845 cpufreq.default_governor=
846 [CPU_FREQ] Name of the default cpufreq governor or
847 policy to use. This governor must be registered in the
848 kernel before the cpufreq driver probes.
851 [X86] Delay for N microsec between assert and de-assert
852 of APIC INIT to start processors. This delay occurs
853 on every CPU online, such as boot, and resume from suspend.
857 [SMP] Enable/disable parallel bringup of secondary CPUs
859 Default is enabled if CONFIG_HOTPLUG_PARALLEL=y. Otherwise
860 the parameter has no effect.
862 crash_kexec_post_notifiers
863 Run kdump after running panic-notifiers and dumping
864 kmsg. This only for the users who doubt kdump always
865 succeeds in any situation.
866 Note that this also increases risks of kdump failure,
867 because some panic notifiers can make the crashed
868 kernel more unstable.
870 crashkernel=size[KMG][@offset[KMG]]
871 [KNL] Using kexec, Linux can switch to a 'crash kernel'
872 upon panic. This parameter reserves the physical
873 memory region [offset, offset + size] for that kernel
874 image. If '@offset' is omitted, then a suitable offset
875 is selected automatically.
876 [KNL, X86-64, ARM64, RISCV] Select a region under 4G first, and
877 fall back to reserve region above 4G when '@offset'
878 hasn't been specified.
879 See Documentation/admin-guide/kdump/kdump.rst for further details.
881 crashkernel=range1:size1[,range2:size2,...][@offset]
882 [KNL] Same as above, but depends on the memory
883 in the running system. The syntax of range is
884 start-[end] where start and end are both
885 a memory unit (amount[KMG]). See also
886 Documentation/admin-guide/kdump/kdump.rst for an example.
888 crashkernel=size[KMG],high
889 [KNL, X86-64, ARM64, RISCV] range could be above 4G.
890 Allow kernel to allocate physical memory region from top,
891 so could be above 4G if system have more than 4G ram
892 installed. Otherwise memory region will be allocated
893 below 4G, if available.
894 It will be ignored if crashkernel=X is specified.
895 crashkernel=size[KMG],low
896 [KNL, X86-64, ARM64, RISCV] range under 4G. When crashkernel=X,high
897 is passed, kernel could allocate physical memory region
898 above 4G, that cause second kernel crash on system
899 that require some amount of low memory, e.g. swiotlb
900 requires at least 64M+32K low memory, also enough extra
901 low memory is needed to make sure DMA buffers for 32-bit
902 devices won't run out. Kernel would try to allocate
903 default size of memory below 4G automatically. The default
904 size is platform dependent.
905 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB)
908 This one lets the user specify own low range under 4G
909 for second kernel instead.
910 0: to disable low allocation.
911 It will be ignored when crashkernel=X,high is not used
912 or memory reserved is below 4G.
915 [KNL] Disable crypto self-tests
920 cs89x0_media= [HW,NET]
921 Format: { rj45 | aui | bnc }
923 csdlock_debug= [KNL] Enable or disable debug add-ons of cross-CPU
924 function call handling. When switched on,
925 additional debug data is printed to the console
926 in case a hanging CPU is detected, and that
927 CPU is pinged again in order to try to resolve
928 the hang situation. The default value of this
929 option depends on the CSD_LOCK_WAIT_DEBUG_DEFAULT
933 See header of drivers/s390/block/dasd_devmap.c.
935 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
936 (one device per port)
937 Format: <port#>,<type>
938 See also Documentation/input/devices/joystick-parport.rst
940 debug [KNL] Enable kernel debugging (events log level).
943 [KNL] Enable printing [hashed] pointers early in the
944 boot sequence. If enabled, we use a weak hash instead
945 of siphash to hash pointers. Use this option if you are
946 seeing instances of '(___ptrval___)') and need to see a
947 value (hashed pointer) instead. Cryptographically
948 insecure, please do not use on production kernels.
951 [KNL] verbose locking self-tests
953 Print debugging info while doing the locking API
955 Bitmask for the various LOCKTYPE_ tests. Defaults to 0
956 (no extra messages), setting it to -1 (all bits set)
957 will print _a_lot_ more information - normally only
958 useful to lockdep developers.
960 debug_objects [KNL] Enable object debugging
962 debug_guardpage_minorder=
963 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
964 parameter allows control of the order of pages that will
965 be intentionally kept free (and hence protected) by the
966 buddy allocator. Bigger value increase the probability
967 of catching random memory corruption, but reduce the
968 amount of memory for normal system use. The maximum
969 possible value is MAX_ORDER/2. Setting this parameter
970 to 1 or 2 should be enough to identify most random
971 memory corruption problems caused by bugs in kernel or
972 driver code when a CPU writes to (or reads from) a
973 random memory location. Note that there exists a class
974 of memory corruptions problems caused by buggy H/W or
975 F/W or by drivers badly programming DMA (basically when
976 memory is written at bus level and the CPU MMU is
977 bypassed) which are not detectable by
978 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
979 tracking down these problems.
982 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter
983 enables the feature at boot time. By default, it is
984 disabled and the system will work mostly the same as a
985 kernel built without CONFIG_DEBUG_PAGEALLOC.
986 Note: to get most of debug_pagealloc error reports, it's
987 useful to also enable the page_owner functionality.
988 on: enable the feature
990 debugfs= [KNL] This parameter enables what is exposed to userspace
991 and debugfs internal clients.
992 Format: { on, no-mount, off }
993 on: All functions are enabled.
995 Filesystem is not registered but kernel clients can
996 access APIs and a crashkernel can be used to read
997 its content. There is nothing to mount.
998 off: Filesystem is not registered and clients
999 get a -EPERM as result when trying to register files
1000 or directories within debugfs.
1001 This is equivalent of the runtime functionality if
1002 debugfs was not enabled in the kernel at all.
1003 Default value is set in build-time with a kernel configuration.
1005 debugpat [X86] Enable PAT debugging
1008 [HW] The size of the default HugeTLB page. This is
1009 the size represented by the legacy /proc/ hugepages
1010 APIs. In addition, this is the default hugetlb size
1011 used for shmget(), mmap() and mounting hugetlbfs
1012 filesystems. If not specified, defaults to the
1013 architecture's default huge page size. Huge page
1014 sizes are architecture dependent. See also
1015 Documentation/admin-guide/mm/hugetlbpage.rst.
1018 deferred_probe_timeout=
1019 [KNL] Debugging option to set a timeout in seconds for
1020 deferred probe to give up waiting on dependencies to
1021 probe. Only specific dependencies (subsystems or
1022 drivers) that have opted in will be ignored. A timeout
1023 of 0 will timeout at the end of initcalls. If the time
1024 out hasn't expired, it'll be restarted by each
1025 successful driver registration. This option will also
1026 dump out devices still on the deferred probe list after
1029 delayacct [KNL] Enable per-task delay accounting
1031 dell_smm_hwmon.ignore_dmi=
1032 [HW] Continue probing hardware even if DMI data
1033 indicates that the driver is running on unsupported
1036 dell_smm_hwmon.force=
1037 [HW] Activate driver even if SMM BIOS signature does
1038 not match list of supported models and enable otherwise
1039 blacklisted features.
1041 dell_smm_hwmon.power_status=
1042 [HW] Report power status in /proc/i8k
1043 (disabled by default).
1045 dell_smm_hwmon.restricted=
1046 [HW] Allow controlling fans only if SYS_ADMIN
1049 dell_smm_hwmon.fan_mult=
1050 [HW] Factor to multiply fan speed with.
1052 dell_smm_hwmon.fan_max=
1053 [HW] Maximum configurable fan speed.
1056 Format: { on | off | def_only | inf_only | always }
1057 on: s390 zlib hardware support for compression on
1058 level 1 and decompression (default)
1059 off: No s390 zlib hardware support
1060 def_only: s390 zlib hardware support for deflate
1061 only (compression on level 1)
1062 inf_only: s390 zlib hardware support for inflate
1063 only (decompression)
1064 always: Same as 'on' but ignores the selected compression
1065 level always using hardware support (used for debugging)
1067 dhash_entries= [KNL]
1068 Set number of hash buckets for dentry cache.
1070 disable_1tb_segments [PPC]
1071 Disables the use of 1TB hash page table segments. This
1072 causes the kernel to fall back to 256MB segments which
1073 can be useful when debugging issues that require an SLB
1077 See Documentation/networking/ipv6.rst.
1080 Disable RADIX MMU mode on POWER9
1083 Disable TLBIE instruction. Currently does not work
1084 with KVM, with HASH MMU, or with coherent accelerators.
1086 disable_cpu_apicid= [X86,APIC,SMP]
1088 The number of initial APIC ID for the
1089 corresponding CPU to be disabled at boot,
1090 mostly used for the kdump 2nd kernel to
1091 disable BSP to wake up multiple CPUs without
1092 causing system reset or hang due to sending
1093 INIT from AP to BSP.
1095 disable_ddw [PPC/PSERIES]
1096 Disable Dynamic DMA Window support. Use this
1097 to workaround buggy firmware.
1099 disable_ipv6= [IPV6]
1100 See Documentation/networking/ipv6.rst.
1102 disable_mtrr_cleanup [X86]
1103 The kernel tries to adjust MTRR layout from continuous
1104 to discrete, to make X server driver able to add WB
1105 entry later. This parameter disables that.
1107 disable_mtrr_trim [X86, Intel and AMD only]
1108 By default the kernel will trim any uncacheable
1109 memory out of your available memory pool based on
1110 MTRR settings. This parameter disables that behavior,
1111 possibly causing your machine to run very slowly.
1113 disable_timer_pin_1 [X86]
1114 Disable PIN 1 of APIC timer
1115 Can be useful to work around chipset bugs.
1117 dis_ucode_ldr [X86] Disable the microcode loader.
1119 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
1120 this option disables the debugging code at boot.
1122 dma_debug_entries=<number>
1123 This option allows to tune the number of preallocated
1124 entries for DMA-API debugging code. One entry is
1125 required per DMA-API allocation. Use this if the
1126 DMA-API debugging code disables itself because the
1127 architectural default is too low.
1129 dma_debug_driver=<driver_name>
1130 With this option the DMA-API debugging driver
1131 filter feature can be enabled at boot time. Just
1132 pass the driver to filter for as the parameter.
1133 The filter can be disabled or changed to another
1134 driver later using sysfs.
1136 driver_async_probe= [KNL]
1137 List of driver names to be probed asynchronously. *
1138 matches with all driver names. If * is specified, the
1139 rest of the listed driver names are those that will NOT
1141 Format: <driver_name1>,<driver_name2>...
1143 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
1144 Broken monitors, graphic adapters, KVMs and EDIDless
1145 panels may send no or incorrect EDID data sets.
1146 This parameter allows to specify an EDID data sets
1147 in the /lib/firmware directory that are used instead.
1148 Generic built-in EDID data sets are used, if one of
1149 edid/1024x768.bin, edid/1280x1024.bin,
1150 edid/1680x1050.bin, or edid/1920x1080.bin is given
1151 and no file with the same name exists. Details and
1152 instructions how to build your own EDID data are
1153 available in Documentation/admin-guide/edid.rst. An EDID
1154 data set will only be used for a particular connector,
1155 if its name and a colon are prepended to the EDID
1156 name. Each connector may use a unique EDID data
1157 set by separating the files with a comma. An EDID
1158 data set with no connector name will be used for
1159 any connectors not explicitly specified.
1164 Format: {"off" | "known"}
1165 Control how the dt_cpu_ftrs device-tree binding is
1166 used for CPU feature discovery and setup (if it
1168 off: Do not use it, fall back to legacy cpu table.
1169 known: Do not pass through unknown features to guests
1170 or userspace, only those that the kernel is aware of.
1172 dump_apple_properties [X86]
1173 Dump name and content of EFI device properties on
1174 x86 Macs. Useful for driver authors to determine
1175 what data is available or for reverse-engineering.
1177 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
1178 <module>.dyndbg[="val"]
1179 Enable debug messages at boot time. See
1180 Documentation/admin-guide/dynamic-debug-howto.rst
1183 early_ioremap_debug [KNL]
1184 Enable debug messages in early_ioremap support. This
1185 is useful for tracking down temporary early mappings
1186 which are not unmapped.
1188 earlycon= [KNL] Output early console device and options.
1190 When used with no options, the early console is
1191 determined by stdout-path property in device tree's
1192 chosen node or the ACPI SPCR table if supported by
1195 cdns,<addr>[,options]
1196 Start an early, polled-mode console on a Cadence
1197 (xuartps) serial port at the specified address. Only
1198 supported option is baud rate. If baud rate is not
1199 specified, the serial port must already be setup and
1202 uart[8250],io,<addr>[,options[,uartclk]]
1203 uart[8250],mmio,<addr>[,options[,uartclk]]
1204 uart[8250],mmio32,<addr>[,options[,uartclk]]
1205 uart[8250],mmio32be,<addr>[,options[,uartclk]]
1206 uart[8250],0x<addr>[,options]
1207 Start an early, polled-mode console on the 8250/16550
1208 UART at the specified I/O port or MMIO address.
1209 MMIO inter-register address stride is either 8-bit
1210 (mmio) or 32-bit (mmio32 or mmio32be).
1211 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
1212 to be equivalent to 'mmio'. 'options' are specified
1213 in the same format described for "console=ttyS<n>"; if
1214 unspecified, the h/w is not initialized. 'uartclk' is
1215 the uart clock frequency; if unspecified, it is set
1216 to 'BASE_BAUD' * 16.
1220 Start an early, polled-mode console on a pl011 serial
1221 port at the specified address. The pl011 serial port
1222 must already be setup and configured. Options are not
1223 yet supported. If 'mmio32' is specified, then only
1224 the driver will use only 32-bit accessors to read/write
1225 the device registers.
1228 Start an early console on a litex serial port at the
1229 specified address. The serial port must already be
1230 setup and configured. Options are not yet supported.
1233 Start an early, polled-mode console on a meson serial
1234 port at the specified address. The serial port must
1235 already be setup and configured. Options are not yet
1239 Start an early, polled-mode console on an msm serial
1240 port at the specified address. The serial port
1241 must already be setup and configured. Options are not
1244 msm_serial_dm,<addr>
1245 Start an early, polled-mode console on an msm serial
1246 dm port at the specified address. The serial port
1247 must already be setup and configured. Options are not
1251 Start an early, polled-mode console on a serial port
1252 of an Actions Semi SoC, such as S500 or S900, at the
1253 specified address. The serial port must already be
1254 setup and configured. Options are not yet supported.
1257 Start an early, polled-mode console on a serial port
1258 of an RDA Micro SoC, such as RDA8810PL, at the
1259 specified address. The serial port must already be
1260 setup and configured. Options are not yet supported.
1263 Use RISC-V SBI (Supervisor Binary Interface) for early
1266 smh Use ARM semihosting calls for early console.
1274 Use early console provided by serial driver available
1275 on Samsung SoCs, requires selecting proper type and
1276 a correct base address of the selected UART port. The
1277 serial port must already be setup and configured.
1278 Options are not yet supported.
1281 Start an early, polled-mode console on a lantiq serial
1282 (lqasc) port at the specified address. The serial port
1283 must already be setup and configured. Options are not
1288 Use early console provided by Freescale LP UART driver
1289 found on Freescale Vybrid and QorIQ LS1021A processors.
1290 A valid base address must be provided, and the serial
1291 port must already be setup and configured.
1295 Start an early, polled-mode, output-only console on the
1296 Freescale i.MX UART at the specified address. The UART
1297 must already be setup and configured.
1300 Start an early, polled-mode console on the
1301 Armada 3700 serial port at the specified
1302 address. The serial port must already be setup
1303 and configured. Options are not yet supported.
1306 Start an early, polled-mode console on a Qualcomm
1307 Generic Interface (GENI) based serial port at the
1308 specified address. The serial port must already be
1309 setup and configured. Options are not yet supported.
1312 Start an early, unaccelerated console on the EFI
1313 memory mapped framebuffer (if available). On cache
1314 coherent non-x86 systems that use system memory for
1315 the framebuffer, pass the 'ram' option so that it is
1316 mapped with the correct attributes.
1319 Use early console provided by Freescale LINFlexD UART
1320 serial driver for NXP S32V234 SoCs. A valid base
1321 address must be provided, and the serial port must
1322 already be setup and configured.
1324 earlyprintk= [X86,SH,ARM,M68k,S390]
1328 earlyprintk=serial[,ttySn[,baudrate]]
1329 earlyprintk=serial[,0x...[,baudrate]]
1330 earlyprintk=ttySn[,baudrate]
1331 earlyprintk=dbgp[debugController#]
1332 earlyprintk=pciserial[,force],bus:device.function[,baudrate]
1333 earlyprintk=xdbc[xhciController#]
1335 earlyprintk is useful when the kernel crashes before
1336 the normal console is initialized. It is not enabled by
1337 default because it has some cosmetic problems.
1339 Append ",keep" to not disable it when the real console
1342 Only one of vga, serial, or usb debug port can
1345 Currently only ttyS0 and ttyS1 may be specified by
1346 name. Other I/O ports may be explicitly specified
1347 on some architectures (x86 and arm at least) by
1348 replacing ttySn with an I/O port address, like this:
1349 earlyprintk=serial,0x1008,115200
1350 You can find the port for a given device in
1351 /proc/tty/driver/serial:
1352 2: uart:ST16650V2 port:00001008 irq:18 ...
1354 Interaction with the standard serial driver is not
1357 The VGA output is eventually overwritten by
1360 The xen option can only be used in Xen domains.
1362 The sclp output can only be used on s390.
1364 The optional "force" to "pciserial" enables use of a
1365 PCI device even when its classcode is not of the
1368 edac_report= [HW,EDAC] Control how to report EDAC event
1369 Format: {"on" | "off" | "force"}
1370 on: enable EDAC to report H/W event. May be overridden
1371 by other higher priority error reporting module.
1372 off: disable H/W event reporting through EDAC.
1373 force: enforce the use of EDAC to report H/W event.
1377 Format: {"off" | "on" | "skip[mbr]"}
1380 Format: { "debug", "disable_early_pci_dma",
1381 "nochunk", "noruntime", "nosoftreserve",
1382 "novamap", "no_disable_early_pci_dma" }
1383 debug: enable misc debug output.
1384 disable_early_pci_dma: disable the busmaster bit on all
1385 PCI bridges while in the EFI boot stub.
1386 nochunk: disable reading files in "chunks" in the EFI
1387 boot stub, as chunking can cause problems with some
1388 firmware implementations.
1389 noruntime : disable EFI runtime services support
1390 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose)
1391 attribute may cause the kernel to reserve the
1392 memory range for a memory mapping driver to
1393 claim. Specify efi=nosoftreserve to disable this
1394 reservation and treat the memory by its base type
1395 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM").
1396 novamap: do not call SetVirtualAddressMap().
1397 no_disable_early_pci_dma: Leave the busmaster bit set
1398 on all PCI bridges while in the EFI boot stub
1400 efi_no_storage_paranoia [EFI; X86]
1401 Using this parameter you can use more than 50% of
1402 your efi variable storage. Use this parameter only if
1403 you are really sure that your UEFI does sane gc and
1404 fulfills the spec otherwise your board may brick.
1406 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1407 Add arbitrary attribute to specific memory range by
1408 updating original EFI memory map.
1409 Region of memory which aa attribute is added to is
1412 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1413 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1414 attribute is added to range 0x100000000-0x180000000 and
1415 0x10a0000000-0x1120000000.
1417 If efi_fake_mem=8G@9G:0x40000 is specified, the
1418 EFI_MEMORY_SP(0x40000) attribute is added to
1419 range 0x240000000-0x43fffffff.
1421 Using this parameter you can do debugging of EFI memmap
1422 related features. For example, you can do debugging of
1423 Address Range Mirroring feature even if your box
1424 doesn't support it, or mark specific memory as
1427 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT
1428 that is to be dynamically loaded by Linux. If there are
1429 multiple variables with the same name but with different
1430 vendor GUIDs, all of them will be loaded. See
1431 Documentation/admin-guide/acpi/ssdt-overlays.rst for details.
1434 eisa_irq_edge= [PARISC,HW]
1435 See header of drivers/parisc/eisa.c.
1437 ekgdboc= [X86,KGDB] Allow early kernel console debugging
1440 This is designed to be used in conjunction with
1441 the boot argument: earlyprintk=vga
1443 This parameter works in place of the kgdboc parameter
1444 but can only be used if the backing tty is available
1445 very early in the boot process. For early debugging
1446 via a serial port see kgdboc_earlycon instead.
1449 See comment before function elanfreq_setup() in
1450 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1452 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1453 Specifies physical address of start of kernel core
1454 image elf header and optionally the size. Generally
1455 kexec loader will pass this option to capture kernel.
1456 See Documentation/admin-guide/kdump/kdump.rst for details.
1458 enable_mtrr_cleanup [X86]
1459 The kernel tries to adjust MTRR layout from continuous
1460 to discrete, to make X server driver able to add WB
1461 entry later. This parameter enables that.
1463 enable_timer_pin_1 [X86]
1464 Enable PIN 1 of APIC timer
1465 Can be useful to work around chipset bugs
1466 (in particular on some ATI chipsets).
1467 The kernel tries to set a reasonable default.
1469 enforcing= [SELINUX] Set initial enforcing status.
1471 See security/selinux/Kconfig help text.
1472 0 -- permissive (log only, no denials).
1473 1 -- enforcing (deny and log).
1475 Value can be changed at runtime via
1476 /sys/fs/selinux/enforce.
1479 Disable Error Record Serialization Table (ERST)
1482 ether= [HW,NET] Ethernet cards parameters
1483 This option is obsoleted by the "netdev=" option, which
1484 has equivalent usage. See its documentation for details.
1488 Permit 'security.evm' to be updated regardless of
1489 current integrity status.
1491 early_page_ext [KNL] Enforces page_ext initialization to earlier
1492 stages so cover more early boot allocations.
1493 Please note that as side effect some optimizations
1494 might be disabled to achieve that (e.g. parallelized
1495 memory initialization is disabled) so the boot process
1496 might take longer, especially on systems with a lot of
1497 memory. Available with CONFIG_PAGE_EXTENSION=y.
1502 fail_make_request=[KNL]
1503 General fault injection mechanism.
1504 Format: <interval>,<probability>,<space>,<times>
1505 See also Documentation/fault-injection/.
1508 Format: { initns | none }
1509 See Documentation/admin-guide/sysctl/net.rst for
1510 fb_tunnels_only_for_init_ns
1513 See Documentation/admin-guide/blockdev/floppy.rst.
1515 force_pal_cache_flush
1516 [IA-64] Avoid check_sal_cache_flush which may hang on
1517 buggy SAL_CACHE_FLUSH implementations. Using this
1518 parameter will force ia64_sal_cache_flush to call
1519 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1522 Forcefully enable Physical Address Extension (PAE).
1523 Many Pentium M systems disable PAE but may have a
1524 functionally usable PAE implementation.
1525 Warning: use of this parameter will taint the kernel
1526 and may cause unknown problems.
1529 [FTRACE] will set and start the specified tracer
1530 as early as possible in order to facilitate early
1533 ftrace_boot_snapshot
1534 [FTRACE] On boot up, a snapshot will be taken of the
1535 ftrace ring buffer that can be read at:
1536 /sys/kernel/tracing/snapshot.
1537 This is useful if you need tracing information from kernel
1538 boot up that is likely to be overridden by user space
1539 start up functionality.
1541 Optionally, the snapshot can also be defined for a tracing
1542 instance that was created by the trace_instance= command
1545 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo
1547 The above will cause the "foo" tracing instance to trigger
1548 a snapshot at the end of boot up.
1550 ftrace_dump_on_oops[=orig_cpu]
1551 [FTRACE] will dump the trace buffers on oops.
1552 If no parameter is passed, ftrace will dump
1553 buffers of all CPUs, but if you pass orig_cpu, it will
1554 dump only the buffer of the CPU that triggered the
1557 ftrace_filter=[function-list]
1558 [FTRACE] Limit the functions traced by the function
1559 tracer at boot up. function-list is a comma-separated
1560 list of functions. This list can be changed at run
1561 time by the set_ftrace_filter file in the debugfs
1564 ftrace_notrace=[function-list]
1565 [FTRACE] Do not trace the functions specified in
1566 function-list. This list can be changed at run time
1567 by the set_ftrace_notrace file in the debugfs
1570 ftrace_graph_filter=[function-list]
1571 [FTRACE] Limit the top level callers functions traced
1572 by the function graph tracer at boot up.
1573 function-list is a comma-separated list of functions
1574 that can be changed at run time by the
1575 set_graph_function file in the debugfs tracing directory.
1577 ftrace_graph_notrace=[function-list]
1578 [FTRACE] Do not trace from the functions specified in
1579 function-list. This list is a comma-separated list of
1580 functions that can be changed at run time by the
1581 set_graph_notrace file in the debugfs tracing directory.
1583 ftrace_graph_max_depth=<uint>
1584 [FTRACE] Used with the function graph tracer. This is
1585 the max depth it will trace into a function. This value
1586 can be changed at run time by the max_graph_depth file
1587 in the tracefs tracing directory. default: 0 (no limit)
1589 fw_devlink= [KNL] Create device links between consumer and supplier
1590 devices by scanning the firmware to infer the
1591 consumer/supplier relationships. This feature is
1592 especially useful when drivers are loaded as modules as
1593 it ensures proper ordering of tasks like device probing
1594 (suppliers first, then consumers), supplier boot state
1595 clean up (only after all consumers have probed),
1596 suspend/resume & runtime PM (consumers first, then
1598 Format: { off | permissive | on | rpm }
1599 off -- Don't create device links from firmware info.
1600 permissive -- Create device links from firmware info
1601 but use it only for ordering boot state clean
1602 up (sync_state() calls).
1603 on -- Create device links from firmware info and use it
1604 to enforce probe and suspend/resume ordering.
1605 rpm -- Like "on", but also use to order runtime PM.
1607 fw_devlink.strict=<bool>
1608 [KNL] Treat all inferred dependencies as mandatory
1609 dependencies. This only applies for fw_devlink=on|rpm.
1612 fw_devlink.sync_state =
1613 [KNL] When all devices that could probe have finished
1614 probing, this parameter controls what to do with
1615 devices that haven't yet received their sync_state()
1617 Format: { strict | timeout }
1618 strict -- Default. Continue waiting on consumers to
1620 timeout -- Give up waiting on consumers and call
1621 sync_state() on any devices that haven't yet
1622 received their sync_state() calls after
1623 deferred_probe_timeout has expired or by
1624 late_initcall() if !CONFIG_MODULES.
1627 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1628 support via parallel port (up to 5 devices per port)
1629 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1630 See also Documentation/input/devices/joystick-parport.rst
1634 gart_fix_e820= [X86-64] disable the fix e820 for K8 GART
1638 gather_data_sampling=
1639 [X86,INTEL] Control the Gather Data Sampling (GDS)
1642 Gather Data Sampling is a hardware vulnerability which
1643 allows unprivileged speculative access to data which was
1644 previously stored in vector registers.
1646 This issue is mitigated by default in updated microcode.
1647 The mitigation may have a performance impact but can be
1648 disabled. On systems without the microcode mitigation
1649 disabling AVX serves as a mitigation.
1651 force: Disable AVX to mitigate systems without
1652 microcode mitigation. No effect if the microcode
1653 mitigation is present. Known to cause crashes in
1654 userspace with buggy AVX enumeration.
1656 off: Disable GDS mitigation.
