1 acpi= [HW,ACPI,X86,ARM64]
2 Advanced Configuration and Power Interface
3 Format: { force | on | off | strict | noirq | rsdt |
5 force -- enable ACPI if default was off
6 on -- enable ACPI but allow fallback to DT [arm64]
7 off -- disable ACPI if default was on
8 noirq -- do not use ACPI for IRQ routing
9 strict -- Be less tolerant of platforms that are not
10 strictly ACPI specification compliant.
11 rsdt -- prefer RSDT over (default) XSDT
12 copy_dsdt -- copy DSDT to memory
13 For ARM64, ONLY "acpi=off", "acpi=on" or "acpi=force"
16 See also Documentation/power/runtime_pm.txt, pci=noacpi
18 acpi_apic_instance= [ACPI, IOAPIC]
20 2: use 2nd APIC table, if available
21 1,0: use 1st APIC table
24 acpi_backlight= [HW,ACPI]
27 If set to vendor, prefer vendor specific driver
28 (e.g. thinkpad_acpi, sony_acpi, etc.) instead
29 of the ACPI video.ko driver.
31 acpi_force_32bit_fadt_addr
32 force FADT to use 32 bit addresses rather than the
33 64 bit X_* addresses. Some firmware have broken 64
34 bit addresses for force ACPI ignore these and use
35 the older legacy 32 bit addresses.
37 acpica_no_return_repair [HW, ACPI]
38 Disable AML predefined validation mechanism
39 This mechanism can repair the evaluation result to make
40 the return objects more ACPI specification compliant.
41 This option is useful for developers to identify the
42 root cause of an AML interpreter issue when the issue
43 has something to do with the repair mechanism.
45 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG]
46 acpi.debug_level= [HW,ACPI,ACPI_DEBUG]
48 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
49 debug output. Bits in debug_layer correspond to a
50 _COMPONENT in an ACPI source file, e.g.,
51 #define _COMPONENT ACPI_PCI_COMPONENT
52 Bits in debug_level correspond to a level in
53 ACPI_DEBUG_PRINT statements, e.g.,
54 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
55 The debug_level mask defaults to "info". See
56 Documentation/acpi/debug.txt for more information about
57 debug layers and levels.
59 Enable processor driver info messages:
60 acpi.debug_layer=0x20000000
61 Enable PCI/PCI interrupt routing info messages:
62 acpi.debug_layer=0x400000
63 Enable AML "Debug" output, i.e., stores to the Debug
64 object while interpreting AML:
65 acpi.debug_layer=0xffffffff acpi.debug_level=0x2
66 Enable all messages related to ACPI hardware:
67 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
69 Some values produce so much output that the system is
70 unusable. The "log_buf_len" parameter may be useful
71 if you need to capture more output.
73 acpi_enforce_resources= [ACPI]
75 Check for resource conflicts between native drivers
76 and ACPI OperationRegions (SystemIO and SystemMemory
77 only). IO ports and memory declared in ACPI might be
78 used by the ACPI subsystem in arbitrary AML code and
79 can interfere with legacy drivers.
80 strict (default): access to resources claimed by ACPI
81 is denied; legacy drivers trying to access reserved
82 resources will fail to bind to device using them.
83 lax: access to resources claimed by ACPI is allowed;
84 legacy drivers trying to access reserved resources
85 will bind successfully but a warning message is logged.
86 no: ACPI OperationRegions are not marked as reserved,
87 no further checks are performed.
89 acpi_force_table_verification [HW,ACPI]
90 Enable table checksum verification during early stage.
91 By default, this is disabled due to x86 early mapping
94 acpi_irq_balance [HW,ACPI]
95 ACPI will balance active IRQs
98 acpi_irq_nobalance [HW,ACPI]
99 ACPI will not move active IRQs (default)
102 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA
103 Format: <irq>,<irq>...
105 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for
107 Format: <irq>,<irq>...
109 acpi_mask_gpe= [HW,ACPI]
110 Due to the existence of _Lxx/_Exx, some GPEs triggered
111 by unsupported hardware/firmware features can result in
112 GPE floodings that cannot be automatically disabled by
114 This facility can be used to prevent such uncontrolled
118 acpi_no_auto_serialize [HW,ACPI]
119 Disable auto-serialization of AML methods
120 AML control methods that contain the opcodes to create
121 named objects will be marked as "Serialized" by the
122 auto-serialization feature.
123 This feature is enabled by default.
124 This option allows to turn off the feature.
126 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump
129 acpi_no_static_ssdt [HW,ACPI]
130 Disable installation of static SSDTs at early boot time
131 By default, SSDTs contained in the RSDT/XSDT will be
132 installed automatically and they will appear under
133 /sys/firmware/acpi/tables.
134 This option turns off this feature.
135 Note that specifying this option does not affect
136 dynamic table installation which will install SSDT
137 tables to /sys/firmware/acpi/tables/dynamic.
139 acpi_rsdp= [ACPI,EFI,KEXEC]
140 Pass the RSDP address to the kernel, mostly used
141 on machines running EFI runtime service to boot the
142 second kernel for kdump.
144 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
145 Format: To spoof as Windows 98: ="Microsoft Windows"
147 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead
148 of 2 which is mandated by ACPI 6) as the supported ACPI
149 specification revision (when using this switch, it may
150 be necessary to carry out a cold reboot _twice_ in a
151 row to make it take effect on the platform firmware).
153 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
154 acpi_osi="string1" # add string1
155 acpi_osi="!string2" # remove string2
156 acpi_osi=!* # remove all strings
157 acpi_osi=! # disable all built-in OS vendor
159 acpi_osi=!! # enable all built-in OS vendor
161 acpi_osi= # disable all strings
163 'acpi_osi=!' can be used in combination with single or
164 multiple 'acpi_osi="string1"' to support specific OS
165 vendor string(s). Note that such command can only
166 affect the default state of the OS vendor strings, thus
167 it cannot affect the default state of the feature group
168 strings and the current state of the OS vendor strings,
169 specifying it multiple times through kernel command line
170 is meaningless. This command is useful when one do not
171 care about the state of the feature group strings which
172 should be controlled by the OSPM.
174 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
175 to 'acpi_osi="Windows 2000" acpi_osi=!', they all
176 can make '_OSI("Windows 2000")' TRUE.
178 'acpi_osi=' cannot be used in combination with other
179 'acpi_osi=' command lines, the _OSI method will not
180 exist in the ACPI namespace. NOTE that such command can
181 only affect the _OSI support state, thus specifying it
182 multiple times through kernel command line is also
185 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
188 'acpi_osi=!*' can be used in combination with single or
189 multiple 'acpi_osi="string1"' to support specific
190 string(s). Note that such command can affect the
191 current state of both the OS vendor strings and the
192 feature group strings, thus specifying it multiple times
193 through kernel command line is meaningful. But it may
194 still not able to affect the final state of a string if
195 there are quirks related to this string. This command
196 is useful when one want to control the state of the
197 feature group strings to debug BIOS issues related to
200 1. 'acpi_osi="Module Device" acpi_osi=!*' can make
201 '_OSI("Module Device")' FALSE.
202 2. 'acpi_osi=!* acpi_osi="Module Device"' can make
203 '_OSI("Module Device")' TRUE.
204 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
206 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
208 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
209 they all will make '_OSI("Windows 2000")' TRUE.
212 Override the pmtimer bug detection: force the kernel
213 to assume that this machine's pmtimer latches its value
214 and always returns good values.
216 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
217 Format: { level | edge | high | low }
219 acpi_skip_timer_override [HW,ACPI]
220 Recognize and ignore IRQ0/pin2 Interrupt Override.
221 For broken nForce2 BIOS resulting in XT-PIC timer.
223 acpi_sleep= [HW,ACPI] Sleep options
224 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
225 old_ordering, nonvs, sci_force_enable, nobl }
226 See Documentation/power/video.txt for information on
228 s3_beep is for debugging; it makes the PC's speaker beep
229 as soon as the kernel's real-mode entry point is called.
230 s4_nohwsig prevents ACPI hardware signature from being
231 used during resume from hibernation.
232 old_ordering causes the ACPI 1.0 ordering of the _PTS
233 control method, with respect to putting devices into
234 low power states, to be enforced (the ACPI 2.0 ordering
235 of _PTS is used by default).
236 nonvs prevents the kernel from saving/restoring the
237 ACPI NVS memory during suspend/hibernation and resume.
238 sci_force_enable causes the kernel to set SCI_EN directly
239 on resume from S1/S3 (which is against the ACPI spec,
240 but some broken systems don't work without it).
241 nobl causes the internal blacklist of systems known to
242 behave incorrectly in some ways with respect to system
243 suspend and resume to be ignored (use wisely).
245 acpi_use_timer_override [HW,ACPI]
246 Use timer override. For some broken Nvidia NF5 boards
247 that require a timer override, but don't have HPET
249 add_efi_memmap [EFI; X86] Include EFI memory map in
250 kernel's map of available physical RAM.
253 { off | try_unsupported }
254 off: disable AGP support
255 try_unsupported: try to drive unsupported chipsets
256 (may crash computer or cause data corruption)
259 See Documentation/sound/alsa-configuration.rst
262 Allow the default userspace alignment fault handler
263 behaviour to be specified. Bit 0 enables warnings,
264 bit 1 enables fixups, and bit 2 sends a segfault.
266 align_va_addr= [X86-64]
267 Align virtual addresses by clearing slice [14:12] when
268 allocating a VMA at process creation time. This option
269 gives you up to 3% performance improvement on AMD F15h
270 machines (where it is enabled by default) for a
271 CPU-intensive style benchmark, and it can vary highly in
272 a microbenchmark depending on workload and compiler.
274 32: only for 32-bit processes
275 64: only for 64-bit processes
276 on: enable for both 32- and 64-bit processes
277 off: disable for both 32- and 64-bit processes
279 alloc_snapshot [FTRACE]
280 Allocate the ftrace snapshot buffer on boot up when the
281 main buffer is allocated. This is handy if debugging
282 and you need to use tracing_snapshot() on boot up, and
283 do not want to use tracing_snapshot_alloc() as it needs
284 to be done where GFP_KERNEL allocations are allowed.
286 amd_iommu= [HW,X86-64]
287 Pass parameters to the AMD IOMMU driver in the system.
289 fullflush - enable flushing of IO/TLB entries when
290 they are unmapped. Otherwise they are
291 flushed before they will be reused, which
293 off - do not initialize any AMD IOMMU found in
295 force_isolation - Force device isolation for all
296 devices. The IOMMU driver is not
297 allowed anymore to lift isolation
298 requirements as needed. This option
299 does not override iommu=pt
301 amd_iommu_dump= [HW,X86-64]
302 Enable AMD IOMMU driver option to dump the ACPI table
303 for AMD IOMMU. With this option enabled, AMD IOMMU
304 driver will print ACPI tables for AMD IOMMU during
305 IOMMU initialization.
307 amd_iommu_intr= [HW,X86-64]
308 Specifies one of the following AMD IOMMU interrupt
310 legacy - Use legacy interrupt remapping mode.
311 vapic - Use virtual APIC mode, which allows IOMMU
312 to inject interrupts directly into guest.
313 This mode requires kvm-amd.avic=1.
314 (Default when IOMMU HW support is present.)
316 amijoy.map= [HW,JOY] Amiga joystick support
317 Map of devices attached to JOY0DAT and JOY1DAT
319 See also Documentation/input/joydev/joystick.rst
321 analog.map= [HW,JOY] Analog joystick and gamepad support
322 Specifies type or capabilities of an analog joystick
323 connected to one of 16 gameports
324 Format: <type1>,<type2>,..<type16>
327 Power management functions (SPARCstation-4/5 + deriv.)
329 Disable APC CPU standby support. SPARCstation-Fox does
330 not play well with APC CPU idle - disable it if you have
331 APC and your system crashes randomly.
333 apic= [APIC,X86] Advanced Programmable Interrupt Controller
334 Change the output verbosity whilst booting
335 Format: { quiet (default) | verbose | debug }
336 Change the amount of debugging information output
337 when initialising the APIC and IO-APIC components.
338 For X86-32, this can also be used to specify an APIC
340 Format: apic=driver_name
341 Examples: apic=bigsmp
343 apic_extnmi= [APIC,X86] External NMI delivery setting
344 Format: { bsp (default) | all | none }
345 bsp: External NMI is delivered only to CPU 0
346 all: External NMIs are broadcast to all CPUs as a
348 none: External NMI is masked for all CPUs. This is
349 useful so that a dump capture kernel won't be
353 See Documentation/networking/ipv6.txt.
355 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
356 Limit apic dumping. The parameter defines the maximal
357 number of local apics being dumped. Also it is possible
358 to set it to "all" by meaning -- no limit here.
359 Format: { 1 (default) | 2 | ... | all }.
360 The parameter valid if only apic=debug or
361 apic=verbose is specified.
362 Example: apic=debug show_lapic=all
364 apm= [APM] Advanced Power Management
365 See header of arch/x86/kernel/apm_32.c.
367 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
368 Format: <io>,<irq>,<nodeID>
372 atarimouse= [HW,MOUSE] Atari Mouse
374 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
375 EzKey and similar keyboards
377 atkbd.reset= [HW] Reset keyboard during initialization
379 atkbd.set= [HW] Select keyboard code set
380 Format: <int> (2 = AT (default), 3 = PS/2)
382 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
385 atkbd.softraw= [HW] Choose between synthetic and real raw mode
386 Format: <bool> (0 = real, 1 = synthetic (default))
388 atkbd.softrepeat= [HW]
389 Use software keyboard repeat
391 audit= [KNL] Enable the audit sub-system
392 Format: { "0" | "1" | "off" | "on" }
393 0 | off - kernel audit is disabled and can not be
394 enabled until the next reboot
395 unset - kernel audit is initialized but disabled and
396 will be fully enabled by the userspace auditd.
397 1 | on - kernel audit is initialized and partially
398 enabled, storing at most audit_backlog_limit
399 messages in RAM until it is fully enabled by the
403 audit_backlog_limit= [KNL] Set the audit queue size limit.
404 Format: <int> (must be >=0)
407 bau= [X86_UV] Enable the BAU on SGI UV. The default
408 behavior is to disable the BAU (i.e. bau=0).
409 Format: { "0" | "1" }
412 unset - Disable the BAU.
414 baycom_epp= [HW,AX25]
417 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
419 See header of drivers/net/hamradio/baycom_par.c.
421 baycom_ser_fdx= [HW,AX25]
422 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
423 Format: <io>,<irq>,<mode>[,<baud>]
424 See header of drivers/net/hamradio/baycom_ser_fdx.c.
426 baycom_ser_hdx= [HW,AX25]
427 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
428 Format: <io>,<irq>,<mode>
429 See header of drivers/net/hamradio/baycom_ser_hdx.c.
431 blkdevparts= Manual partition parsing of block device(s) for
432 embedded devices based on command line input.
433 See Documentation/block/cmdline-partition.txt
435 boot_delay= Milliseconds to delay each printk during boot.
436 Values larger than 10 seconds (10000) are changed to
440 bootmem_debug [KNL] Enable bootmem allocator debug messages.
443 Disable BERT OS support on buggy BIOSes.
445 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
446 bttv.radio= Most important insmod options are available as
448 bttv.pll= See Documentation/media/v4l-drivers/bttv.rst
451 bulk_remove=off [PPC] This parameter disables the use of the pSeries
452 firmware feature for flushing multiple hpte entries
455 c101= [NET] Moxa C101 synchronous serial card
457 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
458 Sometimes CPU hardware bugs make them report the cache
459 size incorrectly. The kernel will attempt work arounds
460 to fix known problems, but for some CPUs it is not
461 possible to determine what the correct size should be.
462 This option provides an override for these situations.
464 ca_keys= [KEYS] This parameter identifies a specific key(s) on
465 the system trusted keyring to be used for certificate
467 format: { id:<keyid> | builtin }
469 cca= [MIPS] Override the kernel pages' cache coherency
470 algorithm. Accepted values range from 0 to 7
471 inclusive. See arch/mips/include/asm/pgtable-bits.h
472 for platform specific values (SB1, Loongson3 and
475 ccw_timeout_log [S390]
476 See Documentation/s390/CommonIO for details.
478 cgroup_disable= [KNL] Disable a particular controller
479 Format: {name of the controller(s) to disable}
480 The effects of cgroup_disable=foo are:
481 - foo isn't auto-mounted if you mount all cgroups in
483 - foo isn't visible as an individually mountable
485 {Currently only "memory" controller deal with this and
486 cut the overhead, others just disable the usage. So
487 only cgroup_disable=memory is actually worthy}
489 cgroup_no_v1= [KNL] Disable one, multiple, all cgroup controllers in v1
490 Format: { controller[,controller...] | "all" }
491 Like cgroup_disable, but only applies to cgroup v1;
492 the blacklisted controllers remain available in cgroup2.
494 cgroup.memory= [KNL] Pass options to the cgroup memory controller.
496 nosocket -- Disable socket memory accounting.
497 nokmem -- Disable kernel memory accounting.
499 checkreqprot [SELINUX] Set initial checkreqprot flag value.
500 Format: { "0" | "1" }
501 See security/selinux/Kconfig help text.
502 0 -- check protection applied by kernel (includes
503 any implied execute protection).
504 1 -- check protection requested by application.
505 Default value is set via a kernel config option.
506 Value can be changed at runtime via
507 /selinux/checkreqprot.
510 See Documentation/s390/CommonIO for details.
513 Prevents the clock framework from automatically gating
514 clocks that have not been explicitly enabled by a Linux
515 device driver but are enabled in hardware at reset or
516 by the bootloader/firmware. Note that this does not
517 force such clocks to be always-on nor does it reserve
518 those clocks in any way. This parameter is useful for
519 debug and development, but should not be needed on a
520 platform with proper driver support. For more
521 information, see Documentation/driver-api/clk.rst.
523 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
525 Forces specified clocksource (if available) to be used
526 when calculating gettimeofday(). If specified
527 clocksource is not available, it defaults to PIT.
528 Format: { pit | tsc | cyclone | pmtmr }
530 clocksource= Override the default clocksource
532 Override the default clocksource and use the clocksource
533 with the name specified.
534 Some clocksource names to choose from, depending on
536 [all] jiffies (this is the base, fallback clocksource)
538 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
539 pxa_timer,timer3,32k_counter,timer0_1
540 [X86-32] pit,hpet,tsc;
541 scx200_hrt on Geode; cyclone on IBM x440
549 clocksource.arm_arch_timer.evtstrm=
552 Enable/disable the eventstream feature of the ARM
553 architected timer so that code using WFE-based polling
554 loops can be debugged more effectively on production
557 clearcpuid=BITNUM [X86]
558 Disable CPUID feature X for the kernel. See
559 arch/x86/include/asm/cpufeatures.h for the valid bit
560 numbers. Note the Linux specific bits are not necessarily
561 stable over kernel options, but the vendor specific
563 Also note that user programs calling CPUID directly
564 or using the feature without checking anything
565 will still see it. This just prevents it from
566 being used by the kernel or shown in /proc/cpuinfo.