1658 gcov_persist= [GCOV] When non-zero (default), profiling data for
1659 kernel modules is saved and remains accessible via
1660 debugfs, even when the module is unloaded/reloaded.
1661 When zero, profiling data is discarded and associated
1662 debugfs files are removed at module unload time.
1664 goldfish [X86] Enable the goldfish android emulator platform.
1665 Don't use this when you are not running on the
1668 gpio-mockup.gpio_mockup_ranges
1669 [HW] Sets the ranges of gpiochip of for this device.
1670 Format: <start1>,<end1>,<start2>,<end2>...
1671 gpio-mockup.gpio_mockup_named_lines
1672 [HW] Let the driver know GPIO lines should be named.
1674 gpt [EFI] Forces disk with valid GPT signature but
1675 invalid Protective MBR to be treated as GPT. If the
1676 primary GPT is corrupted, it enables the backup/alternate
1677 GPT to be used instead.
1679 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1680 the "Enable 0" bit of the configuration register.
1683 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1684 the "Enable 0" bit of the configuration register.
1687 grcan.select= [HW] Select which physical interface to use.
1690 grcan.txsize= [HW] Sets the size of the tx buffer.
1691 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1693 grcan.rxsize= [HW] Sets the size of the rx buffer.
1694 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1698 [KNL] Under CONFIG_HARDENED_USERCOPY, whether
1699 hardening is enabled for this boot. Hardened
1700 usercopy checking is used to protect the kernel
1701 from reading or writing beyond known memory
1702 allocation boundaries as a proactive defense
1703 against bounds-checking flaws in the kernel's
1704 copy_to_user()/copy_from_user() interface.
1705 on Perform hardened usercopy checks (default).
1706 off Disable hardened usercopy checks.
1708 hardlockup_all_cpu_backtrace=
1709 [KNL] Should the hard-lockup detector generate
1710 backtraces on all cpus.
1713 hashdist= [KNL,NUMA] Large hashes allocated during boot
1714 are distributed across NUMA nodes. Defaults on
1715 for 64-bit NUMA, off otherwise.
1716 Format: 0 | 1 (for off | on)
1718 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1720 hd= [EIDE] (E)IDE hard drive subsystem geometry
1721 Format: <cyl>,<head>,<sect>
1724 Disable Hardware Error Source Table (HEST) support;
1725 corresponding firmware-first mode error processing
1726 logic will be disabled.
1728 hibernate= [HIBERNATION]
1729 noresume Don't check if there's a hibernation image
1730 present during boot.
1731 nocompress Don't compress/decompress hibernation images.
1732 no Disable hibernation and resume.
1733 protect_image Turn on image protection during restoration
1734 (that will set all pages holding image data
1735 during restoration read-only).
1737 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1738 size of <nn>. This works even on boxes that have no
1739 highmem otherwise. This also works to reduce highmem
1740 size on bigger boxes.
1742 highres= [KNL] Enable/disable high resolution timer mode.
1743 Valid parameters: "on", "off"
1748 hostname= [KNL] Set the hostname (aka UTS nodename).
1750 This allows setting the system's hostname during early
1751 startup. This sets the name returned by gethostname.
1752 Using this parameter to set the hostname makes it
1753 possible to ensure the hostname is correctly set before
1754 any userspace processes run, avoiding the possibility
1755 that a process may call gethostname before the hostname
1756 has been explicitly set, resulting in the calling
1757 process getting an incorrect result. The string must
1758 not exceed the maximum allowed hostname length (usually
1759 64 characters) and will be truncated otherwise.
1761 hpet= [X86-32,HPET] option to control HPET usage
1762 Format: { enable (default) | disable | force |
1764 disable: disable HPET and use PIT instead
1765 force: allow force enabled of undocumented chips (ICH4,
1767 verbose: show contents of HPET registers during setup
1769 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1770 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1772 hugepages= [HW] Number of HugeTLB pages to allocate at boot.
1773 If this follows hugepagesz (below), it specifies
1774 the number of pages of hugepagesz to be allocated.
1775 If this is the first HugeTLB parameter on the command
1776 line, it specifies the number of pages to allocate for
1777 the default huge page size. If using node format, the
1778 number of pages to allocate per-node can be specified.
1779 See also Documentation/admin-guide/mm/hugetlbpage.rst.
1780 Format: <integer> or (node format)
1781 <node>:<integer>[,<node>:<integer>]
1784 [HW] The size of the HugeTLB pages. This is used in
1785 conjunction with hugepages (above) to allocate huge
1786 pages of a specific size at boot. The pair
1787 hugepagesz=X hugepages=Y can be specified once for
1788 each supported huge page size. Huge page sizes are
1789 architecture dependent. See also
1790 Documentation/admin-guide/mm/hugetlbpage.rst.
1793 hugetlb_cma= [HW,CMA] The size of a CMA area used for allocation
1794 of gigantic hugepages. Or using node format, the size
1795 of a CMA area per node can be specified.
1796 Format: nn[KMGTPE] or (node format)
1797 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]]
1799 Reserve a CMA area of given size and allocate gigantic
1800 hugepages using the CMA allocator. If enabled, the
1801 boot-time allocation of gigantic hugepages is skipped.
1803 hugetlb_free_vmemmap=
1804 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP
1806 Control if HugeTLB Vmemmap Optimization (HVO) is enabled.
1807 Allows heavy hugetlb users to free up some more
1808 memory (7 * PAGE_SIZE for each 2MB hugetlb page).
1809 Format: { on | off (default) }
1814 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y,
1817 Note that the vmemmap pages may be allocated from the added
1818 memory block itself when memory_hotplug.memmap_on_memory is
1819 enabled, those vmemmap pages cannot be optimized even if this
1820 feature is enabled. Other vmemmap pages not allocated from
1821 the added memory block itself do not be affected.
1824 [KNL] Should the hung task detector generate panics.
1827 A value of 1 instructs the kernel to panic when a
1828 hung task is detected. The default value is controlled
1829 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time
1830 option. The value selected by this boot parameter can
1831 be changed later by the kernel.hung_task_panic sysctl.
1833 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1834 terminal devices. Valid values: 0..8
1835 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1836 If specified, z/VM IUCV HVC accepts connections
1837 from listed z/VM user IDs only.
1839 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations
1840 which allow the hypervisor to 'idle' the
1841 guest on lock contention.
1843 i2c_bus= [HW] Override the default board specific I2C bus speed
1844 or register an additional I2C bus that is not
1845 registered from board initialization code.
1849 i8042.debug [HW] Toggle i8042 debug mode
1850 i8042.unmask_kbd_data
1851 [HW] Enable printing of interrupt data from the KBD port
1852 (disabled by default, and as a pre-condition
1853 requires that i8042.debug=1 be enabled)
1854 i8042.direct [HW] Put keyboard port into non-translated mode
1855 i8042.dumbkbd [HW] Pretend that controller can only read data from
1856 keyboard and cannot control its state
1857 (Don't attempt to blink the leds)
1858 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1859 i8042.nokbd [HW] Don't check/create keyboard port
1860 i8042.noloop [HW] Disable the AUX Loopback command while probing
1862 i8042.nomux [HW] Don't check presence of an active multiplexing
1864 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1866 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1867 i8042.reset [HW] Reset the controller during init, cleanup and
1868 suspend-to-ram transitions, only during s2r
1869 transitions, or never reset
1870 Format: { 1 | Y | y | 0 | N | n }
1871 1, Y, y: always reset controller
1872 0, N, n: don't ever reset controller
1873 Default: only on s2r transitions on x86; most other
1874 architectures force reset to be always executed
1875 i8042.unlock [HW] Unlock (ignore) the keylock
1876 i8042.kbdreset [HW] Reset device connected to KBD port
1878 [HW] Allow deferred probing upon i8042 probe errors
1882 i915.invert_brightness=
1883 [DRM] Invert the sense of the variable that is used to
1884 set the brightness of the panel backlight. Normally a
1885 brightness value of 0 indicates backlight switched off,
1886 and the maximum of the brightness value sets the backlight
1887 to maximum brightness. If this parameter is set to 0
1888 (default) and the machine requires it, or this parameter
1889 is set to 1, a brightness value of 0 sets the backlight
1890 to maximum brightness, and the maximum of the brightness
1891 value switches the backlight off.
1892 -1 -- never invert brightness
1893 0 -- machine default
1894 1 -- force brightness inversion
1897 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1901 Format: idle=poll, idle=halt, idle=nomwait
1902 Poll forces a polling idle loop that can slightly
1903 improve the performance of waking up a idle CPU, but
1904 will use a lot of power and make the system run hot.
1906 idle=halt: Halt is forced to be used for CPU idle.
1907 In such case C2/C3 won't be used again.
1908 idle=nomwait: Disable mwait for CPU C-states
1912 Allow force disabling of Shared Virtual Memory (SVA)
1913 support for the idxd driver. By default it is set to
1916 idxd.tc_override= [HW]
1918 Allow override of default traffic class configuration
1919 for the device. By default it is set to false (0).
1921 ieee754= [MIPS] Select IEEE Std 754 conformance mode
1922 Format: { strict | legacy | 2008 | relaxed }
1925 Choose which programs will be accepted for execution
1926 based on the IEEE 754 NaN encoding(s) supported by
1927 the FPU and the NaN encoding requested with the value
1928 of an ELF file header flag individually set by each
1929 binary. Hardware implementations are permitted to
1930 support either or both of the legacy and the 2008 NaN
1933 Available settings are as follows:
1934 strict accept binaries that request a NaN encoding
1935 supported by the FPU
1936 legacy only accept legacy-NaN binaries, if supported
1938 2008 only accept 2008-NaN binaries, if supported
1940 relaxed accept any binaries regardless of whether
1941 supported by the FPU
1943 The FPU emulator is always able to support both NaN
1944 encodings, so if no FPU hardware is present or it has
1945 been disabled with 'nofpu', then the settings of
1946 'legacy' and '2008' strap the emulator accordingly,
1947 'relaxed' straps the emulator for both legacy-NaN and
1948 2008-NaN, whereas 'strict' enables legacy-NaN only on
1949 legacy processors and both NaN encodings on MIPS32 or
1952 The setting for ABS.fmt/NEG.fmt instruction execution
1953 mode generally follows that for the NaN encoding,
1954 except where unsupported by hardware.
1956 ignore_loglevel [KNL]
1957 Ignore loglevel setting - this will print /all/
1958 kernel messages to the console. Useful for debugging.
1959 We also add it as printk module parameter, so users
1960 could change it dynamically, usually by
1961 /sys/module/printk/parameters/ignore_loglevel.
1964 Ignore RLIMIT_DATA setting for data mappings,
1965 print warning at first misuse. Can be changed via
1966 /sys/module/kernel/parameters/ignore_rlimit_data.
1968 ihash_entries= [KNL]
1969 Set number of hash buckets for inode cache.
1971 ima_appraise= [IMA] appraise integrity measurements
1972 Format: { "off" | "enforce" | "fix" | "log" }
1975 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead.
1976 The builtin appraise policy appraises all files
1979 ima_canonical_fmt [IMA]
1980 Use the canonical format for the binary runtime
1981 measurements, instead of host native format.
1984 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1988 The list of supported hash algorithms is defined
1989 in crypto/hash_info.h.
1992 The builtin policies to load during IMA setup.
1993 Format: "tcb | appraise_tcb | secure_boot |
1994 fail_securely | critical_data"
1996 The "tcb" policy measures all programs exec'd, files
1997 mmap'd for exec, and all files opened with the read
1998 mode bit set by either the effective uid (euid=0) or
2001 The "appraise_tcb" policy appraises the integrity of
2002 all files owned by root.
2004 The "secure_boot" policy appraises the integrity
2005 of files (eg. kexec kernel image, kernel modules,
2006 firmware, policy, etc) based on file signatures.
2008 The "fail_securely" policy forces file signature
2009 verification failure also on privileged mounted
2010 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE
2013 The "critical_data" policy measures kernel integrity
2016 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
2017 Load a policy which meets the needs of the Trusted
2018 Computing Base. This means IMA will measure all
2019 programs exec'd, files mmap'd for exec, and all files
2020 opened for read by uid=0.
2023 Select one of defined IMA measurements template formats.
2024 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" |
2029 [IMA] Define a custom template format.
2030 Format: { "field1|...|fieldN" }
2032 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
2033 Format: <min_file_size>
2034 Set the minimal file size for using asynchronous hash.
2035 If left unspecified, ahash usage is disabled.
2037 ahash performance varies for different data sizes on
2038 different crypto accelerators. This option can be used
2039 to achieve the best performance for a particular HW.
2041 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
2043 Set hashing buffer size. Default: 4k.
2045 ahash performance varies for different chunk sizes on
2046 different crypto accelerators. This option can be used
2047 to achieve best performance for particular HW.
2051 Run specified binary instead of /sbin/init as init
2054 initcall_debug [KNL] Trace initcalls as they are executed. Useful
2055 for working out where the kernel is dying during
2058 initcall_blacklist= [KNL] Do not execute a comma-separated list of
2059 initcall functions. Useful for debugging built-in
2060 modules and initcalls.
2062 initramfs_async= [KNL]
2065 This parameter controls whether the initramfs
2066 image is unpacked asynchronously, concurrently
2067 with devices being probed and
2068 initialized. This should normally just work,
2069 but as a debugging aid, one can get the
2070 historical behaviour of the initramfs
2071 unpacking being completed before device_ and
2074 initrd= [BOOT] Specify the location of the initial ramdisk
2076 initrdmem= [KNL] Specify a physical address and size from which to
2077 load the initrd. If an initrd is compiled in or
2078 specified in the bootparams, it takes priority over this
2080 Format: ss[KMG],nn[KMG]
2083 init_on_alloc= [MM] Fill newly allocated pages and heap objects with
2086 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON.
2088 init_on_free= [MM] Fill freed pages and heap objects with zeroes.
2090 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON.
2092 init_pkru= [X86] Specify the default memory protection keys rights
2093 register contents for all processes. 0x55555554 by
2094 default (disallow access to all but pkey 0). Can
2095 override in debugfs after boot.
2097 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
2100 int_pln_enable [X86] Enable power limit notification interrupt
2102 integrity_audit=[IMA]
2103 Format: { "0" | "1" }
2104 0 -- basic integrity auditing messages. (Default)
2105 1 -- additional integrity auditing messages.
2107 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
2109 Enable intel iommu driver.
2111 Disable intel iommu driver.
2112 igfx_off [Default Off]
2113 By default, gfx is mapped as normal device. If a gfx
2114 device has a dedicated DMAR unit, the DMAR unit is
2115 bypassed by not enabling DMAR with this option. In
2116 this case, gfx device will use physical address for
2118 strict [Default Off]
2119 Deprecated, equivalent to iommu.strict=1.
2120 sp_off [Default Off]
2121 By default, super page will be supported if Intel IOMMU
2122 has the capability. With this option, super page will
2125 Enable the Intel IOMMU scalable mode if the hardware
2126 advertises that it has support for the scalable mode
2129 Disallow use of the Intel IOMMU scalable mode.
2130 tboot_noforce [Default Off]
2131 Do not force the Intel IOMMU enabled under tboot.
2132 By default, tboot will force Intel IOMMU on, which
2133 could harm performance of some high-throughput
2134 devices like 40GBit network cards, even if identity
2136 Note that using this option lowers the security
2137 provided by tboot because it makes the system
2138 vulnerable to DMA attacks.
2140 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
2141 0 disables intel_idle and fall back on acpi_idle.
2142 1 to 9 specify maximum depth of C-state.
2146 Do not enable intel_pstate as the default
2147 scaling driver for the supported processors
2149 Use intel_pstate driver to bypass the scaling
2150 governors layer of cpufreq and provides it own
2151 algorithms for p-state selection. There are two
2152 P-state selection algorithms provided by
2153 intel_pstate in the active mode: powersave and
2154 performance. The way they both operate depends
2155 on whether or not the hardware managed P-states
2156 (HWP) feature has been enabled in the processor
2157 and possibly on the processor model.
2159 Use intel_pstate as a scaling driver, but configure it
2160 to work with generic cpufreq governors (instead of
2161 enabling its internal governor). This mode cannot be
2162 used along with the hardware-managed P-states (HWP)
2165 Enable intel_pstate on systems that prohibit it by default
2166 in favor of acpi-cpufreq. Forcing the intel_pstate driver
2167 instead of acpi-cpufreq may disable platform features, such
2168 as thermal controls and power capping, that rely on ACPI
2169 P-States information being indicated to OSPM and therefore
2170 should be used with caution. This option does not work with
2171 processors that aren't supported by the intel_pstate driver
2172 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
2174 Do not enable hardware P state control (HWP)
2177 Only load intel_pstate on systems which support
2178 hardware P state control (HWP) if available.
2180 Enforce ACPI _PPC performance limits. If the Fixed ACPI
2181 Description Table, specifies preferred power management
2182 profile as "Enterprise Server" or "Performance Server",
2183 then this feature is turned on by default.
2185 Allow per-logical-CPU P-State performance control limits using
2186 cpufreq sysfs interface
2188 intremap= [X86-64, Intel-IOMMU]
2189 on enable Interrupt Remapping (default)
2190 off disable Interrupt Remapping
2191 nosid disable Source ID checking
2193 BIOS x2APIC opt-out request will be ignored
2194 nopost disable Interrupt Posting
2196 iomem= Disable strict checking of access to MMIO memory
2197 strict regions from userspace.
2212 nobypass [PPC/POWERNV]
2213 Disable IOMMU bypass, using IOMMU for PCI devices.
2215 iommu.forcedac= [ARM64, X86] Control IOVA allocation for PCI devices.
2216 Format: { "0" | "1" }
2217 0 - Try to allocate a 32-bit DMA address first, before
2218 falling back to the full range if needed.
2219 1 - Allocate directly from the full usable range,
2220 forcing Dual Address Cycle for PCI cards supporting
2221 greater than 32-bit addressing.
2223 iommu.strict= [ARM64, X86] Configure TLB invalidation behaviour
2224 Format: { "0" | "1" }
2226 Request that DMA unmap operations use deferred
2227 invalidation of hardware TLBs, for increased
2228 throughput at the cost of reduced device isolation.
2229 Will fall back to strict mode if not supported by
2230 the relevant IOMMU driver.
2232 DMA unmap operations invalidate IOMMU hardware TLBs
2234 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}.
2235 Note: on x86, strict mode specified via one of the
2236 legacy driver-specific options takes precedence.
2239 [ARM64, X86] Configure DMA to bypass the IOMMU by default.
2240 Format: { "0" | "1" }
2241 0 - Use IOMMU translation for DMA.
2242 1 - Bypass the IOMMU for DMA.
2243 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH.
2245 io7= [HW] IO7 for Marvel-based Alpha systems
2246 See comment before marvel_specify_io7 in
2247 arch/alpha/kernel/core_marvel.c.
2249 io_delay= [X86] I/O delay method
2251 Standard port 0x80 based delay
2253 Alternate port 0xed based delay (needed on some systems)
2255 Simple two microseconds delay
2260 See Documentation/admin-guide/nfs/nfsroot.rst.
2262 ipcmni_extend [KNL] Extend the maximum number of unique System V
2263 IPC identifiers from 32,768 to 16,777,216.
2265 irqaffinity= [SMP] Set the default irq affinity mask
2266 The argument is a cpu list, as described above.
2268 irqchip.gicv2_force_probe=
2271 Force the kernel to look for the second 4kB page
2272 of a GICv2 controller even if the memory range
2273 exposed by the device tree is too small.
2275 irqchip.gicv3_nolpi=
2277 Force the kernel to ignore the availability of
2278 LPIs (and by consequence ITSs). Intended for system
2279 that use the kernel as a bootloader, and thus want
2280 to let secondary kernels in charge of setting up
2283 irqchip.gicv3_pseudo_nmi= [ARM64]
2284 Enables support for pseudo-NMIs in the kernel. This
2285 requires the kernel to be built with
2286 CONFIG_ARM64_PSEUDO_NMI.
2289 When an interrupt is not handled search all handlers
2290 for it. Intended to get systems with badly broken
2294 When an interrupt is not handled search all handlers
2295 for it. Also check all handlers each timer
2296 interrupt. Intended to get systems with badly broken
2300 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
2302 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance.
2303 [Deprecated - use cpusets instead]
2304 Format: [flag-list,]<cpu-list>
2306 Specify one or more CPUs to isolate from disturbances
2307 specified in the flag list (default: domain):
2310 Disable the tick when a single task runs.
2312 A residual 1Hz tick is offloaded to workqueues, which you
2313 need to affine to housekeeping through the global
2314 workqueue's affinity configured via the
2315 /sys/devices/virtual/workqueue/cpumask sysfs file, or
2316 by using the 'domain' flag described below.
2318 NOTE: by default the global workqueue runs on all CPUs,
2319 so to protect individual CPUs the 'cpumask' file has to
2320 be configured manually after bootup.
2323 Isolate from the general SMP balancing and scheduling
2324 algorithms. Note that performing domain isolation this way
2325 is irreversible: it's not possible to bring back a CPU to
2326 the domains once isolated through isolcpus. It's strongly
2327 advised to use cpusets instead to disable scheduler load
2328 balancing through the "cpuset.sched_load_balance" file.
2329 It offers a much more flexible interface where CPUs can
2330 move in and out of an isolated set anytime.
2332 You can move a process onto or off an "isolated" CPU via
2333 the CPU affinity syscalls or cpuset.
2334 <cpu number> begins at 0 and the maximum value is
2335 "number of CPUs in system - 1".
2339 Isolate from being targeted by managed interrupts
2340 which have an interrupt mask containing isolated
2341 CPUs. The affinity of managed interrupts is
2342 handled by the kernel and cannot be changed via
2343 the /proc/irq/* interfaces.
2345 This isolation is best effort and only effective
2346 if the automatically assigned interrupt mask of a
2347 device queue contains isolated and housekeeping
2348 CPUs. If housekeeping CPUs are online then such
2349 interrupts are directed to the housekeeping CPU
2350 so that IO submitted on the housekeeping CPU
2351 cannot disturb the isolated CPU.
2353 If a queue's affinity mask contains only isolated
2354 CPUs then this parameter has no effect on the
2355 interrupt routing decision, though interrupts are
2356 only delivered when tasks running on those
2357 isolated CPUs submit IO. IO submitted on
2358 housekeeping CPUs has no influence on those
2361 The format of <cpu-list> is described above.
2365 ivrs_ioapic [HW,X86-64]
2366 Provide an override to the IOAPIC-ID<->DEVICE-ID
2367 mapping provided in the IVRS ACPI table.
2368 By default, PCI segment is 0, and can be omitted.
2370 For example, to map IOAPIC-ID decimal 10 to
2371 PCI segment 0x1 and PCI device 00:14.0,
2372 write the parameter as:
2373 ivrs_ioapic=10@0001:00:14.0
2376 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0
2377 write the parameter as:
2378 ivrs_ioapic[10]=00:14.0
2379 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and
2380 PCI device 00:14.0 write the parameter as:
2381 ivrs_ioapic[10]=0001:00:14.0
2383 ivrs_hpet [HW,X86-64]
2384 Provide an override to the HPET-ID<->DEVICE-ID
2385 mapping provided in the IVRS ACPI table.
2386 By default, PCI segment is 0, and can be omitted.
2388 For example, to map HPET-ID decimal 10 to
2389 PCI segment 0x1 and PCI device 00:14.0,
2390 write the parameter as:
2391 ivrs_hpet=10@0001:00:14.0
2394 * To map HPET-ID decimal 0 to PCI device 00:14.0
2395 write the parameter as:
2396 ivrs_hpet[0]=00:14.0
2397 * To map HPET-ID decimal 10 to PCI segment 0x1 and
2398 PCI device 00:14.0 write the parameter as:
2399 ivrs_ioapic[10]=0001:00:14.0
2401 ivrs_acpihid [HW,X86-64]
2402 Provide an override to the ACPI-HID:UID<->DEVICE-ID
2403 mapping provided in the IVRS ACPI table.
2404 By default, PCI segment is 0, and can be omitted.
2406 For example, to map UART-HID:UID AMD0020:0 to
2407 PCI segment 0x1 and PCI device ID 00:14.5,
2408 write the parameter as:
2409 ivrs_acpihid=AMD0020:0@0001:00:14.5
2412 * To map UART-HID:UID AMD0020:0 to PCI segment is 0,
2413 PCI device ID 00:14.5, write the parameter as:
2414 ivrs_acpihid[00:14.5]=AMD0020:0
2415 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and
2416 PCI device ID 00:14.5, write the parameter as:
2417 ivrs_acpihid[0001:00:14.5]=AMD0020:0
2419 js= [HW,JOY] Analog joystick
2420 See Documentation/input/joydev/joystick.rst.
2423 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print
2424 report on every invalid memory access. Without this
2425 parameter KASAN will print report only for the first
2429 Do not unregister boot console at start. This is only
2430 useful for debugging when something happens in the window
2431 between unregistering the boot console and initializing
2436 kernelcore= [KNL,X86,IA-64,PPC]
2437 Format: nn[KMGTPE] | nn% | "mirror"
2438 This parameter specifies the amount of memory usable by
2439 the kernel for non-movable allocations. The requested
2440 amount is spread evenly throughout all nodes in the
2441 system as ZONE_NORMAL. The remaining memory is used for
2442 movable memory in its own zone, ZONE_MOVABLE. In the
2443 event, a node is too small to have both ZONE_NORMAL and
2444 ZONE_MOVABLE, kernelcore memory will take priority and
2445 other nodes will have a larger ZONE_MOVABLE.
2447 ZONE_MOVABLE is used for the allocation of pages that
2448 may be reclaimed or moved by the page migration
2449 subsystem. Note that allocations like PTEs-from-HighMem
2450 still use the HighMem zone if it exists, and the Normal
2451 zone if it does not.
2453 It is possible to specify the exact amount of memory in
2454 the form of "nn[KMGTPE]", a percentage of total system
2455 memory in the form of "nn%", or "mirror". If "mirror"
2456 option is specified, mirrored (reliable) memory is used
2457 for non-movable allocations and remaining memory is used
2458 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror"
2459 are exclusive, so you cannot specify multiple forms.
2461 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
2462 Format: <Controller#>[,poll interval]
2463 The controller # is the number of the ehci usb debug
2464 port as it is probed via PCI. The poll interval is
2465 optional and is the number seconds in between
2466 each poll cycle to the debug port in case you need
2467 the functionality for interrupting the kernel with
2468 gdb or control-c on the dbgp connection. When
2469 not using this parameter you use sysrq-g to break into
2470 the kernel debugger.
2472 kgdboc= [KGDB,HW] kgdb over consoles.
2473 Requires a tty driver that supports console polling,
2474 or a supported polling keyboard driver (non-usb).
2475 Serial only format: <serial_device>[,baud]
2476 keyboard only format: kbd
2477 keyboard and serial format: kbd,<serial_device>[,baud]
2478 Optional Kernel mode setting:
2479 kms, kbd format: kms,kbd
2480 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
2482 kgdboc_earlycon= [KGDB,HW]
2483 If the boot console provides the ability to read
2484 characters and can work in polling mode, you can use
2485 this parameter to tell kgdb to use it as a backend
2486 until the normal console is registered. Intended to
2487 be used together with the kgdboc parameter which
2488 specifies the normal console to transition to.
2490 The name of the early console should be specified
2491 as the value of this parameter. Note that the name of
2492 the early console might be different than the tty
2493 name passed to kgdboc. It's OK to leave the value
2494 blank and the first boot console that implements
2495 read() will be picked.
2497 kgdbwait [KGDB] Stop kernel execution and enter the
2498 kernel debugger at the earliest opportunity.
2500 kmac= [MIPS] Korina ethernet MAC address.
2501 Configure the RouterBoard 532 series on-chip
2502 Ethernet adapter MAC address.
2504 kmemleak= [KNL] Boot-time kmemleak enable/disable
2505 Valid arguments: on, off
2507 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
2510 kprobe_event=[probe-list]
2511 [FTRACE] Add kprobe events and enable at boot time.
2512 The probe-list is a semicolon delimited list of probe
2513 definitions. Each definition is same as kprobe_events
2514 interface, but the parameters are comma delimited.
2515 For example, to add a kprobe event on vfs_read with
2516 arg1 and arg2, add to the command line;
2518 kprobe_event=p,vfs_read,$arg1,$arg2
2520 See also Documentation/trace/kprobetrace.rst "Kernel
2521 Boot Parameter" section.
2523 kpti= [ARM64] Control page table isolation of user
2524 and kernel address spaces.
2525 Default: enabled on cores which need mitigation.
2529 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires
2530 CONFIG_KUNIT to be set to be fully enabled. The
2531 default value can be overridden via
2532 KUNIT_DEFAULT_ENABLED.
2533 Default is 1 (enabled)
2535 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
2536 Default is 0 (don't ignore, but inject #GP)
2538 kvm.eager_page_split=
2539 [KVM,X86] Controls whether or not KVM will try to
2540 proactively split all huge pages during dirty logging.
2541 Eager page splitting reduces interruptions to vCPU
2542 execution by eliminating the write-protection faults
2543 and MMU lock contention that would otherwise be
2544 required to split huge pages lazily.
2546 VM workloads that rarely perform writes or that write
2547 only to a small region of VM memory may benefit from
2548 disabling eager page splitting to allow huge pages to
2549 still be used for reads.
2551 The behavior of eager page splitting depends on whether
2552 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If
2553 disabled, all huge pages in a memslot will be eagerly
2554 split when dirty logging is enabled on that memslot. If
2555 enabled, eager page splitting will be performed during
2556 the KVM_CLEAR_DIRTY ioctl, and only for the pages being
2559 Eager page splitting is only supported when kvm.tdp_mmu=Y.
2563 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface.
2564 Default is false (don't support).
2567 [KVM] Controls the software workaround for the
2568 X86_BUG_ITLB_MULTIHIT bug.
2569 force : Always deploy workaround.
2570 off : Never deploy workaround.
2571 auto : Deploy workaround based on the presence of
2572 X86_BUG_ITLB_MULTIHIT.
2576 If the software workaround is enabled for the host,
2577 guests do need not to enable it for nested guests.
2579 kvm.nx_huge_pages_recovery_ratio=
2580 [KVM] Controls how many 4KiB pages are periodically zapped
2581 back to huge pages. 0 disables the recovery, otherwise if
2582 the value is N KVM will zap 1/Nth of the 4KiB pages every
2583 period (see below). The default is 60.
2585 kvm.nx_huge_pages_recovery_period_ms=
2586 [KVM] Controls the time period at which KVM zaps 4KiB pages
2587 back to huge pages. If the value is a non-zero N, KVM will
2588 zap a portion (see ratio above) of the pages every N msecs.
2589 If the value is 0 (the default), KVM will pick a period based
2590 on the ratio, such that a page is zapped after 1 hour on average.
2592 kvm-amd.nested= [KVM,AMD] Control nested virtualization feature in
2593 KVM/SVM. Default is 1 (enabled).
2595 kvm-amd.npt= [KVM,AMD] Control KVM's use of Nested Page Tables,
2596 a.k.a. Two-Dimensional Page Tables. Default is 1
2597 (enabled). Disable by KVM if hardware lacks support
2601 [KVM,ARM] Select one of KVM/arm64's modes of operation.
2603 none: Forcefully disable KVM.
2605 nvhe: Standard nVHE-based mode, without support for
2608 protected: nVHE-based mode with support for guests whose
2609 state is kept private from the host.
2611 nested: VHE-based mode with support for nested
2612 virtualization. Requires at least ARMv8.3
2615 Defaults to VHE/nVHE based on hardware support. Setting
2616 mode to "protected" will disable kexec and hibernation
2617 for the host. "nested" is experimental and should be
2618 used with extreme caution.
2620 kvm-arm.vgic_v3_group0_trap=
2621 [KVM,ARM] Trap guest accesses to GICv3 group-0
2624 kvm-arm.vgic_v3_group1_trap=
2625 [KVM,ARM] Trap guest accesses to GICv3 group-1
2628 kvm-arm.vgic_v3_common_trap=
2629 [KVM,ARM] Trap guest accesses to GICv3 common
2632 kvm-arm.vgic_v4_enable=
2633 [KVM,ARM] Allow use of GICv4 for direct injection of
2636 kvm_cma_resv_ratio=n [PPC]
2637 Reserves given percentage from system memory area for
2638 contiguous memory allocation for KVM hash pagetable
2640 By default it reserves 5% of total system memory.
2644 kvm-intel.ept= [KVM,Intel] Control KVM's use of Extended Page Tables,
2645 a.k.a. Two-Dimensional Page Tables. Default is 1
2646 (enabled). Disable by KVM if hardware lacks support
2649 kvm-intel.emulate_invalid_guest_state=
2650 [KVM,Intel] Control whether to emulate invalid guest
2651 state. Ignored if kvm-intel.enable_unrestricted_guest=1,
2652 as guest state is never invalid for unrestricted
2653 guests. This param doesn't apply to nested guests (L2),
2654 as KVM never emulates invalid L2 guest state.
2655 Default is 1 (enabled).
2657 kvm-intel.flexpriority=
2658 [KVM,Intel] Control KVM's use of FlexPriority feature
2659 (TPR shadow). Default is 1 (enabled). Disable by KVM if
2660 hardware lacks support for it.
2663 [KVM,Intel] Control nested virtualization feature in
2664 KVM/VMX. Default is 1 (enabled).
2666 kvm-intel.unrestricted_guest=
2667 [KVM,Intel] Control KVM's use of unrestricted guest
2668 feature (virtualized real and unpaged mode). Default
2669 is 1 (enabled). Disable by KVM if EPT is disabled or
2670 hardware lacks support for it.
2672 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault
2675 Valid arguments: never, cond, always
2677 always: L1D cache flush on every VMENTER.
2678 cond: Flush L1D on VMENTER only when the code between
2679 VMEXIT and VMENTER can leak host memory.
2680 never: Disables the mitigation
2682 Default is cond (do L1 cache flush in specific instances)
2684 kvm-intel.vpid= [KVM,Intel] Control KVM's use of Virtual Processor
2685 Identification feature (tagged TLBs). Default is 1
2686 (enabled). Disable by KVM if hardware lacks support
2689 l1d_flush= [X86,INTEL]
2690 Control mitigation for L1D based snooping vulnerability.
2692 Certain CPUs are vulnerable to an exploit against CPU
2693 internal buffers which can forward information to a
2694 disclosure gadget under certain conditions.
2696 In vulnerable processors, the speculatively
2697 forwarded data can be used in a cache side channel
2698 attack, to access data to which the attacker does
2699 not have direct access.
2701 This parameter controls the mitigation. The
2704 on - enable the interface for the mitigation
2706 l1tf= [X86] Control mitigation of the L1TF vulnerability on
2709 The kernel PTE inversion protection is unconditionally
2710 enabled and cannot be disabled.
2713 Provides all available mitigations for the
2714 L1TF vulnerability. Disables SMT and
2715 enables all mitigations in the
2716 hypervisors, i.e. unconditional L1D flush.
2718 SMT control and L1D flush control via the
2719 sysfs interface is still possible after
2720 boot. Hypervisors will issue a warning
2721 when the first VM is started in a
2722 potentially insecure configuration,
2723 i.e. SMT enabled or L1D flush disabled.
2726 Same as 'full', but disables SMT and L1D
2727 flush runtime control. Implies the
2728 'nosmt=force' command line option.
2729 (i.e. sysfs control of SMT is disabled.)
2732 Leaves SMT enabled and enables the default
2733 hypervisor mitigation, i.e. conditional
2736 SMT control and L1D flush control via the
2737 sysfs interface is still possible after
2738 boot. Hypervisors will issue a warning
2739 when the first VM is started in a
2740 potentially insecure configuration,
2741 i.e. SMT enabled or L1D flush disabled.
2745 Disables SMT and enables the default
2746 hypervisor mitigation.
2748 SMT control and L1D flush control via the
2749 sysfs interface is still possible after
2750 boot. Hypervisors will issue a warning
2751 when the first VM is started in a
2752 potentially insecure configuration,
2753 i.e. SMT enabled or L1D flush disabled.
2756 Same as 'flush', but hypervisors will not
2757 warn when a VM is started in a potentially
2758 insecure configuration.
2761 Disables hypervisor mitigations and doesn't
2763 It also drops the swap size and available
2764 RAM limit restriction on both hypervisor and
2769 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst
2775 lapic [X86-32,APIC] Enable the local APIC even if BIOS
2778 lapic= [X86,APIC] Do not use TSC deadline
2779 value for LAPIC timer one-shot implementation. Default
2780 back to the programmable timer unit in the LAPIC.
2781 Format: notscdeadline
2783 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
2786 libata.dma= [LIBATA] DMA control
2787 libata.dma=0 Disable all PATA and SATA DMA
2788 libata.dma=1 PATA and SATA Disk DMA only
2789 libata.dma=2 ATAPI (CDROM) DMA only
2790 libata.dma=4 Compact Flash DMA only
2791 Combinations also work, so libata.dma=3 enables DMA
2792 for disks and CDROMs, but not CFs.
2794 libata.ignore_hpa= [LIBATA] Ignore HPA limit
2795 libata.ignore_hpa=0 keep BIOS limits (default)
2796 libata.ignore_hpa=1 ignore limits, using full disk
2798 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
2802 libata.force= [LIBATA] Force configurations. The format is a comma-
2803 separated list of "[ID:]VAL" where ID is PORT[.DEVICE].
2804 PORT and DEVICE are decimal numbers matching port, link
2805 or device. Basically, it matches the ATA ID string
2806 printed on console by libata. If the whole ID part is
2807 omitted, the last PORT and DEVICE values are used. If
2808 ID hasn't been specified yet, the configuration applies
2809 to all ports, links and devices.
2811 If only DEVICE is omitted, the parameter applies to
2812 the port and all links and devices behind it. DEVICE
2813 number of 0 either selects the first device or the
2814 first fan-out link behind PMP device. It does not
2815 select the host link. DEVICE number of 15 selects the
2816 host link and device attached to it.
2818 The VAL specifies the configuration to force. As long
2819 as there is no ambiguity, shortcut notation is allowed.
2820 For example, both 1.5 and 1.5G would work for 1.5Gbps.
2821 The following configurations can be forced.
2823 * Cable type: 40c, 80c, short40c, unk, ign or sata.
2824 Any ID with matching PORT is used.
2826 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
2828 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
2829 udma[/][16,25,33,44,66,100,133] notation is also
2832 * nohrst, nosrst, norst: suppress hard, soft and both
2835 * rstonce: only attempt one reset during hot-unplug
2838 * [no]dbdelay: Enable or disable the extra 200ms delay
2839 before debouncing a link PHY and device presence
2842 * [no]ncq: Turn on or off NCQ.
2844 * [no]ncqtrim: Enable or disable queued DSM TRIM.
2846 * [no]ncqati: Enable or disable NCQ trim on ATI chipset.
2848 * [no]trim: Enable or disable (unqueued) TRIM.
2850 * trim_zero: Indicate that TRIM command zeroes data.
2852 * max_trim_128m: Set 128M maximum trim size limit.
2854 * [no]dma: Turn on or off DMA transfers.
2856 * atapi_dmadir: Enable ATAPI DMADIR bridge support.
2858 * atapi_mod16_dma: Enable the use of ATAPI DMA for
2859 commands that are not a multiple of 16 bytes.
2861 * [no]dmalog: Enable or disable the use of the
2862 READ LOG DMA EXT command to access logs.
2864 * [no]iddevlog: Enable or disable access to the
2865 identify device data log.
2867 * [no]logdir: Enable or disable access to the general
2868 purpose log directory.
2870 * max_sec_128: Set transfer size limit to 128 sectors.
2872 * max_sec_1024: Set or clear transfer size limit to
2875 * max_sec_lba48: Set or clear transfer size limit to
2878 * [no]lpm: Enable or disable link power management.
2880 * [no]setxfer: Indicate if transfer speed mode setting
2883 * [no]fua: Disable or enable FUA (Force Unit Access)
2884 support for devices supporting this feature.
2886 * dump_id: Dump IDENTIFY data.
2888 * disable: Disable this device.
2890 If there are multiple matching configurations changing
2891 the same attribute, the last one is used.
2893 load_ramdisk= [RAM] [Deprecated]
2895 lockd.nlm_grace_period=P [NFS] Assign grace period.
2898 lockd.nlm_tcpport=N [NFS] Assign TCP port.
2901 lockd.nlm_timeout=T [NFS] Assign timeout value.
2904 lockd.nlm_udpport=M [NFS] Assign UDP port.
2907 lockdown= [SECURITY]
2908 { integrity | confidentiality }
2909 Enable the kernel lockdown feature. If set to
2910 integrity, kernel features that allow userland to
2911 modify the running kernel are disabled. If set to
2912 confidentiality, kernel features that allow userland
2913 to extract confidential information from the kernel
2916 locktorture.nreaders_stress= [KNL]
2917 Set the number of locking read-acquisition kthreads.
2918 Defaults to being automatically set based on the
2919 number of online CPUs.
2921 locktorture.nwriters_stress= [KNL]
2922 Set the number of locking write-acquisition kthreads.
2924 locktorture.onoff_holdoff= [KNL]
2925 Set time (s) after boot for CPU-hotplug testing.
2927 locktorture.onoff_interval= [KNL]
2928 Set time (s) between CPU-hotplug operations, or
2929 zero to disable CPU-hotplug testing.
2931 locktorture.shuffle_interval= [KNL]
2932 Set task-shuffle interval (jiffies). Shuffling
2933 tasks allows some CPUs to go into dyntick-idle
2934 mode during the locktorture test.
2936 locktorture.shutdown_secs= [KNL]
2937 Set time (s) after boot system shutdown. This
2938 is useful for hands-off automated testing.
2940 locktorture.stat_interval= [KNL]
2941 Time (s) between statistics printk()s.
2943 locktorture.stutter= [KNL]
2944 Time (s) to stutter testing, for example,
2945 specifying five seconds causes the test to run for
2946 five seconds, wait for five seconds, and so on.
2947 This tests the locking primitive's ability to
2948 transition abruptly to and from idle.
2950 locktorture.torture_type= [KNL]
2951 Specify the locking implementation to test.
2953 locktorture.writer_fifo= [KNL]
2954 Run the write-side locktorture kthreads at
2955 sched_set_fifo() real-time priority.
2957 locktorture.verbose= [KNL]
2958 Enable additional printk() statements.
2960 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
2963 loglevel= All Kernel Messages with a loglevel smaller than the
2964 console loglevel will be printed to the console. It can
2965 also be changed with klogd or other programs. The
2966 loglevels are defined as follows:
2968 0 (KERN_EMERG) system is unusable
2969 1 (KERN_ALERT) action must be taken immediately
2970 2 (KERN_CRIT) critical conditions
2971 3 (KERN_ERR) error conditions
2972 4 (KERN_WARNING) warning conditions
2973 5 (KERN_NOTICE) normal but significant condition
2974 6 (KERN_INFO) informational
2975 7 (KERN_DEBUG) debug-level messages
2977 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
2978 in bytes. n must be a power of two and greater
2979 than the minimal size. The minimal size is defined
2980 by LOG_BUF_SHIFT kernel config parameter. There is
2981 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
2982 that allows to increase the default size depending on
2983 the number of CPUs. See init/Kconfig for more details.
2985 logo.nologo [FB] Disables display of the built-in Linux logo.
2986 This may be used to provide more screen space for
2987 kernel log messages and is useful when debugging
2988 kernel boot problems.
2990 lp=0 [LP] Specify parallel ports to use, e.g,
2991 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
2992 lp=reset first parallel port). 'lp=0' disables the
2993 lp=auto printer driver. 'lp=reset' (which can be
2994 specified in addition to the ports) causes
2995 attached printers to be reset. Using
2996 lp=port1,port2,... specifies the parallel ports
2997 to associate lp devices with, starting with
2998 lp0. A port specification may be 'none' to skip
2999 that lp device, or a parport name such as
3000 'parport0'. Specifying 'lp=auto' instead of a
3001 port specification list means that device IDs
3002 from each port should be examined, to see if
3003 an IEEE 1284-compliant printer is attached; if
3004 so, the driver will manage that printer.
3005 See also header of drivers/char/lp.c.
3008 Sets loops_per_jiffy to given constant, thus avoiding
3009 time-consuming boot-time autodetection (up to 250 ms per
3010 CPU). 0 enables autodetection (default). To determine
3011 the correct value for your kernel, boot with normal
3012 autodetection and see what value is printed. Note that
3013 on SMP systems the preset will be applied to all CPUs,
3014 which is likely to cause problems if your CPUs need
3015 significantly divergent settings. An incorrect value
3016 will cause delays in the kernel to be wrong, leading to
3017 unpredictable I/O errors and other breakage. Although
3018 unlikely, in the extreme case this might damage your
3022 Format: <io>,<irq>,<dma>
3024 lsm.debug [SECURITY] Enable LSM initialization debugging output.
3027 [SECURITY] Choose order of LSM initialization. This
3028 overrides CONFIG_LSM, and the "security=" parameter.
3030 machvec= [IA-64] Force the use of a particular machine-vector
3031 (machvec) in a generic kernel.
3032 Example: machvec=hpzx1
3034 machtype= [Loongson] Share the same kernel image file between
3035 different yeeloong laptops.
3036 Example: machtype=lemote-yeeloong-2f-7inch
3038 max_addr=nn[KMG] [KNL,BOOT,IA-64] All physical memory greater
3039 than or equal to this physical address is ignored.
3041 maxcpus= [SMP] Maximum number of processors that an SMP kernel
3042 will bring up during bootup. maxcpus=n : n >= 0 limits
3043 the kernel to bring up 'n' processors. Surely after
3044 bootup you can bring up the other plugged cpu by executing
3045 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus
3046 only takes effect during system bootup.
3047 While n=0 is a special case, it is equivalent to "nosmp",
3048 which also disables the IO APIC.
3050 max_loop= [LOOP] The number of loop block devices that get
3051 (loop.max_loop) unconditionally pre-created at init time. The default
3052 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
3053 of statically allocating a predefined number, loop
3054 devices can be requested on-demand with the
3055 /dev/loop-control interface.
3057 mce [X86-32] Machine Check Exception
3059 mce=option [X86-64] See Documentation/arch/x86/x86_64/boot-options.rst
3061 md= [HW] RAID subsystems devices and level
3062 See Documentation/admin-guide/md.rst.
3065 Format: <first>,<last>
3066 Specifies range of consoles to be captured by the MDA.
3069 Control mitigation for the Micro-architectural Data
3070 Sampling (MDS) vulnerability.
3072 Certain CPUs are vulnerable to an exploit against CPU
3073 internal buffers which can forward information to a
3074 disclosure gadget under certain conditions.
3076 In vulnerable processors, the speculatively
3077 forwarded data can be used in a cache side channel
3078 attack, to access data to which the attacker does
3079 not have direct access.
3081 This parameter controls the MDS mitigation. The
3084 full - Enable MDS mitigation on vulnerable CPUs
3085 full,nosmt - Enable MDS mitigation and disable
3086 SMT on vulnerable CPUs
3087 off - Unconditionally disable MDS mitigation
3089 On TAA-affected machines, mds=off can be prevented by
3090 an active TAA mitigation as both vulnerabilities are
3091 mitigated with the same mechanism so in order to disable
3092 this mitigation, you need to specify tsx_async_abort=off
3095 Not specifying this option is equivalent to
3098 For details see: Documentation/admin-guide/hw-vuln/mds.rst
3100 mem=nn[KMG] [HEXAGON] Set the memory size.
3101 Must be specified, otherwise memory size will be 0.
3103 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
3104 Amount of memory to be used in cases as follows:
3107 2 when the kernel is not able to see the whole system memory;
3108 3 memory that lies after 'mem=' boundary is excluded from
3109 the hypervisor, then assigned to KVM guests.
3110 4 to limit the memory available for kdump kernel.
3112 [ARC,MICROBLAZE] - the limit applies only to low memory,
3113 high memory is not affected.
3115 [ARM64] - only limits memory covered by the linear
3116 mapping. The NOMAP regions are not affected.
3118 [X86] Work as limiting max address. Use together
3119 with memmap= to avoid physical address space collisions.
3120 Without memmap= PCI devices could be placed at addresses
3121 belonging to unused RAM.
3123 Note that this only takes effects during boot time since
3124 in above case 3, memory may need be hot added after boot
3125 if system memory of hypervisor is not sufficient.
3128 [ARM,MIPS] - override the memory layout reported by
3130 Define a memory region of size nn[KMG] starting at
3132 Multiple different regions can be specified with
3133 multiple mem= parameters on the command line.
3135 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
3138 memblock=debug [KNL] Enable memblock debug messages.
3141 [KNL,SH] Allow user to override the default size for
3142 per-device physically contiguous DMA buffers.
3144 memhp_default_state=online/offline/online_kernel/online_movable
3145 [KNL] Set the initial state for the memory hotplug
3146 onlining policy. If not specified, the default value is
3147 set according to the
3148 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config
3150 See Documentation/admin-guide/mm/memory-hotplug.rst.
3152 memmap=exactmap [KNL,X86] Enable setting of an exact
3153 E820 memory map, as specified by the user.
3154 Such memmap=exactmap lines can be constructed based on
3155 BIOS output or other requirements. See the memmap=nn@ss
3158 memmap=nn[KMG]@ss[KMG]
3159 [KNL, X86, MIPS, XTENSA] Force usage of a specific region of memory.
3160 Region of memory to be used is from ss to ss+nn.
3161 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG],
3162 which limits max address to nn[KMG].
3163 Multiple different regions can be specified,
3166 memmap=100M@2G,100M#3G,1G!1024G
3168 memmap=nn[KMG]#ss[KMG]
3169 [KNL,ACPI] Mark specific memory as ACPI data.
3170 Region of memory to be marked is from ss to ss+nn.
3172 memmap=nn[KMG]$ss[KMG]
3173 [KNL,ACPI] Mark specific memory as reserved.
3174 Region of memory to be reserved is from ss to ss+nn.
3175 Example: Exclude memory from 0x18690000-0x1869ffff
3176 memmap=64K$0x18690000
3178 memmap=0x10000$0x18690000
3179 Some bootloaders may need an escape character before '$',
3180 like Grub2, otherwise '$' and the following number
3183 memmap=nn[KMG]!ss[KMG]
3184 [KNL,X86] Mark specific memory as protected.
3185 Region of memory to be used, from ss to ss+nn.
3186 The memory region may be marked as e820 type 12 (0xc)
3187 and is NVDIMM or ADR memory.
3189 memmap=<size>%<offset>-<oldtype>+<newtype>
3190 [KNL,ACPI] Convert memory within the specified region
3191 from <oldtype> to <newtype>. If "-<oldtype>" is left
3192 out, the whole region will be marked as <newtype>,
3193 even if previously unavailable. If "+<newtype>" is left
3194 out, matching memory will be removed. Types are
3195 specified as e820 types, e.g., 1 = RAM, 2 = reserved,
3196 3 = ACPI, 12 = PRAM.
3198 memory_corruption_check=0/1 [X86]
3199 Some BIOSes seem to corrupt the first 64k of
3200 memory when doing things like suspend/resume.
3201 Setting this option will scan the memory
3202 looking for corruption. Enabling this will
3203 both detect corruption and prevent the kernel
3204 from using the memory being corrupted.
3205 However, its intended as a diagnostic tool; if
3206 repeatable BIOS-originated corruption always
3207 affects the same memory, you can use memmap=
3208 to prevent the kernel from using that memory.
3210 memory_corruption_check_size=size [X86]
3211 By default it checks for corruption in the low
3212 64k, making this memory unavailable for normal
3213 use. Use this parameter to scan for
3214 corruption in more or less memory.
3216 memory_corruption_check_period=seconds [X86]
3217 By default it checks for corruption every 60
3218 seconds. Use this parameter to check at some
3219 other rate. 0 disables periodic checking.
3221 memory_hotplug.memmap_on_memory
3222 [KNL,X86,ARM] Boolean flag to enable this feature.
3223 Format: {on | off (default)}
3224 When enabled, runtime hotplugged memory will
3225 allocate its internal metadata (struct pages,
3226 those vmemmap pages cannot be optimized even
3227 if hugetlb_free_vmemmap is enabled) from the
3228 hotadded memory which will allow to hotadd a
3229 lot of memory without requiring additional
3231 This feature is disabled by default because it
3232 has some implication on large (e.g. GB)
3233 allocations in some configurations (e.g. small
3235 The state of the flag can be read in
3236 /sys/module/memory_hotplug/parameters/memmap_on_memory.
3237 Note that even when enabled, there are a few cases where
3238 the feature is not effective.
3240 memtest= [KNL,X86,ARM,M68K,PPC,RISCV] Enable memtest
3242 default : 0 <disable>
3243 Specifies the number of memtest passes to be
3244 performed. Each pass selects another test
3245 pattern from a given set of patterns. Memtest
3246 fills the memory with this pattern, validates
3247 memory contents and reserves bad memory
3248 regions that are detected.
3250 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control
3251 Valid arguments: on, off
3252 Default (depends on kernel configuration option):
3253 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y)
3254 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n)
3255 mem_encrypt=on: Activate SME
3256 mem_encrypt=off: Do not activate SME
3258 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst
3259 for details on when memory encryption can be activated.
3261 mem_sleep_default= [SUSPEND] Default system suspend mode:
3262 s2idle - Suspend-To-Idle
3263 shallow - Power-On Suspend or equivalent (if supported)
3264 deep - Suspend-To-RAM or equivalent (if supported)
3265 See Documentation/admin-guide/pm/sleep-states.rst.
3267 mfgpt_irq= [IA-32] Specify the IRQ to use for the
3268 Multi-Function General Purpose Timers on AMD Geode
3271 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
3272 the BIOS has incorrectly applied a workaround. TinyBIOS
3273 version 0.98 is known to be affected, 0.99 fixes the
3274 problem by letting the user disable the workaround.
3278 min_addr=nn[KMG] [KNL,BOOT,IA-64] All physical memory below this
3279 physical address is ignored.
3281 mini2440= [ARM,HW,KNL]
3282 Format:[0..2][b][c][t]
3284 MINI2440 configuration specification:
3285 0 - The attached screen is the 3.5" TFT
3286 1 - The attached screen is the 7" TFT
3287 2 - The VGA Shield is attached (1024x768)
3288 Leaving out the screen size parameter will not load
3289 the TFT driver, and the framebuffer will be left
3291 b - Enable backlight. The TFT backlight pin will be
3292 linked to the kernel VESA blanking code and a GPIO
3293 LED. This parameter is not necessary when using the
3295 c - Enable the s3c camera interface.
3296 t - Reserved for enabling touchscreen support. The
3297 touchscreen support is not enabled in the mainstream
3298 kernel as of 2.6.30, a preliminary port can be found
3299 in the "bleeding edge" mini2440 support kernel at
3300 https://repo.or.cz/w/linux-2.6/mini2440.git
3303 [X86,PPC,S390,ARM64] Control optional mitigations for
3304 CPU vulnerabilities. This is a set of curated,
3305 arch-independent options, each of which is an
3306 aggregation of existing arch-specific options.