567 Also note the kernel might malfunction if you disable
570 cma=nn[MG]@[start[MG][-end[MG]]]
572 Sets the size of kernel global memory area for
573 contiguous memory allocations and optionally the
574 placement constraint by the physical address range of
575 memory allocations. A value of 0 disables CMA
576 altogether. For more information, see
577 include/linux/dma-contiguous.h
579 cmo_free_hint= [PPC] Format: { yes | no }
580 Specify whether pages are marked as being inactive
581 when they are freed. This is used in CMO environments
582 to determine OS memory pressure for page stealing by
586 coherent_pool=nn[KMG] [ARM,KNL]
587 Sets the size of memory pool for coherent, atomic dma
588 allocations, by default set to 256K.
590 com20020= [HW,NET] ARCnet - COM20020 chipset
592 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
594 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
598 ARCnet - COM90xx chipset (memory-mapped buffers)
599 Format: <io>[,<irq>[,<memstart>]]
601 condev= [HW,S390] console device
604 console= [KNL] Output console device and options.
606 tty<n> Use the virtual console device <n>.
610 Use the specified serial port. The options are of
611 the form "bbbbpnf", where "bbbb" is the baud rate,
612 "p" is parity ("n", "o", or "e"), "n" is number of
613 bits, and "f" is flow control ("r" for RTS or
614 omit it). Default is "9600n8".
616 See Documentation/admin-guide/serial-console.rst for more
618 Documentation/networking/netconsole.txt for an
621 uart[8250],io,<addr>[,options]
622 uart[8250],mmio,<addr>[,options]
623 uart[8250],mmio16,<addr>[,options]
624 uart[8250],mmio32,<addr>[,options]
625 uart[8250],0x<addr>[,options]
626 Start an early, polled-mode console on the 8250/16550
627 UART at the specified I/O port or MMIO address,
628 switching to the matching ttyS device later.
629 MMIO inter-register address stride is either 8-bit
630 (mmio), 16-bit (mmio16), or 32-bit (mmio32).
631 If none of [io|mmio|mmio16|mmio32], <addr> is assumed
632 to be equivalent to 'mmio'. 'options' are specified in
633 the same format described for ttyS above; if unspecified,
634 the h/w is not re-initialized.
636 hvc<n> Use the hypervisor console device <n>. This is for
637 both Xen and PowerPC hypervisors.
639 If the device connected to the port is not a TTY but a braille
640 device, prepend "brl," before the device type, for instance
642 For now, only VisioBraille is supported.
645 [KNL] Change console messages format
647 By default we print messages on consoles in
648 "[time stamp] text\n" format (time stamp may not be
649 printed, depending on CONFIG_PRINTK_TIME or
650 `printk_time' param).
652 Switch to syslog format: "<%u>[time stamp] text\n"
653 IOW, each message will have a facility and loglevel
654 prefix. The format is similar to one used by syslog()
655 syscall, or to executing "dmesg -S --raw" or to reading
658 consoleblank= [KNL] The console blank (screen saver) timeout in
659 seconds. A value of 0 disables the blank timer.
663 [KNL] Change the default value for
664 /proc/<pid>/coredump_filter.
665 See also Documentation/filesystems/proc.txt.
667 coresight_cpu_debug.enable
670 Enable/disable the CPU sampling based debugging.
671 0: default value, disable debugging
672 1: enable debugging at boot time
674 cpuidle.off=1 [CPU_IDLE]
675 disable the cpuidle sub-system
678 [CPU_IDLE] Name of the cpuidle governor to use.
680 cpufreq.off=1 [CPU_FREQ]
681 disable the cpufreq sub-system
684 [X86] Delay for N microsec between assert and de-assert
685 of APIC INIT to start processors. This delay occurs
686 on every CPU online, such as boot, and resume from suspend.
689 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
691 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
693 crashkernel=size[KMG][@offset[KMG]]
694 [KNL] Using kexec, Linux can switch to a 'crash kernel'
695 upon panic. This parameter reserves the physical
696 memory region [offset, offset + size] for that kernel
697 image. If '@offset' is omitted, then a suitable offset
698 is selected automatically. Check
699 Documentation/kdump/kdump.txt for further details.
701 crashkernel=range1:size1[,range2:size2,...][@offset]
702 [KNL] Same as above, but depends on the memory
703 in the running system. The syntax of range is
704 start-[end] where start and end are both
705 a memory unit (amount[KMG]). See also
706 Documentation/kdump/kdump.txt for an example.
708 crashkernel=size[KMG],high
709 [KNL, x86_64] range could be above 4G. Allow kernel
710 to allocate physical memory region from top, so could
711 be above 4G if system have more than 4G ram installed.
712 Otherwise memory region will be allocated below 4G, if
714 It will be ignored if crashkernel=X is specified.
715 crashkernel=size[KMG],low
716 [KNL, x86_64] range under 4G. When crashkernel=X,high
717 is passed, kernel could allocate physical memory region
718 above 4G, that cause second kernel crash on system
719 that require some amount of low memory, e.g. swiotlb
720 requires at least 64M+32K low memory, also enough extra
721 low memory is needed to make sure DMA buffers for 32-bit
722 devices won't run out. Kernel would try to allocate at
723 at least 256M below 4G automatically.
724 This one let user to specify own low range under 4G
725 for second kernel instead.
726 0: to disable low allocation.
727 It will be ignored when crashkernel=X,high is not used
728 or memory reserved is below 4G.
731 [KNL] Disable crypto self-tests
736 cs89x0_media= [HW,NET]
737 Format: { rj45 | aui | bnc }
740 See header of drivers/s390/block/dasd_devmap.c.
742 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
743 (one device per port)
744 Format: <port#>,<type>
745 See also Documentation/input/devices/joystick-parport.rst
747 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
749 Documentation/admin-guide/dynamic-debug-howto.rst for
750 details. Deprecated, see dyndbg.
752 debug [KNL] Enable kernel debugging (events log level).
755 [KNL] Enable printing [hashed] pointers early in the
756 boot sequence. If enabled, we use a weak hash instead
757 of siphash to hash pointers. Use this option if you are
758 seeing instances of '(___ptrval___)') and need to see a
759 value (hashed pointer) instead. Cryptographically
760 insecure, please do not use on production kernels.
763 [KNL] verbose self-tests
765 Print debugging info while doing the locking API
767 We default to 0 (no extra messages), setting it to
768 1 will print _a lot_ more information - normally
769 only useful to kernel developers.
771 debug_objects [KNL] Enable object debugging
774 [KNL] Disable object debugging
776 debug_guardpage_minorder=
777 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
778 parameter allows control of the order of pages that will
779 be intentionally kept free (and hence protected) by the
780 buddy allocator. Bigger value increase the probability
781 of catching random memory corruption, but reduce the
782 amount of memory for normal system use. The maximum
783 possible value is MAX_ORDER/2. Setting this parameter
784 to 1 or 2 should be enough to identify most random
785 memory corruption problems caused by bugs in kernel or
786 driver code when a CPU writes to (or reads from) a
787 random memory location. Note that there exists a class
788 of memory corruptions problems caused by buggy H/W or
789 F/W or by drivers badly programing DMA (basically when
790 memory is written at bus level and the CPU MMU is
791 bypassed) which are not detectable by
792 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
793 tracking down these problems.
796 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
797 parameter enables the feature at boot time. In
798 default, it is disabled. We can avoid allocating huge
799 chunk of memory for debug pagealloc if we don't enable
800 it at boot time and the system will work mostly same
801 with the kernel built without CONFIG_DEBUG_PAGEALLOC.
802 on: enable the feature
804 debugpat [X86] Enable PAT debugging
806 decnet.addr= [HW,NET]
807 Format: <area>[,<node>]
808 See also Documentation/networking/decnet.txt.
811 [same as hugepagesz=] The size of the default
812 HugeTLB page size. This is the size represented by
813 the legacy /proc/ hugepages APIs, used for SHM, and
814 default size when mounting hugetlbfs filesystems.
815 Defaults to the default architecture's huge page size
818 deferred_probe_timeout=
819 [KNL] Debugging option to set a timeout in seconds for
820 deferred probe to give up waiting on dependencies to
821 probe. Only specific dependencies (subsystems or
822 drivers) that have opted in will be ignored. A timeout of 0
823 will timeout at the end of initcalls. This option will also
824 dump out devices still on the deferred probe list after
828 Set number of hash buckets for dentry cache.
830 disable_1tb_segments [PPC]
831 Disables the use of 1TB hash page table segments. This
832 causes the kernel to fall back to 256MB segments which
833 can be useful when debugging issues that require an SLB
837 See Documentation/networking/ipv6.txt.
840 [KNL] Under CONFIG_HARDENED_USERCOPY, whether
841 hardening is enabled for this boot. Hardened
842 usercopy checking is used to protect the kernel
843 from reading or writing beyond known memory
844 allocation boundaries as a proactive defense
845 against bounds-checking flaws in the kernel's
846 copy_to_user()/copy_from_user() interface.
847 on Perform hardened usercopy checks (default).
848 off Disable hardened usercopy checks.
851 Disable RADIX MMU mode on POWER9
853 disable_cpu_apicid= [X86,APIC,SMP]
855 The number of initial APIC ID for the
856 corresponding CPU to be disabled at boot,
857 mostly used for the kdump 2nd kernel to
858 disable BSP to wake up multiple CPUs without
859 causing system reset or hang due to sending
862 perf_v4_pmi= [X86,INTEL]
864 Disable Intel PMU counter freezing feature.
865 The feature only exists starting from
866 Arch Perfmon v4 (Skylake and newer).
868 disable_ddw [PPC/PSERIES]
869 Disable Dynamic DMA Window support. Use this if
870 to workaround buggy firmware.
873 See Documentation/networking/ipv6.txt.
875 disable_mtrr_cleanup [X86]
876 The kernel tries to adjust MTRR layout from continuous
877 to discrete, to make X server driver able to add WB
878 entry later. This parameter disables that.
880 disable_mtrr_trim [X86, Intel and AMD only]
881 By default the kernel will trim any uncacheable
882 memory out of your available memory pool based on
883 MTRR settings. This parameter disables that behavior,
884 possibly causing your machine to run very slowly.
886 disable_timer_pin_1 [X86]
887 Disable PIN 1 of APIC timer
888 Can be useful to work around chipset bugs.
890 dis_ucode_ldr [X86] Disable the microcode loader.
892 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
893 this option disables the debugging code at boot.
895 dma_debug_entries=<number>
896 This option allows to tune the number of preallocated
897 entries for DMA-API debugging code. One entry is
898 required per DMA-API allocation. Use this if the
899 DMA-API debugging code disables itself because the
900 architectural default is too low.
902 dma_debug_driver=<driver_name>
903 With this option the DMA-API debugging driver
904 filter feature can be enabled at boot time. Just
905 pass the driver to filter for as the parameter.
906 The filter can be disabled or changed to another
907 driver later using sysfs.
909 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
910 Broken monitors, graphic adapters, KVMs and EDIDless
911 panels may send no or incorrect EDID data sets.
912 This parameter allows to specify an EDID data sets
913 in the /lib/firmware directory that are used instead.
914 Generic built-in EDID data sets are used, if one of
915 edid/1024x768.bin, edid/1280x1024.bin,
916 edid/1680x1050.bin, or edid/1920x1080.bin is given
917 and no file with the same name exists. Details and
918 instructions how to build your own EDID data are
919 available in Documentation/EDID/HOWTO.txt. An EDID
920 data set will only be used for a particular connector,
921 if its name and a colon are prepended to the EDID
922 name. Each connector may use a unique EDID data
923 set by separating the files with a comma. An EDID
924 data set with no connector name will be used for
925 any connectors not explicitly specified.
930 Format: {"off" | "known"}
931 Control how the dt_cpu_ftrs device-tree binding is
932 used for CPU feature discovery and setup (if it
934 off: Do not use it, fall back to legacy cpu table.
935 known: Do not pass through unknown features to guests
936 or userspace, only those that the kernel is aware of.
938 dump_apple_properties [X86]
939 Dump name and content of EFI device properties on
940 x86 Macs. Useful for driver authors to determine
941 what data is available or for reverse-engineering.
943 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
944 module.dyndbg[="val"]
945 Enable debug messages at boot time. See
946 Documentation/admin-guide/dynamic-debug-howto.rst
949 nompx [X86] Disables Intel Memory Protection Extensions.
950 See Documentation/x86/intel_mpx.txt for more
951 information about the feature.
953 nopku [X86] Disable Memory Protection Keys CPU feature found
956 module.async_probe [KNL]
957 Enable asynchronous probe on this module.
959 early_ioremap_debug [KNL]
960 Enable debug messages in early_ioremap support. This
961 is useful for tracking down temporary early mappings
962 which are not unmapped.
964 earlycon= [KNL] Output early console device and options.
966 [ARM64] The early console is determined by the
967 stdout-path property in device tree's chosen node,
968 or determined by the ACPI SPCR table.
970 [X86] When used with no options the early console is
971 determined by the ACPI SPCR table.
973 cdns,<addr>[,options]
974 Start an early, polled-mode console on a Cadence
975 (xuartps) serial port at the specified address. Only
976 supported option is baud rate. If baud rate is not
977 specified, the serial port must already be setup and
980 uart[8250],io,<addr>[,options]
981 uart[8250],mmio,<addr>[,options]
982 uart[8250],mmio32,<addr>[,options]
983 uart[8250],mmio32be,<addr>[,options]
984 uart[8250],0x<addr>[,options]
985 Start an early, polled-mode console on the 8250/16550
986 UART at the specified I/O port or MMIO address.
987 MMIO inter-register address stride is either 8-bit
988 (mmio) or 32-bit (mmio32 or mmio32be).
989 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
990 to be equivalent to 'mmio'. 'options' are specified
991 in the same format described for "console=ttyS<n>"; if
992 unspecified, the h/w is not initialized.
996 Start an early, polled-mode console on a pl011 serial
997 port at the specified address. The pl011 serial port
998 must already be setup and configured. Options are not
999 yet supported. If 'mmio32' is specified, then only
1000 the driver will use only 32-bit accessors to read/write
1001 the device registers.
1004 Start an early, polled-mode console on a meson serial
1005 port at the specified address. The serial port must
1006 already be setup and configured. Options are not yet
1010 Start an early, polled-mode console on an msm serial
1011 port at the specified address. The serial port
1012 must already be setup and configured. Options are not
1015 msm_serial_dm,<addr>
1016 Start an early, polled-mode console on an msm serial
1017 dm port at the specified address. The serial port
1018 must already be setup and configured. Options are not
1022 Start an early, polled-mode console on a serial port
1023 of an Actions Semi SoC, such as S500 or S900, at the
1024 specified address. The serial port must already be
1025 setup and configured. Options are not yet supported.
1027 smh Use ARM semihosting calls for early console.
1035 Use early console provided by serial driver available
1036 on Samsung SoCs, requires selecting proper type and
1037 a correct base address of the selected UART port. The
1038 serial port must already be setup and configured.
1039 Options are not yet supported.
1042 Start an early, polled-mode console on a lantiq serial
1043 (lqasc) port at the specified address. The serial port
1044 must already be setup and configured. Options are not
1049 Use early console provided by Freescale LP UART driver
1050 found on Freescale Vybrid and QorIQ LS1021A processors.
1051 A valid base address must be provided, and the serial
1052 port must already be setup and configured.
1055 Start an early, polled-mode console on the
1056 Armada 3700 serial port at the specified
1057 address. The serial port must already be setup
1058 and configured. Options are not yet supported.
1061 Start an early, polled-mode console on a Qualcomm
1062 Generic Interface (GENI) based serial port at the
1063 specified address. The serial port must already be
1064 setup and configured. Options are not yet supported.
1066 earlyprintk= [X86,SH,ARM,M68k,S390]
1071 earlyprintk=serial[,ttySn[,baudrate]]
1072 earlyprintk=serial[,0x...[,baudrate]]
1073 earlyprintk=ttySn[,baudrate]
1074 earlyprintk=dbgp[debugController#]
1075 earlyprintk=pciserial[,force],bus:device.function[,baudrate]
1076 earlyprintk=xdbc[xhciController#]
1078 earlyprintk is useful when the kernel crashes before
1079 the normal console is initialized. It is not enabled by
1080 default because it has some cosmetic problems.
1082 Append ",keep" to not disable it when the real console
1085 Only one of vga, efi, serial, or usb debug port can
1088 Currently only ttyS0 and ttyS1 may be specified by
1089 name. Other I/O ports may be explicitly specified
1090 on some architectures (x86 and arm at least) by
1091 replacing ttySn with an I/O port address, like this:
1092 earlyprintk=serial,0x1008,115200
1093 You can find the port for a given device in
1094 /proc/tty/driver/serial:
1095 2: uart:ST16650V2 port:00001008 irq:18 ...
1097 Interaction with the standard serial driver is not
1100 The VGA and EFI output is eventually overwritten by
1103 The xen output can only be used by Xen PV guests.
1105 The sclp output can only be used on s390.
1107 The optional "force" to "pciserial" enables use of a
1108 PCI device even when its classcode is not of the
1111 edac_report= [HW,EDAC] Control how to report EDAC event
1112 Format: {"on" | "off" | "force"}
1113 on: enable EDAC to report H/W event. May be overridden
1114 by other higher priority error reporting module.
1115 off: disable H/W event reporting through EDAC.
1116 force: enforce the use of EDAC to report H/W event.
1119 ekgdboc= [X86,KGDB] Allow early kernel console debugging
1122 This is designed to be used in conjunction with
1123 the boot argument: earlyprintk=vga
1126 Format: {"off" | "on" | "skip[mbr]"}
1129 Format: { "old_map", "nochunk", "noruntime", "debug" }
1130 old_map [X86-64]: switch to the old ioremap-based EFI
1131 runtime services mapping. 32-bit still uses this one by
1133 nochunk: disable reading files in "chunks" in the EFI
1134 boot stub, as chunking can cause problems with some
1135 firmware implementations.
1136 noruntime : disable EFI runtime services support
1137 debug: enable misc debug output
1139 efi_no_storage_paranoia [EFI; X86]
1140 Using this parameter you can use more than 50% of
1141 your efi variable storage. Use this parameter only if
1142 you are really sure that your UEFI does sane gc and
1143 fulfills the spec otherwise your board may brick.
1145 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1146 Add arbitrary attribute to specific memory range by
1147 updating original EFI memory map.
1148 Region of memory which aa attribute is added to is
1150 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1151 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1152 attribute is added to range 0x100000000-0x180000000 and
1153 0x10a0000000-0x1120000000.
1155 Using this parameter you can do debugging of EFI memmap
1156 related feature. For example, you can do debugging of
1157 Address Range Mirroring feature even if your box
1160 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT
1161 that is to be dynamically loaded by Linux. If there are
1162 multiple variables with the same name but with different
1163 vendor GUIDs, all of them will be loaded. See
1164 Documentation/acpi/ssdt-overlays.txt for details.
1167 eisa_irq_edge= [PARISC,HW]
1168 See header of drivers/parisc/eisa.c.
1171 See comment before function elanfreq_setup() in
1172 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1175 Format: {"cfq" | "deadline" | "noop"}
1176 See Documentation/block/cfq-iosched.txt and
1177 Documentation/block/deadline-iosched.txt for details.
1179 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1180 Specifies physical address of start of kernel core
1181 image elf header and optionally the size. Generally
1182 kexec loader will pass this option to capture kernel.
1183 See Documentation/kdump/kdump.txt for details.
1185 enable_mtrr_cleanup [X86]
1186 The kernel tries to adjust MTRR layout from continuous
1187 to discrete, to make X server driver able to add WB
1188 entry later. This parameter enables that.