3309 Disable all optional CPU mitigations. This
3310 improves system performance, but it may also
3311 expose users to several CPU vulnerabilities.
3312 Equivalent to: if nokaslr then kpti=0 [ARM64]
3313 gather_data_sampling=off [X86]
3314 kvm.nx_huge_pages=off [X86]
3317 mmio_stale_data=off [X86]
3318 no_entry_flush [PPC]
3319 no_uaccess_flush [PPC]
3322 nospectre_bhb [ARM64]
3323 nospectre_v1 [X86,PPC]
3324 nospectre_v2 [X86,PPC,S390,ARM64]
3326 spec_store_bypass_disable=off [X86,PPC]
3327 spectre_v2_user=off [X86]
3328 srbds=off [X86,INTEL]
3329 ssbd=force-off [ARM64]
3330 tsx_async_abort=off [X86]
3333 This does not have any effect on
3334 kvm.nx_huge_pages when
3335 kvm.nx_huge_pages=force.
3338 Mitigate all CPU vulnerabilities, but leave SMT
3339 enabled, even if it's vulnerable. This is for
3340 users who don't want to be surprised by SMT
3341 getting disabled across kernel upgrades, or who
3342 have other ways of avoiding SMT-based attacks.
3343 Equivalent to: (default behavior)
3346 Mitigate all CPU vulnerabilities, disabling SMT
3347 if needed. This is for users who always want to
3348 be fully mitigated, even if it means losing SMT.
3349 Equivalent to: l1tf=flush,nosmt [X86]
3350 mds=full,nosmt [X86]
3351 tsx_async_abort=full,nosmt [X86]
3352 mmio_stale_data=full,nosmt [X86]
3353 retbleed=auto,nosmt [X86]
3356 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
3357 parameter allows control of the logging verbosity for
3358 the additional memory initialisation checks. A value
3359 of 0 disables mminit logging and a level of 4 will
3360 log everything. Information is printed at KERN_DEBUG
3361 so loglevel=8 may also need to be specified.
3364 [X86,INTEL] Control mitigation for the Processor
3365 MMIO Stale Data vulnerabilities.
3367 Processor MMIO Stale Data is a class of
3368 vulnerabilities that may expose data after an MMIO
3369 operation. Exposed data could originate or end in
3370 the same CPU buffers as affected by MDS and TAA.
3371 Therefore, similar to MDS and TAA, the mitigation
3372 is to clear the affected CPU buffers.
3374 This parameter controls the mitigation. The
3377 full - Enable mitigation on vulnerable CPUs
3379 full,nosmt - Enable mitigation and disable SMT on
3382 off - Unconditionally disable mitigation
3384 On MDS or TAA affected machines,
3385 mmio_stale_data=off can be prevented by an active
3386 MDS or TAA mitigation as these vulnerabilities are
3387 mitigated with the same mechanism so in order to
3388 disable this mitigation, you need to specify
3389 mds=off and tsx_async_abort=off too.
3391 Not specifying this option is equivalent to
3392 mmio_stale_data=full.
3395 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst
3397 <module>.async_probe[=<bool>] [KNL]
3398 If no <bool> value is specified or if the value
3399 specified is not a valid <bool>, enable asynchronous
3400 probe on this module. Otherwise, enable/disable
3401 asynchronous probe on this module as indicated by the
3402 <bool> value. See also: module.async_probe
3404 module.async_probe=<bool>
3405 [KNL] When set to true, modules will use async probing
3406 by default. To enable/disable async probing for a
3407 specific module, use the module specific control that
3408 is documented under <module>.async_probe. When both
3409 module.async_probe and <module>.async_probe are
3410 specified, <module>.async_probe takes precedence for
3411 the specific module.
3413 module.enable_dups_trace
3414 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set,
3415 this means that duplicate request_module() calls will
3416 trigger a WARN_ON() instead of a pr_warn(). Note that
3417 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s
3418 will always be issued and this option does nothing.
3420 [KNL] When CONFIG_MODULE_SIG is set, this means that
3421 modules without (valid) signatures will fail to load.
3422 Note that if CONFIG_MODULE_SIG_FORCE is set, that
3423 is always true, so this option does nothing.
3425 module_blacklist= [KNL] Do not load a comma-separated list of
3426 modules. Useful for debugging problem modules.
3429 [MOUSE] Maximum time between finger touching and
3430 leaving touchpad surface for touch to be considered
3431 a tap and be reported as a left button click (for
3432 touchpads working in absolute mode only).
3434 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
3435 reporting absolute coordinates, such as tablets
3436 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
3437 reporting absolute coordinates, such as tablets
3439 movablecore= [KNL,X86,IA-64,PPC]
3440 Format: nn[KMGTPE] | nn%
3441 This parameter is the complement to kernelcore=, it
3442 specifies the amount of memory used for migratable
3443 allocations. If both kernelcore and movablecore is
3444 specified, then kernelcore will be at *least* the
3445 specified value but may be more. If movablecore on its
3446 own is specified, the administrator must be careful
3447 that the amount of memory usable for all allocations
3450 movable_node [KNL] Boot-time switch to make hotplugable memory
3451 NUMA nodes to be movable. This means that the memory
3452 of such nodes will be usable only for movable
3453 allocations which rules out almost all kernel
3454 allocations. Use with caution!
3456 MTD_Partition= [MTD]
3457 Format: <name>,<region-number>,<size>,<offset>
3459 MTD_Region= [MTD] Format:
3460 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
3463 See drivers/mtd/parsers/cmdlinepart.c
3466 ARM/S3C2412 JIVE boot control
3468 See arch/arm/mach-s3c/mach-jive.c
3470 mtouchusb.raw_coordinates=
3471 [HW] Make the MicroTouch USB driver use raw coordinates
3472 ('y', default) or cooked coordinates ('n')
3475 Enable printing debug information related to MTRR
3476 registers at boot time.
3478 mtrr_chunk_size=nn[KMG] [X86]
3479 used for mtrr cleanup. It is largest continuous chunk
3480 that could hold holes aka. UC entries.
3482 mtrr_gran_size=nn[KMG] [X86]
3483 Used for mtrr cleanup. It is granularity of mtrr block.
3485 Large value could prevent small alignment from
3488 mtrr_spare_reg_nr=n [X86]
3490 Range: 0,7 : spare reg number
3492 Used for mtrr cleanup. It is spare mtrr entries number.
3493 Set to 2 or more if your graphical card needs more.
3495 multitce=off [PPC] This parameter disables the use of the pSeries
3496 firmware feature for updating multiple TCE entries
3499 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
3501 netdev= [NET] Network devices parameters
3502 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
3503 Note that mem_start is often overloaded to mean
3504 something different and driver-specific.
3505 This usage is only documented in each driver source
3508 netpoll.carrier_timeout=
3509 [NET] Specifies amount of time (in seconds) that
3510 netpoll should wait for a carrier. By default netpoll
3514 [NETFILTER] Enable connection tracking flow accounting
3515 0 to disable accounting
3516 1 to enable accounting
3520 [NFS] sets the pathname to the program which is used
3521 to update the NFS client cache entries.
3523 nfs.cache_getent_timeout=
3524 [NFS] sets the timeout after which an attempt to
3525 update a cache entry is deemed to have failed.
3527 nfs.callback_nr_threads=
3528 [NFSv4] set the total number of threads that the
3529 NFS client will assign to service NFSv4 callback
3532 nfs.callback_tcpport=
3533 [NFS] set the TCP port on which the NFSv4 callback
3534 channel should listen.
3537 [NFS] enable 64-bit inode numbers.
3538 If zero, the NFS client will fake up a 32-bit inode
3539 number for the readdir() and stat() syscalls instead
3540 of returning the full 64-bit number.
3541 The default is to return 64-bit inode numbers.
3543 nfs.idmap_cache_timeout=
3544 [NFS] set the maximum lifetime for idmapper cache
3547 nfs.max_session_cb_slots=
3548 [NFSv4.1] Sets the maximum number of session
3549 slots the client will assign to the callback
3550 channel. This determines the maximum number of
3551 callbacks the client will process in parallel for
3552 a particular server.
3554 nfs.max_session_slots=
3555 [NFSv4.1] Sets the maximum number of session slots
3556 the client will attempt to negotiate with the server.
3557 This limits the number of simultaneous RPC requests
3558 that the client can send to the NFSv4.1 server.
3559 Note that there is little point in setting this
3560 value higher than the max_tcp_slot_table_limit.
3562 nfs.nfs4_disable_idmapping=
3563 [NFSv4] When set to the default of '1', this option
3564 ensures that both the RPC level authentication
3565 scheme and the NFS level operations agree to use
3566 numeric uids/gids if the mount is using the
3567 'sec=sys' security flavour. In effect it is
3568 disabling idmapping, which can make migration from
3569 legacy NFSv2/v3 systems to NFSv4 easier.
3570 Servers that do not support this mode of operation
3571 will be autodetected by the client, and it will fall
3572 back to using the idmapper.
3573 To turn off this behaviour, set the value to '0'.
3576 [NFS4] Specify an additional fixed unique ident-
3577 ification string that NFSv4 clients can insert into
3578 their nfs_client_id4 string. This is typically a
3579 UUID that is generated at system install time.
3581 nfs.recover_lost_locks=
3582 [NFSv4] Attempt to recover locks that were lost due
3583 to a lease timeout on the server. Please note that
3584 doing this risks data corruption, since there are
3585 no guarantees that the file will remain unchanged
3586 after the locks are lost.
3587 If you want to enable the kernel legacy behaviour of
3588 attempting to recover these locks, then set this
3590 The default parameter value of '0' causes the kernel
3591 not to attempt recovery of lost locks.
3593 nfs.send_implementation_id=
3594 [NFSv4.1] Send client implementation identification
3595 information in exchange_id requests.
3596 If zero, no implementation identification information
3598 The default is to send the implementation identification
3601 nfs4.layoutstats_timer=
3602 [NFSv4.2] Change the rate at which the kernel sends
3603 layoutstats to the pNFS metadata server.
3605 Setting this to value to 0 causes the kernel to use
3606 whatever value is the default set by the layout
3607 driver. A non-zero value sets the minimum interval
3608 in seconds between layoutstats transmissions.
3610 nfsd.inter_copy_offload_enable=
3611 [NFSv4.2] When set to 1, the server will support
3612 server-to-server copies for which this server is
3613 the destination of the copy.
3615 nfsd.nfs4_disable_idmapping=
3616 [NFSv4] When set to the default of '1', the NFSv4
3617 server will return only numeric uids and gids to
3618 clients using auth_sys, and will accept numeric uids
3619 and gids from such clients. This is intended to ease
3620 migration from NFSv2/v3.
3622 nfsd.nfsd4_ssc_umount_timeout=
3623 [NFSv4.2] When used as the destination of a
3624 server-to-server copy, knfsd temporarily mounts
3625 the source server. It caches the mount in case
3626 it will be needed again, and discards it if not
3627 used for the number of milliseconds specified by
3630 nfsaddrs= [NFS] Deprecated. Use ip= instead.
3631 See Documentation/admin-guide/nfs/nfsroot.rst.
3633 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
3634 See Documentation/admin-guide/nfs/nfsroot.rst.
3636 nfsrootdebug [NFS] enable nfsroot debugging messages.
3637 See Documentation/admin-guide/nfs/nfsroot.rst.
3639 nmi_backtrace.backtrace_idle [KNL]
3640 Dump stacks even of idle CPUs in response to an
3641 NMI stack-backtrace request.
3643 nmi_debug= [KNL,SH] Specify one or more actions to take
3644 when a NMI is triggered.
3645 Format: [state][,regs][,debounce][,die]
3647 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
3648 Format: [panic,][nopanic,][num]
3650 0 - turn hardlockup detector in nmi_watchdog off
3651 1 - turn hardlockup detector in nmi_watchdog on
3652 When panic is specified, panic when an NMI watchdog
3653 timeout occurs (or 'nopanic' to not panic on an NMI
3654 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set)
3655 To disable both hard and soft lockup detectors,
3656 please see 'nowatchdog'.
3657 This is useful when you use a panic=... timeout and
3658 need the box quickly up again.
3660 These settings can be accessed at runtime via
3661 the nmi_watchdog and hardlockup_panic sysctls.
3663 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
3664 emulation library even if a 387 maths coprocessor
3667 no4lvl [RISCV] Disable 4-level and 5-level paging modes. Forces
3668 kernel to use 3-level paging instead.
3670 no5lvl [X86-64,RISCV] Disable 5-level paging mode. Forces
3671 kernel to use 4-level paging instead.
3673 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
3674 caches in the slab allocator. Saves per-node memory,
3675 but will impact performance.
3679 noaltinstr [S390] Disables alternative instructions patching
3680 (CPU alternatives feature).
3682 noapic [SMP,APIC] Tells the kernel to not make use of any
3683 IOAPICs that may be present in the system.
3685 noautogroup Disable scheduler automatic task group creation.
3690 [HW] Never suspend the console
3691 Disable suspending of consoles during suspend and
3692 hibernate operations. Once disabled, debugging
3693 messages can reach various consoles while the rest
3694 of the system is being put to sleep (ie, while
3695 debugging driver suspend/resume hooks). This may
3696 not work reliably with all consoles, but is known
3697 to work with serial and VGA consoles.
3698 To facilitate more flexible debugging, we also add
3699 console_suspend, a printk module parameter to control
3700 it. Users could use console_suspend (usually
3701 /sys/module/printk/parameters/console_suspend) to
3702 turn on/off it dynamically.
3705 [KNL] Disable object debugging
3707 nodsp [SH] Disable hardware DSP at boot time.
3709 noefi Disable EFI runtime services support.
3711 no_entry_flush [PPC] Don't flush the L1-D cache when entering the kernel.
3716 This affects only 32-bit executables.
3717 noexec32=on: enable non-executable mappings (default)
3718 read doesn't imply executable mappings
3719 noexec32=off: disable non-executable mappings
3720 read implies executable mappings
3722 no_file_caps Tells the kernel not to honor file capabilities. The
3723 only way then for a file to be executed with privilege
3724 is to be setuid root or executed by root.
3726 nofpu [MIPS,SH] Disable hardware FPU at boot time.
3728 nofsgsbase [X86] Disables FSGSBASE instructions.
3730 nofxsr [BUGS=X86-32] Disables x86 floating point extended
3731 register save and restore. The kernel will only save
3732 legacy floating-point registers on task switch.
3734 nohalt [IA-64] Tells the kernel not to use the power saving
3735 function PAL_HALT_LIGHT when idle. This increases
3736 power-consumption. On the positive side, it reduces
3737 interrupt wake-up latency, which may improve performance
3738 in certain environments such as networked servers or
3742 Force pointers printed to the console or buffers to be
3743 unhashed. By default, when a pointer is printed via %p
3744 format string, that pointer is "hashed", i.e. obscured
3745 by hashing the pointer value. This is a security feature
3746 that hides actual kernel addresses from unprivileged
3747 users, but it also makes debugging the kernel more
3748 difficult since unequal pointers can no longer be
3749 compared. However, if this command-line option is
3750 specified, then all normal pointers will have their true
3751 value printed. This option should only be specified when
3752 debugging the kernel. Please do not use on production
3755 nohibernate [HIBERNATION] Disable hibernation and resume.
3757 nohlt [ARM,ARM64,MICROBLAZE,MIPS,PPC,SH] Forces the kernel to
3758 busy wait in do_idle() and not use the arch_cpu_idle()
3759 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP
3760 to be effective. This is useful on platforms where the
3761 sleep(SH) or wfi(ARM,ARM64) instructions do not work
3762 correctly or when doing power measurements to evaluate
3763 the impact of the sleep instructions. This is also
3764 useful when using JTAG debugger.
3766 nohugeiomap [KNL,X86,PPC,ARM64] Disable kernel huge I/O mappings.
3768 nohugevmalloc [KNL,X86,PPC,ARM64] Disable kernel huge vmalloc mappings.
3770 nohz= [KNL] Boottime enable/disable dynamic ticks
3771 Valid arguments: on, off
3774 nohz_full= [KNL,BOOT,SMP,ISOL]
3775 The argument is a cpu list, as described above.
3776 In kernels built with CONFIG_NO_HZ_FULL=y, set
3777 the specified list of CPUs whose tick will be stopped
3778 whenever possible. The boot CPU will be forced outside
3779 the range to maintain the timekeeping. Any CPUs
3780 in this list will have their RCU callbacks offloaded,
3781 just as if they had also been called out in the
3782 rcu_nocbs= boot parameter.
3784 Note that this argument takes precedence over
3785 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option.
3787 noinitrd [RAM] Tells the kernel not to load any configured
3790 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
3792 [Deprecated - use intremap=off]
3796 noinvpcid [X86] Disable the INVPCID cpu feature.
3798 noiotrap [SH] Disables trapped I/O port accesses.
3800 noirqdebug [X86-32] Disables the code which attempts to detect and
3801 disable unhandled interrupt sources.
3803 noisapnp [ISAPNP] Disables ISA PnP code.
3805 nojitter [IA-64] Disables jitter checking for ITC timers.
3808 When CONFIG_RANDOMIZE_BASE is set, this disables
3809 kernel and module base offset ASLR (Address Space
3810 Layout Randomization).
3812 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
3815 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
3817 nolapic [X86-32,APIC] Do not enable or use the local APIC.
3819 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
3821 nomca [IA-64] Disable machine check abort handling
3823 nomce [X86-32] Disable Machine Check Exception
3825 nomfgpt [X86-32] Disable Multi-Function General Purpose
3826 Timer usage (for AMD Geode machines).
3828 nomodeset Disable kernel modesetting. Most systems' firmware
3829 sets up a display mode and provides framebuffer memory
3830 for output. With nomodeset, DRM and fbdev drivers will
3831 not load if they could possibly displace the pre-
3832 initialized output. Only the system framebuffer will
3833 be available for use. The respective drivers will not
3834 perform display-mode changes or accelerated rendering.
3836 Useful as error fallback, or for testing and debugging.
3838 nomodule Disable module load
3840 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
3841 shutdown the other cpus. Instead use the REBOOT_VECTOR
3844 nopat [X86] Disable PAT (page attribute table extension of
3845 pagetables) support.
3847 nopcid [X86-64] Disable the PCID cpu feature.
3849 nopku [X86] Disable Memory Protection Keys CPU feature found
3853 Equivalent to pti=off
3855 nopv= [X86,XEN,KVM,HYPER_V,VMWARE]
3856 Disables the PV optimizations forcing the guest to run
3857 as generic guest with no PV drivers. Currently support
3858 XEN HVM, KVM, HYPER_V and VMWARE guest.
3860 nopvspin [X86,XEN,KVM]
3861 Disables the qspinlock slow path using PV optimizations
3862 which allow the hypervisor to 'idle' the guest on lock
3865 norandmaps Don't use address space randomization. Equivalent to
3866 echo 0 > /proc/sys/kernel/randomize_va_space
3868 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
3869 with UP alternatives
3871 noresume [SWSUSP] Disables resume and restores original swap
3876 no-scroll [VGA] Disables scrollback.
3877 This is required for the Braillex ib80-piezo Braille
3878 reader made by F.H. Papenmeier (Germany).
3880 nosgx [X86-64,SGX] Disables Intel SGX kernel support.
3883 Disable SMAP (Supervisor Mode Access Prevention)
3884 even if it is supported by processor.
3887 Disable SMEP (Supervisor Mode Execution Prevention)
3888 even if it is supported by processor.
3890 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
3891 and disable the IO APIC. legacy for "maxcpus=0".
3893 nosmt [KNL,MIPS,PPC,S390] Disable symmetric multithreading (SMT).
3894 Equivalent to smt=1.
3896 [KNL,X86,PPC] Disable symmetric multithreading (SMT).
3897 nosmt=force: Force disable SMT, cannot be undone
3898 via the sysfs control file.
3900 nosoftlockup [KNL] Disable the soft-lockup detector.
3902 nospec_store_bypass_disable
3903 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability
3905 nospectre_bhb [ARM64] Disable all mitigations for Spectre-BHB (branch
3906 history injection) vulnerability. System may allow data leaks
3909 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1
3910 (bounds check bypass). With this option data leaks are
3911 possible in the system.
3913 nospectre_v2 [X86,PPC_E500,ARM64] Disable all mitigations for
3914 the Spectre variant 2 (indirect branch prediction)
3915 vulnerability. System may allow data leaks with this
3918 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES] Disable paravirtualized
3919 steal time accounting. steal time is computed, but
3920 won't influence scheduler behaviour
3922 nosync [HW,M68K] Disables sync negotiation for all devices.
3924 no_timer_check [X86,APIC] Disables the code which tests for
3925 broken timer IRQ sources.
3928 [PPC] Don't flush the L1-D cache after accessing user data.
3930 novmcoredd [KNL,KDUMP]
3931 Disable device dump. Device dump allows drivers to
3932 append dump data to vmcore so you can collect driver
3933 specified debug info. Drivers can append the data
3934 without any limit and this data is stored in memory,
3935 so this may cause significant memory stress. Disabling
3936 device dump can help save memory but the driver debug
3937 data will be no longer available. This parameter
3938 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP
3942 [X86,PV_OPS] Disable paravirtualized VMware scheduler
3943 clock and use the default one.
3945 nowatchdog [KNL] Disable both lockup detectors, i.e.
3946 soft-lockup and NMI watchdog (hard-lockup).
3950 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
3952 NOTE: this parameter will be ignored on systems with the
3953 LEGACY_XAPIC_DISABLED bit set in the
3954 IA32_XAPIC_DISABLE_STATUS MSR.
3956 noxsave [BUGS=X86] Disables x86 extended register state save
3957 and restore using xsave. The kernel will fallback to
3958 enabling legacy floating-point and sse state.
3960 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
3961 register states. The kernel will fall back to use
3962 xsave to save the states. By using this parameter,
3963 performance of saving the states is degraded because
3964 xsave doesn't support modified optimization while
3965 xsaveopt supports it on xsaveopt enabled systems.
3967 noxsaves [X86] Disables xsaves and xrstors used in saving and
3968 restoring x86 extended register state in compacted
3969 form of xsave area. The kernel will fall back to use
3970 xsaveopt and xrstor to save and restore the states
3971 in standard form of xsave area. By using this
3972 parameter, xsave area per process might occupy more
3973 memory on xsaves enabled systems.
3975 nps_mtm_hs_ctr= [KNL,ARC]
3976 This parameter sets the maximum duration, in
3977 cycles, each HW thread of the CTOP can run
3978 without interruptions, before HW switches it.
3979 The actual maximum duration is 16 times this
3981 Format: integer between 1 and 255
3984 nptcg= [IA-64] Override max number of concurrent global TLB
3985 purges which is reported from either PAL_VM_SUMMARY or
3988 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
3989 could support. nr_cpus=n : n >= 1 limits the kernel to
3990 support 'n' processors. It could be larger than the
3991 number of already plugged CPU during bootup, later in
3992 runtime you can physically add extra cpu until it reaches
3993 n. So during boot up some boot time memory for per-cpu
3994 variables need be pre-allocated for later physical cpu
3997 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
3999 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86] Disable NUMA, Only
4000 set up a single NUMA node spanning all memory.
4002 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic
4004 Allowed values are enable and disable
4006 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
4007 'node', 'default' can be specified
4008 This can be set from sysctl after boot.
4009 See Documentation/admin-guide/sysctl/vm.rst for details.
4011 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
4012 See Documentation/core-api/debugging-via-ohci1394.rst for more
4015 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
4016 Rather than timing out after 20 ms if an EC
4017 command is not properly ACKed, override the length
4018 of the timeout. We have interrupts disabled while
4019 waiting for the ACK, so if this is set too high
4020 interrupts *may* be lost!
4022 omap_mux= [OMAP] Override bootloader pin multiplexing.
4023 Format: <mux_mode0.mode_name=value>...
4024 For example, to override I2C bus2:
4025 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
4027 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
4029 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
4031 boundary - index of last SLC block on Flex-OneNAND.
4032 The remaining blocks are configured as MLC blocks.
4033 lock - Configure if Flex-OneNAND boundary should be locked.
4034 Once locked, the boundary cannot be changed.
4035 1 indicates lock status, 0 indicates unlock status.
4037 oops=panic Always panic on oopses. Default is to just kill the
4038 process, but there is a small probability of
4039 deadlocking the machine.
4040 This will also cause panics on machine check exceptions.
4041 Useful together with panic=30 to trigger a reboot.
4044 [KNL] Boolean flag to control whether the page allocator
4045 should randomize its free lists. The randomization may
4046 be automatically enabled if the kernel detects it is
4047 running on a platform with a direct-mapped memory-side
4048 cache, and this parameter can be used to
4049 override/disable that behavior. The state of the flag
4050 can be read from sysfs at:
4051 /sys/module/page_alloc/parameters/shuffle.
4053 page_owner= [KNL] Boot-time page_owner enabling option.
4054 Storage of the information about who allocated
4055 each page is disabled in default. With this switch,
4057 on: enable the feature
4059 page_poison= [KNL] Boot-time parameter changing the state of
4060 poisoning on the buddy allocator, available with
4061 CONFIG_PAGE_POISONING=y.
4062 off: turn off poisoning (default)
4063 on: turn on poisoning
4065 page_reporting.page_reporting_order=
4066 [KNL] Minimal page reporting order
4068 Adjust the minimal page reporting order. The page
4069 reporting is disabled when it exceeds MAX_ORDER.
4071 panic= [KNL] Kernel behaviour on panic: delay <timeout>
4072 timeout > 0: seconds before rebooting
4073 timeout = 0: wait forever
4074 timeout < 0: reboot immediately
4077 panic_on_taint= Bitmask for conditionally calling panic() in add_taint()
4078 Format: <hex>[,nousertaint]
4079 Hexadecimal bitmask representing the set of TAINT flags
4080 that will cause the kernel to panic when add_taint() is
4081 called with any of the flags in this set.
4082 The optional switch "nousertaint" can be utilized to
4083 prevent userspace forced crashes by writing to sysctl
4084 /proc/sys/kernel/tainted any flagset matching with the
4085 bitmask set on panic_on_taint.
4086 See Documentation/admin-guide/tainted-kernels.rst for
4087 extra details on the taint flags that users can pick
4088 to compose the bitmask to assign to panic_on_taint.
4090 panic_on_warn=1 panic() instead of WARN(). Useful to cause kdump
4093 panic_print= Bitmask for printing system info when panic happens.
4094 User can chose combination of the following bits:
4095 bit 0: print all tasks info
4096 bit 1: print system memory info
4097 bit 2: print timer info
4098 bit 3: print locks info if CONFIG_LOCKDEP is on
4099 bit 4: print ftrace buffer
4100 bit 5: print all printk messages in buffer
4101 bit 6: print all CPUs backtrace (if available in the arch)
4102 *Be aware* that this option may print a _lot_ of lines,
4103 so there are risks of losing older messages in the log.
4104 Use this option carefully, maybe worth to setup a
4105 bigger log buffer with "log_buf_len" along with this.
4107 parkbd.port= [HW] Parallel port number the keyboard adapter is
4108 connected to, default is 0.
4110 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
4111 0 for XT, 1 for AT (default is AT).
4114 parport= [HW,PPT] Specify parallel ports. 0 disables.
4115 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
4116 Use 'auto' to force the driver to use any
4117 IRQ/DMA settings detected (the default is to
4118 ignore detected IRQ/DMA settings because of
4119 possible conflicts). You can specify the base
4120 address, IRQ, and DMA settings; IRQ and DMA
4121 should be numbers, or 'auto' (for using detected
4122 settings on that particular port), or 'nofifo'
4123 (to avoid using a FIFO even if it is detected).
4124 Parallel ports are assigned in the order they
4125 are specified on the command line, starting
4128 parport_init_mode= [HW,PPT]
4129 Configure VIA parallel port to operate in
4130 a specific mode. This is necessary on Pegasos
4131 computer where firmware has no options for setting
4132 up parallel port mode and sets it to spp.
4133 Currently this function knows 686a and 8231 chips.
4134 Format: [spp|ps2|epp|ecp|ecpepp]
4136 pata_legacy.all= [HW,LIBATA]
4138 Set to non-zero to probe primary and secondary ISA
4139 port ranges on PCI systems where no PCI PATA device
4140 has been found at either range. Disabled by default.
4142 pata_legacy.autospeed= [HW,LIBATA]
4144 Set to non-zero if a chip is present that snoops speed
4145 changes. Disabled by default.
4147 pata_legacy.ht6560a= [HW,LIBATA]
4149 Set to 1, 2, or 3 for HT 6560A on the primary channel,
4150 the secondary channel, or both channels respectively.
4151 Disabled by default.
4153 pata_legacy.ht6560b= [HW,LIBATA]
4155 Set to 1, 2, or 3 for HT 6560B on the primary channel,
4156 the secondary channel, or both channels respectively.
4157 Disabled by default.
4159 pata_legacy.iordy_mask= [HW,LIBATA]
4161 IORDY enable mask. Set individual bits to allow IORDY
4162 for the respective channel. Bit 0 is for the first
4163 legacy channel handled by this driver, bit 1 is for
4164 the second channel, and so on. The sequence will often
4165 correspond to the primary legacy channel, the secondary
4166 legacy channel, and so on, but the handling of a PCI
4167 bus and the use of other driver options may interfere
4168 with the sequence. By default IORDY is allowed across
4171 pata_legacy.opti82c46x= [HW,LIBATA]
4173 Set to 1, 2, or 3 for Opti 82c611A on the primary
4174 channel, the secondary channel, or both channels
4175 respectively. Disabled by default.
4177 pata_legacy.opti82c611a= [HW,LIBATA]
4179 Set to 1, 2, or 3 for Opti 82c465MV on the primary
4180 channel, the secondary channel, or both channels
4181 respectively. Disabled by default.
4183 pata_legacy.pio_mask= [HW,LIBATA]
4185 PIO mode mask for autospeed devices. Set individual
4186 bits to allow the use of the respective PIO modes.
4187 Bit 0 is for mode 0, bit 1 is for mode 1, and so on.
4188 All modes allowed by default.
4190 pata_legacy.probe_all= [HW,LIBATA]
4192 Set to non-zero to probe tertiary and further ISA
4193 port ranges on PCI systems. Disabled by default.
4195 pata_legacy.probe_mask= [HW,LIBATA]
4197 Probe mask for legacy ISA PATA ports. Depending on
4198 platform configuration and the use of other driver
4199 options up to 6 legacy ports are supported: 0x1f0,
4200 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing
4201 of individual ports can be disabled by setting the
4202 corresponding bits in the mask to 1. Bit 0 is for
4203 the first port in the list above (0x1f0), and so on.
4204 By default all supported ports are probed.
4206 pata_legacy.qdi= [HW,LIBATA]
4208 Set to non-zero to probe QDI controllers. By default
4209 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise.
4211 pata_legacy.winbond= [HW,LIBATA]
4213 Set to non-zero to probe Winbond controllers. Use
4214 the standard I/O port (0x130) if 1, otherwise the
4215 value given is the I/O port to use (typically 0x1b0).
4216 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE,
4219 pata_platform.pio_mask= [HW,LIBATA]
4221 Supported PIO mode mask. Set individual bits to allow
4222 the use of the respective PIO modes. Bit 0 is for
4223 mode 0, bit 1 is for mode 1, and so on. Mode 0 only
4227 Halt all CPUs after the first oops has been printed for
4228 the specified number of seconds. This is to be used if
4229 your oopses keep scrolling off the screen.
4233 pci=option[,option...] [PCI] various PCI subsystem options.
4235 Some options herein operate on a specific device
4236 or a set of devices (<pci_dev>). These are
4237 specified in one of the following formats:
4239 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]*
4240 pci:<vendor>:<device>[:<subvendor>:<subdevice>]
4242 Note: the first format specifies a PCI
4243 bus/device/function address which may change
4244 if new hardware is inserted, if motherboard
4245 firmware changes, or due to changes caused
4246 by other kernel parameters. If the
4247 domain is left unspecified, it is
4248 taken to be zero. Optionally, a path
4249 to a device through multiple device/function
4250 addresses can be specified after the base
4251 address (this is more robust against
4252 renumbering issues). The second format
4253 selects devices using IDs from the
4254 configuration space which may match multiple
4255 devices in the system.
4257 earlydump dump PCI config space before the kernel
4259 off [X86] don't probe for the PCI bus
4260 bios [X86-32] force use of PCI BIOS, don't access
4261 the hardware directly. Use this if your machine
4262 has a non-standard PCI host bridge.
4263 nobios [X86-32] disallow use of PCI BIOS, only direct
4264 hardware access methods are allowed. Use this
4265 if you experience crashes upon bootup and you
4266 suspect they are caused by the BIOS.
4267 conf1 [X86] Force use of PCI Configuration Access
4268 Mechanism 1 (config address in IO port 0xCF8,
4269 data in IO port 0xCFC, both 32-bit).
4270 conf2 [X86] Force use of PCI Configuration Access
4271 Mechanism 2 (IO port 0xCF8 is an 8-bit port for
4272 the function, IO port 0xCFA, also 8-bit, sets
4273 bus number. The config space is then accessed
4274 through ports 0xC000-0xCFFF).
4275 See http://wiki.osdev.org/PCI for more info
4276 on the configuration access mechanisms.
4277 noaer [PCIE] If the PCIEAER kernel config parameter is
4278 enabled, this kernel boot option can be used to
4279 disable the use of PCIE advanced error reporting.
4280 nodomains [PCI] Disable support for multiple PCI
4281 root domains (aka PCI segments, in ACPI-speak).
4282 nommconf [X86] Disable use of MMCONFIG for PCI
4284 check_enable_amd_mmconf [X86] check for and enable
4285 properly configured MMIO access to PCI
4286 config space on AMD family 10h CPU
4287 nomsi [MSI] If the PCI_MSI kernel config parameter is
4288 enabled, this kernel boot option can be used to
4289 disable the use of MSI interrupts system-wide.
4290 noioapicquirk [APIC] Disable all boot interrupt quirks.
4291 Safety option to keep boot IRQs enabled. This
4292 should never be necessary.
4293 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
4294 primary IO-APIC for bridges that cannot disable
4295 boot IRQs. This fixes a source of spurious IRQs
4296 when the system masks IRQs.
4297 noioapicreroute [APIC] Disable workaround that uses the
4298 boot IRQ equivalent of an IRQ that connects to
4299 a chipset where boot IRQs cannot be disabled.
4300 The opposite of ioapicreroute.
4301 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
4302 routing table. These calls are known to be buggy
4303 on several machines and they hang the machine
4304 when used, but on other computers it's the only
4305 way to get the interrupt routing table. Try
4306 this option if the kernel is unable to allocate
4307 IRQs or discover secondary PCI buses on your
4309 rom [X86] Assign address space to expansion ROMs.
4310 Use with caution as certain devices share
4311 address decoders between ROMs and other
4313 norom [X86] Do not assign address space to
4314 expansion ROMs that do not already have
4315 BIOS assigned address ranges.
4316 nobar [X86] Do not assign address space to the
4317 BARs that weren't assigned by the BIOS.
4318 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
4319 assigned automatically to PCI devices. You can
4320 make the kernel exclude IRQs of your ISA cards
4322 pirqaddr=0xAAAAA [X86] Specify the physical address
4323 of the PIRQ table (normally generated
4324 by the BIOS) if it is outside the
4325 F0000h-100000h range.
4326 lastbus=N [X86] Scan all buses thru bus #N. Can be
4327 useful if the kernel is unable to find your
4328 secondary buses and you want to tell it
4329 explicitly which ones they are.
4330 assign-busses [X86] Always assign all PCI bus
4331 numbers ourselves, overriding
4332 whatever the firmware may have done.
4333 usepirqmask [X86] Honor the possible IRQ mask stored
4334 in the BIOS $PIR table. This is needed on
4335 some systems with broken BIOSes, notably
4336 some HP Pavilion N5400 and Omnibook XE3
4337 notebooks. This will have no effect if ACPI
4338 IRQ routing is enabled.
4339 noacpi [X86] Do not use ACPI for IRQ routing
4340 or for PCI scanning.
4341 use_crs [X86] Use PCI host bridge window information
4342 from ACPI. On BIOSes from 2008 or later, this
4343 is enabled by default. If you need to use this,
4344 please report a bug.
4345 nocrs [X86] Ignore PCI host bridge windows from ACPI.
4346 If you need to use this, please report a bug.
4347 use_e820 [X86] Use E820 reservations to exclude parts of
4348 PCI host bridge windows. This is a workaround
4349 for BIOS defects in host bridge _CRS methods.
4350 If you need to use this, please report a bug to
4351 <linux-pci@vger.kernel.org>.
4352 no_e820 [X86] Ignore E820 reservations for PCI host
4353 bridge windows. This is the default on modern
4354 hardware. If you need to use this, please report
4355 a bug to <linux-pci@vger.kernel.org>.
4356 routeirq Do IRQ routing for all PCI devices.
4357 This is normally done in pci_enable_device(),
4358 so this option is a temporary workaround
4359 for broken drivers that don't call it.
4360 skip_isa_align [X86] do not align io start addr, so can
4361 handle more pci cards
4362 noearly [X86] Don't do any early type 1 scanning.
4363 This might help on some broken boards which
4364 machine check when some devices' config space
4365 is read. But various workarounds are disabled
4366 and some IOMMU drivers will not work.
4367 bfsort Sort PCI devices into breadth-first order.
4368 This sorting is done to get a device
4369 order compatible with older (<= 2.4) kernels.
4370 nobfsort Don't sort PCI devices into breadth-first order.
4371 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
4372 tuning and use the BIOS-configured MPS defaults.
4373 pcie_bus_safe Set every device's MPS to the largest value
4374 supported by all devices below the root complex.
4375 pcie_bus_perf Set device MPS to the largest allowable MPS
4376 based on its parent bus. Also set MRRS (Max
4377 Read Request Size) to the largest supported
4378 value (no larger than the MPS that the device
4379 or bus can support) for best performance.
4380 pcie_bus_peer2peer Set every device's MPS to 128B, which
4381 every device is guaranteed to support. This
4382 configuration allows peer-to-peer DMA between
4383 any pair of devices, possibly at the cost of
4384 reduced performance. This also guarantees
4385 that hot-added devices will work.
4386 cbiosize=nn[KMG] The fixed amount of bus space which is
4387 reserved for the CardBus bridge's IO window.
4388 The default value is 256 bytes.
4389 cbmemsize=nn[KMG] The fixed amount of bus space which is
4390 reserved for the CardBus bridge's memory
4391 window. The default value is 64 megabytes.
4394 [<order of align>@]<pci_dev>[; ...]
4395 Specifies alignment and device to reassign
4396 aligned memory resources. How to
4397 specify the device is described above.
4398 If <order of align> is not specified,
4399 PAGE_SIZE is used as alignment.
4400 A PCI-PCI bridge can be specified if resource
4401 windows need to be expanded.
4402 To specify the alignment for several
4403 instances of a device, the PCI vendor,
4404 device, subvendor, and subdevice may be
4405 specified, e.g., 12@pci:8086:9c22:103c:198f
4406 for 4096-byte alignment.
4407 ecrc= Enable/disable PCIe ECRC (transaction layer
4408 end-to-end CRC checking). Only effective if
4409 OS has native AER control (either granted by
4410 ACPI _OSC or forced via "pcie_ports=native")
4411 bios: Use BIOS/firmware settings. This is the
4415 hpiosize=nn[KMG] The fixed amount of bus space which is
4416 reserved for hotplug bridge's IO window.
4417 Default size is 256 bytes.
4418 hpmmiosize=nn[KMG] The fixed amount of bus space which is
4419 reserved for hotplug bridge's MMIO window.
4420 Default size is 2 megabytes.
4421 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is
4422 reserved for hotplug bridge's MMIO_PREF window.
4423 Default size is 2 megabytes.
4424 hpmemsize=nn[KMG] The fixed amount of bus space which is
4425 reserved for hotplug bridge's MMIO and
4427 Default size is 2 megabytes.
4428 hpbussize=nn The minimum amount of additional bus numbers
4429 reserved for buses below a hotplug bridge.
4431 realloc= Enable/disable reallocating PCI bridge resources
4432 if allocations done by BIOS are too small to
4433 accommodate resources required by all child
4435 off: Turn realloc off
4437 realloc same as realloc=on
4438 noari do not use PCIe ARI.
4439 noats [PCIE, Intel-IOMMU, AMD-IOMMU]
4440 do not use PCIe ATS (and IOMMU device IOTLB).
4441 pcie_scan_all Scan all possible PCIe devices. Otherwise we
4442 only look for one device below a PCIe downstream
4444 big_root_window Try to add a big 64bit memory window to the PCIe
4445 root complex on AMD CPUs. Some GFX hardware
4446 can resize a BAR to allow access to all VRAM.
4447 Adding the window is slightly risky (it may
4448 conflict with unreported devices), so this
4450 disable_acs_redir=<pci_dev>[; ...]
4451 Specify one or more PCI devices (in the format
4452 specified above) separated by semicolons.
4453 Each device specified will have the PCI ACS
4454 redirect capabilities forced off which will
4455 allow P2P traffic between devices through
4456 bridges without forcing it upstream. Note:
4457 this removes isolation between devices and
4458 may put more devices in an IOMMU group.
4459 force_floating [S390] Force usage of floating interrupts.
4460 nomio [S390] Do not use MIO instructions.
4461 norid [S390] ignore the RID field and force use of
4462 one PCI domain per PCI function
4464 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
4467 force Enable ASPM even on devices that claim not to support it.
4468 WARNING: Forcing ASPM on may cause system lockups.
4470 pcie_ports= [PCIE] PCIe port services handling:
4471 native Use native PCIe services (PME, AER, DPC, PCIe hotplug)
4472 even if the platform doesn't give the OS permission to
4473 use them. This may cause conflicts if the platform
4474 also tries to use these services.
4475 dpc-native Use native PCIe service for DPC only. May
4476 cause conflicts if firmware uses AER or DPC.
4477 compat Disable native PCIe services (PME, AER, DPC, PCIe
4480 pcie_port_pm= [PCIE] PCIe port power management handling:
4481 off Disable power management of all PCIe ports
4482 force Forcibly enable power management of all PCIe ports
4484 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
4485 nomsi Do not use MSI for native PCIe PME signaling (this makes
4486 all PCIe root ports use INTx for all services).
4488 pcmv= [HW,PCMCIA] BadgePAD 4
4492 Keep all power-domains already enabled by bootloader on,
4493 even if no driver has claimed them. This is useful
4494 for debug and development, but should not be
4495 needed on a platform with proper driver support.
4497 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
4500 See arch/parisc/kernel/pdc_chassis.c
4502 percpu_alloc= Select which percpu first chunk allocator to use.
4503 Currently supported values are "embed" and "page".
4504 Archs may support subset or none of the selections.
4505 See comments in mm/percpu.c for details on each
4506 allocator. This parameter is primarily for debugging
4507 and performance comparison.
4509 pirq= [SMP,APIC] Manual mp-table setup
4510 See Documentation/arch/x86/i386/IO-APIC.rst.
4512 plip= [PPT,NET] Parallel port network link
4513 Format: { parport<nr> | timid | 0 }
4514 See also Documentation/admin-guide/parport.rst.
4516 pmtmr= [X86] Manual setup of pmtmr I/O Port.
4517 Override pmtimer IOPort with a hex value.
4520 pmu_override= [PPC] Override the PMU.
4521 This option takes over the PMU facility, so it is no
4522 longer usable by perf. Setting this option starts the
4523 PMU counters by setting MMCR0 to 0 (the FC bit is
4524 cleared). If a number is given, then MMCR1 is set to
4525 that number, otherwise (e.g., 'pmu_override=on'), MMCR1
4528 pm_debug_messages [SUSPEND,KNL]
4529 Enable suspend/resume debug messages during boot up.
4532 Enable PNP debug messages (depends on the
4533 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
4534 via /sys/module/pnp/parameters/debug. We always show
4535 current resource usage; turning this on also shows
4536 possible settings and some assignment information.
4542 { on | off | curr | res | no-curr | no-res }
4545 [ISAPNP] Exclude IRQs for the autoconfiguration
4548 [ISAPNP] Exclude DMAs for the autoconfiguration
4550 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
4551 Ranges are in pairs (I/O port base and size).
4554 [ISAPNP] Exclude memory regions for the
4556 Ranges are in pairs (memory base and size).
4558 ports= [IP_VS_FTP] IPVS ftp helper module
4560 Up to 8 (IP_VS_APP_MAX_PORTS) ports
4562 Format: <port>,<port>....
4564 powersave=off [PPC] This option disables power saving features.
4565 It specifically disables cpuidle and sets the
4566 platform machine description specific power_save
4567 function to NULL. On Idle the CPU just reduces
4570 ppc_strict_facility_enable
4571 [PPC] This option catches any kernel floating point,
4572 Altivec, VSX and SPE outside of regions specifically
4573 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
4574 There is some performance impact when enabling this.
4578 Disable Hardware Transactional Memory
4581 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC
4582 none - Limited to cond_resched() calls
4583 voluntary - Limited to cond_resched() and might_sleep() calls
4584 full - Any section that isn't explicitly preempt disabled
4585 can be preempted anytime.
4587 print-fatal-signals=
4588 [KNL] debug: print fatal signals
4590 If enabled, warn about various signal handling
4591 related application anomalies: too many signals,
4592 too many POSIX.1 timers, fatal signals causing a
4595 If you hit the warning due to signal overflow,
4596 you might want to try "ulimit -i unlimited".
4600 printk.always_kmsg_dump=
4601 Trigger kmsg_dump for cases other than kernel oops or
4603 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
4606 printk.console_no_auto_verbose=
4607 Disable console loglevel raise on oops, panic
4608 or lockdep-detected issues (only if lock debug is on).
4609 With an exception to setups with low baudrate on
4610 serial console, keeping this 0 is a good choice
4611 in order to provide more debug information.
4613 default: 0 (auto_verbose is enabled)
4615 printk.devkmsg={on,off,ratelimit}
4616 Control writing to /dev/kmsg.
4617 on - unlimited logging to /dev/kmsg from userspace
4618 off - logging to /dev/kmsg disabled
4619 ratelimit - ratelimit the logging
4622 printk.time= Show timing data prefixed to each printk message line
4623 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
4625 processor.max_cstate= [HW,ACPI]
4626 Limit processor to maximum C-state
4627 max_cstate=9 overrides any DMI blacklist limit.
4629 processor.nocst [HW,ACPI]
4630 Ignore the _CST method to determine C-states,
4631 instead using the legacy FADT method
4633 profile= [KNL] Enable kernel profiling via /proc/profile
4634 Format: [<profiletype>,]<number>
4635 Param: <profiletype>: "schedule", "sleep", or "kvm"
4636 [defaults to kernel profiling]
4637 Param: "schedule" - profile schedule points.
4638 Param: "sleep" - profile D-state sleeping (millisecs).
4639 Requires CONFIG_SCHEDSTATS
4640 Param: "kvm" - profile VM exits.
4641 Param: <number> - step/bucket size as a power of 2 for
4642 statistical time based profiling.
4644 prompt_ramdisk= [RAM] [Deprecated]
4646 prot_virt= [S390] enable hosting protected virtual machines
4647 isolated from the hypervisor (if hardware supports
4651 psi= [KNL] Enable or disable pressure stall information
4655 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
4656 probe for; one of (bare|imps|exps|lifebook|any).
4657 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
4659 psmouse.resetafter= [HW,MOUSE]
4660 Try to reset the device after so many bad packets
4663 [HW,MOUSE] Set desired mouse resolution, in dpi.
4664 psmouse.smartscroll=
4665 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
4666 0 = disabled, 1 = enabled (default).
4668 pstore.backend= Specify the name of the pstore backend to use
4670 pti= [X86-64] Control Page Table Isolation of user and
4671 kernel address spaces. Disabling this feature
4672 removes hardening, but improves performance of
4673 system calls and interrupts.
4675 on - unconditionally enable
4676 off - unconditionally disable
4677 auto - kernel detects whether your CPU model is
4678 vulnerable to issues that PTI mitigates
4680 Not specifying this option is equivalent to pti=auto.
4683 [KNL] Number of legacy pty's. Overwrites compiled-in
4686 quiet [KNL] Disable most log messages
4690 radix_hcall_invalidate=on [PPC/PSERIES]
4691 Disable RADIX GTSE feature and use hcall for TLB
4695 See Documentation/admin-guide/md.rst.
4697 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
4698 See Documentation/admin-guide/blockdev/ramdisk.rst.
4700 ramdisk_start= [RAM] RAM disk image start address
4702 random.trust_cpu=off
4703 [KNL] Disable trusting the use of the CPU's
4704 random number generator (if available) to
4705 initialize the kernel's RNG.
4707 random.trust_bootloader=off
4708 [KNL] Disable trusting the use of the a seed
4709 passed by the bootloader (if available) to
4710 initialize the kernel's RNG.
4712 randomize_kstack_offset=
4713 [KNL] Enable or disable kernel stack offset
4714 randomization, which provides roughly 5 bits of
4715 entropy, frustrating memory corruption attacks
4716 that depend on stack address determinism or
4717 cross-syscall address exposures. This is only
4718 available on architectures that have defined
4719 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET.
4720 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
4721 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT.
4723 ras=option[,option,...] [KNL] RAS-specific options
4726 Disable the Correctable Errors Collector,
4727 see CONFIG_RAS_CEC help text.
4729 rcu_nocbs[=cpu-list]
4730 [KNL] The optional argument is a cpu list,
4733 In kernels built with CONFIG_RCU_NOCB_CPU=y,
4734 enable the no-callback CPU mode, which prevents
4735 such CPUs' callbacks from being invoked in
4736 softirq context. Invocation of such CPUs' RCU
4737 callbacks will instead be offloaded to "rcuox/N"
4738 kthreads created for that purpose, where "x" is
4739 "p" for RCU-preempt, "s" for RCU-sched, and "g"
4740 for the kthreads that mediate grace periods; and
4741 "N" is the CPU number. This reduces OS jitter on
4742 the offloaded CPUs, which can be useful for HPC
4743 and real-time workloads. It can also improve
4744 energy efficiency for asymmetric multiprocessors.
4746 If a cpulist is passed as an argument, the specified
4747 list of CPUs is set to no-callback mode from boot.
4749 Otherwise, if the '=' sign and the cpulist
4750 arguments are omitted, no CPU will be set to
4751 no-callback mode from boot but the mode may be
4752 toggled at runtime via cpusets.
4754 Note that this argument takes precedence over
4755 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option.
4758 Rather than requiring that offloaded CPUs
4759 (specified by rcu_nocbs= above) explicitly
4760 awaken the corresponding "rcuoN" kthreads,
4761 make these kthreads poll for callbacks.
4762 This improves the real-time response for the
4763 offloaded CPUs by relieving them of the need to
4764 wake up the corresponding kthread, but degrades
4765 energy efficiency by requiring that the kthreads
4766 periodically wake up to do the polling.
4768 rcutree.blimit= [KNL]
4769 Set maximum number of finished RCU callbacks to
4770 process in one batch.
4772 rcutree.dump_tree= [KNL]
4773 Dump the structure of the rcu_node combining tree
4774 out at early boot. This is used for diagnostic
4775 purposes, to verify correct tree setup.
4777 rcutree.gp_cleanup_delay= [KNL]
4778 Set the number of jiffies to delay each step of
4779 RCU grace-period cleanup.
4781 rcutree.gp_init_delay= [KNL]
4782 Set the number of jiffies to delay each step of
4783 RCU grace-period initialization.
4785 rcutree.gp_preinit_delay= [KNL]
4786 Set the number of jiffies to delay each step of
4787 RCU grace-period pre-initialization, that is,
4788 the propagation of recent CPU-hotplug changes up
4789 the rcu_node combining tree.
4791 rcutree.jiffies_till_first_fqs= [KNL]
4792 Set delay from grace-period initialization to
4793 first attempt to force quiescent states.
4794 Units are jiffies, minimum value is zero,
4795 and maximum value is HZ.
4797 rcutree.jiffies_till_next_fqs= [KNL]
4798 Set delay between subsequent attempts to force
4799 quiescent states. Units are jiffies, minimum
4800 value is one, and maximum value is HZ.
4802 rcutree.jiffies_till_sched_qs= [KNL]
4803 Set required age in jiffies for a
4804 given grace period before RCU starts
4805 soliciting quiescent-state help from
4806 rcu_note_context_switch() and cond_resched().
4807 If not specified, the kernel will calculate
4808 a value based on the most recent settings
4809 of rcutree.jiffies_till_first_fqs
4810 and rcutree.jiffies_till_next_fqs.
4811 This calculated value may be viewed in
4812 rcutree.jiffies_to_sched_qs. Any attempt to set
4813 rcutree.jiffies_to_sched_qs will be cheerfully
4816 rcutree.kthread_prio= [KNL,BOOT]
4817 Set the SCHED_FIFO priority of the RCU per-CPU
4818 kthreads (rcuc/N). This value is also used for
4819 the priority of the RCU boost threads (rcub/N)
4820 and for the RCU grace-period kthreads (rcu_bh,
4821 rcu_preempt, and rcu_sched). If RCU_BOOST is
4822 set, valid values are 1-99 and the default is 1
4823 (the least-favored priority). Otherwise, when
4824 RCU_BOOST is not set, valid values are 0-99 and
4825 the default is zero (non-realtime operation).
4826 When RCU_NOCB_CPU is set, also adjust the
4827 priority of NOCB callback kthreads.
4829 rcutree.nocb_nobypass_lim_per_jiffy= [KNL]
4830 On callback-offloaded (rcu_nocbs) CPUs,
4831 RCU reduces the lock contention that would
4832 otherwise be caused by callback floods through
4833 use of the ->nocb_bypass list. However, in the
4834 common non-flooded case, RCU queues directly to
4835 the main ->cblist in order to avoid the extra
4836 overhead of the ->nocb_bypass list and its lock.
4837 But if there are too many callbacks queued during
4838 a single jiffy, RCU pre-queues the callbacks into
4839 the ->nocb_bypass queue. The definition of "too
4840 many" is supplied by this kernel boot parameter.
4842 rcutree.qhimark= [KNL]
4843 Set threshold of queued RCU callbacks beyond which
4844 batch limiting is disabled.
4846 rcutree.qlowmark= [KNL]
4847 Set threshold of queued RCU callbacks below which
4848 batch limiting is re-enabled.
4850 rcutree.qovld= [KNL]
4851 Set threshold of queued RCU callbacks beyond which
4852 RCU's force-quiescent-state scan will aggressively
4853 enlist help from cond_resched() and sched IPIs to
4854 help CPUs more quickly reach quiescent states.
4855 Set to less than zero to make this be set based
4856 on rcutree.qhimark at boot time and to zero to
4857 disable more aggressive help enlistment.