1190 enable_timer_pin_1 [X86]
1191 Enable PIN 1 of APIC timer
1192 Can be useful to work around chipset bugs
1193 (in particular on some ATI chipsets).
1194 The kernel tries to set a reasonable default.
1196 enforcing [SELINUX] Set initial enforcing status.
1198 See security/selinux/Kconfig help text.
1199 0 -- permissive (log only, no denials).
1200 1 -- enforcing (deny and log).
1202 Value can be changed at runtime via /selinux/enforce.
1205 Disable Error Record Serialization Table (ERST)
1208 ether= [HW,NET] Ethernet cards parameters
1209 This option is obsoleted by the "netdev=" option, which
1210 has equivalent usage. See its documentation for details.
1214 Permit 'security.evm' to be updated regardless of
1215 current integrity status.
1219 fail_make_request=[KNL]
1220 General fault injection mechanism.
1221 Format: <interval>,<probability>,<space>,<times>
1222 See also Documentation/fault-injection/.
1225 See Documentation/blockdev/floppy.txt.
1227 force_pal_cache_flush
1228 [IA-64] Avoid check_sal_cache_flush which may hang on
1229 buggy SAL_CACHE_FLUSH implementations. Using this
1230 parameter will force ia64_sal_cache_flush to call
1231 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1234 Forcefully enable Physical Address Extension (PAE).
1235 Many Pentium M systems disable PAE but may have a
1236 functionally usable PAE implementation.
1237 Warning: use of this parameter will taint the kernel
1238 and may cause unknown problems.
1241 [FTRACE] will set and start the specified tracer
1242 as early as possible in order to facilitate early
1245 ftrace_dump_on_oops[=orig_cpu]
1246 [FTRACE] will dump the trace buffers on oops.
1247 If no parameter is passed, ftrace will dump
1248 buffers of all CPUs, but if you pass orig_cpu, it will
1249 dump only the buffer of the CPU that triggered the
1252 ftrace_filter=[function-list]
1253 [FTRACE] Limit the functions traced by the function
1254 tracer at boot up. function-list is a comma separated
1255 list of functions. This list can be changed at run
1256 time by the set_ftrace_filter file in the debugfs
1259 ftrace_notrace=[function-list]
1260 [FTRACE] Do not trace the functions specified in
1261 function-list. This list can be changed at run time
1262 by the set_ftrace_notrace file in the debugfs
1265 ftrace_graph_filter=[function-list]
1266 [FTRACE] Limit the top level callers functions traced
1267 by the function graph tracer at boot up.
1268 function-list is a comma separated list of functions
1269 that can be changed at run time by the
1270 set_graph_function file in the debugfs tracing directory.
1272 ftrace_graph_notrace=[function-list]
1273 [FTRACE] Do not trace from the functions specified in
1274 function-list. This list is a comma separated list of
1275 functions that can be changed at run time by the
1276 set_graph_notrace file in the debugfs tracing directory.
1278 ftrace_graph_max_depth=<uint>
1279 [FTRACE] Used with the function graph tracer. This is
1280 the max depth it will trace into a function. This value
1281 can be changed at run time by the max_graph_depth file
1282 in the tracefs tracing directory. default: 0 (no limit)
1285 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1286 support via parallel port (up to 5 devices per port)
1287 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1288 See also Documentation/input/devices/joystick-parport.rst
1292 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART
1296 gcov_persist= [GCOV] When non-zero (default), profiling data for
1297 kernel modules is saved and remains accessible via
1298 debugfs, even when the module is unloaded/reloaded.
1299 When zero, profiling data is discarded and associated
1300 debugfs files are removed at module unload time.
1302 goldfish [X86] Enable the goldfish android emulator platform.
1303 Don't use this when you are not running on the
1306 gpt [EFI] Forces disk with valid GPT signature but
1307 invalid Protective MBR to be treated as GPT. If the
1308 primary GPT is corrupted, it enables the backup/alternate
1309 GPT to be used instead.
1311 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1312 the "Enable 0" bit of the configuration register.
1315 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1316 the "Enable 0" bit of the configuration register.
1319 grcan.select= [HW] Select which physical interface to use.
1322 grcan.txsize= [HW] Sets the size of the tx buffer.
1323 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1325 grcan.rxsize= [HW] Sets the size of the rx buffer.
1326 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1329 gpio-mockup.gpio_mockup_ranges
1330 [HW] Sets the ranges of gpiochip of for this device.
1331 Format: <start1>,<end1>,<start2>,<end2>...
1333 hardlockup_all_cpu_backtrace=
1334 [KNL] Should the hard-lockup detector generate
1335 backtraces on all cpus.
1338 hashdist= [KNL,NUMA] Large hashes allocated during boot
1339 are distributed across NUMA nodes. Defaults on
1340 for 64-bit NUMA, off otherwise.
1341 Format: 0 | 1 (for off | on)
1343 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1345 hd= [EIDE] (E)IDE hard drive subsystem geometry
1346 Format: <cyl>,<head>,<sect>
1349 Disable Hardware Error Source Table (HEST) support;
1350 corresponding firmware-first mode error processing
1351 logic will be disabled.
1353 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1354 size of <nn>. This works even on boxes that have no
1355 highmem otherwise. This also works to reduce highmem
1356 size on bigger boxes.
1358 highres= [KNL] Enable/disable high resolution timer mode.
1359 Valid parameters: "on", "off"
1363 See Documentation/isdn/README.HiSax.
1367 hpet= [X86-32,HPET] option to control HPET usage
1368 Format: { enable (default) | disable | force |
1370 disable: disable HPET and use PIT instead
1371 force: allow force enabled of undocumented chips (ICH4,
1373 verbose: show contents of HPET registers during setup
1375 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1376 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1378 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
1379 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
1380 On x86-64 and powerpc, this option can be specified
1381 multiple times interleaved with hugepages= to reserve
1382 huge pages of different sizes. Valid pages sizes on
1383 x86-64 are 2M (when the CPU supports "pse") and 1G
1384 (when the CPU supports the "pdpe1gb" cpuinfo flag).
1387 [KNL] Should the hung task detector generate panics.
1390 A nonzero value instructs the kernel to panic when a
1391 hung task is detected. The default value is controlled
1392 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time
1393 option. The value selected by this boot parameter can
1394 be changed later by the kernel.hung_task_panic sysctl.
1396 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1397 terminal devices. Valid values: 0..8
1398 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1399 If specified, z/VM IUCV HVC accepts connections
1400 from listed z/VM user IDs only.
1402 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations
1403 which allow the hypervisor to 'idle' the
1404 guest on lock contention.
1407 Do not unregister boot console at start. This is only
1408 useful for debugging when something happens in the window
1409 between unregistering the boot console and initializing
1412 i2c_bus= [HW] Override the default board specific I2C bus speed
1413 or register an additional I2C bus that is not
1414 registered from board initialization code.
1418 i8042.debug [HW] Toggle i8042 debug mode
1419 i8042.unmask_kbd_data
1420 [HW] Enable printing of interrupt data from the KBD port
1421 (disabled by default, and as a pre-condition
1422 requires that i8042.debug=1 be enabled)
1423 i8042.direct [HW] Put keyboard port into non-translated mode
1424 i8042.dumbkbd [HW] Pretend that controller can only read data from
1425 keyboard and cannot control its state
1426 (Don't attempt to blink the leds)
1427 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1428 i8042.nokbd [HW] Don't check/create keyboard port
1429 i8042.noloop [HW] Disable the AUX Loopback command while probing
1431 i8042.nomux [HW] Don't check presence of an active multiplexing
1433 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1435 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1436 i8042.reset [HW] Reset the controller during init, cleanup and
1437 suspend-to-ram transitions, only during s2r
1438 transitions, or never reset
1439 Format: { 1 | Y | y | 0 | N | n }
1440 1, Y, y: always reset controller
1441 0, N, n: don't ever reset controller
1442 Default: only on s2r transitions on x86; most other
1443 architectures force reset to be always executed
1444 i8042.unlock [HW] Unlock (ignore) the keylock
1445 i8042.kbdreset [HW] Reset device connected to KBD port
1449 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
1450 indicates that the driver is running on unsupported
1452 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
1453 does not match list of supported models.
1455 [HW] Report power status in /proc/i8k
1456 (disabled by default)
1457 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
1460 i915.invert_brightness=
1461 [DRM] Invert the sense of the variable that is used to
1462 set the brightness of the panel backlight. Normally a
1463 brightness value of 0 indicates backlight switched off,
1464 and the maximum of the brightness value sets the backlight
1465 to maximum brightness. If this parameter is set to 0
1466 (default) and the machine requires it, or this parameter
1467 is set to 1, a brightness value of 0 sets the backlight
1468 to maximum brightness, and the maximum of the brightness
1469 value switches the backlight off.
1470 -1 -- never invert brightness
1471 0 -- machine default
1472 1 -- force brightness inversion
1475 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1477 ide-core.nodma= [HW] (E)IDE subsystem
1478 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1479 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1480 .cdrom .chs .ignore_cable are additional options
1481 See Documentation/ide/ide.txt.
1483 ide-generic.probe-mask= [HW] (E)IDE subsystem
1485 Probe mask for legacy ISA IDE ports. Depending on
1486 platform up to 6 ports are supported, enabled by
1487 setting corresponding bits in the mask to 1. The
1488 default value is 0x0, which has a special meaning.
1489 On systems that have PCI, it triggers scanning the
1490 PCI bus for the first and the second port, which
1491 are then probed. On systems without PCI the value
1492 of 0x0 enables probing the two first ports as if it
1495 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1496 Claim all unknown PCI IDE storage controllers.
1499 Format: idle=poll, idle=halt, idle=nomwait
1500 Poll forces a polling idle loop that can slightly
1501 improve the performance of waking up a idle CPU, but
1502 will use a lot of power and make the system run hot.
1504 idle=halt: Halt is forced to be used for CPU idle.
1505 In such case C2/C3 won't be used again.
1506 idle=nomwait: Disable mwait for CPU C-states
1508 ieee754= [MIPS] Select IEEE Std 754 conformance mode
1509 Format: { strict | legacy | 2008 | relaxed }
1512 Choose which programs will be accepted for execution
1513 based on the IEEE 754 NaN encoding(s) supported by
1514 the FPU and the NaN encoding requested with the value
1515 of an ELF file header flag individually set by each
1516 binary. Hardware implementations are permitted to
1517 support either or both of the legacy and the 2008 NaN
1520 Available settings are as follows:
1521 strict accept binaries that request a NaN encoding
1522 supported by the FPU
1523 legacy only accept legacy-NaN binaries, if supported
1525 2008 only accept 2008-NaN binaries, if supported
1527 relaxed accept any binaries regardless of whether
1528 supported by the FPU
1530 The FPU emulator is always able to support both NaN
1531 encodings, so if no FPU hardware is present or it has
1532 been disabled with 'nofpu', then the settings of
1533 'legacy' and '2008' strap the emulator accordingly,
1534 'relaxed' straps the emulator for both legacy-NaN and
1535 2008-NaN, whereas 'strict' enables legacy-NaN only on
1536 legacy processors and both NaN encodings on MIPS32 or
1539 The setting for ABS.fmt/NEG.fmt instruction execution
1540 mode generally follows that for the NaN encoding,
1541 except where unsupported by hardware.
1543 ignore_loglevel [KNL]
1544 Ignore loglevel setting - this will print /all/
1545 kernel messages to the console. Useful for debugging.
1546 We also add it as printk module parameter, so users
1547 could change it dynamically, usually by
1548 /sys/module/printk/parameters/ignore_loglevel.
1551 Ignore RLIMIT_DATA setting for data mappings,
1552 print warning at first misuse. Can be changed via
1553 /sys/module/kernel/parameters/ignore_rlimit_data.
1555 ihash_entries= [KNL]
1556 Set number of hash buckets for inode cache.
1558 ima_appraise= [IMA] appraise integrity measurements
1559 Format: { "off" | "enforce" | "fix" | "log" }
1562 ima_appraise_tcb [IMA]
1563 The builtin appraise policy appraises all files
1566 ima_canonical_fmt [IMA]
1567 Use the canonical format for the binary runtime
1568 measurements, instead of host native format.
1571 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1575 The list of supported hash algorithms is defined
1576 in crypto/hash_info.h.
1579 The builtin policies to load during IMA setup.
1580 Format: "tcb | appraise_tcb | secure_boot |
1583 The "tcb" policy measures all programs exec'd, files
1584 mmap'd for exec, and all files opened with the read
1585 mode bit set by either the effective uid (euid=0) or
1588 The "appraise_tcb" policy appraises the integrity of
1589 all files owned by root. (This is the equivalent
1590 of ima_appraise_tcb.)
1592 The "secure_boot" policy appraises the integrity
1593 of files (eg. kexec kernel image, kernel modules,
1594 firmware, policy, etc) based on file signatures.
1596 The "fail_securely" policy forces file signature
1597 verification failure also on privileged mounted
1598 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE
1601 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
1602 Load a policy which meets the needs of the Trusted
1603 Computing Base. This means IMA will measure all
1604 programs exec'd, files mmap'd for exec, and all files
1605 opened for read by uid=0.
1608 Select one of defined IMA measurements template formats.
1609 Formats: { "ima" | "ima-ng" | "ima-sig" }
1613 [IMA] Define a custom template format.
1614 Format: { "field1|...|fieldN" }
1616 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1617 Format: <min_file_size>
1618 Set the minimal file size for using asynchronous hash.
1619 If left unspecified, ahash usage is disabled.
1621 ahash performance varies for different data sizes on
1622 different crypto accelerators. This option can be used
1623 to achieve the best performance for a particular HW.
1625 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1627 Set hashing buffer size. Default: 4k.
1629 ahash performance varies for different chunk sizes on
1630 different crypto accelerators. This option can be used
1631 to achieve best performance for particular HW.
1635 Run specified binary instead of /sbin/init as init
1638 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1639 for working out where the kernel is dying during
1642 initcall_blacklist= [KNL] Do not execute a comma-separated list of
1643 initcall functions. Useful for debugging built-in
1644 modules and initcalls.
1646 initrd= [BOOT] Specify the location of the initial ramdisk
1648 init_pkru= [x86] Specify the default memory protection keys rights
1649 register contents for all processes. 0x55555554 by
1650 default (disallow access to all but pkey 0). Can
1651 override in debugfs after boot.
1653 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1656 int_pln_enable [x86] Enable power limit notification interrupt
1658 integrity_audit=[IMA]
1659 Format: { "0" | "1" }
1660 0 -- basic integrity auditing messages. (Default)
1661 1 -- additional integrity auditing messages.
1663 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
1665 Enable intel iommu driver.
1667 Disable intel iommu driver.
1668 igfx_off [Default Off]
1669 By default, gfx is mapped as normal device. If a gfx
1670 device has a dedicated DMAR unit, the DMAR unit is
1671 bypassed by not enabling DMAR with this option. In
1672 this case, gfx device will use physical address for
1675 With this option iommu will not optimize to look
1676 for io virtual address below 32-bit forcing dual
1677 address cycle on pci bus for cards supporting greater
1678 than 32-bit addressing. The default is to look
1679 for translation below 32-bit and if not available
1680 then look in the higher range.
1681 strict [Default Off]
1682 With this option on every unmap_single operation will
1683 result in a hardware IOTLB flush operation as opposed
1684 to batching them for performance.
1685 sp_off [Default Off]
1686 By default, super page will be supported if Intel IOMMU
1687 has the capability. With this option, super page will
1689 ecs_off [Default Off]
1690 By default, extended context tables will be supported if
1691 the hardware advertises that it has support both for the
1692 extended tables themselves, and also PASID support. With
1693 this option set, extended tables will not be used even
1694 on hardware which claims to support them.
1695 tboot_noforce [Default Off]
1696 Do not force the Intel IOMMU enabled under tboot.
1697 By default, tboot will force Intel IOMMU on, which
1698 could harm performance of some high-throughput
1699 devices like 40GBit network cards, even if identity
1701 Note that using this option lowers the security
1702 provided by tboot because it makes the system
1703 vulnerable to DMA attacks.
1705 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
1706 0 disables intel_idle and fall back on acpi_idle.
1707 1 to 9 specify maximum depth of C-state.
1711 Do not enable intel_pstate as the default
1712 scaling driver for the supported processors
1714 Use intel_pstate as a scaling driver, but configure it
1715 to work with generic cpufreq governors (instead of
1716 enabling its internal governor). This mode cannot be
1717 used along with the hardware-managed P-states (HWP)
1720 Enable intel_pstate on systems that prohibit it by default
1721 in favor of acpi-cpufreq. Forcing the intel_pstate driver
1722 instead of acpi-cpufreq may disable platform features, such
1723 as thermal controls and power capping, that rely on ACPI
1724 P-States information being indicated to OSPM and therefore
1725 should be used with caution. This option does not work with
1726 processors that aren't supported by the intel_pstate driver
1727 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
1729 Do not enable hardware P state control (HWP)
1732 Only load intel_pstate on systems which support
1733 hardware P state control (HWP) if available.
1735 Enforce ACPI _PPC performance limits. If the Fixed ACPI
1736 Description Table, specifies preferred power management
1737 profile as "Enterprise Server" or "Performance Server",
1738 then this feature is turned on by default.
1740 Allow per-logical-CPU P-State performance control limits using
1741 cpufreq sysfs interface
1743 intremap= [X86-64, Intel-IOMMU]
1744 on enable Interrupt Remapping (default)
1745 off disable Interrupt Remapping
1746 nosid disable Source ID checking
1748 BIOS x2APIC opt-out request will be ignored
1749 nopost disable Interrupt Posting
1751 iomem= Disable strict checking of access to MMIO memory
1752 strict regions from userspace.
1767 nobypass [PPC/POWERNV]
1768 Disable IOMMU bypass, using IOMMU for PCI devices.
1770 iommu.strict= [ARM64] Configure TLB invalidation behaviour
1771 Format: { "0" | "1" }
1773 Request that DMA unmap operations use deferred
1774 invalidation of hardware TLBs, for increased
1775 throughput at the cost of reduced device isolation.
1776 Will fall back to strict mode if not supported by
1777 the relevant IOMMU driver.
1778 1 - Strict mode (default).
1779 DMA unmap operations invalidate IOMMU hardware TLBs
1783 [ARM64] Configure DMA to bypass the IOMMU by default.
1784 Format: { "0" | "1" }
1785 0 - Use IOMMU translation for DMA.
1786 1 - Bypass the IOMMU for DMA.
1787 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH.
1789 io7= [HW] IO7 for Marvel based alpha systems
1790 See comment before marvel_specify_io7 in
1791 arch/alpha/kernel/core_marvel.c.
1793 io_delay= [X86] I/O delay method
1795 Standard port 0x80 based delay
1797 Alternate port 0xed based delay (needed on some systems)
1799 Simple two microseconds delay
1804 See Documentation/filesystems/nfs/nfsroot.txt.
1806 irqaffinity= [SMP] Set the default irq affinity mask
1807 The argument is a cpu list, as described above.
1809 irqchip.gicv2_force_probe=
1812 Force the kernel to look for the second 4kB page
1813 of a GICv2 controller even if the memory range
1814 exposed by the device tree is too small.