4859 rcutree.rcu_delay_page_cache_fill_msec= [KNL]
4860 Set the page-cache refill delay (in milliseconds)
4861 in response to low-memory conditions. The range
4862 of permitted values is in the range 0:100000.
4864 rcutree.rcu_divisor= [KNL]
4865 Set the shift-right count to use to compute
4866 the callback-invocation batch limit bl from
4867 the number of callbacks queued on this CPU.
4868 The result will be bounded below by the value of
4869 the rcutree.blimit kernel parameter. Every bl
4870 callbacks, the softirq handler will exit in
4871 order to allow the CPU to do other work.
4873 Please note that this callback-invocation batch
4874 limit applies only to non-offloaded callback
4875 invocation. Offloaded callbacks are instead
4876 invoked in the context of an rcuoc kthread, which
4877 scheduler will preempt as it does any other task.
4879 rcutree.rcu_fanout_exact= [KNL]
4880 Disable autobalancing of the rcu_node combining
4881 tree. This is used by rcutorture, and might
4882 possibly be useful for architectures having high
4883 cache-to-cache transfer latencies.
4885 rcutree.rcu_fanout_leaf= [KNL]
4886 Change the number of CPUs assigned to each
4887 leaf rcu_node structure. Useful for very
4888 large systems, which will choose the value 64,
4889 and for NUMA systems with large remote-access
4890 latencies, which will choose a value aligned
4891 with the appropriate hardware boundaries.
4893 rcutree.rcu_min_cached_objs= [KNL]
4894 Minimum number of objects which are cached and
4895 maintained per one CPU. Object size is equal
4896 to PAGE_SIZE. The cache allows to reduce the
4897 pressure to page allocator, also it makes the
4898 whole algorithm to behave better in low memory
4901 rcutree.rcu_nocb_gp_stride= [KNL]
4902 Set the number of NOCB callback kthreads in
4903 each group, which defaults to the square root
4904 of the number of CPUs. Larger numbers reduce
4905 the wakeup overhead on the global grace-period
4906 kthread, but increases that same overhead on
4907 each group's NOCB grace-period kthread.
4909 rcutree.rcu_kick_kthreads= [KNL]
4910 Cause the grace-period kthread to get an extra
4911 wake_up() if it sleeps three times longer than
4912 it should at force-quiescent-state time.
4913 This wake_up() will be accompanied by a
4914 WARN_ONCE() splat and an ftrace_dump().
4916 rcutree.rcu_resched_ns= [KNL]
4917 Limit the time spend invoking a batch of RCU
4918 callbacks to the specified number of nanoseconds.
4919 By default, this limit is checked only once
4920 every 32 callbacks in order to limit the pain
4921 inflicted by local_clock() overhead.
4923 rcutree.rcu_unlock_delay= [KNL]
4924 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels,
4925 this specifies an rcu_read_unlock()-time delay
4926 in microseconds. This defaults to zero.
4927 Larger delays increase the probability of
4928 catching RCU pointer leaks, that is, buggy use
4929 of RCU-protected pointers after the relevant
4930 rcu_read_unlock() has completed.
4932 rcutree.sysrq_rcu= [KNL]
4933 Commandeer a sysrq key to dump out Tree RCU's
4934 rcu_node tree with an eye towards determining
4935 why a new grace period has not yet started.
4937 rcutree.use_softirq= [KNL]
4938 If set to zero, move all RCU_SOFTIRQ processing to
4939 per-CPU rcuc kthreads. Defaults to a non-zero
4940 value, meaning that RCU_SOFTIRQ is used by default.
4941 Specify rcutree.use_softirq=0 to use rcuc kthreads.
4943 But note that CONFIG_PREEMPT_RT=y kernels disable
4944 this kernel boot parameter, forcibly setting it
4947 rcuscale.gp_async= [KNL]
4948 Measure performance of asynchronous
4949 grace-period primitives such as call_rcu().
4951 rcuscale.gp_async_max= [KNL]
4952 Specify the maximum number of outstanding
4953 callbacks per writer thread. When a writer
4954 thread exceeds this limit, it invokes the
4955 corresponding flavor of rcu_barrier() to allow
4956 previously posted callbacks to drain.
4958 rcuscale.gp_exp= [KNL]
4959 Measure performance of expedited synchronous
4960 grace-period primitives.
4962 rcuscale.holdoff= [KNL]
4963 Set test-start holdoff period. The purpose of
4964 this parameter is to delay the start of the
4965 test until boot completes in order to avoid
4968 rcuscale.kfree_by_call_rcu= [KNL]
4969 In kernels built with CONFIG_RCU_LAZY=y, test
4970 call_rcu() instead of kfree_rcu().
4972 rcuscale.kfree_mult= [KNL]
4973 Instead of allocating an object of size kfree_obj,
4974 allocate one of kfree_mult * sizeof(kfree_obj).
4977 rcuscale.kfree_rcu_test= [KNL]
4978 Set to measure performance of kfree_rcu() flooding.
4980 rcuscale.kfree_rcu_test_double= [KNL]
4981 Test the double-argument variant of kfree_rcu().
4982 If this parameter has the same value as
4983 rcuscale.kfree_rcu_test_single, both the single-
4984 and double-argument variants are tested.
4986 rcuscale.kfree_rcu_test_single= [KNL]
4987 Test the single-argument variant of kfree_rcu().
4988 If this parameter has the same value as
4989 rcuscale.kfree_rcu_test_double, both the single-
4990 and double-argument variants are tested.
4992 rcuscale.kfree_nthreads= [KNL]
4993 The number of threads running loops of kfree_rcu().
4995 rcuscale.kfree_alloc_num= [KNL]
4996 Number of allocations and frees done in an iteration.
4998 rcuscale.kfree_loops= [KNL]
4999 Number of loops doing rcuscale.kfree_alloc_num number
5000 of allocations and frees.
5002 rcuscale.minruntime= [KNL]
5003 Set the minimum test run time in seconds. This
5004 does not affect the data-collection interval,
5005 but instead allows better measurement of things
5006 like CPU consumption.
5008 rcuscale.nreaders= [KNL]
5009 Set number of RCU readers. The value -1 selects
5010 N, where N is the number of CPUs. A value
5011 "n" less than -1 selects N-n+1, where N is again
5012 the number of CPUs. For example, -2 selects N
5013 (the number of CPUs), -3 selects N+1, and so on.
5014 A value of "n" less than or equal to -N selects
5017 rcuscale.nwriters= [KNL]
5018 Set number of RCU writers. The values operate
5019 the same as for rcuscale.nreaders.
5020 N, where N is the number of CPUs
5022 rcuscale.scale_type= [KNL]
5023 Specify the RCU implementation to test.
5025 rcuscale.shutdown= [KNL]
5026 Shut the system down after performance tests
5027 complete. This is useful for hands-off automated
5030 rcuscale.verbose= [KNL]
5031 Enable additional printk() statements.
5033 rcuscale.writer_holdoff= [KNL]
5034 Write-side holdoff between grace periods,
5035 in microseconds. The default of zero says
5038 rcuscale.writer_holdoff_jiffies= [KNL]
5039 Additional write-side holdoff between grace
5040 periods, but in jiffies. The default of zero
5043 rcutorture.fqs_duration= [KNL]
5044 Set duration of force_quiescent_state bursts
5047 rcutorture.fqs_holdoff= [KNL]
5048 Set holdoff time within force_quiescent_state bursts
5051 rcutorture.fqs_stutter= [KNL]
5052 Set wait time between force_quiescent_state bursts
5055 rcutorture.fwd_progress= [KNL]
5056 Specifies the number of kthreads to be used
5057 for RCU grace-period forward-progress testing
5058 for the types of RCU supporting this notion.
5059 Defaults to 1 kthread, values less than zero or
5060 greater than the number of CPUs cause the number
5063 rcutorture.fwd_progress_div= [KNL]
5064 Specify the fraction of a CPU-stall-warning
5065 period to do tight-loop forward-progress testing.
5067 rcutorture.fwd_progress_holdoff= [KNL]
5068 Number of seconds to wait between successive
5069 forward-progress tests.
5071 rcutorture.fwd_progress_need_resched= [KNL]
5072 Enclose cond_resched() calls within checks for
5073 need_resched() during tight-loop forward-progress
5076 rcutorture.gp_cond= [KNL]
5077 Use conditional/asynchronous update-side
5078 primitives, if available.
5080 rcutorture.gp_exp= [KNL]
5081 Use expedited update-side primitives, if available.
5083 rcutorture.gp_normal= [KNL]
5084 Use normal (non-expedited) asynchronous
5085 update-side primitives, if available.
5087 rcutorture.gp_sync= [KNL]
5088 Use normal (non-expedited) synchronous
5089 update-side primitives, if available. If all
5090 of rcutorture.gp_cond=, rcutorture.gp_exp=,
5091 rcutorture.gp_normal=, and rcutorture.gp_sync=
5092 are zero, rcutorture acts as if is interpreted
5093 they are all non-zero.
5095 rcutorture.irqreader= [KNL]
5096 Run RCU readers from irq handlers, or, more
5097 accurately, from a timer handler. Not all RCU
5098 flavors take kindly to this sort of thing.
5100 rcutorture.leakpointer= [KNL]
5101 Leak an RCU-protected pointer out of the reader.
5102 This can of course result in splats, and is
5103 intended to test the ability of things like
5104 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect
5107 rcutorture.n_barrier_cbs= [KNL]
5108 Set callbacks/threads for rcu_barrier() testing.
5110 rcutorture.nfakewriters= [KNL]
5111 Set number of concurrent RCU writers. These just
5112 stress RCU, they don't participate in the actual
5113 test, hence the "fake".
5115 rcutorture.nocbs_nthreads= [KNL]
5116 Set number of RCU callback-offload togglers.
5117 Zero (the default) disables toggling.
5119 rcutorture.nocbs_toggle= [KNL]
5120 Set the delay in milliseconds between successive
5121 callback-offload toggling attempts.
5123 rcutorture.nreaders= [KNL]
5124 Set number of RCU readers. The value -1 selects
5125 N-1, where N is the number of CPUs. A value
5126 "n" less than -1 selects N-n-2, where N is again
5127 the number of CPUs. For example, -2 selects N
5128 (the number of CPUs), -3 selects N+1, and so on.
5130 rcutorture.object_debug= [KNL]
5131 Enable debug-object double-call_rcu() testing.
5133 rcutorture.onoff_holdoff= [KNL]
5134 Set time (s) after boot for CPU-hotplug testing.
5136 rcutorture.onoff_interval= [KNL]
5137 Set time (jiffies) between CPU-hotplug operations,
5138 or zero to disable CPU-hotplug testing.
5140 rcutorture.read_exit= [KNL]
5141 Set the number of read-then-exit kthreads used
5142 to test the interaction of RCU updaters and
5143 task-exit processing.
5145 rcutorture.read_exit_burst= [KNL]
5146 The number of times in a given read-then-exit
5147 episode that a set of read-then-exit kthreads
5150 rcutorture.read_exit_delay= [KNL]
5151 The delay, in seconds, between successive
5152 read-then-exit testing episodes.
5154 rcutorture.shuffle_interval= [KNL]
5155 Set task-shuffle interval (s). Shuffling tasks
5156 allows some CPUs to go into dyntick-idle mode
5157 during the rcutorture test.
5159 rcutorture.shutdown_secs= [KNL]
5160 Set time (s) after boot system shutdown. This
5161 is useful for hands-off automated testing.
5163 rcutorture.stall_cpu= [KNL]
5164 Duration of CPU stall (s) to test RCU CPU stall
5165 warnings, zero to disable.
5167 rcutorture.stall_cpu_block= [KNL]
5168 Sleep while stalling if set. This will result
5169 in warnings from preemptible RCU in addition to
5170 any other stall-related activity. Note that
5171 in kernels built with CONFIG_PREEMPTION=n and
5172 CONFIG_PREEMPT_COUNT=y, this parameter will
5173 cause the CPU to pass through a quiescent state.
5174 Given CONFIG_PREEMPTION=n, this will suppress
5175 RCU CPU stall warnings, but will instead result
5176 in scheduling-while-atomic splats.
5178 Use of this module parameter results in splats.
5181 rcutorture.stall_cpu_holdoff= [KNL]
5182 Time to wait (s) after boot before inducing stall.
5184 rcutorture.stall_cpu_irqsoff= [KNL]
5185 Disable interrupts while stalling if set.
5187 rcutorture.stall_gp_kthread= [KNL]
5188 Duration (s) of forced sleep within RCU
5189 grace-period kthread to test RCU CPU stall
5190 warnings, zero to disable. If both stall_cpu
5191 and stall_gp_kthread are specified, the
5192 kthread is starved first, then the CPU.
5194 rcutorture.stat_interval= [KNL]
5195 Time (s) between statistics printk()s.
5197 rcutorture.stutter= [KNL]
5198 Time (s) to stutter testing, for example, specifying
5199 five seconds causes the test to run for five seconds,
5200 wait for five seconds, and so on. This tests RCU's
5201 ability to transition abruptly to and from idle.
5203 rcutorture.test_boost= [KNL]
5204 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
5205 "Maybe" means test if the RCU implementation
5206 under test support RCU priority boosting.
5208 rcutorture.test_boost_duration= [KNL]
5209 Duration (s) of each individual boost test.
5211 rcutorture.test_boost_interval= [KNL]
5212 Interval (s) between each boost test.
5214 rcutorture.test_no_idle_hz= [KNL]
5215 Test RCU's dyntick-idle handling. See also the
5216 rcutorture.shuffle_interval parameter.
5218 rcutorture.torture_type= [KNL]
5219 Specify the RCU implementation to test.
5221 rcutorture.verbose= [KNL]
5222 Enable additional printk() statements.
5224 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL]
5225 Dump ftrace buffer after reporting RCU CPU
5228 rcupdate.rcu_cpu_stall_suppress= [KNL]
5229 Suppress RCU CPU stall warning messages.
5231 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL]
5232 Suppress RCU CPU stall warning messages and
5233 rcutorture writer stall warnings that occur
5234 during early boot, that is, during the time
5235 before the init task is spawned.
5237 rcupdate.rcu_cpu_stall_timeout= [KNL]
5238 Set timeout for RCU CPU stall warning messages.
5239 The value is in seconds and the maximum allowed
5240 value is 300 seconds.
5242 rcupdate.rcu_exp_cpu_stall_timeout= [KNL]
5243 Set timeout for expedited RCU CPU stall warning
5244 messages. The value is in milliseconds
5245 and the maximum allowed value is 21000
5246 milliseconds. Please note that this value is
5247 adjusted to an arch timer tick resolution.
5248 Setting this to zero causes the value from
5249 rcupdate.rcu_cpu_stall_timeout to be used (after
5250 conversion from seconds to milliseconds).
5252 rcupdate.rcu_cpu_stall_cputime= [KNL]
5253 Provide statistics on the cputime and count of
5254 interrupts and tasks during the sampling period. For
5255 multiple continuous RCU stalls, all sampling periods
5256 begin at half of the first RCU stall timeout.
5258 rcupdate.rcu_exp_stall_task_details= [KNL]
5259 Print stack dumps of any tasks blocking the
5260 current expedited RCU grace period during an
5261 expedited RCU CPU stall warning.
5263 rcupdate.rcu_expedited= [KNL]
5264 Use expedited grace-period primitives, for
5265 example, synchronize_rcu_expedited() instead
5266 of synchronize_rcu(). This reduces latency,
5267 but can increase CPU utilization, degrade
5268 real-time latency, and degrade energy efficiency.
5269 No effect on CONFIG_TINY_RCU kernels.
5271 rcupdate.rcu_normal= [KNL]
5272 Use only normal grace-period primitives,
5273 for example, synchronize_rcu() instead of
5274 synchronize_rcu_expedited(). This improves
5275 real-time latency, CPU utilization, and
5276 energy efficiency, but can expose users to
5277 increased grace-period latency. This parameter
5278 overrides rcupdate.rcu_expedited. No effect on
5279 CONFIG_TINY_RCU kernels.
5281 rcupdate.rcu_normal_after_boot= [KNL]
5282 Once boot has completed (that is, after
5283 rcu_end_inkernel_boot() has been invoked), use
5284 only normal grace-period primitives. No effect
5285 on CONFIG_TINY_RCU kernels.
5287 But note that CONFIG_PREEMPT_RT=y kernels enables
5288 this kernel boot parameter, forcibly setting
5289 it to the value one, that is, converting any
5290 post-boot attempt at an expedited RCU grace
5291 period to instead use normal non-expedited
5292 grace-period processing.
5294 rcupdate.rcu_task_collapse_lim= [KNL]
5295 Set the maximum number of callbacks present
5296 at the beginning of a grace period that allows
5297 the RCU Tasks flavors to collapse back to using
5298 a single callback queue. This switching only
5299 occurs when rcupdate.rcu_task_enqueue_lim is
5300 set to the default value of -1.
5302 rcupdate.rcu_task_contend_lim= [KNL]
5303 Set the minimum number of callback-queuing-time
5304 lock-contention events per jiffy required to
5305 cause the RCU Tasks flavors to switch to per-CPU
5306 callback queuing. This switching only occurs
5307 when rcupdate.rcu_task_enqueue_lim is set to
5308 the default value of -1.
5310 rcupdate.rcu_task_enqueue_lim= [KNL]
5311 Set the number of callback queues to use for the
5312 RCU Tasks family of RCU flavors. The default
5313 of -1 allows this to be automatically (and
5314 dynamically) adjusted. This parameter is intended
5317 rcupdate.rcu_task_ipi_delay= [KNL]
5318 Set time in jiffies during which RCU tasks will
5319 avoid sending IPIs, starting with the beginning
5320 of a given grace period. Setting a large
5321 number avoids disturbing real-time workloads,
5322 but lengthens grace periods.
5324 rcupdate.rcu_task_lazy_lim= [KNL]
5325 Number of callbacks on a given CPU that will
5326 cancel laziness on that CPU. Use -1 to disable
5327 cancellation of laziness, but be advised that
5328 doing so increases the danger of OOM due to
5331 rcupdate.rcu_task_stall_info= [KNL]
5332 Set initial timeout in jiffies for RCU task stall
5333 informational messages, which give some indication
5334 of the problem for those not patient enough to
5335 wait for ten minutes. Informational messages are
5336 only printed prior to the stall-warning message
5337 for a given grace period. Disable with a value
5338 less than or equal to zero. Defaults to ten
5339 seconds. A change in value does not take effect
5340 until the beginning of the next grace period.
5342 rcupdate.rcu_task_stall_info_mult= [KNL]
5343 Multiplier for time interval between successive
5344 RCU task stall informational messages for a given
5345 RCU tasks grace period. This value is clamped
5346 to one through ten, inclusive. It defaults to
5347 the value three, so that the first informational
5348 message is printed 10 seconds into the grace
5349 period, the second at 40 seconds, the third at
5350 160 seconds, and then the stall warning at 600
5351 seconds would prevent a fourth at 640 seconds.
5353 rcupdate.rcu_task_stall_timeout= [KNL]
5354 Set timeout in jiffies for RCU task stall
5355 warning messages. Disable with a value less
5356 than or equal to zero. Defaults to ten minutes.
5357 A change in value does not take effect until
5358 the beginning of the next grace period.
5360 rcupdate.rcu_tasks_lazy_ms= [KNL]
5361 Set timeout in milliseconds RCU Tasks asynchronous
5362 callback batching for call_rcu_tasks().
5363 A negative value will take the default. A value
5364 of zero will disable batching. Batching is
5365 always disabled for synchronize_rcu_tasks().
5367 rcupdate.rcu_tasks_rude_lazy_ms= [KNL]
5368 Set timeout in milliseconds RCU Tasks
5369 Rude asynchronous callback batching for
5370 call_rcu_tasks_rude(). A negative value
5371 will take the default. A value of zero will
5372 disable batching. Batching is always disabled
5373 for synchronize_rcu_tasks_rude().
5375 rcupdate.rcu_tasks_trace_lazy_ms= [KNL]
5376 Set timeout in milliseconds RCU Tasks
5377 Trace asynchronous callback batching for
5378 call_rcu_tasks_trace(). A negative value
5379 will take the default. A value of zero will
5380 disable batching. Batching is always disabled
5381 for synchronize_rcu_tasks_trace().
5383 rcupdate.rcu_self_test= [KNL]
5384 Run the RCU early boot self tests
5388 Run specified binary instead of /init from the ramdisk,
5389 used for early userspace startup. See initrd.
5392 force - Override the decision by the kernel to hide the
5393 advertisement of RDRAND support (this affects
5394 certain AMD processors because of buggy BIOS
5395 support, specifically around the suspend/resume
5399 Turn on/off individual RDT features. List is:
5400 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp,
5402 E.g. to turn on cmt and turn off mba use:
5406 Format (x86 or x86_64):
5407 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \
5409 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
5411 Where reboot_mode is one of warm (soft) or cold (hard) or gpio
5412 (prefix with 'panic_' to set mode for panic
5414 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
5415 reboot_force is either force or not specified,
5416 reboot_cpu is s[mp]#### with #### being the processor
5417 to be used for rebooting.
5419 refscale.holdoff= [KNL]
5420 Set test-start holdoff period. The purpose of
5421 this parameter is to delay the start of the
5422 test until boot completes in order to avoid
5425 refscale.loops= [KNL]
5426 Set the number of loops over the synchronization
5427 primitive under test. Increasing this number
5428 reduces noise due to loop start/end overhead,
5429 but the default has already reduced the per-pass
5430 noise to a handful of picoseconds on ca. 2020
5433 refscale.nreaders= [KNL]
5434 Set number of readers. The default value of -1
5435 selects N, where N is roughly 75% of the number
5436 of CPUs. A value of zero is an interesting choice.
5438 refscale.nruns= [KNL]
5439 Set number of runs, each of which is dumped onto
5442 refscale.readdelay= [KNL]
5443 Set the read-side critical-section duration,
5444 measured in microseconds.
5446 refscale.scale_type= [KNL]
5447 Specify the read-protection implementation to test.
5449 refscale.shutdown= [KNL]
5450 Shut down the system at the end of the performance
5451 test. This defaults to 1 (shut it down) when
5452 refscale is built into the kernel and to 0 (leave
5453 it running) when refscale is built as a module.
5455 refscale.verbose= [KNL]
5456 Enable additional printk() statements.
5458 refscale.verbose_batched= [KNL]
5459 Batch the additional printk() statements. If zero
5460 (the default) or negative, print everything. Otherwise,
5461 print every Nth verbose statement, where N is the value
5465 [KNL, SMP] Set scheduler's default relax_domain_level.
5466 See Documentation/admin-guide/cgroup-v1/cpusets.rst.
5468 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory
5469 Format: <base1>,<size1>[,<base2>,<size2>,...]
5470 Reserve I/O ports or memory so the kernel won't use
5471 them. If <base> is less than 0x10000, the region
5472 is assumed to be I/O ports; otherwise it is memory.
5474 reservetop= [X86-32]
5476 Reserves a hole at the top of the kernel virtual
5479 reset_devices [KNL] Force drivers to reset the underlying device
5480 during initialization.
5483 Specify the partition device for software suspend
5485 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
5487 resume_offset= [SWSUSP]
5488 Specify the offset from the beginning of the partition
5489 given by "resume=" at which the swap header is located,
5490 in <PAGE_SIZE> units (needed only for swap files).
5491 See Documentation/power/swsusp-and-swap-files.rst
5493 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
5494 read the resume files
5496 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
5497 Useful for devices that are detected asynchronously
5498 (e.g. USB and MMC devices).
5500 retain_initrd [RAM] Keep initrd memory after extraction
5502 retbleed= [X86] Control mitigation of RETBleed (Arbitrary
5503 Speculative Code Execution with Return Instructions)
5506 AMD-based UNRET and IBPB mitigations alone do not stop
5507 sibling threads from influencing the predictions of other
5508 sibling threads. For that reason, STIBP is used on pro-
5509 cessors that support it, and mitigate SMT on processors
5513 auto - automatically select a migitation
5514 auto,nosmt - automatically select a mitigation,
5515 disabling SMT if necessary for
5516 the full mitigation (only on Zen1
5517 and older without STIBP).
5518 ibpb - On AMD, mitigate short speculation
5519 windows on basic block boundaries too.
5520 Safe, highest perf impact. It also
5521 enables STIBP if present. Not suitable
5523 ibpb,nosmt - Like "ibpb" above but will disable SMT
5524 when STIBP is not available. This is
5525 the alternative for systems which do not
5527 unret - Force enable untrained return thunks,
5528 only effective on AMD f15h-f17h based
5530 unret,nosmt - Like unret, but will disable SMT when STIBP
5531 is not available. This is the alternative for
5532 systems which do not have STIBP.
5534 Selecting 'auto' will choose a mitigation method at run
5535 time according to the CPU.
5537 Not specifying this option is equivalent to retbleed=auto.
5539 rfkill.default_state=
5540 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
5541 etc. communication is blocked by default.
5544 rfkill.master_switch_mode=
5545 0 The "airplane mode" button does nothing.
5546 1 The "airplane mode" button toggles between everything
5547 blocked and the previous configuration.
5548 2 The "airplane mode" button toggles between everything
5549 blocked and everything unblocked.
5551 rhash_entries= [KNL,NET]
5552 Set number of hash buckets for route cache
5555 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported
5558 riscv_isa_fallback [RISCV]
5559 When CONFIG_RISCV_ISA_FALLBACK is not enabled, permit
5560 falling back to detecting extension support by parsing
5561 "riscv,isa" property on devicetree systems when the
5562 replacement properties are not found. See the Kconfig
5563 entry for RISCV_ISA_FALLBACK.
5565 ro [KNL] Mount root device read-only on boot
5568 on Mark read-only kernel memory as read-only (default).
5569 off Leave read-only kernel memory writable for debugging.
5570 full Mark read-only kernel memory and aliases as read-only
5574 Enable the uart passthrough on the designated usb port
5575 on Rockchip SoCs. When active, the signals of the
5576 debug-uart get routed to the D+ and D- pins of the usb
5577 port and the regular usb controller gets disabled.
5579 root= [KNL] Root filesystem
5580 Usually this a a block device specifier of some kind,
5581 see the early_lookup_bdev comment in
5582 block/early-lookup.c for details.
5583 Alternatively this can be "ram" for the legacy initial
5584 ramdisk, "nfs" and "cifs" for root on a network file
5585 system, or "mtd" and "ubi" for mounting from raw flash.
5587 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
5588 mount the root filesystem
5590 rootflags= [KNL] Set root filesystem mount option string
5592 rootfstype= [KNL] Set root filesystem type
5594 rootwait [KNL] Wait (indefinitely) for root device to show up.
5595 Useful for devices that are detected asynchronously
5596 (e.g. USB and MMC devices).
5598 rootwait= [KNL] Maximum time (in seconds) to wait for root device
5599 to show up before attempting to mount the root
5602 rproc_mem=nn[KMG][@address]
5603 [KNL,ARM,CMA] Remoteproc physical memory block.
5604 Memory area to be used by remote processor image,
5607 rw [KNL] Mount root device read-write on boot
5609 S [KNL] Run init in single mode
5611 s390_iommu= [HW,S390]
5612 Set s390 IOTLB flushing mode
5614 With strict flushing every unmap operation will result in
5615 an IOTLB flush. Default is lazy flushing before reuse,
5618 s390_iommu_aperture= [KNL,S390]
5619 Specifies the size of the per device DMA address space
5620 accessible through the DMA and IOMMU APIs as a decimal
5621 factor of the size of main memory.
5622 The default is 1 meaning that one can concurrently use
5623 as many DMA addresses as physical memory is installed,
5624 if supported by hardware, and thus map all of memory
5625 once. With a value of 2 one can map all of memory twice
5626 and so on. As a special case a factor of 0 imposes no
5627 restrictions other than those given by hardware at the
5628 cost of significant additional memory use for tables.
5631 See drivers/net/irda/sa1100_ir.c.
5633 sched_verbose [KNL] Enables verbose scheduler debug messages.
5635 schedstats= [KNL,X86] Enable or disable scheduled statistics.
5636 Allowed values are enable and disable. This feature
5637 incurs a small amount of overhead in the scheduler
5638 but is useful for debugging and performance tuning.
5640 sched_thermal_decay_shift=
5641 [KNL, SMP] Set a decay shift for scheduler thermal
5642 pressure signal. Thermal pressure signal follows the
5643 default decay period of other scheduler pelt
5644 signals(usually 32 ms but configurable). Setting
5645 sched_thermal_decay_shift will left shift the decay
5646 period for the thermal pressure signal by the shift
5648 i.e. with the default pelt decay period of 32 ms
5649 sched_thermal_decay_shift thermal pressure decay pr
5653 Format: integer between 0 and 10
5656 scftorture.holdoff= [KNL]
5657 Number of seconds to hold off before starting
5658 test. Defaults to zero for module insertion and
5659 to 10 seconds for built-in smp_call_function()
5662 scftorture.longwait= [KNL]
5663 Request ridiculously long waits randomly selected
5664 up to the chosen limit in seconds. Zero (the
5665 default) disables this feature. Please note
5666 that requesting even small non-zero numbers of
5667 seconds can result in RCU CPU stall warnings,
5668 softlockup complaints, and so on.
5670 scftorture.nthreads= [KNL]
5671 Number of kthreads to spawn to invoke the
5672 smp_call_function() family of functions.
5673 The default of -1 specifies a number of kthreads
5674 equal to the number of CPUs.
5676 scftorture.onoff_holdoff= [KNL]
5677 Number seconds to wait after the start of the
5678 test before initiating CPU-hotplug operations.
5680 scftorture.onoff_interval= [KNL]
5681 Number seconds to wait between successive
5682 CPU-hotplug operations. Specifying zero (which
5683 is the default) disables CPU-hotplug operations.
5685 scftorture.shutdown_secs= [KNL]
5686 The number of seconds following the start of the
5687 test after which to shut down the system. The
5688 default of zero avoids shutting down the system.
5689 Non-zero values are useful for automated tests.
5691 scftorture.stat_interval= [KNL]
5692 The number of seconds between outputting the
5693 current test statistics to the console. A value
5694 of zero disables statistics output.
5696 scftorture.stutter_cpus= [KNL]
5697 The number of jiffies to wait between each change
5698 to the set of CPUs under test.
5700 scftorture.use_cpus_read_lock= [KNL]
5701 Use use_cpus_read_lock() instead of the default
5702 preempt_disable() to disable CPU hotplug
5703 while invoking one of the smp_call_function*()
5706 scftorture.verbose= [KNL]
5707 Enable additional printk() statements.
5709 scftorture.weight_single= [KNL]
5710 The probability weighting to use for the
5711 smp_call_function_single() function with a zero
5712 "wait" parameter. A value of -1 selects the
5713 default if all other weights are -1. However,
5714 if at least one weight has some other value, a
5715 value of -1 will instead select a weight of zero.
5717 scftorture.weight_single_wait= [KNL]
5718 The probability weighting to use for the
5719 smp_call_function_single() function with a
5720 non-zero "wait" parameter. See weight_single.
5722 scftorture.weight_many= [KNL]
5723 The probability weighting to use for the
5724 smp_call_function_many() function with a zero
5725 "wait" parameter. See weight_single.
5726 Note well that setting a high probability for
5727 this weighting can place serious IPI load
5730 scftorture.weight_many_wait= [KNL]
5731 The probability weighting to use for the
5732 smp_call_function_many() function with a
5733 non-zero "wait" parameter. See weight_single
5736 scftorture.weight_all= [KNL]
5737 The probability weighting to use for the
5738 smp_call_function_all() function with a zero
5739 "wait" parameter. See weight_single and
5742 scftorture.weight_all_wait= [KNL]
5743 The probability weighting to use for the
5744 smp_call_function_all() function with a
5745 non-zero "wait" parameter. See weight_single
5748 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
5749 xtime_lock contention on larger systems, and/or RCU lock
5750 contention on all systems with CONFIG_MAXSMP set.
5751 Format: { "0" | "1" }
5752 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
5754 Note: increases power consumption, thus should only be
5755 enabled if running jitter sensitive (HPC/RT) workloads.
5757 security= [SECURITY] Choose a legacy "major" security module to
5758 enable at boot. This has been deprecated by the
5761 selinux= [SELINUX] Disable or enable SELinux at boot time.
5762 Format: { "0" | "1" }
5763 See security/selinux/Kconfig help text.
5768 serialnumber [BUGS=X86-32]
5770 sev=option[,option...] [X86-64] See Documentation/arch/x86/x86_64/boot-options.rst
5773 Maximal number of shapers.
5775 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
5776 Limit apic dumping. The parameter defines the maximal
5777 number of local apics being dumped. Also it is possible
5778 to set it to "all" by meaning -- no limit here.
5779 Format: { 1 (default) | 2 | ... | all }.
5780 The parameter valid if only apic=debug or
5781 apic=verbose is specified.
5782 Example: apic=debug show_lapic=all
5790 Enable merging of slabs with similar size when the
5791 kernel is built without CONFIG_SLAB_MERGE_DEFAULT.
5794 Disable merging of slabs with similar size. May be
5795 necessary if there is some reason to distinguish
5796 allocs to different slabs, especially in hardened
5797 environments where the risk of heap overflows and
5798 layout control by attackers can usually be
5799 frustrated by disabling merging. This will reduce
5800 most of the exposure of a heap attack to a single
5801 cache (risks via metadata attacks are mostly
5802 unchanged). Debug options disable merging on their
5804 For more information see Documentation/mm/slub.rst.
5806 slab_max_order= [MM, SLAB]
5807 Determines the maximum allowed order for slabs.
5808 A high setting may cause OOMs due to memory
5809 fragmentation. Defaults to 1 for systems with
5810 more than 32MB of RAM, 0 otherwise.
5812 slub_debug[=options[,slabs][;[options[,slabs]]...] [MM, SLUB]
5813 Enabling slub_debug allows one to determine the
5814 culprit if slab objects become corrupted. Enabling
5815 slub_debug can create guard zones around objects and
5816 may poison objects when not in use. Also tracks the
5817 last alloc / free. For more information see
5818 Documentation/mm/slub.rst.
5820 slub_max_order= [MM, SLUB]
5821 Determines the maximum allowed order for slabs.
5822 A high setting may cause OOMs due to memory
5823 fragmentation. For more information see
5824 Documentation/mm/slub.rst.
5826 slub_min_objects= [MM, SLUB]
5827 The minimum number of objects per slab. SLUB will
5828 increase the slab order up to slub_max_order to
5829 generate a sufficiently large slab able to contain
5830 the number of objects indicated. The higher the number
5831 of objects the smaller the overhead of tracking slabs
5832 and the less frequently locks need to be acquired.
5833 For more information see Documentation/mm/slub.rst.
5835 slub_min_order= [MM, SLUB]
5836 Determines the minimum page order for slabs. Must be
5837 lower than slub_max_order.
5838 For more information see Documentation/mm/slub.rst.
5840 slub_merge [MM, SLUB]
5841 Same with slab_merge.
5843 slub_nomerge [MM, SLUB]
5844 Same with slab_nomerge. This is supported for legacy.
5845 See slab_nomerge for more information.
5848 Format: <io1>[,<io2>[,...,<io8>]]
5850 smp.csd_lock_timeout= [KNL]
5851 Specify the period of time in milliseconds
5852 that smp_call_function() and friends will wait
5853 for a CPU to release the CSD lock. This is
5854 useful when diagnosing bugs involving CPUs
5855 disabling interrupts for extended periods
5856 of time. Defaults to 5,000 milliseconds, and
5857 setting a value of zero disables this feature.
5858 This feature may be more efficiently disabled
5859 using the csdlock_debug- kernel parameter.
5861 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
5862 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
5863 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
5864 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
5865 smsc-ircc2.ircc_irq= [HW] IRQ line
5866 smsc-ircc2.ircc_dma= [HW] DMA channel
5867 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
5868 0: Toshiba Satellite 1800 (GP data pin select)
5869 1: Fast pin select (default)
5872 smt= [KNL,MIPS,S390] Set the maximum number of threads (logical
5873 CPUs) to use per physical CPU on systems capable of
5874 symmetric multithreading (SMT). Will be capped to the
5875 actual hardware limit.
5877 Default: -1 (no limit)
5880 [KNL] Should the soft-lockup detector generate panics.
5883 A value of 1 instructs the soft-lockup detector
5884 to panic the machine when a soft-lockup occurs. It is
5885 also controlled by the kernel.softlockup_panic sysctl
5886 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the
5887 respective build-time switch to that functionality.
5889 softlockup_all_cpu_backtrace=
5890 [KNL] Should the soft-lockup detector generate
5891 backtraces on all cpus.
5894 sonypi.*= [HW] Sony Programmable I/O Control Device driver
5895 See Documentation/admin-guide/laptops/sonypi.rst
5897 spectre_v2= [X86] Control mitigation of Spectre variant 2
5898 (indirect branch speculation) vulnerability.
5899 The default operation protects the kernel from
5902 on - unconditionally enable, implies
5904 off - unconditionally disable, implies
5906 auto - kernel detects whether your CPU model is
5909 Selecting 'on' will, and 'auto' may, choose a
5910 mitigation method at run time according to the
5911 CPU, the available microcode, the setting of the
5912 CONFIG_RETPOLINE configuration option, and the
5913 compiler with which the kernel was built.
5915 Selecting 'on' will also enable the mitigation
5916 against user space to user space task attacks.
5918 Selecting 'off' will disable both the kernel and
5919 the user space protections.
5921 Specific mitigations can also be selected manually:
5923 retpoline - replace indirect branches
5924 retpoline,generic - Retpolines
5925 retpoline,lfence - LFENCE; indirect branch
5926 retpoline,amd - alias for retpoline,lfence
5927 eibrs - Enhanced/Auto IBRS
5928 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines
5929 eibrs,lfence - Enhanced/Auto IBRS + LFENCE
5930 ibrs - use IBRS to protect kernel
5932 Not specifying this option is equivalent to
5936 [X86] Control mitigation of Spectre variant 2
5937 (indirect branch speculation) vulnerability between
5940 on - Unconditionally enable mitigations. Is
5941 enforced by spectre_v2=on
5943 off - Unconditionally disable mitigations. Is
5944 enforced by spectre_v2=off
5946 prctl - Indirect branch speculation is enabled,
5947 but mitigation can be enabled via prctl
5948 per thread. The mitigation control state
5949 is inherited on fork.
5952 - Like "prctl" above, but only STIBP is
5953 controlled per thread. IBPB is issued
5954 always when switching between different user
5958 - Same as "prctl" above, but all seccomp
5959 threads will enable the mitigation unless
5960 they explicitly opt out.
5963 - Like "seccomp" above, but only STIBP is
5964 controlled per thread. IBPB is issued
5965 always when switching between different
5966 user space processes.
5968 auto - Kernel selects the mitigation depending on
5969 the available CPU features and vulnerability.
5971 Default mitigation: "prctl"
5973 Not specifying this option is equivalent to
5974 spectre_v2_user=auto.
5976 spec_rstack_overflow=
5977 [X86] Control RAS overflow mitigation on AMD Zen CPUs
5979 off - Disable mitigation
5980 microcode - Enable microcode mitigation only
5981 safe-ret - Enable sw-only safe RET mitigation (default)
5982 ibpb - Enable mitigation by issuing IBPB on
5984 ibpb-vmexit - Issue IBPB only on VMEXIT
5985 (cloud-specific mitigation)
5987 spec_store_bypass_disable=
5988 [HW] Control Speculative Store Bypass (SSB) Disable mitigation
5989 (Speculative Store Bypass vulnerability)
5991 Certain CPUs are vulnerable to an exploit against a
5992 a common industry wide performance optimization known
5993 as "Speculative Store Bypass" in which recent stores
5994 to the same memory location may not be observed by
5995 later loads during speculative execution. The idea
5996 is that such stores are unlikely and that they can
5997 be detected prior to instruction retirement at the
5998 end of a particular speculation execution window.
6000 In vulnerable processors, the speculatively forwarded
6001 store can be used in a cache side channel attack, for
6002 example to read memory to which the attacker does not
6003 directly have access (e.g. inside sandboxed code).
6005 This parameter controls whether the Speculative Store
6006 Bypass optimization is used.
6008 On x86 the options are:
6010 on - Unconditionally disable Speculative Store Bypass
6011 off - Unconditionally enable Speculative Store Bypass
6012 auto - Kernel detects whether the CPU model contains an
6013 implementation of Speculative Store Bypass and
6014 picks the most appropriate mitigation. If the
6015 CPU is not vulnerable, "off" is selected. If the
6016 CPU is vulnerable the default mitigation is
6017 architecture and Kconfig dependent. See below.
6018 prctl - Control Speculative Store Bypass per thread
6019 via prctl. Speculative Store Bypass is enabled
6020 for a process by default. The state of the control
6021 is inherited on fork.
6022 seccomp - Same as "prctl" above, but all seccomp threads
6023 will disable SSB unless they explicitly opt out.
6025 Default mitigations:
6028 On powerpc the options are:
6030 on,auto - On Power8 and Power9 insert a store-forwarding
6031 barrier on kernel entry and exit. On Power7
6032 perform a software flush on kernel entry and
6036 Not specifying this option is equivalent to
6037 spec_store_bypass_disable=auto.
6039 spia_io_base= [HW,MTD]
6045 [X86] Enable split lock detection or bus lock detection
6047 When enabled (and if hardware support is present), atomic
6048 instructions that access data across cache line
6049 boundaries will result in an alignment check exception
6050 for split lock detection or a debug exception for
6055 warn - the kernel will emit rate-limited warnings
6056 about applications triggering the #AC
6057 exception or the #DB exception. This mode is
6058 the default on CPUs that support split lock
6059 detection or bus lock detection. Default
6060 behavior is by #AC if both features are
6061 enabled in hardware.
6063 fatal - the kernel will send SIGBUS to applications
6064 that trigger the #AC exception or the #DB
6065 exception. Default behavior is by #AC if
6066 both features are enabled in hardware.
6069 Set system wide rate limit to N bus locks
6070 per second for bus lock detection.
6073 N/A for split lock detection.
6076 If an #AC exception is hit in the kernel or in
6077 firmware (i.e. not while executing in user mode)
6078 the kernel will oops in either "warn" or "fatal"
6081 #DB exception for bus lock is triggered only when
6085 Control the Special Register Buffer Data Sampling
6088 Certain CPUs are vulnerable to an MDS-like
6089 exploit which can leak bits from the random
6092 By default, this issue is mitigated by
6093 microcode. However, the microcode fix can cause
6094 the RDRAND and RDSEED instructions to become
6095 much slower. Among other effects, this will
6096 result in reduced throughput from /dev/urandom.
6098 The microcode mitigation can be disabled with
6099 the following option:
6101 off: Disable mitigation and remove
6102 performance impact to RDRAND and RDSEED
6104 srcutree.big_cpu_lim [KNL]
6105 Specifies the number of CPUs constituting a
6106 large system, such that srcu_struct structures
6107 should immediately allocate an srcu_node array.
6108 This kernel-boot parameter defaults to 128,
6109 but takes effect only when the low-order four
6110 bits of srcutree.convert_to_big is equal to 3
6113 srcutree.convert_to_big [KNL]
6114 Specifies under what conditions an SRCU tree
6115 srcu_struct structure will be converted to big
6116 form, that is, with an rcu_node tree:
6119 1: At init_srcu_struct() time.
6120 2: When rcutorture decides to.
6121 3: Decide at boot time (default).
6122 0x1X: Above plus if high contention.
6124 Either way, the srcu_node tree will be sized based
6125 on the actual runtime number of CPUs (nr_cpu_ids)
6126 instead of the compile-time CONFIG_NR_CPUS.
6128 srcutree.counter_wrap_check [KNL]
6129 Specifies how frequently to check for
6130 grace-period sequence counter wrap for the
6131 srcu_data structure's ->srcu_gp_seq_needed field.
6132 The greater the number of bits set in this kernel
6133 parameter, the less frequently counter wrap will
6134 be checked for. Note that the bottom two bits
6137 srcutree.exp_holdoff [KNL]
6138 Specifies how many nanoseconds must elapse
6139 since the end of the last SRCU grace period for
6140 a given srcu_struct until the next normal SRCU
6141 grace period will be considered for automatic
6142 expediting. Set to zero to disable automatic
6145 srcutree.srcu_max_nodelay [KNL]
6146 Specifies the number of no-delay instances
6147 per jiffy for which the SRCU grace period
6148 worker thread will be rescheduled with zero
6149 delay. Beyond this limit, worker thread will
6150 be rescheduled with a sleep delay of one jiffy.
6152 srcutree.srcu_max_nodelay_phase [KNL]
6153 Specifies the per-grace-period phase, number of
6154 non-sleeping polls of readers. Beyond this limit,
6155 grace period worker thread will be rescheduled
6156 with a sleep delay of one jiffy, between each
6157 rescan of the readers, for a grace period phase.
6159 srcutree.srcu_retry_check_delay [KNL]
6160 Specifies number of microseconds of non-sleeping
6161 delay between each non-sleeping poll of readers.
6163 srcutree.small_contention_lim [KNL]
6164 Specifies the number of update-side contention
6165 events per jiffy will be tolerated before
6166 initiating a conversion of an srcu_struct
6167 structure to big form. Note that the value of
6168 srcutree.convert_to_big must have the 0x10 bit
6169 set for contention-based conversions to occur.
6172 Speculative Store Bypass Disable control
6174 On CPUs that are vulnerable to the Speculative
6175 Store Bypass vulnerability and offer a
6176 firmware based mitigation, this parameter
6177 indicates how the mitigation should be used:
6179 force-on: Unconditionally enable mitigation for
6180 for both kernel and userspace
6181 force-off: Unconditionally disable mitigation for
6182 for both kernel and userspace
6183 kernel: Always enable mitigation in the
6184 kernel, and offer a prctl interface
6185 to allow userspace to register its
6186 interest in being mitigated too.
6188 stack_guard_gap= [MM]
6189 override the default stack gap protection. The value
6190 is in page units and it defines how many pages prior
6191 to (for stacks growing down) resp. after (for stacks
6192 growing up) the main stack are reserved for no other
6193 mapping. Default value is 256 pages.
6195 stack_depot_disable= [KNL]
6196 Setting this to true through kernel command line will
6197 disable the stack depot thereby saving the static memory
6198 consumed by the stack hash table. By default this is set
6202 Enabled the stack tracer on boot up.
6204 stacktrace_filter=[function-list]
6205 [FTRACE] Limit the functions that the stack tracer
6206 will trace at boot up. function-list is a comma-separated
6207 list of functions. This list can be changed at run
6208 time by the stack_trace_filter file in the debugfs
6209 tracing directory. Note, this enables stack tracing
6210 and the stacktrace above is not needed.
6214 Set the STI (builtin display/keyboard on the HP-PARISC
6215 machines) console (graphic card) which should be used
6216 as the initial boot-console.
6217 See also comment in drivers/video/console/sticore.c.
6220 See comment in drivers/video/console/sticore.c.
6223 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
6228 Enable or disable strict sigaltstack size checks
6229 against the required signal frame size which
6230 depends on the supported FPU features. This can
6231 be used to filter out binaries which have
6232 not yet been made aware of AT_MINSIGSTKSZ.
6235 Limits the number of kernel HPT entries in the hash
6236 page table to increase the rate of hash page table
6237 faults on kernel addresses.
6240 Limits the number of kernel SLB entries, and flushes
6241 them frequently to increase the rate of SLB faults
6242 on kernel addresses.
6244 sunrpc.min_resvport=
6245 sunrpc.max_resvport=
6247 SunRPC servers often require that client requests
6248 originate from a privileged port (i.e. a port in the
6249 range 0 < portnr < 1024).
6250 An administrator who wishes to reserve some of these
6251 ports for other uses may adjust the range that the
6252 kernel's sunrpc client considers to be privileged
6253 using these two parameters to set the minimum and
6254 maximum port values.
6256 sunrpc.svc_rpc_per_connection_limit=
6258 Limit the number of requests that the server will
6259 process in parallel from a single connection.
6260 The default value is 0 (no limit).
6264 Control how the NFS server code allocates CPUs to
6265 service thread pools. Depending on how many NICs
6266 you have and where their interrupts are bound, this
6267 option will affect which CPUs will do NFS serving.
6268 Note: this parameter cannot be changed while the
6269 NFS server is running.
6271 auto the server chooses an appropriate mode
6272 automatically using heuristics
6273 global a single global pool contains all CPUs
6274 percpu one pool for each CPU
6275 pernode one pool for each NUMA node (equivalent
6276 to global on non-NUMA machines)
6278 sunrpc.tcp_slot_table_entries=
6279 sunrpc.udp_slot_table_entries=
6281 Sets the upper limit on the number of simultaneous
6282 RPC calls that can be sent from the client to a
6283 server. Increasing these values may allow you to
6284 improve throughput, but will also increase the
6285 amount of memory reserved for use by the client.
6287 suspend.pm_test_delay=
6289 Sets the number of seconds to remain in a suspend test
6290 mode before resuming the system (see
6291 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
6292 is set. Default value is 5.
6295 Format: { on | off | y | n | 1 | 0 }
6296 This parameter controls use of the Protected
6297 Execution Facility on pSeries.
6299 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
6300 Format: { <int> [,<int>] | force | noforce }
6301 <int> -- Number of I/O TLB slabs
6302 <int> -- Second integer after comma. Number of swiotlb
6303 areas with their own lock. Will be rounded up
6305 force -- force using of bounce buffers even if they
6306 wouldn't be automatically used by the kernel
6307 noforce -- Never use bounce buffers (for debugging)
6312 Set a sysctl parameter, right before loading the init
6313 process, as if the value was written to the respective
6314 /proc/sys/... file. Both '.' and '/' are recognized as
6315 separators. Unrecognized parameters and invalid values
6316 are reported in the kernel log. Sysctls registered
6317 later by a loaded module cannot be set this way.
6318 Example: sysctl.vm.swappiness=40
6320 sysrq_always_enabled
6322 Ignore sysrq setting - this boot parameter will
6323 neutralize any effect of /proc/sys/kernel/sysrq.
6324 Useful for debugging.
6326 tcpmhash_entries= [KNL,NET]
6327 Set the number of tcp_metrics_hash slots.
6328 Default value is 8192 or 16384 depending on total
6329 ram pages. This is used to specify the TCP metrics
6330 cache size. See Documentation/networking/ip-sysctl.rst
6331 "tcp_no_metrics_save" section for more details.
6335 test_suspend= [SUSPEND]
6336 Format: { "mem" | "standby" | "freeze" }[,N]
6337 Specify "mem" (for Suspend-to-RAM) or "standby" (for
6338 standby suspend) or "freeze" (for suspend type freeze)
6339 as the system sleep state during system startup with
6340 the optional capability to repeat N number of times.
6341 The system is woken from this state using a
6342 wakeup-capable RTC alarm.
6344 thash_entries= [KNL,NET]
6345 Set number of hash buckets for TCP connection
6347 thermal.act= [HW,ACPI]
6348 -1: disable all active trip points in all thermal zones
6349 <degrees C>: override all lowest active trip points
6351 thermal.crt= [HW,ACPI]
6352 -1: disable all critical trip points in all thermal zones
6353 <degrees C>: override all critical trip points
6355 thermal.off= [HW,ACPI]
6356 1: disable ACPI thermal control
6358 thermal.psv= [HW,ACPI]
6359 -1: disable all passive trip points
6360 <degrees C>: override all passive trip points to this
6363 thermal.tzp= [HW,ACPI]
6364 Specify global default ACPI thermal zone polling rate
6365 <deci-seconds>: poll all this frequency
6366 0: no polling (default)
6369 Force threading of all interrupt handlers except those
6370 marked explicitly IRQF_NO_THREAD.
6374 Specify if the kernel should make use of the cpu
6375 topology information if the hardware supports this.
6376 The scheduler will make use of this information and
6377 e.g. base its process migration decisions on it.
6380 topology_updates= [KNL, PPC, NUMA]
6382 Specify if the kernel should ignore (off)
6383 topology updates sent by the hypervisor to this
6386 torture.disable_onoff_at_boot= [KNL]
6387 Prevent the CPU-hotplug component of torturing
6388 until after init has spawned.
6390 torture.ftrace_dump_at_shutdown= [KNL]
6391 Dump the ftrace buffer at torture-test shutdown,
6392 even if there were no errors. This can be a
6393 very costly operation when many torture tests
6394 are running concurrently, especially on systems
6395 with rotating-rust storage.
6397 torture.verbose_sleep_frequency= [KNL]
6398 Specifies how many verbose printk()s should be
6399 emitted between each sleep. The default of zero
6400 disables verbose-printk() sleeping.
6402 torture.verbose_sleep_duration= [KNL]
6403 Duration of each verbose-printk() sleep in jiffies.
6407 tpm_suspend_pcr=[HW,TPM]
6408 Format: integer pcr id
6409 Specify that at suspend time, the tpm driver
6410 should extend the specified pcr with zeros,
6411 as a workaround for some chips which fail to
6412 flush the last written pcr on TPM_SaveState.
6413 This will guarantee that all the other pcrs
6416 tpm_tis.interrupts= [HW,TPM]
6417 Enable interrupts for the MMIO based physical layer
6418 for the FIFO interface. By default it is set to false
6419 (0). For more information about TPM hardware interfaces
6420 defined by Trusted Computing Group (TCG) see
6421 https://trustedcomputinggroup.org/resource/pc-client-platform-tpm-profile-ptp-specification/
6424 Have the tracepoints sent to printk as well as the
6425 tracing ring buffer. This is useful for early boot up
6426 where the system hangs or reboots and does not give the
6427 option for reading the tracing buffer or performing a
6428 ftrace_dump_on_oops.
6430 To turn off having tracepoints sent to printk,
6431 echo 0 > /proc/sys/kernel/tracepoint_printk
6432 Note, echoing 1 into this file without the
6433 tracepoint_printk kernel cmdline option has no effect.
6435 The tp_printk_stop_on_boot (see below) can also be used
6436 to stop the printing of events to console at
6441 Having tracepoints sent to printk() and activating high
6442 frequency tracepoints such as irq or sched, can cause
6443 the system to live lock.
6445 tp_printk_stop_on_boot [FTRACE]
6446 When tp_printk (above) is set, it can cause a lot of noise
6447 on the console. It may be useful to only include the
6448 printing of events during boot up, as user space may
6449 make the system inoperable.
6451 This command line option will stop the printing of events
6452 to console at the late_initcall_sync() time frame.
6454 trace_buf_size=nn[KMG]
6455 [FTRACE] will set tracing buffer size on each cpu.
6457 trace_clock= [FTRACE] Set the clock used for tracing events
6459 local - Use the per CPU time stamp counter
6460 (converted into nanoseconds). Fast, but
6461 depending on the architecture, may not be
6462 in sync between CPUs.