1816 irqchip.gicv3_nolpi=
1818 Force the kernel to ignore the availability of
1819 LPIs (and by consequence ITSs). Intended for system
1820 that use the kernel as a bootloader, and thus want
1821 to let secondary kernels in charge of setting up
1825 When an interrupt is not handled search all handlers
1826 for it. Intended to get systems with badly broken
1830 When an interrupt is not handled search all handlers
1831 for it. Also check all handlers each timer
1832 interrupt. Intended to get systems with badly broken
1836 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
1838 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance.
1839 [Deprecated - use cpusets instead]
1840 Format: [flag-list,]<cpu-list>
1842 Specify one or more CPUs to isolate from disturbances
1843 specified in the flag list (default: domain):
1846 Disable the tick when a single task runs.
1848 A residual 1Hz tick is offloaded to workqueues, which you
1849 need to affine to housekeeping through the global
1850 workqueue's affinity configured via the
1851 /sys/devices/virtual/workqueue/cpumask sysfs file, or
1852 by using the 'domain' flag described below.
1854 NOTE: by default the global workqueue runs on all CPUs,
1855 so to protect individual CPUs the 'cpumask' file has to
1856 be configured manually after bootup.
1859 Isolate from the general SMP balancing and scheduling
1860 algorithms. Note that performing domain isolation this way
1861 is irreversible: it's not possible to bring back a CPU to
1862 the domains once isolated through isolcpus. It's strongly
1863 advised to use cpusets instead to disable scheduler load
1864 balancing through the "cpuset.sched_load_balance" file.
1865 It offers a much more flexible interface where CPUs can
1866 move in and out of an isolated set anytime.
1868 You can move a process onto or off an "isolated" CPU via
1869 the CPU affinity syscalls or cpuset.
1870 <cpu number> begins at 0 and the maximum value is
1871 "number of CPUs in system - 1".
1873 The format of <cpu-list> is described above.
1879 ivrs_ioapic [HW,X86_64]
1880 Provide an override to the IOAPIC-ID<->DEVICE-ID
1881 mapping provided in the IVRS ACPI table. For
1882 example, to map IOAPIC-ID decimal 10 to
1883 PCI device 00:14.0 write the parameter as:
1884 ivrs_ioapic[10]=00:14.0
1886 ivrs_hpet [HW,X86_64]
1887 Provide an override to the HPET-ID<->DEVICE-ID
1888 mapping provided in the IVRS ACPI table. For
1889 example, to map HPET-ID decimal 0 to
1890 PCI device 00:14.0 write the parameter as:
1891 ivrs_hpet[0]=00:14.0
1893 ivrs_acpihid [HW,X86_64]
1894 Provide an override to the ACPI-HID:UID<->DEVICE-ID
1895 mapping provided in the IVRS ACPI table. For
1896 example, to map UART-HID:UID AMD0020:0 to
1897 PCI device 00:14.5 write the parameter as:
1898 ivrs_acpihid[00:14.5]=AMD0020:0
1900 js= [HW,JOY] Analog joystick
1901 See Documentation/input/joydev/joystick.rst.
1904 When CONFIG_RANDOMIZE_BASE is set, this disables
1905 kernel and module base offset ASLR (Address Space
1906 Layout Randomization).
1909 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print
1910 report on every invalid memory access. Without this
1911 parameter KASAN will print report only for the first
1916 kernelcore= [KNL,X86,IA-64,PPC]
1917 Format: nn[KMGTPE] | nn% | "mirror"
1918 This parameter specifies the amount of memory usable by
1919 the kernel for non-movable allocations. The requested
1920 amount is spread evenly throughout all nodes in the
1921 system as ZONE_NORMAL. The remaining memory is used for
1922 movable memory in its own zone, ZONE_MOVABLE. In the
1923 event, a node is too small to have both ZONE_NORMAL and
1924 ZONE_MOVABLE, kernelcore memory will take priority and
1925 other nodes will have a larger ZONE_MOVABLE.
1927 ZONE_MOVABLE is used for the allocation of pages that
1928 may be reclaimed or moved by the page migration
1929 subsystem. Note that allocations like PTEs-from-HighMem
1930 still use the HighMem zone if it exists, and the Normal
1931 zone if it does not.
1933 It is possible to specify the exact amount of memory in
1934 the form of "nn[KMGTPE]", a percentage of total system
1935 memory in the form of "nn%", or "mirror". If "mirror"
1936 option is specified, mirrored (reliable) memory is used
1937 for non-movable allocations and remaining memory is used
1938 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror"
1939 are exclusive, so you cannot specify multiple forms.
1941 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
1942 Format: <Controller#>[,poll interval]
1943 The controller # is the number of the ehci usb debug
1944 port as it is probed via PCI. The poll interval is
1945 optional and is the number seconds in between
1946 each poll cycle to the debug port in case you need
1947 the functionality for interrupting the kernel with
1948 gdb or control-c on the dbgp connection. When
1949 not using this parameter you use sysrq-g to break into
1950 the kernel debugger.
1952 kgdboc= [KGDB,HW] kgdb over consoles.
1953 Requires a tty driver that supports console polling,
1954 or a supported polling keyboard driver (non-usb).
1955 Serial only format: <serial_device>[,baud]
1956 keyboard only format: kbd
1957 keyboard and serial format: kbd,<serial_device>[,baud]
1958 Optional Kernel mode setting:
1959 kms, kbd format: kms,kbd
1960 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
1962 kgdbwait [KGDB] Stop kernel execution and enter the
1963 kernel debugger at the earliest opportunity.
1965 kmac= [MIPS] korina ethernet MAC address.
1966 Configure the RouterBoard 532 series on-chip
1967 Ethernet adapter MAC address.
1969 kmemleak= [KNL] Boot-time kmemleak enable/disable
1970 Valid arguments: on, off
1972 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
1975 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
1976 Default is 0 (don't ignore, but inject #GP)
1978 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface.
1979 Default is false (don't support).
1981 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit
1985 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
1986 Default is 1 (enabled)
1988 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
1990 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
1992 kvm-arm.vgic_v3_group0_trap=
1993 [KVM,ARM] Trap guest accesses to GICv3 group-0
1996 kvm-arm.vgic_v3_group1_trap=
1997 [KVM,ARM] Trap guest accesses to GICv3 group-1
2000 kvm-arm.vgic_v3_common_trap=
2001 [KVM,ARM] Trap guest accesses to GICv3 common
2004 kvm-arm.vgic_v4_enable=
2005 [KVM,ARM] Allow use of GICv4 for direct injection of
2008 kvm-intel.ept= [KVM,Intel] Disable extended page tables
2009 (virtualized MMU) support on capable Intel chips.
2010 Default is 1 (enabled)
2012 kvm-intel.emulate_invalid_guest_state=
2013 [KVM,Intel] Enable emulation of invalid guest states
2014 Default is 0 (disabled)
2016 kvm-intel.flexpriority=
2017 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
2018 Default is 1 (enabled)
2021 [KVM,Intel] Enable VMX nesting (nVMX).
2022 Default is 0 (disabled)
2024 kvm-intel.unrestricted_guest=
2025 [KVM,Intel] Disable unrestricted guest feature
2026 (virtualized real and unpaged mode) on capable
2027 Intel chips. Default is 1 (enabled)
2029 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault
2032 Valid arguments: never, cond, always
2034 always: L1D cache flush on every VMENTER.
2035 cond: Flush L1D on VMENTER only when the code between
2036 VMEXIT and VMENTER can leak host memory.
2037 never: Disables the mitigation
2039 Default is cond (do L1 cache flush in specific instances)
2041 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
2042 feature (tagged TLBs) on capable Intel chips.
2043 Default is 1 (enabled)
2045 l1tf= [X86] Control mitigation of the L1TF vulnerability on
2048 The kernel PTE inversion protection is unconditionally
2049 enabled and cannot be disabled.
2052 Provides all available mitigations for the
2053 L1TF vulnerability. Disables SMT and
2054 enables all mitigations in the
2055 hypervisors, i.e. unconditional L1D flush.
2057 SMT control and L1D flush control via the
2058 sysfs interface is still possible after
2059 boot. Hypervisors will issue a warning
2060 when the first VM is started in a
2061 potentially insecure configuration,
2062 i.e. SMT enabled or L1D flush disabled.
2065 Same as 'full', but disables SMT and L1D
2066 flush runtime control. Implies the
2067 'nosmt=force' command line option.
2068 (i.e. sysfs control of SMT is disabled.)
2071 Leaves SMT enabled and enables the default
2072 hypervisor mitigation, i.e. conditional
2075 SMT control and L1D flush control via the
2076 sysfs interface is still possible after
2077 boot. Hypervisors will issue a warning
2078 when the first VM is started in a
2079 potentially insecure configuration,
2080 i.e. SMT enabled or L1D flush disabled.
2084 Disables SMT and enables the default
2085 hypervisor mitigation.
2087 SMT control and L1D flush control via the
2088 sysfs interface is still possible after
2089 boot. Hypervisors will issue a warning
2090 when the first VM is started in a
2091 potentially insecure configuration,
2092 i.e. SMT enabled or L1D flush disabled.
2095 Same as 'flush', but hypervisors will not
2096 warn when a VM is started in a potentially
2097 insecure configuration.
2100 Disables hypervisor mitigations and doesn't
2102 It also drops the swap size and available
2103 RAM limit restriction on both hypervisor and
2108 For details see: Documentation/admin-guide/l1tf.rst
2114 lapic [X86-32,APIC] Enable the local APIC even if BIOS
2117 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline
2118 value for LAPIC timer one-shot implementation. Default
2119 back to the programmable timer unit in the LAPIC.
2121 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
2124 libata.dma= [LIBATA] DMA control
2125 libata.dma=0 Disable all PATA and SATA DMA
2126 libata.dma=1 PATA and SATA Disk DMA only
2127 libata.dma=2 ATAPI (CDROM) DMA only
2128 libata.dma=4 Compact Flash DMA only
2129 Combinations also work, so libata.dma=3 enables DMA
2130 for disks and CDROMs, but not CFs.
2132 libata.ignore_hpa= [LIBATA] Ignore HPA limit
2133 libata.ignore_hpa=0 keep BIOS limits (default)
2134 libata.ignore_hpa=1 ignore limits, using full disk
2136 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
2140 libata.force= [LIBATA] Force configurations. The format is comma
2141 separated list of "[ID:]VAL" where ID is
2142 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
2143 matching port, link or device. Basically, it matches
2144 the ATA ID string printed on console by libata. If
2145 the whole ID part is omitted, the last PORT and DEVICE
2146 values are used. If ID hasn't been specified yet, the
2147 configuration applies to all ports, links and devices.
2149 If only DEVICE is omitted, the parameter applies to
2150 the port and all links and devices behind it. DEVICE
2151 number of 0 either selects the first device or the
2152 first fan-out link behind PMP device. It does not
2153 select the host link. DEVICE number of 15 selects the
2154 host link and device attached to it.
2156 The VAL specifies the configuration to force. As long
2157 as there's no ambiguity shortcut notation is allowed.
2158 For example, both 1.5 and 1.5G would work for 1.5Gbps.
2159 The following configurations can be forced.
2161 * Cable type: 40c, 80c, short40c, unk, ign or sata.
2162 Any ID with matching PORT is used.
2164 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
2166 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
2167 udma[/][16,25,33,44,66,100,133] notation is also
2170 * [no]ncq: Turn on or off NCQ.
2172 * [no]ncqtrim: Turn off queued DSM TRIM.
2174 * nohrst, nosrst, norst: suppress hard, soft
2177 * rstonce: only attempt one reset during
2178 hot-unplug link recovery
2180 * dump_id: dump IDENTIFY data.
2182 * atapi_dmadir: Enable ATAPI DMADIR bridge support
2184 * disable: Disable this device.
2186 If there are multiple matching configurations changing
2187 the same attribute, the last one is used.
2189 memblock=debug [KNL] Enable memblock debug messages.
2191 load_ramdisk= [RAM] List of ramdisks to load from floppy
2192 See Documentation/blockdev/ramdisk.txt.
2194 lockd.nlm_grace_period=P [NFS] Assign grace period.
2197 lockd.nlm_tcpport=N [NFS] Assign TCP port.
2200 lockd.nlm_timeout=T [NFS] Assign timeout value.
2203 lockd.nlm_udpport=M [NFS] Assign UDP port.
2206 locktorture.nreaders_stress= [KNL]
2207 Set the number of locking read-acquisition kthreads.
2208 Defaults to being automatically set based on the
2209 number of online CPUs.
2211 locktorture.nwriters_stress= [KNL]
2212 Set the number of locking write-acquisition kthreads.
2214 locktorture.onoff_holdoff= [KNL]
2215 Set time (s) after boot for CPU-hotplug testing.
2217 locktorture.onoff_interval= [KNL]
2218 Set time (s) between CPU-hotplug operations, or
2219 zero to disable CPU-hotplug testing.
2221 locktorture.shuffle_interval= [KNL]
2222 Set task-shuffle interval (jiffies). Shuffling
2223 tasks allows some CPUs to go into dyntick-idle
2224 mode during the locktorture test.
2226 locktorture.shutdown_secs= [KNL]
2227 Set time (s) after boot system shutdown. This
2228 is useful for hands-off automated testing.
2230 locktorture.stat_interval= [KNL]
2231 Time (s) between statistics printk()s.
2233 locktorture.stutter= [KNL]
2234 Time (s) to stutter testing, for example,
2235 specifying five seconds causes the test to run for
2236 five seconds, wait for five seconds, and so on.
2237 This tests the locking primitive's ability to
2238 transition abruptly to and from idle.
2240 locktorture.torture_type= [KNL]
2241 Specify the locking implementation to test.
2243 locktorture.verbose= [KNL]
2244 Enable additional printk() statements.
2246 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
2249 loglevel= All Kernel Messages with a loglevel smaller than the
2250 console loglevel will be printed to the console. It can
2251 also be changed with klogd or other programs. The
2252 loglevels are defined as follows:
2254 0 (KERN_EMERG) system is unusable
2255 1 (KERN_ALERT) action must be taken immediately
2256 2 (KERN_CRIT) critical conditions
2257 3 (KERN_ERR) error conditions
2258 4 (KERN_WARNING) warning conditions
2259 5 (KERN_NOTICE) normal but significant condition
2260 6 (KERN_INFO) informational
2261 7 (KERN_DEBUG) debug-level messages
2263 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
2264 in bytes. n must be a power of two and greater
2265 than the minimal size. The minimal size is defined
2266 by LOG_BUF_SHIFT kernel config parameter. There is
2267 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
2268 that allows to increase the default size depending on
2269 the number of CPUs. See init/Kconfig for more details.
2271 logo.nologo [FB] Disables display of the built-in Linux logo.
2272 This may be used to provide more screen space for
2273 kernel log messages and is useful when debugging
2274 kernel boot problems.
2276 lp=0 [LP] Specify parallel ports to use, e.g,
2277 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
2278 lp=reset first parallel port). 'lp=0' disables the
2279 lp=auto printer driver. 'lp=reset' (which can be
2280 specified in addition to the ports) causes
2281 attached printers to be reset. Using
2282 lp=port1,port2,... specifies the parallel ports
2283 to associate lp devices with, starting with
2284 lp0. A port specification may be 'none' to skip
2285 that lp device, or a parport name such as
2286 'parport0'. Specifying 'lp=auto' instead of a
2287 port specification list means that device IDs
2288 from each port should be examined, to see if
2289 an IEEE 1284-compliant printer is attached; if
2290 so, the driver will manage that printer.
2291 See also header of drivers/char/lp.c.
2294 Sets loops_per_jiffy to given constant, thus avoiding
2295 time-consuming boot-time autodetection (up to 250 ms per
2296 CPU). 0 enables autodetection (default). To determine
2297 the correct value for your kernel, boot with normal
2298 autodetection and see what value is printed. Note that
2299 on SMP systems the preset will be applied to all CPUs,
2300 which is likely to cause problems if your CPUs need
2301 significantly divergent settings. An incorrect value
2302 will cause delays in the kernel to be wrong, leading to
2303 unpredictable I/O errors and other breakage. Although
2304 unlikely, in the extreme case this might damage your
2308 Format: <io>,<irq>,<dma>
2310 lsm.debug [SECURITY] Enable LSM initialization debugging output.
2312 machvec= [IA-64] Force the use of a particular machine-vector
2313 (machvec) in a generic kernel.
2314 Example: machvec=hpzx1_swiotlb
2316 machtype= [Loongson] Share the same kernel image file between different
2318 Example: machtype=lemote-yeeloong-2f-7inch
2320 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
2321 than or equal to this physical address is ignored.
2323 maxcpus= [SMP] Maximum number of processors that an SMP kernel
2324 will bring up during bootup. maxcpus=n : n >= 0 limits
2325 the kernel to bring up 'n' processors. Surely after
2326 bootup you can bring up the other plugged cpu by executing
2327 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus
2328 only takes effect during system bootup.
2329 While n=0 is a special case, it is equivalent to "nosmp",
2330 which also disables the IO APIC.
2332 max_loop= [LOOP] The number of loop block devices that get
2333 (loop.max_loop) unconditionally pre-created at init time. The default
2334 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
2335 of statically allocating a predefined number, loop
2336 devices can be requested on-demand with the
2337 /dev/loop-control interface.
2339 mce [X86-32] Machine Check Exception
2341 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.txt
2343 md= [HW] RAID subsystems devices and level
2344 See Documentation/admin-guide/md.rst.
2347 Format: <first>,<last>
2348 Specifies range of consoles to be captured by the MDA.
2350 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
2351 Amount of memory to be used when the kernel is not able
2352 to see the whole system memory or for test.
2353 [X86] Work as limiting max address. Use together
2354 with memmap= to avoid physical address space collisions.
2355 Without memmap= PCI devices could be placed at addresses
2356 belonging to unused RAM.
2358 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
2362 [KNL,SH] Allow user to override the default size for
2363 per-device physically contiguous DMA buffers.
2365 memhp_default_state=online/offline
2366 [KNL] Set the initial state for the memory hotplug
2367 onlining policy. If not specified, the default value is
2368 set according to the
2369 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config
2371 See Documentation/memory-hotplug.txt.
2373 memmap=exactmap [KNL,X86] Enable setting of an exact
2374 E820 memory map, as specified by the user.
2375 Such memmap=exactmap lines can be constructed based on
2376 BIOS output or other requirements. See the memmap=nn@ss
2379 memmap=nn[KMG]@ss[KMG]
2380 [KNL] Force usage of a specific region of memory.
2381 Region of memory to be used is from ss to ss+nn.
2382 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG],
2383 which limits max address to nn[KMG].
2384 Multiple different regions can be specified,
2387 memmap=100M@2G,100M#3G,1G!1024G
2389 memmap=nn[KMG]#ss[KMG]
2390 [KNL,ACPI] Mark specific memory as ACPI data.
2391 Region of memory to be marked is from ss to ss+nn.
2393 memmap=nn[KMG]$ss[KMG]
2394 [KNL,ACPI] Mark specific memory as reserved.
2395 Region of memory to be reserved is from ss to ss+nn.
2396 Example: Exclude memory from 0x18690000-0x1869ffff
2397 memmap=64K$0x18690000
2399 memmap=0x10000$0x18690000
2400 Some bootloaders may need an escape character before '$',
2401 like Grub2, otherwise '$' and the following number
2404 memmap=nn[KMG]!ss[KMG]
2405 [KNL,X86] Mark specific memory as protected.
2406 Region of memory to be used, from ss to ss+nn.
2407 The memory region may be marked as e820 type 12 (0xc)
2408 and is NVDIMM or ADR memory.