6463 global - Event time stamps are synchronize across
6464 CPUs. May be slower than the local clock,
6465 but better for some race conditions.
6466 counter - Simple counting of events (1, 2, ..)
6467 note, some counts may be skipped due to the
6468 infrastructure grabbing the clock more than
6470 uptime - Use jiffies as the time stamp.
6471 perf - Use the same clock that perf uses.
6472 mono - Use ktime_get_mono_fast_ns() for time stamps.
6473 mono_raw - Use ktime_get_raw_fast_ns() for time
6475 boot - Use ktime_get_boot_fast_ns() for time stamps.
6476 Architectures may add more clocks. See
6477 Documentation/trace/ftrace.rst for more details.
6479 trace_event=[event-list]
6480 [FTRACE] Set and start specified trace events in order
6481 to facilitate early boot debugging. The event-list is a
6482 comma-separated list of trace events to enable. See
6483 also Documentation/trace/events.rst
6485 trace_instance=[instance-info]
6486 [FTRACE] Create a ring buffer instance early in boot up.
6487 This will be listed in:
6489 /sys/kernel/tracing/instances
6491 Events can be enabled at the time the instance is created
6494 trace_instance=<name>,<system1>:<event1>,<system2>:<event2>
6496 Note, the "<system*>:" portion is optional if the event is
6499 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall
6501 will enable the "sched_switch" event (note, the "sched:" is optional, and
6502 the same thing would happen if it was left off). The irq_handler_entry
6503 event, and all events under the "initcall" system.
6505 trace_options=[option-list]
6506 [FTRACE] Enable or disable tracer options at boot.
6507 The option-list is a comma delimited list of options
6508 that can be enabled or disabled just as if you were
6509 to echo the option name into
6511 /sys/kernel/tracing/trace_options
6513 For example, to enable stacktrace option (to dump the
6514 stack trace of each event), add to the command line:
6516 trace_options=stacktrace
6518 See also Documentation/trace/ftrace.rst "trace options"
6521 trace_trigger=[trigger-list]
6522 [FTRACE] Add a event trigger on specific events.
6523 Set a trigger on top of a specific event, with an optional
6526 The format is is "trace_trigger=<event>.<trigger>[ if <filter>],..."
6527 Where more than one trigger may be specified that are comma deliminated.
6531 trace_trigger="sched_switch.stacktrace if prev_state == 2"
6533 The above will enable the "stacktrace" trigger on the "sched_switch"
6534 event but only trigger it if the "prev_state" of the "sched_switch"
6535 event is "2" (TASK_UNINTERUPTIBLE).
6537 See also "Event triggers" in Documentation/trace/events.rst
6541 [FTRACE] enable this option to disable tracing when a
6542 warning is hit. This turns off "tracing_on". Tracing can
6543 be enabled again by echoing '1' into the "tracing_on"
6544 file located in /sys/kernel/tracing/
6546 This option is useful, as it disables the trace before
6547 the WARNING dump is called, which prevents the trace to
6548 be filled with content caused by the warning output.
6550 This option can also be set at run time via the sysctl
6551 option: kernel/traceoff_on_warning
6553 transparent_hugepage=
6555 Format: [always|madvise|never]
6556 Can be used to control the default behavior of the system
6557 with respect to transparent hugepages.
6558 See Documentation/admin-guide/mm/transhuge.rst
6561 trusted.source= [KEYS]
6563 This parameter identifies the trust source as a backend
6564 for trusted keys implementation. Supported trust
6569 If not specified then it defaults to iterating through
6570 the trust source list starting with TPM and assigns the
6571 first trust source as a backend which is initialized
6572 successfully during iteration.
6576 The RNG used to generate key material for trusted keys.
6579 - the same value as trusted.source: "tpm" or "tee"
6581 If not specified, "default" is used. In this case,
6582 the RNG's choice is left to each individual trust source.
6584 tsc= Disable clocksource stability checks for TSC.
6586 [x86] reliable: mark tsc clocksource as reliable, this
6587 disables clocksource verification at runtime, as well
6588 as the stability checks done at bootup. Used to enable
6589 high-resolution timer mode on older hardware, and in
6590 virtualized environment.
6591 [x86] noirqtime: Do not use TSC to do irq accounting.
6592 Used to run time disable IRQ_TIME_ACCOUNTING on any
6593 platforms where RDTSC is slow and this accounting
6595 [x86] unstable: mark the TSC clocksource as unstable, this
6596 marks the TSC unconditionally unstable at bootup and
6597 avoids any further wobbles once the TSC watchdog notices.
6598 [x86] nowatchdog: disable clocksource watchdog. Used
6599 in situations with strict latency requirements (where
6600 interruptions from clocksource watchdog are not
6602 [x86] recalibrate: force recalibration against a HW timer
6603 (HPET or PM timer) on systems whose TSC frequency was
6604 obtained from HW or FW using either an MSR or CPUID(0x15).
6605 Warn if the difference is more than 500 ppm.
6606 [x86] watchdog: Use TSC as the watchdog clocksource with
6607 which to check other HW timers (HPET or PM timer), but
6608 only on systems where TSC has been deemed trustworthy.
6609 This will be suppressed by an earlier tsc=nowatchdog and
6610 can be overridden by a later tsc=nowatchdog. A console
6611 message will flag any such suppression or overriding.
6613 tsc_early_khz= [X86] Skip early TSC calibration and use the given
6614 value instead. Useful when the early TSC frequency discovery
6615 procedure is not reliable, such as on overclocked systems
6616 with CPUID.16h support and partial CPUID.15h support.
6617 Format: <unsigned int>
6619 tsx= [X86] Control Transactional Synchronization
6620 Extensions (TSX) feature in Intel processors that
6621 support TSX control.
6623 This parameter controls the TSX feature. The options are:
6625 on - Enable TSX on the system. Although there are
6626 mitigations for all known security vulnerabilities,
6627 TSX has been known to be an accelerator for
6628 several previous speculation-related CVEs, and
6629 so there may be unknown security risks associated
6630 with leaving it enabled.
6632 off - Disable TSX on the system. (Note that this
6633 option takes effect only on newer CPUs which are
6634 not vulnerable to MDS, i.e., have
6635 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get
6636 the new IA32_TSX_CTRL MSR through a microcode
6637 update. This new MSR allows for the reliable
6638 deactivation of the TSX functionality.)
6640 auto - Disable TSX if X86_BUG_TAA is present,
6641 otherwise enable TSX on the system.
6643 Not specifying this option is equivalent to tsx=off.
6645 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
6648 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async
6649 Abort (TAA) vulnerability.
6651 Similar to Micro-architectural Data Sampling (MDS)
6652 certain CPUs that support Transactional
6653 Synchronization Extensions (TSX) are vulnerable to an
6654 exploit against CPU internal buffers which can forward
6655 information to a disclosure gadget under certain
6658 In vulnerable processors, the speculatively forwarded
6659 data can be used in a cache side channel attack, to
6660 access data to which the attacker does not have direct
6663 This parameter controls the TAA mitigation. The
6666 full - Enable TAA mitigation on vulnerable CPUs
6669 full,nosmt - Enable TAA mitigation and disable SMT on
6670 vulnerable CPUs. If TSX is disabled, SMT
6671 is not disabled because CPU is not
6672 vulnerable to cross-thread TAA attacks.
6673 off - Unconditionally disable TAA mitigation
6675 On MDS-affected machines, tsx_async_abort=off can be
6676 prevented by an active MDS mitigation as both vulnerabilities
6677 are mitigated with the same mechanism so in order to disable
6678 this mitigation, you need to specify mds=off too.
6680 Not specifying this option is equivalent to
6681 tsx_async_abort=full. On CPUs which are MDS affected
6682 and deploy MDS mitigation, TAA mitigation is not
6683 required and doesn't provide any additional
6687 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
6689 turbografx.map[2|3]= [HW,JOY]
6690 TurboGraFX parallel port interface
6692 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
6693 See also Documentation/input/devices/joystick-parport.rst
6695 udbg-immortal [PPC] When debugging early kernel crashes that
6696 happen after console_init() and before a proper
6697 console driver takes over, this boot options might
6698 help "seeing" what's going on.
6700 uhash_entries= [KNL,NET]
6701 Set number of hash buckets for UDP/UDP-Lite connections
6704 [USB] Ignore overcurrent events (default N).
6705 Some badly-designed motherboards generate lots of
6706 bogus events, for ports that aren't wired to
6707 anything. Set this parameter to avoid log spamming.
6708 Note that genuine overcurrent events won't be
6712 [X86] Cause panic on unknown NMI.
6714 unwind_debug [X86-64]
6715 Enable unwinder debug output. This can be
6716 useful for debugging certain unwinder error
6717 conditions, including corrupt stacks and
6718 bad/missing unwinder metadata.
6720 usbcore.authorized_default=
6721 [USB] Default USB device authorization:
6722 (default -1 = authorized (same as 1),
6723 0 = not authorized, 1 = authorized, 2 = authorized
6724 if device connected to internal port)
6726 usbcore.autosuspend=
6727 [USB] The autosuspend time delay (in seconds) used
6728 for newly-detected USB devices (default 2). This
6729 is the time required before an idle device will be
6730 autosuspended. Devices for which the delay is set
6731 to a negative value won't be autosuspended at all.
6733 usbcore.usbfs_snoop=
6734 [USB] Set to log all usbfs traffic (default 0 = off).
6736 usbcore.usbfs_snoop_max=
6737 [USB] Maximum number of bytes to snoop in each URB
6740 usbcore.blinkenlights=
6741 [USB] Set to cycle leds on hubs (default 0 = off).
6743 usbcore.old_scheme_first=
6744 [USB] Start with the old device initialization
6745 scheme (default 0 = off).
6747 usbcore.usbfs_memory_mb=
6748 [USB] Memory limit (in MB) for buffers allocated by
6749 usbfs (default = 16, 0 = max = 2047).
6751 usbcore.use_both_schemes=
6752 [USB] Try the other device initialization scheme
6753 if the first one fails (default 1 = enabled).
6755 usbcore.initial_descriptor_timeout=
6756 [USB] Specifies timeout for the initial 64-byte
6757 USB_REQ_GET_DESCRIPTOR request in milliseconds
6758 (default 5000 = 5.0 seconds).
6760 usbcore.nousb [USB] Disable the USB subsystem
6763 [USB] A list of quirk entries to augment the built-in
6764 usb core quirk list. List entries are separated by
6765 commas. Each entry has the form
6766 VendorID:ProductID:Flags. The IDs are 4-digit hex
6767 numbers and Flags is a set of letters. Each letter
6768 will change the built-in quirk; setting it if it is
6769 clear and clearing it if it is set. The letters have
6770 the following meanings:
6771 a = USB_QUIRK_STRING_FETCH_255 (string
6772 descriptors must not be fetched using
6774 b = USB_QUIRK_RESET_RESUME (device can't resume
6775 correctly so reset it instead);
6776 c = USB_QUIRK_NO_SET_INTF (device can't handle
6777 Set-Interface requests);
6778 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't
6779 handle its Configuration or Interface
6781 e = USB_QUIRK_RESET (device can't be reset
6782 (e.g morph devices), don't use reset);
6783 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has
6784 more interface descriptions than the
6785 bNumInterfaces count, and can't handle
6786 talking to these interfaces);
6787 g = USB_QUIRK_DELAY_INIT (device needs a pause
6788 during initialization, after we read
6789 the device descriptor);
6790 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For
6791 high speed and super speed interrupt
6792 endpoints, the USB 2.0 and USB 3.0 spec
6793 require the interval in microframes (1
6794 microframe = 125 microseconds) to be
6795 calculated as interval = 2 ^
6797 Devices with this quirk report their
6798 bInterval as the result of this
6799 calculation instead of the exponent
6800 variable used in the calculation);
6801 i = USB_QUIRK_DEVICE_QUALIFIER (device can't
6802 handle device_qualifier descriptor
6804 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device
6805 generates spurious wakeup, ignore
6806 remote wakeup capability);
6807 k = USB_QUIRK_NO_LPM (device can't handle Link
6809 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL
6810 (Device reports its bInterval as linear
6811 frames instead of the USB 2.0
6813 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs
6814 to be disconnected before suspend to
6815 prevent spurious wakeup);
6816 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a
6817 pause after every control message);
6818 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra
6819 delay after resetting its port);
6820 Example: quirks=0781:5580:bk,0a5c:5834:gij
6823 [USBHID] The interval which mice are to be polled at.
6826 [USBHID] The interval which joysticks are to be polled at.
6829 [USBHID] The interval which keyboards are to be polled at.
6831 usb-storage.delay_use=
6832 [UMS] The delay in seconds before a new device is
6833 scanned for Logical Units (default 1).
6836 [UMS] A list of quirks entries to supplement or
6837 override the built-in unusual_devs list. List
6838 entries are separated by commas. Each entry has
6839 the form VID:PID:Flags where VID and PID are Vendor
6840 and Product ID values (4-digit hex numbers) and
6841 Flags is a set of characters, each corresponding
6842 to a common usb-storage quirk flag as follows:
6843 a = SANE_SENSE (collect more than 18 bytes
6844 of sense data, not on uas);
6845 b = BAD_SENSE (don't collect more than 18
6846 bytes of sense data, not on uas);
6847 c = FIX_CAPACITY (decrease the reported
6848 device capacity by one sector);
6849 d = NO_READ_DISC_INFO (don't use
6850 READ_DISC_INFO command, not on uas);
6851 e = NO_READ_CAPACITY_16 (don't use
6852 READ_CAPACITY_16 command);
6853 f = NO_REPORT_OPCODES (don't use report opcodes
6855 g = MAX_SECTORS_240 (don't transfer more than
6856 240 sectors at a time, uas only);
6857 h = CAPACITY_HEURISTICS (decrease the
6858 reported device capacity by one
6859 sector if the number is odd);
6860 i = IGNORE_DEVICE (don't bind to this
6862 j = NO_REPORT_LUNS (don't use report luns
6864 k = NO_SAME (do not use WRITE_SAME, uas only)
6865 l = NOT_LOCKABLE (don't try to lock and
6866 unlock ejectable media, not on uas);
6867 m = MAX_SECTORS_64 (don't transfer more
6868 than 64 sectors = 32 KB at a time,
6870 n = INITIAL_READ10 (force a retry of the
6871 initial READ(10) command, not on uas);
6872 o = CAPACITY_OK (accept the capacity
6873 reported by the device, not on uas);
6874 p = WRITE_CACHE (the device cache is ON
6875 by default, not on uas);
6876 r = IGNORE_RESIDUE (the device reports
6877 bogus residue values, not on uas);
6878 s = SINGLE_LUN (the device has only one
6880 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
6881 commands, uas only);
6882 u = IGNORE_UAS (don't bind to the uas driver);
6883 w = NO_WP_DETECT (don't test whether the
6884 medium is write-protected).
6885 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE
6886 even if the device claims no cache,
6888 Example: quirks=0419:aaf5:rl,0421:0433:rc
6890 user_debug= [KNL,ARM]
6892 See arch/arm/Kconfig.debug help text.
6893 1 - undefined instruction events
6895 4 - invalid data aborts
6898 Example: user_debug=31
6901 [X86] Flags controlling user PTE allocations.
6903 nohigh = do not allocate PTE pages in
6904 HIGHMEM regardless of setting
6907 vdso= [X86,SH,SPARC]
6908 On X86_32, this is an alias for vdso32=. Otherwise:
6910 vdso=1: enable VDSO (the default)
6911 vdso=0: disable VDSO mapping
6913 vdso32= [X86] Control the 32-bit vDSO
6914 vdso32=1: enable 32-bit VDSO
6915 vdso32=0 or vdso32=2: disable 32-bit VDSO
6917 See the help text for CONFIG_COMPAT_VDSO for more
6918 details. If CONFIG_COMPAT_VDSO is set, the default is
6919 vdso32=0; otherwise, the default is vdso32=1.
6921 For compatibility with older kernels, vdso32=2 is an
6924 Try vdso32=0 if you encounter an error that says:
6925 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
6928 vector=percpu: enable percpu vector domain
6930 video= [FB] Frame buffer configuration
6931 See Documentation/fb/modedb.rst.
6933 video.brightness_switch_enabled= [ACPI]
6935 If set to 1, on receiving an ACPI notify event
6936 generated by hotkey, video driver will adjust brightness
6937 level and then send out the event to user space through
6938 the allocated input device. If set to 0, video driver
6939 will only send out the event without touching backlight
6944 [VMMIO] Memory mapped virtio (platform) device.
6946 <size>@<baseaddr>:<irq>[:<id>]
6948 <size> := size (can use standard suffixes
6950 <baseaddr> := physical base address
6951 <irq> := interrupt number (as passed to
6953 <id> := (optional) platform device id
6955 virtio_mmio.device=1K@0x100b0000:48:7
6957 Can be used multiple times for multiple devices.
6959 vga= [BOOT,X86-32] Select a particular video mode
6960 See Documentation/arch/x86/boot.rst and
6961 Documentation/admin-guide/svga.rst.
6962 Use vga=ask for menu.
6963 This is actually a boot loader parameter; the value is
6964 passed to the kernel using a special protocol.
6966 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y.
6967 May slow down system boot speed, especially when
6968 enabled on systems with a large amount of memory.
6969 All options are enabled by default, and this
6970 interface is meant to allow for selectively
6971 enabling or disabling specific virtual memory
6974 Available options are:
6975 P Enable page structure init time poisoning
6976 - Disable all of the above options
6978 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
6979 size of <nn>. This can be used to increase the
6980 minimum size (128MB on x86). It can also be used to
6981 decrease the size and leave more room for directly
6984 vmcp_cma=nn[MG] [KNL,S390]
6985 Sets the memory size reserved for contiguous memory
6986 allocations for the vmcp device driver.
6988 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
6991 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
6994 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
6998 Controls the behavior of vsyscalls (i.e. calls to
6999 fixed addresses of 0xffffffffff600x00 from legacy
7000 code). Most statically-linked binaries and older
7001 versions of glibc use these calls. Because these
7002 functions are at fixed addresses, they make nice
7003 targets for exploits that can control RIP.
7005 emulate Vsyscalls turn into traps and are emulated
7006 reasonably safely. The vsyscall page is
7009 xonly [default] Vsyscalls turn into traps and are
7010 emulated reasonably safely. The vsyscall
7011 page is not readable.
7013 none Vsyscalls don't work at all. This makes
7014 them quite hard to use for exploits but
7015 might break your system.
7017 vt.color= [VT] Default text color.
7018 Format: 0xYX, X = foreground, Y = background.
7019 Default: 0x07 = light gray on black.
7021 vt.cur_default= [VT] Default cursor shape.
7022 Format: 0xCCBBAA, where AA, BB, and CC are the same as
7023 the parameters of the <Esc>[?A;B;Cc escape sequence;
7024 see VGA-softcursor.txt. Default: 2 = underline.
7026 vt.default_blu= [VT]
7027 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
7028 Change the default blue palette of the console.
7029 This is a 16-member array composed of values
7032 vt.default_grn= [VT]
7033 Format: <green0>,<green1>,<green2>,...,<green15>
7034 Change the default green palette of the console.
7035 This is a 16-member array composed of values
7038 vt.default_red= [VT]
7039 Format: <red0>,<red1>,<red2>,...,<red15>
7040 Change the default red palette of the console.
7041 This is a 16-member array composed of values
7047 Set system-wide default UTF-8 mode for all tty's.
7048 Default is 1, i.e. UTF-8 mode is enabled for all
7049 newly opened terminals.
7051 vt.global_cursor_default=
7054 Set system-wide default for whether a cursor
7055 is shown on new VTs. Default is -1,
7056 i.e. cursors will be created by default unless
7057 overridden by individual drivers. 0 will hide
7058 cursors, 1 will display them.
7060 vt.italic= [VT] Default color for italic text; 0-15.
7063 vt.underline= [VT] Default color for underlined text; 0-15.
7066 watchdog timers [HW,WDT] For information on watchdog timers,
7067 see Documentation/watchdog/watchdog-parameters.rst
7068 or other driver-specific files in the
7069 Documentation/watchdog/ directory.
7073 Set the hard lockup detector stall duration
7074 threshold in seconds. The soft lockup detector
7075 threshold is set to twice the value. A value of 0
7076 disables both lockup detectors. Default is 10
7079 workqueue.unbound_cpus=
7080 [KNL,SMP] Specify to constrain one or some CPUs
7081 to use in unbound workqueues.
7083 By default, all online CPUs are available for
7086 workqueue.watchdog_thresh=
7087 If CONFIG_WQ_WATCHDOG is configured, workqueue can
7088 warn stall conditions and dump internal state to
7089 help debugging. 0 disables workqueue stall
7090 detection; otherwise, it's the stall threshold
7091 duration in seconds. The default value is 30 and
7092 it can be updated at runtime by writing to the
7093 corresponding sysfs file.
7095 workqueue.cpu_intensive_thresh_us=
7096 Per-cpu work items which run for longer than this
7097 threshold are automatically considered CPU intensive
7098 and excluded from concurrency management to prevent
7099 them from noticeably delaying other per-cpu work
7100 items. Default is 10000 (10ms).
7102 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel
7103 will report the work functions which violate this
7104 threshold repeatedly. They are likely good
7105 candidates for using WQ_UNBOUND workqueues instead.
7107 workqueue.power_efficient
7108 Per-cpu workqueues are generally preferred because
7109 they show better performance thanks to cache
7110 locality; unfortunately, per-cpu workqueues tend to
7111 be more power hungry than unbound workqueues.
7113 Enabling this makes the per-cpu workqueues which
7114 were observed to contribute significantly to power
7115 consumption unbound, leading to measurably lower
7116 power usage at the cost of small performance
7119 The default value of this parameter is determined by
7120 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
7122 workqueue.default_affinity_scope=
7123 Select the default affinity scope to use for unbound
7124 workqueues. Can be one of "cpu", "smt", "cache",
7125 "numa" and "system". Default is "cache". For more
7126 information, see the Affinity Scopes section in
7127 Documentation/core-api/workqueue.rst.
7129 This can be changed after boot by writing to the
7130 matching /sys/module/workqueue/parameters file. All
7131 workqueues with the "default" affinity scope will be
7132 updated accordignly.
7134 workqueue.debug_force_rr_cpu
7135 Workqueue used to implicitly guarantee that work
7136 items queued without explicit CPU specified are put
7137 on the local CPU. This guarantee is no longer true
7138 and while local CPU is still preferred work items
7139 may be put on foreign CPUs. This debug option
7140 forces round-robin CPU selection to flush out
7141 usages which depend on the now broken guarantee.
7142 When enabled, memory and cache locality will be
7145 writecombine= [LOONGARCH] Control the MAT (Memory Access Type) of
7148 on - Enable writecombine, use WUC for ioremap_wc()
7149 off - Disable writecombine, use SUC for ioremap_wc()
7151 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
7152 default x2apic cluster mode on platforms
7155 xen_512gb_limit [KNL,X86-64,XEN]
7156 Restricts the kernel running paravirtualized under Xen
7157 to use only up to 512 GB of RAM. The reason to do so is
7158 crash analysis tools and Xen tools for doing domain
7159 save/restore/migration must be enabled to handle larger
7162 xen_emul_unplug= [HW,X86,XEN]
7163 Unplug Xen emulated devices
7164 Format: [unplug0,][unplug1]
7165 ide-disks -- unplug primary master IDE devices
7166 aux-ide-disks -- unplug non-primary-master IDE devices
7167 nics -- unplug network devices
7168 all -- unplug all emulated devices (NICs and IDE disks)
7169 unnecessary -- unplugging emulated devices is
7170 unnecessary even if the host did not respond to
7172 never -- do not unplug even if version check succeeds
7174 xen_legacy_crash [X86,XEN]
7175 Crash from Xen panic notifier, without executing late
7176 panic() code such as dumping handler.
7178 xen_msr_safe= [X86,XEN]
7180 Select whether to always use non-faulting (safe) MSR
7181 access functions when running as Xen PV guest. The
7182 default value is controlled by CONFIG_XEN_PV_MSR_SAFE.
7184 xen_nopvspin [X86,XEN]
7185 Disables the qspinlock slowpath using Xen PV optimizations.
7186 This parameter is obsoleted by "nopvspin" parameter, which
7187 has equivalent effect for XEN platform.
7190 Disables the PV optimizations forcing the HVM guest to
7191 run as generic HVM guest with no PV drivers.
7192 This option is obsoleted by the "nopv" option, which
7193 has equivalent effect for XEN platform.
7195 xen_no_vector_callback
7196 [KNL,X86,XEN] Disable the vector callback for Xen
7197 event channel interrupts.
7199 xen_scrub_pages= [XEN]
7200 Boolean option to control scrubbing pages before giving them back
7201 to Xen, for use by other domains. Can be also changed at runtime
7202 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages.
7203 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT.
7205 xen_timer_slop= [X86-64,XEN]
7206 Set the timer slop (in nanoseconds) for the virtual Xen
7207 timers (default is 100000). This adjusts the minimum
7208 delta of virtualized Xen timers, where lower values
7209 improve timer resolution at the expense of processing
7210 more timer interrupts.
7212 xen.balloon_boot_timeout= [XEN]
7213 The time (in seconds) to wait before giving up to boot
7214 in case initial ballooning fails to free enough memory.
7215 Applies only when running as HVM or PVH guest and
7216 started with less memory configured than allowed at
7217 max. Default is 180.
7219 xen.event_eoi_delay= [XEN]
7220 How long to delay EOI handling in case of event
7221 storms (jiffies). Default is 10.
7223 xen.event_loop_timeout= [XEN]
7224 After which time (jiffies) the event handling loop
7225 should start to delay EOI handling. Default is 2.
7227 xen.fifo_events= [XEN]
7228 Boolean parameter to disable using fifo event handling
7229 even if available. Normally fifo event handling is
7230 preferred over the 2-level event handling, as it is
7231 fairer and the number of possible event channels is
7232 much higher. Default is on (use fifo events).
7234 xirc2ps_cs= [NET,PCMCIA]
7236 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
7239 By default on POWER9 and above, the kernel will
7240 natively use the XIVE interrupt controller. This option
7241 allows the fallback firmware mode to be used:
7243 off Fallback to firmware control of XIVE interrupt
7244 controller on both pseries and powernv
7245 platforms. Only useful on POWER9 and above.
7247 xive.store-eoi=off [PPC]
7248 By default on POWER10 and above, the kernel will use
7249 stores for EOI handling when the XIVE interrupt mode
7250 is active. This option allows the XIVE driver to use
7251 loads instead, as on POWER9.
7253 xhci-hcd.quirks [USB,KNL]
7254 A hex value specifying bitmask with supplemental xhci
7255 host controller quirks. Meaning of each bit can be
7256 consulted in header drivers/usb/host/xhci.h.
7259 Format: { early | on | rw | ro | off }
7260 Controls if xmon debugger is enabled. Default is off.
7261 Passing only "xmon" is equivalent to "xmon=early".
7262 early Call xmon as early as possible on boot; xmon
7263 debugger is called from setup_arch().
7264 on xmon debugger hooks will be installed so xmon
7265 is only called on a kernel crash. Default mode,
7266 i.e. either "ro" or "rw" mode, is controlled
7267 with CONFIG_XMON_DEFAULT_RO_MODE.
7268 rw xmon debugger hooks will be installed so xmon
7269 is called only on a kernel crash, mode is write,
7270 meaning SPR registers, memory and, other data
7271 can be written using xmon commands.
7272 ro same as "rw" option above but SPR registers,
7273 memory, and other data can't be written using
7275 off xmon is disabled.
projects / linux-2.6-microblaze.git / blob