2410 memmap=<size>%<offset>-<oldtype>+<newtype>
2411 [KNL,ACPI] Convert memory within the specified region
2412 from <oldtype> to <newtype>. If "-<oldtype>" is left
2413 out, the whole region will be marked as <newtype>,
2414 even if previously unavailable. If "+<newtype>" is left
2415 out, matching memory will be removed. Types are
2416 specified as e820 types, e.g., 1 = RAM, 2 = reserved,
2417 3 = ACPI, 12 = PRAM.
2419 memory_corruption_check=0/1 [X86]
2420 Some BIOSes seem to corrupt the first 64k of
2421 memory when doing things like suspend/resume.
2422 Setting this option will scan the memory
2423 looking for corruption. Enabling this will
2424 both detect corruption and prevent the kernel
2425 from using the memory being corrupted.
2426 However, its intended as a diagnostic tool; if
2427 repeatable BIOS-originated corruption always
2428 affects the same memory, you can use memmap=
2429 to prevent the kernel from using that memory.
2431 memory_corruption_check_size=size [X86]
2432 By default it checks for corruption in the low
2433 64k, making this memory unavailable for normal
2434 use. Use this parameter to scan for
2435 corruption in more or less memory.
2437 memory_corruption_check_period=seconds [X86]
2438 By default it checks for corruption every 60
2439 seconds. Use this parameter to check at some
2440 other rate. 0 disables periodic checking.
2442 memtest= [KNL,X86,ARM,PPC] Enable memtest
2444 default : 0 <disable>
2445 Specifies the number of memtest passes to be
2446 performed. Each pass selects another test
2447 pattern from a given set of patterns. Memtest
2448 fills the memory with this pattern, validates
2449 memory contents and reserves bad memory
2450 regions that are detected.
2452 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control
2453 Valid arguments: on, off
2454 Default (depends on kernel configuration option):
2455 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y)
2456 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n)
2457 mem_encrypt=on: Activate SME
2458 mem_encrypt=off: Do not activate SME
2460 Refer to Documentation/x86/amd-memory-encryption.txt
2461 for details on when memory encryption can be activated.
2463 mem_sleep_default= [SUSPEND] Default system suspend mode:
2464 s2idle - Suspend-To-Idle
2465 shallow - Power-On Suspend or equivalent (if supported)
2466 deep - Suspend-To-RAM or equivalent (if supported)
2467 See Documentation/admin-guide/pm/sleep-states.rst.
2469 meye.*= [HW] Set MotionEye Camera parameters
2470 See Documentation/media/v4l-drivers/meye.rst.
2472 mfgpt_irq= [IA-32] Specify the IRQ to use for the
2473 Multi-Function General Purpose Timers on AMD Geode
2476 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
2477 the BIOS has incorrectly applied a workaround. TinyBIOS
2478 version 0.98 is known to be affected, 0.99 fixes the
2479 problem by letting the user disable the workaround.
2483 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
2484 physical address is ignored.
2486 mini2440= [ARM,HW,KNL]
2487 Format:[0..2][b][c][t]
2489 MINI2440 configuration specification:
2490 0 - The attached screen is the 3.5" TFT
2491 1 - The attached screen is the 7" TFT
2492 2 - The VGA Shield is attached (1024x768)
2493 Leaving out the screen size parameter will not load
2494 the TFT driver, and the framebuffer will be left
2496 b - Enable backlight. The TFT backlight pin will be
2497 linked to the kernel VESA blanking code and a GPIO
2498 LED. This parameter is not necessary when using the
2500 c - Enable the s3c camera interface.
2501 t - Reserved for enabling touchscreen support. The
2502 touchscreen support is not enabled in the mainstream
2503 kernel as of 2.6.30, a preliminary port can be found
2504 in the "bleeding edge" mini2440 support kernel at
2505 http://repo.or.cz/w/linux-2.6/mini2440.git
2508 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
2509 parameter allows control of the logging verbosity for
2510 the additional memory initialisation checks. A value
2511 of 0 disables mminit logging and a level of 4 will
2512 log everything. Information is printed at KERN_DEBUG
2513 so loglevel=8 may also need to be specified.
2516 [KNL] When CONFIG_MODULE_SIG is set, this means that
2517 modules without (valid) signatures will fail to load.
2518 Note that if CONFIG_MODULE_SIG_FORCE is set, that
2519 is always true, so this option does nothing.
2521 module_blacklist= [KNL] Do not load a comma-separated list of
2522 modules. Useful for debugging problem modules.
2525 [MOUSE] Maximum time between finger touching and
2526 leaving touchpad surface for touch to be considered
2527 a tap and be reported as a left button click (for
2528 touchpads working in absolute mode only).
2530 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
2531 reporting absolute coordinates, such as tablets
2532 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
2533 reporting absolute coordinates, such as tablets
2535 movablecore= [KNL,X86,IA-64,PPC]
2536 Format: nn[KMGTPE] | nn%
2537 This parameter is the complement to kernelcore=, it
2538 specifies the amount of memory used for migratable
2539 allocations. If both kernelcore and movablecore is
2540 specified, then kernelcore will be at *least* the
2541 specified value but may be more. If movablecore on its
2542 own is specified, the administrator must be careful
2543 that the amount of memory usable for all allocations
2546 movable_node [KNL] Boot-time switch to make hotplugable memory
2547 NUMA nodes to be movable. This means that the memory
2548 of such nodes will be usable only for movable
2549 allocations which rules out almost all kernel
2550 allocations. Use with caution!
2552 MTD_Partition= [MTD]
2553 Format: <name>,<region-number>,<size>,<offset>
2555 MTD_Region= [MTD] Format:
2556 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
2559 See drivers/mtd/cmdlinepart.c.
2561 multitce=off [PPC] This parameter disables the use of the pSeries
2562 firmware feature for updating multiple TCE entries
2565 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
2567 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
2569 boundary - index of last SLC block on Flex-OneNAND.
2570 The remaining blocks are configured as MLC blocks.
2571 lock - Configure if Flex-OneNAND boundary should be locked.
2572 Once locked, the boundary cannot be changed.
2573 1 indicates lock status, 0 indicates unlock status.
2576 ARM/S3C2412 JIVE boot control
2578 See arch/arm/mach-s3c2412/mach-jive.c
2580 mtouchusb.raw_coordinates=
2581 [HW] Make the MicroTouch USB driver use raw coordinates
2582 ('y', default) or cooked coordinates ('n')
2584 mtrr_chunk_size=nn[KMG] [X86]
2585 used for mtrr cleanup. It is largest continuous chunk
2586 that could hold holes aka. UC entries.
2588 mtrr_gran_size=nn[KMG] [X86]
2589 Used for mtrr cleanup. It is granularity of mtrr block.
2591 Large value could prevent small alignment from
2594 mtrr_spare_reg_nr=n [X86]
2596 Range: 0,7 : spare reg number
2598 Used for mtrr cleanup. It is spare mtrr entries number.
2599 Set to 2 or more if your graphical card needs more.
2601 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
2603 netdev= [NET] Network devices parameters
2604 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
2605 Note that mem_start is often overloaded to mean
2606 something different and driver-specific.
2607 This usage is only documented in each driver source
2611 [NETFILTER] Enable connection tracking flow accounting
2612 0 to disable accounting
2613 1 to enable accounting
2616 nfsaddrs= [NFS] Deprecated. Use ip= instead.
2617 See Documentation/filesystems/nfs/nfsroot.txt.
2619 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
2620 See Documentation/filesystems/nfs/nfsroot.txt.
2622 nfsrootdebug [NFS] enable nfsroot debugging messages.
2623 See Documentation/filesystems/nfs/nfsroot.txt.
2625 nfs.callback_nr_threads=
2626 [NFSv4] set the total number of threads that the
2627 NFS client will assign to service NFSv4 callback
2630 nfs.callback_tcpport=
2631 [NFS] set the TCP port on which the NFSv4 callback
2632 channel should listen.
2635 [NFS] sets the pathname to the program which is used
2636 to update the NFS client cache entries.
2638 nfs.cache_getent_timeout=
2639 [NFS] sets the timeout after which an attempt to
2640 update a cache entry is deemed to have failed.
2642 nfs.idmap_cache_timeout=
2643 [NFS] set the maximum lifetime for idmapper cache
2647 [NFS] enable 64-bit inode numbers.
2648 If zero, the NFS client will fake up a 32-bit inode
2649 number for the readdir() and stat() syscalls instead
2650 of returning the full 64-bit number.
2651 The default is to return 64-bit inode numbers.
2653 nfs.max_session_cb_slots=
2654 [NFSv4.1] Sets the maximum number of session
2655 slots the client will assign to the callback
2656 channel. This determines the maximum number of
2657 callbacks the client will process in parallel for
2658 a particular server.
2660 nfs.max_session_slots=
2661 [NFSv4.1] Sets the maximum number of session slots
2662 the client will attempt to negotiate with the server.
2663 This limits the number of simultaneous RPC requests
2664 that the client can send to the NFSv4.1 server.
2665 Note that there is little point in setting this
2666 value higher than the max_tcp_slot_table_limit.
2668 nfs.nfs4_disable_idmapping=
2669 [NFSv4] When set to the default of '1', this option
2670 ensures that both the RPC level authentication
2671 scheme and the NFS level operations agree to use
2672 numeric uids/gids if the mount is using the
2673 'sec=sys' security flavour. In effect it is
2674 disabling idmapping, which can make migration from
2675 legacy NFSv2/v3 systems to NFSv4 easier.
2676 Servers that do not support this mode of operation
2677 will be autodetected by the client, and it will fall
2678 back to using the idmapper.
2679 To turn off this behaviour, set the value to '0'.
2681 [NFS4] Specify an additional fixed unique ident-
2682 ification string that NFSv4 clients can insert into
2683 their nfs_client_id4 string. This is typically a
2684 UUID that is generated at system install time.
2686 nfs.send_implementation_id =
2687 [NFSv4.1] Send client implementation identification
2688 information in exchange_id requests.
2689 If zero, no implementation identification information
2691 The default is to send the implementation identification
2694 nfs.recover_lost_locks =
2695 [NFSv4] Attempt to recover locks that were lost due
2696 to a lease timeout on the server. Please note that
2697 doing this risks data corruption, since there are
2698 no guarantees that the file will remain unchanged
2699 after the locks are lost.
2700 If you want to enable the kernel legacy behaviour of
2701 attempting to recover these locks, then set this
2703 The default parameter value of '0' causes the kernel
2704 not to attempt recovery of lost locks.
2706 nfs4.layoutstats_timer =
2707 [NFSv4.2] Change the rate at which the kernel sends
2708 layoutstats to the pNFS metadata server.
2710 Setting this to value to 0 causes the kernel to use
2711 whatever value is the default set by the layout
2712 driver. A non-zero value sets the minimum interval
2713 in seconds between layoutstats transmissions.
2715 nfsd.nfs4_disable_idmapping=
2716 [NFSv4] When set to the default of '1', the NFSv4
2717 server will return only numeric uids and gids to
2718 clients using auth_sys, and will accept numeric uids
2719 and gids from such clients. This is intended to ease
2720 migration from NFSv2/v3.
2722 nmi_debug= [KNL,SH] Specify one or more actions to take
2723 when a NMI is triggered.
2724 Format: [state][,regs][,debounce][,die]
2726 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
2727 Format: [panic,][nopanic,][num]
2729 0 - turn hardlockup detector in nmi_watchdog off
2730 1 - turn hardlockup detector in nmi_watchdog on
2731 When panic is specified, panic when an NMI watchdog
2732 timeout occurs (or 'nopanic' to override the opposite
2733 default). To disable both hard and soft lockup detectors,
2734 please see 'nowatchdog'.
2735 This is useful when you use a panic=... timeout and
2736 need the box quickly up again.
2738 These settings can be accessed at runtime via
2739 the nmi_watchdog and hardlockup_panic sysctls.
2741 netpoll.carrier_timeout=
2742 [NET] Specifies amount of time (in seconds) that
2743 netpoll should wait for a carrier. By default netpoll
2746 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
2747 emulation library even if a 387 maths coprocessor
2750 no5lvl [X86-64] Disable 5-level paging mode. Forces
2751 kernel to use 4-level paging instead.
2754 [HW] Never suspend the console
2755 Disable suspending of consoles during suspend and
2756 hibernate operations. Once disabled, debugging
2757 messages can reach various consoles while the rest
2758 of the system is being put to sleep (ie, while
2759 debugging driver suspend/resume hooks). This may
2760 not work reliably with all consoles, but is known
2761 to work with serial and VGA consoles.
2762 To facilitate more flexible debugging, we also add
2763 console_suspend, a printk module parameter to control
2764 it. Users could use console_suspend (usually
2765 /sys/module/printk/parameters/console_suspend) to
2766 turn on/off it dynamically.
2768 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
2769 caches in the slab allocator. Saves per-node memory,
2770 but will impact performance.
2774 noaltinstr [S390] Disables alternative instructions patching
2775 (CPU alternatives feature).
2777 noapic [SMP,APIC] Tells the kernel to not make use of any
2778 IOAPICs that may be present in the system.
2780 noautogroup Disable scheduler automatic task group creation.
2782 nobats [PPC] Do not use BATs for mapping kernel lowmem
2783 on "Classic" PPC cores.
2787 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
2789 nodelayacct [KNL] Disable per-task delay accounting
2791 nodsp [SH] Disable hardware DSP at boot time.
2793 noefi Disable EFI runtime services support.
2798 On X86-32 available only on PAE configured kernels.
2799 noexec=on: enable non-executable mappings (default)
2800 noexec=off: disable non-executable mappings
2803 Disable SMAP (Supervisor Mode Access Prevention)
2804 even if it is supported by processor.
2807 Disable SMEP (Supervisor Mode Execution Prevention)
2808 even if it is supported by processor.
2811 This affects only 32-bit executables.
2812 noexec32=on: enable non-executable mappings (default)
2813 read doesn't imply executable mappings
2814 noexec32=off: disable non-executable mappings
2815 read implies executable mappings
2817 nofpu [MIPS,SH] Disable hardware FPU at boot time.
2819 nofxsr [BUGS=X86-32] Disables x86 floating point extended
2820 register save and restore. The kernel will only save
2821 legacy floating-point registers on task switch.
2823 nohugeiomap [KNL,x86] Disable kernel huge I/O mappings.
2825 nosmt [KNL,S390] Disable symmetric multithreading (SMT).
2826 Equivalent to smt=1.
2828 [KNL,x86] Disable symmetric multithreading (SMT).
2829 nosmt=force: Force disable SMT, cannot be undone
2830 via the sysfs control file.
2832 nospectre_v1 [PPC] Disable mitigations for Spectre Variant 1 (bounds
2833 check bypass). With this option data leaks are possible
2836 nospectre_v2 [X86] Disable all mitigations for the Spectre variant 2
2837 (indirect branch prediction) vulnerability. System may
2838 allow data leaks with this option, which is equivalent
2841 nospec_store_bypass_disable
2842 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability
2844 noxsave [BUGS=X86] Disables x86 extended register state save
2845 and restore using xsave. The kernel will fallback to
2846 enabling legacy floating-point and sse state.
2848 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
2849 register states. The kernel will fall back to use
2850 xsave to save the states. By using this parameter,
2851 performance of saving the states is degraded because
2852 xsave doesn't support modified optimization while
2853 xsaveopt supports it on xsaveopt enabled systems.
2855 noxsaves [X86] Disables xsaves and xrstors used in saving and
2856 restoring x86 extended register state in compacted
2857 form of xsave area. The kernel will fall back to use
2858 xsaveopt and xrstor to save and restore the states
2859 in standard form of xsave area. By using this
2860 parameter, xsave area per process might occupy more
2861 memory on xsaves enabled systems.
2863 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
2864 wfi(ARM) instruction doesn't work correctly and not to
2865 use it. This is also useful when using JTAG debugger.
2867 no_file_caps Tells the kernel not to honor file capabilities. The
2868 only way then for a file to be executed with privilege
2869 is to be setuid root or executed by root.
2871 nohalt [IA-64] Tells the kernel not to use the power saving
2872 function PAL_HALT_LIGHT when idle. This increases
2873 power-consumption. On the positive side, it reduces
2874 interrupt wake-up latency, which may improve performance
2875 in certain environments such as networked servers or
2878 nohibernate [HIBERNATION] Disable hibernation and resume.
2880 nohz= [KNL] Boottime enable/disable dynamic ticks
2881 Valid arguments: on, off
2884 nohz_full= [KNL,BOOT,SMP,ISOL]
2885 The argument is a cpu list, as described above.
2886 In kernels built with CONFIG_NO_HZ_FULL=y, set
2887 the specified list of CPUs whose tick will be stopped
2888 whenever possible. The boot CPU will be forced outside
2889 the range to maintain the timekeeping. Any CPUs
2890 in this list will have their RCU callbacks offloaded,
2891 just as if they had also been called out in the
2892 rcu_nocbs= boot parameter.
2894 noiotrap [SH] Disables trapped I/O port accesses.
2896 noirqdebug [X86-32] Disables the code which attempts to detect and
2897 disable unhandled interrupt sources.
2899 no_timer_check [X86,APIC] Disables the code which tests for
2900 broken timer IRQ sources.
2902 noisapnp [ISAPNP] Disables ISA PnP code.
2904 noinitrd [RAM] Tells the kernel not to load any configured
2907 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
2909 [Deprecated - use intremap=off]
2913 noinvpcid [X86] Disable the INVPCID cpu feature.
2915 nojitter [IA-64] Disables jitter checking for ITC timers.
2917 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
2919 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
2923 [X86,PV_OPS] Disable paravirtualized VMware scheduler
2924 clock and use the default one.
2926 no-steal-acc [X86,KVM] Disable paravirtualized steal time accounting.
2927 steal time is computed, but won't influence scheduler
2930 nolapic [X86-32,APIC] Do not enable or use the local APIC.
2932 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
2934 noltlbs [PPC] Do not use large page/tlb entries for kernel
2935 lowmem mapping on PPC40x and PPC8xx
2937 nomca [IA-64] Disable machine check abort handling
2939 nomce [X86-32] Disable Machine Check Exception
2941 nomfgpt [X86-32] Disable Multi-Function General Purpose
2942 Timer usage (for AMD Geode machines).
2944 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
2945 shutdown the other cpus. Instead use the REBOOT_VECTOR
2948 nomodule Disable module load
2950 nopat [X86] Disable PAT (page attribute table extension of
2951 pagetables) support.
2953 nopcid [X86-64] Disable the PCID cpu feature.
2955 norandmaps Don't use address space randomization. Equivalent to
2956 echo 0 > /proc/sys/kernel/randomize_va_space
2958 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
2959 with UP alternatives
2961 nordrand [X86] Disable kernel use of the RDRAND and
2962 RDSEED instructions even if they are supported
2963 by the processor. RDRAND and RDSEED are still
2964 available to user space applications.
2966 noresume [SWSUSP] Disables resume and restores original swap
2969 no-scroll [VGA] Disables scrollback.
2970 This is required for the Braillex ib80-piezo Braille
2971 reader made by F.H. Papenmeier (Germany).
2975 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
2977 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
2978 and disable the IO APIC. legacy for "maxcpus=0".
2980 nosoftlockup [KNL] Disable the soft-lockup detector.
2982 nosync [HW,M68K] Disables sync negotiation for all devices.
2984 nowatchdog [KNL] Disable both lockup detectors, i.e.
2985 soft-lockup and NMI watchdog (hard-lockup).
2989 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
2991 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
2992 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
2993 Some features depend on CPU0. Known dependencies are:
2994 1. Resume from suspend/hibernate depends on CPU0.
2995 Suspend/hibernate will fail if CPU0 is offline and you
2996 need to online CPU0 before suspend/hibernate.
2997 2. PIC interrupts also depend on CPU0. CPU0 can't be
2998 removed if a PIC interrupt is detected.
2999 It's said poweroff/reboot may depend on CPU0 on some
3000 machines although I haven't seen such issues so far
3001 after CPU0 is offline on a few tested machines.
3002 If the dependencies are under your control, you can
3003 turn on cpu0_hotplug.
3005 nps_mtm_hs_ctr= [KNL,ARC]
3006 This parameter sets the maximum duration, in
3007 cycles, each HW thread of the CTOP can run
3008 without interruptions, before HW switches it.
3009 The actual maximum duration is 16 times this
3011 Format: integer between 1 and 255
3014 nptcg= [IA-64] Override max number of concurrent global TLB
3015 purges which is reported from either PAL_VM_SUMMARY or
3018 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
3019 could support. nr_cpus=n : n >= 1 limits the kernel to
3020 support 'n' processors. It could be larger than the
3021 number of already plugged CPU during bootup, later in
3022 runtime you can physically add extra cpu until it reaches
3023 n. So during boot up some boot time memory for per-cpu
3024 variables need be pre-allocated for later physical cpu
3027 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
3029 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing.
3030 Allowed values are enable and disable
3032 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
3033 'node', 'default' can be specified
3034 This can be set from sysctl after boot.
3035 See Documentation/sysctl/vm.txt for details.
3037 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
3038 See Documentation/debugging-via-ohci1394.txt for more
3041 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
3042 Rather than timing out after 20 ms if an EC
3043 command is not properly ACKed, override the length
3044 of the timeout. We have interrupts disabled while
3045 waiting for the ACK, so if this is set too high
3046 interrupts *may* be lost!
3048 omap_mux= [OMAP] Override bootloader pin multiplexing.
3049 Format: <mux_mode0.mode_name=value>...
3050 For example, to override I2C bus2:
3051 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
3053 oprofile.timer= [HW]
3054 Use timer interrupt instead of performance counters
3056 oprofile.cpu_type= Force an oprofile cpu type
3057 This might be useful if you have an older oprofile
3058 userland or if you want common events.
3059 Format: { arch_perfmon }
3060 arch_perfmon: [X86] Force use of architectural
3061 perfmon on Intel CPUs instead of the
3062 CPU specific event set.
3063 timer: [X86] Force use of architectural NMI
3064 timer mode (see also oprofile.timer
3065 for generic hr timer mode)
3067 oops=panic Always panic on oopses. Default is to just kill the
3068 process, but there is a small probability of
3069 deadlocking the machine.
3070 This will also cause panics on machine check exceptions.
3071 Useful together with panic=30 to trigger a reboot.
3073 page_owner= [KNL] Boot-time page_owner enabling option.
3074 Storage of the information about who allocated
3075 each page is disabled in default. With this switch,
3077 on: enable the feature
3079 page_poison= [KNL] Boot-time parameter changing the state of
3080 poisoning on the buddy allocator, available with
3081 CONFIG_PAGE_POISONING=y.
3082 off: turn off poisoning (default)
3083 on: turn on poisoning
3085 panic= [KNL] Kernel behaviour on panic: delay <timeout>
3086 timeout > 0: seconds before rebooting
3087 timeout = 0: wait forever
3088 timeout < 0: reboot immediately
3091 panic_on_warn panic() instead of WARN(). Useful to cause kdump
3094 crash_kexec_post_notifiers
3095 Run kdump after running panic-notifiers and dumping
3096 kmsg. This only for the users who doubt kdump always
3097 succeeds in any situation.
3098 Note that this also increases risks of kdump failure,
3099 because some panic notifiers can make the crashed
3100 kernel more unstable.
3102 parkbd.port= [HW] Parallel port number the keyboard adapter is
3103 connected to, default is 0.
3105 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
3106 0 for XT, 1 for AT (default is AT).
3109 parport= [HW,PPT] Specify parallel ports. 0 disables.
3110 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
3111 Use 'auto' to force the driver to use any
3112 IRQ/DMA settings detected (the default is to
3113 ignore detected IRQ/DMA settings because of
3114 possible conflicts). You can specify the base
3115 address, IRQ, and DMA settings; IRQ and DMA
3116 should be numbers, or 'auto' (for using detected
3117 settings on that particular port), or 'nofifo'
3118 (to avoid using a FIFO even if it is detected).
3119 Parallel ports are assigned in the order they
3120 are specified on the command line, starting
3123 parport_init_mode= [HW,PPT]
3124 Configure VIA parallel port to operate in
3125 a specific mode. This is necessary on Pegasos
3126 computer where firmware has no options for setting
3127 up parallel port mode and sets it to spp.
3128 Currently this function knows 686a and 8231 chips.
3129 Format: [spp|ps2|epp|ecp|ecpepp]
3132 Halt all CPUs after the first oops has been printed for
3133 the specified number of seconds. This is to be used if
3134 your oopses keep scrolling off the screen.
3139 See header of drivers/block/paride/pcd.c.
3140 See also Documentation/blockdev/paride.txt.
3142 pci=option[,option...] [PCI] various PCI subsystem options.
3144 Some options herein operate on a specific device
3145 or a set of devices (<pci_dev>). These are
3146 specified in one of the following formats:
3148 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]*
3149 pci:<vendor>:<device>[:<subvendor>:<subdevice>]
3151 Note: the first format specifies a PCI
3152 bus/device/function address which may change
3153 if new hardware is inserted, if motherboard
3154 firmware changes, or due to changes caused
3155 by other kernel parameters. If the
3156 domain is left unspecified, it is
3157 taken to be zero. Optionally, a path
3158 to a device through multiple device/function
3159 addresses can be specified after the base
3160 address (this is more robust against
3161 renumbering issues). The second format
3162 selects devices using IDs from the
3163 configuration space which may match multiple
3164 devices in the system.
3166 earlydump dump PCI config space before the kernel
3168 off [X86] don't probe for the PCI bus
3169 bios [X86-32] force use of PCI BIOS, don't access
3170 the hardware directly. Use this if your machine
3171 has a non-standard PCI host bridge.
3172 nobios [X86-32] disallow use of PCI BIOS, only direct
3173 hardware access methods are allowed. Use this
3174 if you experience crashes upon bootup and you
3175 suspect they are caused by the BIOS.
3176 conf1 [X86] Force use of PCI Configuration Access
3177 Mechanism 1 (config address in IO port 0xCF8,
3178 data in IO port 0xCFC, both 32-bit).
3179 conf2 [X86] Force use of PCI Configuration Access
3180 Mechanism 2 (IO port 0xCF8 is an 8-bit port for
3181 the function, IO port 0xCFA, also 8-bit, sets
3182 bus number. The config space is then accessed
3183 through ports 0xC000-0xCFFF).
3184 See http://wiki.osdev.org/PCI for more info
3185 on the configuration access mechanisms.
3186 noaer [PCIE] If the PCIEAER kernel config parameter is
3187 enabled, this kernel boot option can be used to
3188 disable the use of PCIE advanced error reporting.
3189 nodomains [PCI] Disable support for multiple PCI
3190 root domains (aka PCI segments, in ACPI-speak).
3191 nommconf [X86] Disable use of MMCONFIG for PCI
3193 check_enable_amd_mmconf [X86] check for and enable
3194 properly configured MMIO access to PCI
3195 config space on AMD family 10h CPU
3196 nomsi [MSI] If the PCI_MSI kernel config parameter is
3197 enabled, this kernel boot option can be used to
3198 disable the use of MSI interrupts system-wide.
3199 noioapicquirk [APIC] Disable all boot interrupt quirks.
3200 Safety option to keep boot IRQs enabled. This
3201 should never be necessary.
3202 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
3203 primary IO-APIC for bridges that cannot disable
3204 boot IRQs. This fixes a source of spurious IRQs
3205 when the system masks IRQs.
3206 noioapicreroute [APIC] Disable workaround that uses the
3207 boot IRQ equivalent of an IRQ that connects to
3208 a chipset where boot IRQs cannot be disabled.
3209 The opposite of ioapicreroute.
3210 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
3211 routing table. These calls are known to be buggy
3212 on several machines and they hang the machine
3213 when used, but on other computers it's the only
3214 way to get the interrupt routing table. Try
3215 this option if the kernel is unable to allocate
3216 IRQs or discover secondary PCI buses on your
3218 rom [X86] Assign address space to expansion ROMs.
3219 Use with caution as certain devices share
3220 address decoders between ROMs and other
3222 norom [X86] Do not assign address space to
3223 expansion ROMs that do not already have
3224 BIOS assigned address ranges.
3225 nobar [X86] Do not assign address space to the
3226 BARs that weren't assigned by the BIOS.
3227 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
3228 assigned automatically to PCI devices. You can
3229 make the kernel exclude IRQs of your ISA cards
3231 pirqaddr=0xAAAAA [X86] Specify the physical address
3232 of the PIRQ table (normally generated
3233 by the BIOS) if it is outside the
3234 F0000h-100000h range.
3235 lastbus=N [X86] Scan all buses thru bus #N. Can be
3236 useful if the kernel is unable to find your
3237 secondary buses and you want to tell it
3238 explicitly which ones they are.
3239 assign-busses [X86] Always assign all PCI bus
3240 numbers ourselves, overriding
3241 whatever the firmware may have done.
3242 usepirqmask [X86] Honor the possible IRQ mask stored
3243 in the BIOS $PIR table. This is needed on
3244 some systems with broken BIOSes, notably
3245 some HP Pavilion N5400 and Omnibook XE3
3246 notebooks. This will have no effect if ACPI
3247 IRQ routing is enabled.
3248 noacpi [X86] Do not use ACPI for IRQ routing
3249 or for PCI scanning.
3250 use_crs [X86] Use PCI host bridge window information
3251 from ACPI. On BIOSes from 2008 or later, this
3252 is enabled by default. If you need to use this,
3253 please report a bug.
3254 nocrs [X86] Ignore PCI host bridge windows from ACPI.
3255 If you need to use this, please report a bug.
3256 routeirq Do IRQ routing for all PCI devices.
3257 This is normally done in pci_enable_device(),
3258 so this option is a temporary workaround
3259 for broken drivers that don't call it.
3260 skip_isa_align [X86] do not align io start addr, so can
3261 handle more pci cards
3262 noearly [X86] Don't do any early type 1 scanning.
3263 This might help on some broken boards which
3264 machine check when some devices' config space
3265 is read. But various workarounds are disabled
3266 and some IOMMU drivers will not work.
3267 bfsort Sort PCI devices into breadth-first order.
3268 This sorting is done to get a device
3269 order compatible with older (<= 2.4) kernels.
3270 nobfsort Don't sort PCI devices into breadth-first order.
3271 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
3272 tuning and use the BIOS-configured MPS defaults.
3273 pcie_bus_safe Set every device's MPS to the largest value
3274 supported by all devices below the root complex.
3275 pcie_bus_perf Set device MPS to the largest allowable MPS
3276 based on its parent bus. Also set MRRS (Max
3277 Read Request Size) to the largest supported
3278 value (no larger than the MPS that the device
3279 or bus can support) for best performance.
3280 pcie_bus_peer2peer Set every device's MPS to 128B, which
3281 every device is guaranteed to support. This
3282 configuration allows peer-to-peer DMA between
3283 any pair of devices, possibly at the cost of
3284 reduced performance. This also guarantees
3285 that hot-added devices will work.
3286 cbiosize=nn[KMG] The fixed amount of bus space which is
3287 reserved for the CardBus bridge's IO window.
3288 The default value is 256 bytes.
3289 cbmemsize=nn[KMG] The fixed amount of bus space which is
3290 reserved for the CardBus bridge's memory
3291 window. The default value is 64 megabytes.
3294 [<order of align>@]<pci_dev>[; ...]
3295 Specifies alignment and device to reassign
3296 aligned memory resources. How to
3297 specify the device is described above.
3298 If <order of align> is not specified,
3299 PAGE_SIZE is used as alignment.
3300 PCI-PCI bridge can be specified, if resource
3301 windows need to be expanded.
3302 To specify the alignment for several
3303 instances of a device, the PCI vendor,
3304 device, subvendor, and subdevice may be
3305 specified, e.g., 4096@pci:8086:9c22:103c:198f
3306 ecrc= Enable/disable PCIe ECRC (transaction layer
3307 end-to-end CRC checking).
3308 bios: Use BIOS/firmware settings. This is the
3312 hpiosize=nn[KMG] The fixed amount of bus space which is
3313 reserved for hotplug bridge's IO window.
3314 Default size is 256 bytes.
3315 hpmemsize=nn[KMG] The fixed amount of bus space which is
3316 reserved for hotplug bridge's memory window.
3317 Default size is 2 megabytes.
3318 hpbussize=nn The minimum amount of additional bus numbers
3319 reserved for buses below a hotplug bridge.
3321 realloc= Enable/disable reallocating PCI bridge resources
3322 if allocations done by BIOS are too small to
3323 accommodate resources required by all child
3325 off: Turn realloc off
3327 realloc same as realloc=on
3328 noari do not use PCIe ARI.
3329 noats [PCIE, Intel-IOMMU, AMD-IOMMU]
3330 do not use PCIe ATS (and IOMMU device IOTLB).
3331 pcie_scan_all Scan all possible PCIe devices. Otherwise we
3332 only look for one device below a PCIe downstream
3334 big_root_window Try to add a big 64bit memory window to the PCIe
3335 root complex on AMD CPUs. Some GFX hardware
3336 can resize a BAR to allow access to all VRAM.
3337 Adding the window is slightly risky (it may
3338 conflict with unreported devices), so this
3340 disable_acs_redir=<pci_dev>[; ...]
3341 Specify one or more PCI devices (in the format
3342 specified above) separated by semicolons.
3343 Each device specified will have the PCI ACS
3344 redirect capabilities forced off which will
3345 allow P2P traffic between devices through
3346 bridges without forcing it upstream. Note:
3347 this removes isolation between devices and
3348 may put more devices in an IOMMU group.
3350 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
3353 force Enable ASPM even on devices that claim not to support it.
3354 WARNING: Forcing ASPM on may cause system lockups.
3356 pcie_ports= [PCIE] PCIe port services handling:
3357 native Use native PCIe services (PME, AER, DPC, PCIe hotplug)
3358 even if the platform doesn't give the OS permission to
3359 use them. This may cause conflicts if the platform
3360 also tries to use these services.
3361 compat Disable native PCIe services (PME, AER, DPC, PCIe
3364 pcie_port_pm= [PCIE] PCIe port power management handling:
3365 off Disable power management of all PCIe ports
3366 force Forcibly enable power management of all PCIe ports
3368 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
3369 nomsi Do not use MSI for native PCIe PME signaling (this makes
3370 all PCIe root ports use INTx for all services).
3372 pcmv= [HW,PCMCIA] BadgePAD 4
3376 Keep all power-domains already enabled by bootloader on,
3377 even if no driver has claimed them. This is useful
3378 for debug and development, but should not be
3379 needed on a platform with proper driver support.
3382 See Documentation/blockdev/paride.txt.
3384 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
3387 See arch/parisc/kernel/pdc_chassis.c
3389 percpu_alloc= Select which percpu first chunk allocator to use.
3390 Currently supported values are "embed" and "page".
3391 Archs may support subset or none of the selections.
3392 See comments in mm/percpu.c for details on each
3393 allocator. This parameter is primarily for debugging
3394 and performance comparison.
3397 See Documentation/blockdev/paride.txt.
3400 See Documentation/blockdev/paride.txt.
3402 pirq= [SMP,APIC] Manual mp-table setup
3403 See Documentation/x86/i386/IO-APIC.txt.
3405 plip= [PPT,NET] Parallel port network link
3406 Format: { parport<nr> | timid | 0 }
3407 See also Documentation/admin-guide/parport.rst.
3409 pmtmr= [X86] Manual setup of pmtmr I/O Port.
3410 Override pmtimer IOPort with a hex value.
3414 Enable PNP debug messages (depends on the
3415 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
3416 via /sys/module/pnp/parameters/debug. We always show
3417 current resource usage; turning this on also shows
3418 possible settings and some assignment information.
3424 { on | off | curr | res | no-curr | no-res }
3427 [ISAPNP] Exclude IRQs for the autoconfiguration
3430 [ISAPNP] Exclude DMAs for the autoconfiguration
3432 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
3433 Ranges are in pairs (I/O port base and size).
3436 [ISAPNP] Exclude memory regions for the
3438 Ranges are in pairs (memory base and size).
3440 ports= [IP_VS_FTP] IPVS ftp helper module
3442 Up to 8 (IP_VS_APP_MAX_PORTS) ports
3444 Format: <port>,<port>....
3446 powersave=off [PPC] This option disables power saving features.
3447 It specifically disables cpuidle and sets the
3448 platform machine description specific power_save
3449 function to NULL. On Idle the CPU just reduces
3452 ppc_strict_facility_enable
3453 [PPC] This option catches any kernel floating point,
3454 Altivec, VSX and SPE outside of regions specifically
3455 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
3456 There is some performance impact when enabling this.
3460 Disable Hardware Transactional Memory
3462 print-fatal-signals=
3463 [KNL] debug: print fatal signals
3465 If enabled, warn about various signal handling
3466 related application anomalies: too many signals,
3467 too many POSIX.1 timers, fatal signals causing a
3470 If you hit the warning due to signal overflow,
3471 you might want to try "ulimit -i unlimited".
3475 printk.always_kmsg_dump=
3476 Trigger kmsg_dump for cases other than kernel oops or
3478 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3481 printk.devkmsg={on,off,ratelimit}
3482 Control writing to /dev/kmsg.
3483 on - unlimited logging to /dev/kmsg from userspace
3484 off - logging to /dev/kmsg disabled
3485 ratelimit - ratelimit the logging
3488 printk.time= Show timing data prefixed to each printk message line
3489 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3491 processor.max_cstate= [HW,ACPI]
3492 Limit processor to maximum C-state
3493 max_cstate=9 overrides any DMI blacklist limit.
3495 processor.nocst [HW,ACPI]
3496 Ignore the _CST method to determine C-states,
3497 instead using the legacy FADT method
3499 profile= [KNL] Enable kernel profiling via /proc/profile
3500 Format: [<profiletype>,]<number>
3501 Param: <profiletype>: "schedule", "sleep", or "kvm"
3502 [defaults to kernel profiling]
3503 Param: "schedule" - profile schedule points.
3504 Param: "sleep" - profile D-state sleeping (millisecs).
3505 Requires CONFIG_SCHEDSTATS
3506 Param: "kvm" - profile VM exits.
3507 Param: <number> - step/bucket size as a power of 2 for
3508 statistical time based profiling.
3510 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
3512 See Documentation/blockdev/ramdisk.txt.
3514 psi= [KNL] Enable or disable pressure stall information
3518 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
3519 probe for; one of (bare|imps|exps|lifebook|any).
3520 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
3522 psmouse.resetafter= [HW,MOUSE]
3523 Try to reset the device after so many bad packets
3526 [HW,MOUSE] Set desired mouse resolution, in dpi.
3527 psmouse.smartscroll=
3528 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
3529 0 = disabled, 1 = enabled (default).
3531 pstore.backend= Specify the name of the pstore backend to use
3534 See Documentation/blockdev/paride.txt.
3536 pti= [X86_64] Control Page Table Isolation of user and
3537 kernel address spaces. Disabling this feature
3538 removes hardening, but improves performance of
3539 system calls and interrupts.
3541 on - unconditionally enable
3542 off - unconditionally disable
3543 auto - kernel detects whether your CPU model is
3544 vulnerable to issues that PTI mitigates
3546 Not specifying this option is equivalent to pti=auto.
3549 Equivalent to pti=off
3552 [KNL] Number of legacy pty's. Overwrites compiled-in
3555 quiet [KNL] Disable most log messages
3560 See Documentation/admin-guide/md.rst.
3562 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
3563 See Documentation/blockdev/ramdisk.txt.
3565 random.trust_cpu={on,off}
3566 [KNL] Enable or disable trusting the use of the
3567 CPU's random number generator (if available) to
3568 fully seed the kernel's CRNG. Default is controlled
3569 by CONFIG_RANDOM_TRUST_CPU.
3571 ras=option[,option,...] [KNL] RAS-specific options
3574 Disable the Correctable Errors Collector,
3575 see CONFIG_RAS_CEC help text.
3578 The argument is a cpu list, as described above.
3580 In kernels built with CONFIG_RCU_NOCB_CPU=y, set
3581 the specified list of CPUs to be no-callback CPUs.
3582 Invocation of these CPUs' RCU callbacks will be
3583 offloaded to "rcuox/N" kthreads created for that
3584 purpose, where "x" is "p" for RCU-preempt, and
3585 "s" for RCU-sched, and "N" is the CPU number.
3586 This reduces OS jitter on the offloaded CPUs,
3587 which can be useful for HPC and real-time
3588 workloads. It can also improve energy efficiency
3589 for asymmetric multiprocessors.
3592 Rather than requiring that offloaded CPUs
3593 (specified by rcu_nocbs= above) explicitly
3594 awaken the corresponding "rcuoN" kthreads,
3595 make these kthreads poll for callbacks.
3596 This improves the real-time response for the
3597 offloaded CPUs by relieving them of the need to
3598 wake up the corresponding kthread, but degrades
3599 energy efficiency by requiring that the kthreads
3600 periodically wake up to do the polling.
3602 rcutree.blimit= [KNL]
3603 Set maximum number of finished RCU callbacks to
3604 process in one batch.
3606 rcutree.dump_tree= [KNL]
3607 Dump the structure of the rcu_node combining tree
3608 out at early boot. This is used for diagnostic
3609 purposes, to verify correct tree setup.
3611 rcutree.gp_cleanup_delay= [KNL]
3612 Set the number of jiffies to delay each step of
3613 RCU grace-period cleanup.
3615 rcutree.gp_init_delay= [KNL]
3616 Set the number of jiffies to delay each step of
3617 RCU grace-period initialization.
3619 rcutree.gp_preinit_delay= [KNL]
3620 Set the number of jiffies to delay each step of
3621 RCU grace-period pre-initialization, that is,
3622 the propagation of recent CPU-hotplug changes up
3623 the rcu_node combining tree.
3625 rcutree.rcu_fanout_exact= [KNL]
3626 Disable autobalancing of the rcu_node combining
3627 tree. This is used by rcutorture, and might
3628 possibly be useful for architectures having high
3629 cache-to-cache transfer latencies.
3631 rcutree.rcu_fanout_leaf= [KNL]
3632 Change the number of CPUs assigned to each
3633 leaf rcu_node structure. Useful for very
3634 large systems, which will choose the value 64,
3635 and for NUMA systems with large remote-access
3636 latencies, which will choose a value aligned
3637 with the appropriate hardware boundaries.
3639 rcutree.jiffies_till_sched_qs= [KNL]
3640 Set required age in jiffies for a
3641 given grace period before RCU starts
3642 soliciting quiescent-state help from
3643 rcu_note_context_switch(). If not specified, the
3644 kernel will calculate a value based on the most
3645 recent settings of rcutree.jiffies_till_first_fqs
3646 and rcutree.jiffies_till_next_fqs.
3647 This calculated value may be viewed in
3648 rcutree.jiffies_to_sched_qs. Any attempt to
3649 set rcutree.jiffies_to_sched_qs will be
3650 cheerfully overwritten.
3652 rcutree.jiffies_till_first_fqs= [KNL]
3653 Set delay from grace-period initialization to
3654 first attempt to force quiescent states.
3655 Units are jiffies, minimum value is zero,
3656 and maximum value is HZ.
3658 rcutree.jiffies_till_next_fqs= [KNL]
3659 Set delay between subsequent attempts to force
3660 quiescent states. Units are jiffies, minimum
3661 value is one, and maximum value is HZ.
3663 rcutree.kthread_prio= [KNL,BOOT]
3664 Set the SCHED_FIFO priority of the RCU per-CPU
3665 kthreads (rcuc/N). This value is also used for
3666 the priority of the RCU boost threads (rcub/N)
3667 and for the RCU grace-period kthreads (rcu_bh,
3668 rcu_preempt, and rcu_sched). If RCU_BOOST is
3669 set, valid values are 1-99 and the default is 1
3670 (the least-favored priority). Otherwise, when
3671 RCU_BOOST is not set, valid values are 0-99 and
3672 the default is zero (non-realtime operation).
3674 rcutree.rcu_nocb_leader_stride= [KNL]
3675 Set the number of NOCB kthread groups, which
3676 defaults to the square root of the number of
3677 CPUs. Larger numbers reduces the wakeup overhead
3678 on the per-CPU grace-period kthreads, but increases
3679 that same overhead on each group's leader.
3681 rcutree.qhimark= [KNL]
3682 Set threshold of queued RCU callbacks beyond which
3683 batch limiting is disabled.
3685 rcutree.qlowmark= [KNL]
3686 Set threshold of queued RCU callbacks below which
3687 batch limiting is re-enabled.
3689 rcutree.rcu_idle_gp_delay= [KNL]
3690 Set wakeup interval for idle CPUs that have
3691 RCU callbacks (RCU_FAST_NO_HZ=y).
3693 rcutree.rcu_idle_lazy_gp_delay= [KNL]
3694 Set wakeup interval for idle CPUs that have
3695 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
3696 Lazy RCU callbacks are those which RCU can
3697 prove do nothing more than free memory.
3699 rcutree.rcu_kick_kthreads= [KNL]
3700 Cause the grace-period kthread to get an extra
3701 wake_up() if it sleeps three times longer than
3702 it should at force-quiescent-state time.
3703 This wake_up() will be accompanied by a
3704 WARN_ONCE() splat and an ftrace_dump().
3706 rcuperf.gp_async= [KNL]
3707 Measure performance of asynchronous
3708 grace-period primitives such as call_rcu().
3710 rcuperf.gp_async_max= [KNL]
3711 Specify the maximum number of outstanding
3712 callbacks per writer thread. When a writer
3713 thread exceeds this limit, it invokes the
3714 corresponding flavor of rcu_barrier() to allow
3715 previously posted callbacks to drain.
3717 rcuperf.gp_exp= [KNL]
3718 Measure performance of expedited synchronous
3719 grace-period primitives.
3721 rcuperf.holdoff= [KNL]
3722 Set test-start holdoff period. The purpose of
3723 this parameter is to delay the start of the
3724 test until boot completes in order to avoid
3727 rcuperf.nreaders= [KNL]
3728 Set number of RCU readers. The value -1 selects
3729 N, where N is the number of CPUs. A value
3730 "n" less than -1 selects N-n+1, where N is again
3731 the number of CPUs. For example, -2 selects N
3732 (the number of CPUs), -3 selects N+1, and so on.
3733 A value of "n" less than or equal to -N selects
3736 rcuperf.nwriters= [KNL]
3737 Set number of RCU writers. The values operate
3738 the same as for rcuperf.nreaders.
3739 N, where N is the number of CPUs
3741 rcuperf.perf_type= [KNL]
3742 Specify the RCU implementation to test.
3744 rcuperf.shutdown= [KNL]
3745 Shut the system down after performance tests
3746 complete. This is useful for hands-off automated
3749 rcuperf.verbose= [KNL]
3750 Enable additional printk() statements.
3752 rcuperf.writer_holdoff= [KNL]
3753 Write-side holdoff between grace periods,
3754 in microseconds. The default of zero says
3757 rcutorture.cbflood_inter_holdoff= [KNL]
3758 Set holdoff time (jiffies) between successive
3759 callback-flood tests.
3761 rcutorture.cbflood_intra_holdoff= [KNL]
3762 Set holdoff time (jiffies) between successive
3763 bursts of callbacks within a given callback-flood
3766 rcutorture.cbflood_n_burst= [KNL]
3767 Set the number of bursts making up a given
3768 callback-flood test. Set this to zero to
3769 disable callback-flood testing.
3771 rcutorture.cbflood_n_per_burst= [KNL]
3772 Set the number of callbacks to be registered
3773 in a given burst of a callback-flood test.
3775 rcutorture.fqs_duration= [KNL]
3776 Set duration of force_quiescent_state bursts
3779 rcutorture.fqs_holdoff= [KNL]
3780 Set holdoff time within force_quiescent_state bursts
3783 rcutorture.fqs_stutter= [KNL]
3784 Set wait time between force_quiescent_state bursts
3787 rcutorture.gp_cond= [KNL]
3788 Use conditional/asynchronous update-side
3789 primitives, if available.
3791 rcutorture.gp_exp= [KNL]
3792 Use expedited update-side primitives, if available.
3794 rcutorture.gp_normal= [KNL]
3795 Use normal (non-expedited) asynchronous
3796 update-side primitives, if available.
3798 rcutorture.gp_sync= [KNL]
3799 Use normal (non-expedited) synchronous
3800 update-side primitives, if available. If all
3801 of rcutorture.gp_cond=, rcutorture.gp_exp=,
3802 rcutorture.gp_normal=, and rcutorture.gp_sync=
3803 are zero, rcutorture acts as if is interpreted
3804 they are all non-zero.
3806 rcutorture.n_barrier_cbs= [KNL]
3807 Set callbacks/threads for rcu_barrier() testing.
3809 rcutorture.nfakewriters= [KNL]
3810 Set number of concurrent RCU writers. These just
3811 stress RCU, they don't participate in the actual
3812 test, hence the "fake".
3814 rcutorture.nreaders= [KNL]
3815 Set number of RCU readers. The value -1 selects
3816 N-1, where N is the number of CPUs. A value
3817 "n" less than -1 selects N-n-2, where N is again
3818 the number of CPUs. For example, -2 selects N
3819 (the number of CPUs), -3 selects N+1, and so on.
3821 rcutorture.object_debug= [KNL]
3822 Enable debug-object double-call_rcu() testing.
3824 rcutorture.onoff_holdoff= [KNL]
3825 Set time (s) after boot for CPU-hotplug testing.
3827 rcutorture.onoff_interval= [KNL]
3828 Set time (jiffies) between CPU-hotplug operations,
3829 or zero to disable CPU-hotplug testing.
3831 rcutorture.shuffle_interval= [KNL]
3832 Set task-shuffle interval (s). Shuffling tasks
3833 allows some CPUs to go into dyntick-idle mode
3834 during the rcutorture test.
3836 rcutorture.shutdown_secs= [KNL]
3837 Set time (s) after boot system shutdown. This
3838 is useful for hands-off automated testing.
3840 rcutorture.stall_cpu= [KNL]
3841 Duration of CPU stall (s) to test RCU CPU stall
3842 warnings, zero to disable.
3844 rcutorture.stall_cpu_holdoff= [KNL]
3845 Time to wait (s) after boot before inducing stall.
3847 rcutorture.stall_cpu_irqsoff= [KNL]
3848 Disable interrupts while stalling if set.
3850 rcutorture.stat_interval= [KNL]
3851 Time (s) between statistics printk()s.
3853 rcutorture.stutter= [KNL]
3854 Time (s) to stutter testing, for example, specifying
3855 five seconds causes the test to run for five seconds,
3856 wait for five seconds, and so on. This tests RCU's
3857 ability to transition abruptly to and from idle.
3859 rcutorture.test_boost= [KNL]
3860 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
3861 "Maybe" means test if the RCU implementation
3862 under test support RCU priority boosting.
3864 rcutorture.test_boost_duration= [KNL]
3865 Duration (s) of each individual boost test.
3867 rcutorture.test_boost_interval= [KNL]
3868 Interval (s) between each boost test.
3870 rcutorture.test_no_idle_hz= [KNL]
3871 Test RCU's dyntick-idle handling. See also the
3872 rcutorture.shuffle_interval parameter.
3874 rcutorture.torture_type= [KNL]
3875 Specify the RCU implementation to test.
3877 rcutorture.verbose= [KNL]
3878 Enable additional printk() statements.
3880 rcupdate.rcu_cpu_stall_suppress= [KNL]
3881 Suppress RCU CPU stall warning messages.
3883 rcupdate.rcu_cpu_stall_timeout= [KNL]
3884 Set timeout for RCU CPU stall warning messages.
3886 rcupdate.rcu_expedited= [KNL]
3887 Use expedited grace-period primitives, for
3888 example, synchronize_rcu_expedited() instead
3889 of synchronize_rcu(). This reduces latency,
3890 but can increase CPU utilization, degrade
3891 real-time latency, and degrade energy efficiency.
3892 No effect on CONFIG_TINY_RCU kernels.
3894 rcupdate.rcu_normal= [KNL]
3895 Use only normal grace-period primitives,
3896 for example, synchronize_rcu() instead of
3897 synchronize_rcu_expedited(). This improves
3898 real-time latency, CPU utilization, and
3899 energy efficiency, but can expose users to
3900 increased grace-period latency. This parameter
3901 overrides rcupdate.rcu_expedited. No effect on
3902 CONFIG_TINY_RCU kernels.
3904 rcupdate.rcu_normal_after_boot= [KNL]
3905 Once boot has completed (that is, after
3906 rcu_end_inkernel_boot() has been invoked), use
3907 only normal grace-period primitives. No effect
3908 on CONFIG_TINY_RCU kernels.
3910 rcupdate.rcu_task_stall_timeout= [KNL]
3911 Set timeout in jiffies for RCU task stall warning
3912 messages. Disable with a value less than or equal
3915 rcupdate.rcu_self_test= [KNL]
3916 Run the RCU early boot self tests
3920 Run specified binary instead of /init from the ramdisk,
3921 used for early userspace startup. See initrd.
3924 Turn on/off individual RDT features. List is:
3925 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp,
3927 E.g. to turn on cmt and turn off mba use:
3931 Format (x86 or x86_64):
3932 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
3934 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
3936 Where reboot_mode is one of warm (soft) or cold (hard) or gpio,
3937 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
3938 reboot_force is either force or not specified,
3939 reboot_cpu is s[mp]#### with #### being the processor
3940 to be used for rebooting.
3943 [KNL, SMP] Set scheduler's default relax_domain_level.
3944 See Documentation/cgroup-v1/cpusets.txt.
3946 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory
3947 Format: <base1>,<size1>[,<base2>,<size2>,...]
3948 Reserve I/O ports or memory so the kernel won't use
3949 them. If <base> is less than 0x10000, the region
3950 is assumed to be I/O ports; otherwise it is memory.
3952 reservetop= [X86-32]
3954 Reserves a hole at the top of the kernel virtual
3959 Set the amount of memory to reserve for BIOS at
3960 the bottom of the address space.
3962 reset_devices [KNL] Force drivers to reset the underlying device
3963 during initialization.
3966 Specify the partition device for software suspend
3968 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
3970 resume_offset= [SWSUSP]
3971 Specify the offset from the beginning of the partition
3972 given by "resume=" at which the swap header is located,
3973 in <PAGE_SIZE> units (needed only for swap files).
3974 See Documentation/power/swsusp-and-swap-files.txt
3976 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
3977 read the resume files
3979 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
3980 Useful for devices that are detected asynchronously
3981 (e.g. USB and MMC devices).
3983 hibernate= [HIBERNATION]
3984 noresume Don't check if there's a hibernation image
3985 present during boot.
3986 nocompress Don't compress/decompress hibernation images.
3987 no Disable hibernation and resume.
3988 protect_image Turn on image protection during restoration
3989 (that will set all pages holding image data
3990 during restoration read-only).
3992 retain_initrd [RAM] Keep initrd memory after extraction
3994 rfkill.default_state=
3995 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
3996 etc. communication is blocked by default.
3999 rfkill.master_switch_mode=
4000 0 The "airplane mode" button does nothing.
4001 1 The "airplane mode" button toggles between everything
4002 blocked and the previous configuration.
4003 2 The "airplane mode" button toggles between everything
4004 blocked and everything unblocked.
4006 rhash_entries= [KNL,NET]
4007 Set number of hash buckets for route cache
4010 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported
4013 ro [KNL] Mount root device read-only on boot
4016 on Mark read-only kernel memory as read-only (default).
4017 off Leave read-only kernel memory writable for debugging.
4020 Enable the uart passthrough on the designated usb port
4021 on Rockchip SoCs. When active, the signals of the
4022 debug-uart get routed to the D+ and D- pins of the usb
4023 port and the regular usb controller gets disabled.
4025 root= [KNL] Root filesystem
4026 See name_to_dev_t comment in init/do_mounts.c.
4028 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
4029 mount the root filesystem
4031 rootflags= [KNL] Set root filesystem mount option string
4033 rootfstype= [KNL] Set root filesystem type
4035 rootwait [KNL] Wait (indefinitely) for root device to show up.
4036 Useful for devices that are detected asynchronously
4037 (e.g. USB and MMC devices).
4039 rproc_mem=nn[KMG][@address]
4040 [KNL,ARM,CMA] Remoteproc physical memory block.
4041 Memory area to be used by remote processor image,
4044 rw [KNL] Mount root device read-write on boot
4046 S [KNL] Run init in single mode
4048 s390_iommu= [HW,S390]
4049 Set s390 IOTLB flushing mode
4051 With strict flushing every unmap operation will result in
4052 an IOTLB flush. Default is lazy flushing before reuse,
4056 See drivers/net/irda/sa1100_ir.c.
4058 sbni= [NET] Granch SBNI12 leased line adapter
4060 sched_debug [KNL] Enables verbose scheduler debug messages.
4062 schedstats= [KNL,X86] Enable or disable scheduled statistics.
4063 Allowed values are enable and disable. This feature
4064 incurs a small amount of overhead in the scheduler
4065 but is useful for debugging and performance tuning.
4067 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
4068 xtime_lock contention on larger systems, and/or RCU lock
4069 contention on all systems with CONFIG_MAXSMP set.
4070 Format: { "0" | "1" }
4071 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
4073 Note: increases power consumption, thus should only be
4074 enabled if running jitter sensitive (HPC/RT) workloads.
4076 security= [SECURITY] Choose a security module to enable at boot.
4077 If this boot parameter is not specified, only the first
4078 security module asking for security registration will be
4079 loaded. An invalid security module name will be treated
4080 as if no module has been chosen.
4082 selinux= [SELINUX] Disable or enable SELinux at boot time.
4083 Format: { "0" | "1" }
4084 See security/selinux/Kconfig help text.
4087 Default value is set via kernel config option.
4088 If enabled at boot time, /selinux/disable can be used
4089 later to disable prior to initial policy load.
4091 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
4092 Format: { "0" | "1" }
4093 See security/apparmor/Kconfig help text
4096 Default value is set via kernel config option.
4098 serialnumber [BUGS=X86-32]
4101 Maximal number of shapers.
4109 Disable merging of slabs with similar size. May be
4110 necessary if there is some reason to distinguish
4111 allocs to different slabs, especially in hardened
4112 environments where the risk of heap overflows and
4113 layout control by attackers can usually be
4114 frustrated by disabling merging. This will reduce
4115 most of the exposure of a heap attack to a single
4116 cache (risks via metadata attacks are mostly
4117 unchanged). Debug options disable merging on their
4119 For more information see Documentation/vm/slub.rst.
4121 slab_max_order= [MM, SLAB]
4122 Determines the maximum allowed order for slabs.
4123 A high setting may cause OOMs due to memory
4124 fragmentation. Defaults to 1 for systems with
4125 more than 32MB of RAM, 0 otherwise.
4127 slub_debug[=options[,slabs]] [MM, SLUB]
4128 Enabling slub_debug allows one to determine the
4129 culprit if slab objects become corrupted. Enabling
4130 slub_debug can create guard zones around objects and
4131 may poison objects when not in use. Also tracks the
4132 last alloc / free. For more information see
4133 Documentation/vm/slub.rst.
4135 slub_memcg_sysfs= [MM, SLUB]
4136 Determines whether to enable sysfs directories for
4137 memory cgroup sub-caches. 1 to enable, 0 to disable.
4138 The default is determined by CONFIG_SLUB_MEMCG_SYSFS_ON.
4139 Enabling this can lead to a very high number of debug
4140 directories and files being created under
4143 slub_max_order= [MM, SLUB]
4144 Determines the maximum allowed order for slabs.
4145 A high setting may cause OOMs due to memory
4146 fragmentation. For more information see
4147 Documentation/vm/slub.rst.
4149 slub_min_objects= [MM, SLUB]
4150 The minimum number of objects per slab. SLUB will
4151 increase the slab order up to slub_max_order to
4152 generate a sufficiently large slab able to contain
4153 the number of objects indicated. The higher the number
4154 of objects the smaller the overhead of tracking slabs
4155 and the less frequently locks need to be acquired.
4156 For more information see Documentation/vm/slub.rst.
4158 slub_min_order= [MM, SLUB]
4159 Determines the minimum page order for slabs. Must be
4160 lower than slub_max_order.
4161 For more information see Documentation/vm/slub.rst.
4163 slub_nomerge [MM, SLUB]
4164 Same with slab_nomerge. This is supported for legacy.
4165 See slab_nomerge for more information.
4168 Format: <io1>[,<io2>[,...,<io8>]]
4170 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
4171 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
4172 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
4173 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
4174 smsc-ircc2.ircc_irq= [HW] IRQ line
4175 smsc-ircc2.ircc_dma= [HW] DMA channel
4176 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
4177 0: Toshiba Satellite 1800 (GP data pin select)
4178 1: Fast pin select (default)
4181 smt [KNL,S390] Set the maximum number of threads (logical
4182 CPUs) to use per physical CPU on systems capable of
4183 symmetric multithreading (SMT). Will be capped to the
4184 actual hardware limit.
4186 Default: -1 (no limit)
4189 [KNL] Should the soft-lockup detector generate panics.
4192 A nonzero value instructs the soft-lockup detector
4193 to panic the machine when a soft-lockup occurs. This
4194 is also controlled by CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC
4195 which is the respective build-time switch to that
4198 softlockup_all_cpu_backtrace=
4199 [KNL] Should the soft-lockup detector generate
4200 backtraces on all cpus.
4203 sonypi.*= [HW] Sony Programmable I/O Control Device driver
4204 See Documentation/laptops/sonypi.txt
4206 spectre_v2= [X86] Control mitigation of Spectre variant 2
4207 (indirect branch speculation) vulnerability.
4208 The default operation protects the kernel from
4211 on - unconditionally enable, implies
4213 off - unconditionally disable, implies
4215 auto - kernel detects whether your CPU model is
4218 Selecting 'on' will, and 'auto' may, choose a
4219 mitigation method at run time according to the
4220 CPU, the available microcode, the setting of the
4221 CONFIG_RETPOLINE configuration option, and the
4222 compiler with which the kernel was built.
4224 Selecting 'on' will also enable the mitigation
4225 against user space to user space task attacks.
4227 Selecting 'off' will disable both the kernel and
4228 the user space protections.
4230 Specific mitigations can also be selected manually:
4232 retpoline - replace indirect branches
4233 retpoline,generic - google's original retpoline
4234 retpoline,amd - AMD-specific minimal thunk
4236 Not specifying this option is equivalent to
4240 [X86] Control mitigation of Spectre variant 2
4241 (indirect branch speculation) vulnerability between
4244 on - Unconditionally enable mitigations. Is
4245 enforced by spectre_v2=on
4247 off - Unconditionally disable mitigations. Is
4248 enforced by spectre_v2=off
4250 prctl - Indirect branch speculation is enabled,
4251 but mitigation can be enabled via prctl
4252 per thread. The mitigation control state
4253 is inherited on fork.
4256 - Like "prctl" above, but only STIBP is
4257 controlled per thread. IBPB is issued
4258 always when switching between different user
4262 - Same as "prctl" above, but all seccomp
4263 threads will enable the mitigation unless
4264 they explicitly opt out.
4267 - Like "seccomp" above, but only STIBP is
4268 controlled per thread. IBPB is issued
4269 always when switching between different
4270 user space processes.
4272 auto - Kernel selects the mitigation depending on
4273 the available CPU features and vulnerability.
4276 If CONFIG_SECCOMP=y then "seccomp", otherwise "prctl"
4278 Not specifying this option is equivalent to
4279 spectre_v2_user=auto.
4281 spec_store_bypass_disable=
4282 [HW] Control Speculative Store Bypass (SSB) Disable mitigation
4283 (Speculative Store Bypass vulnerability)
4285 Certain CPUs are vulnerable to an exploit against a
4286 a common industry wide performance optimization known
4287 as "Speculative Store Bypass" in which recent stores
4288 to the same memory location may not be observed by
4289 later loads during speculative execution. The idea
4290 is that such stores are unlikely and that they can
4291 be detected prior to instruction retirement at the
4292 end of a particular speculation execution window.
4294 In vulnerable processors, the speculatively forwarded
4295 store can be used in a cache side channel attack, for
4296 example to read memory to which the attacker does not
4297 directly have access (e.g. inside sandboxed code).
4299 This parameter controls whether the Speculative Store
4300 Bypass optimization is used.
4302 On x86 the options are:
4304 on - Unconditionally disable Speculative Store Bypass
4305 off - Unconditionally enable Speculative Store Bypass
4306 auto - Kernel detects whether the CPU model contains an
4307 implementation of Speculative Store Bypass and
4308 picks the most appropriate mitigation. If the
4309 CPU is not vulnerable, "off" is selected. If the
4310 CPU is vulnerable the default mitigation is
4311 architecture and Kconfig dependent. See below.
4312 prctl - Control Speculative Store Bypass per thread
4313 via prctl. Speculative Store Bypass is enabled
4314 for a process by default. The state of the control
4315 is inherited on fork.
4316 seccomp - Same as "prctl" above, but all seccomp threads
4317 will disable SSB unless they explicitly opt out.
4319 Default mitigations:
4320 X86: If CONFIG_SECCOMP=y "seccomp", otherwise "prctl"
4322 On powerpc the options are:
4324 on,auto - On Power8 and Power9 insert a store-forwarding
4325 barrier on kernel entry and exit. On Power7
4326 perform a software flush on kernel entry and
4330 Not specifying this option is equivalent to
4331 spec_store_bypass_disable=auto.
4333 spia_io_base= [HW,MTD]
4338 srcutree.counter_wrap_check [KNL]
4339 Specifies how frequently to check for
4340 grace-period sequence counter wrap for the
4341 srcu_data structure's ->srcu_gp_seq_needed field.
4342 The greater the number of bits set in this kernel
4343 parameter, the less frequently counter wrap will
4344 be checked for. Note that the bottom two bits
4347 srcutree.exp_holdoff [KNL]
4348 Specifies how many nanoseconds must elapse
4349 since the end of the last SRCU grace period for
4350 a given srcu_struct until the next normal SRCU
4351 grace period will be considered for automatic
4352 expediting. Set to zero to disable automatic
4356 Speculative Store Bypass Disable control
4358 On CPUs that are vulnerable to the Speculative
4359 Store Bypass vulnerability and offer a
4360 firmware based mitigation, this parameter
4361 indicates how the mitigation should be used:
4363 force-on: Unconditionally enable mitigation for
4364 for both kernel and userspace
4365 force-off: Unconditionally disable mitigation for
4366 for both kernel and userspace
4367 kernel: Always enable mitigation in the
4368 kernel, and offer a prctl interface
4369 to allow userspace to register its
4370 interest in being mitigated too.
4372 stack_guard_gap= [MM]
4373 override the default stack gap protection. The value
4374 is in page units and it defines how many pages prior
4375 to (for stacks growing down) resp. after (for stacks
4376 growing up) the main stack are reserved for no other
4377 mapping. Default value is 256 pages.
4380 Enabled the stack tracer on boot up.
4382 stacktrace_filter=[function-list]
4383 [FTRACE] Limit the functions that the stack tracer
4384 will trace at boot up. function-list is a comma separated
4385 list of functions. This list can be changed at run
4386 time by the stack_trace_filter file in the debugfs
4387 tracing directory. Note, this enables stack tracing
4388 and the stacktrace above is not needed.
4392 Set the STI (builtin display/keyboard on the HP-PARISC
4393 machines) console (graphic card) which should be used
4394 as the initial boot-console.
4395 See also comment in drivers/video/console/sticore.c.
4398 See comment in drivers/video/console/sticore.c.
4401 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
4403 sunrpc.min_resvport=
4404 sunrpc.max_resvport=
4406 SunRPC servers often require that client requests
4407 originate from a privileged port (i.e. a port in the
4408 range 0 < portnr < 1024).
4409 An administrator who wishes to reserve some of these
4410 ports for other uses may adjust the range that the
4411 kernel's sunrpc client considers to be privileged
4412 using these two parameters to set the minimum and
4413 maximum port values.
4415 sunrpc.svc_rpc_per_connection_limit=
4417 Limit the number of requests that the server will
4418 process in parallel from a single connection.
4419 The default value is 0 (no limit).
4423 Control how the NFS server code allocates CPUs to
4424 service thread pools. Depending on how many NICs
4425 you have and where their interrupts are bound, this
4426 option will affect which CPUs will do NFS serving.
4427 Note: this parameter cannot be changed while the
4428 NFS server is running.
4430 auto the server chooses an appropriate mode
4431 automatically using heuristics
4432 global a single global pool contains all CPUs
4433 percpu one pool for each CPU
4434 pernode one pool for each NUMA node (equivalent
4435 to global on non-NUMA machines)
4437 sunrpc.tcp_slot_table_entries=
4438 sunrpc.udp_slot_table_entries=
4440 Sets the upper limit on the number of simultaneous
4441 RPC calls that can be sent from the client to a
4442 server. Increasing these values may allow you to
4443 improve throughput, but will also increase the
4444 amount of memory reserved for use by the client.
4446 suspend.pm_test_delay=
4448 Sets the number of seconds to remain in a suspend test
4449 mode before resuming the system (see
4450 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
4451 is set. Default value is 5.
4454 [KNL] Enable accounting of swap in memory resource
4455 controller if no parameter or 1 is given or disable
4456 it if 0 is given (See Documentation/cgroup-v1/memory.txt)
4458 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
4459 Format: { <int> | force | noforce }
4460 <int> -- Number of I/O TLB slabs
4461 force -- force using of bounce buffers even if they
4462 wouldn't be automatically used by the kernel
4463 noforce -- Never use bounce buffers (for debugging)
4467 sysfs.deprecated=0|1 [KNL]
4468 Enable/disable old style sysfs layout for old udev
4469 on older distributions. When this option is enabled
4470 very new udev will not work anymore. When this option
4471 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
4472 in older udev will not work anymore.
4473 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
4474 the kernel configuration.
4476 sysrq_always_enabled
4478 Ignore sysrq setting - this boot parameter will
4479 neutralize any effect of /proc/sys/kernel/sysrq.
4480 Useful for debugging.
4482 tcpmhash_entries= [KNL,NET]
4483 Set the number of tcp_metrics_hash slots.
4484 Default value is 8192 or 16384 depending on total
4485 ram pages. This is used to specify the TCP metrics
4486 cache size. See Documentation/networking/ip-sysctl.txt
4487 "tcp_no_metrics_save" section for more details.
4491 test_suspend= [SUSPEND][,N]
4492 Specify "mem" (for Suspend-to-RAM) or "standby" (for
4493 standby suspend) or "freeze" (for suspend type freeze)
4494 as the system sleep state during system startup with
4495 the optional capability to repeat N number of times.
4496 The system is woken from this state using a
4497 wakeup-capable RTC alarm.
4499 thash_entries= [KNL,NET]
4500 Set number of hash buckets for TCP connection
4502 thermal.act= [HW,ACPI]
4503 -1: disable all active trip points in all thermal zones
4504 <degrees C>: override all lowest active trip points
4506 thermal.crt= [HW,ACPI]
4507 -1: disable all critical trip points in all thermal zones
4508 <degrees C>: override all critical trip points
4510 thermal.nocrt= [HW,ACPI]
4511 Set to disable actions on ACPI thermal zone
4512 critical and hot trip points.
4514 thermal.off= [HW,ACPI]
4515 1: disable ACPI thermal control
4517 thermal.psv= [HW,ACPI]
4518 -1: disable all passive trip points
4519 <degrees C>: override all passive trip points to this
4522 thermal.tzp= [HW,ACPI]
4523 Specify global default ACPI thermal zone polling rate
4524 <deci-seconds>: poll all this frequency
4525 0: no polling (default)
4528 Force threading of all interrupt handlers except those
4529 marked explicitly IRQF_NO_THREAD.
4532 Enable the Transcendent memory driver if built-in.
4534 tmem.cleancache=0|1 [KNL, XEN]
4535 Default is on (1). Disable the usage of the cleancache
4536 API to send anonymous pages to the hypervisor.
4538 tmem.frontswap=0|1 [KNL, XEN]
4539 Default is on (1). Disable the usage of the frontswap
4540 API to send swap pages to the hypervisor. If disabled
4541 the selfballooning and selfshrinking are force disabled.
4543 tmem.selfballooning=0|1 [KNL, XEN]
4544 Default is on (1). Disable the driving of swap pages
4547 tmem.selfshrinking=0|1 [KNL, XEN]
4548 Default is on (1). Partial swapoff that immediately
4549 transfers pages from Xen hypervisor back to the
4550 kernel based on different criteria.
4554 Specify if the kernel should make use of the cpu
4555 topology information if the hardware supports this.
4556 The scheduler will make use of this information and
4557 e.g. base its process migration decisions on it.
4560 topology_updates= [KNL, PPC, NUMA]
4562 Specify if the kernel should ignore (off)
4563 topology updates sent by the hypervisor to this
4568 tpm_suspend_pcr=[HW,TPM]
4569 Format: integer pcr id
4570 Specify that at suspend time, the tpm driver
4571 should extend the specified pcr with zeros,
4572 as a workaround for some chips which fail to
4573 flush the last written pcr on TPM_SaveState.
4574 This will guarantee that all the other pcrs
4577 trace_buf_size=nn[KMG]
4578 [FTRACE] will set tracing buffer size on each cpu.
4580 trace_event=[event-list]
4581 [FTRACE] Set and start specified trace events in order
4582 to facilitate early boot debugging. The event-list is a
4583 comma separated list of trace events to enable. See
4584 also Documentation/trace/events.rst
4586 trace_options=[option-list]
4587 [FTRACE] Enable or disable tracer options at boot.
4588 The option-list is a comma delimited list of options
4589 that can be enabled or disabled just as if you were
4590 to echo the option name into
4592 /sys/kernel/debug/tracing/trace_options
4594 For example, to enable stacktrace option (to dump the
4595 stack trace of each event), add to the command line:
4597 trace_options=stacktrace
4599 See also Documentation/trace/ftrace.rst "trace options"
4603 Have the tracepoints sent to printk as well as the
4604 tracing ring buffer. This is useful for early boot up
4605 where the system hangs or reboots and does not give the
4606 option for reading the tracing buffer or performing a
4607 ftrace_dump_on_oops.
4609 To turn off having tracepoints sent to printk,
4610 echo 0 > /proc/sys/kernel/tracepoint_printk
4611 Note, echoing 1 into this file without the
4612 tracepoint_printk kernel cmdline option has no effect.
4616 Having tracepoints sent to printk() and activating high
4617 frequency tracepoints such as irq or sched, can cause
4618 the system to live lock.
4621 [FTRACE] enable this option to disable tracing when a
4622 warning is hit. This turns off "tracing_on". Tracing can
4623 be enabled again by echoing '1' into the "tracing_on"
4624 file located in /sys/kernel/debug/tracing/
4626 This option is useful, as it disables the trace before
4627 the WARNING dump is called, which prevents the trace to
4628 be filled with content caused by the warning output.
4630 This option can also be set at run time via the sysctl
4631 option: kernel/traceoff_on_warning
4633 transparent_hugepage=
4635 Format: [always|madvise|never]
4636 Can be used to control the default behavior of the system
4637 with respect to transparent hugepages.
4638 See Documentation/admin-guide/mm/transhuge.rst
4641 tsc= Disable clocksource stability checks for TSC.
4643 [x86] reliable: mark tsc clocksource as reliable, this
4644 disables clocksource verification at runtime, as well
4645 as the stability checks done at bootup. Used to enable
4646 high-resolution timer mode on older hardware, and in
4647 virtualized environment.
4648 [x86] noirqtime: Do not use TSC to do irq accounting.
4649 Used to run time disable IRQ_TIME_ACCOUNTING on any
4650 platforms where RDTSC is slow and this accounting
4652 [x86] unstable: mark the TSC clocksource as unstable, this
4653 marks the TSC unconditionally unstable at bootup and
4654 avoids any further wobbles once the TSC watchdog notices.
4656 turbografx.map[2|3]= [HW,JOY]
4657 TurboGraFX parallel port interface
4659 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
4660 See also Documentation/input/devices/joystick-parport.rst
4662 udbg-immortal [PPC] When debugging early kernel crashes that
4663 happen after console_init() and before a proper
4664 console driver takes over, this boot options might
4665 help "seeing" what's going on.
4667 uhash_entries= [KNL,NET]
4668 Set number of hash buckets for UDP/UDP-Lite connections
4671 [USB] Ignore overcurrent events (default N).
4672 Some badly-designed motherboards generate lots of
4673 bogus events, for ports that aren't wired to
4674 anything. Set this parameter to avoid log spamming.
4675 Note that genuine overcurrent events won't be
4679 [X86] Cause panic on unknown NMI.
4681 usbcore.authorized_default=
4682 [USB] Default USB device authorization:
4683 (default -1 = authorized except for wireless USB,
4684 0 = not authorized, 1 = authorized)
4686 usbcore.autosuspend=
4687 [USB] The autosuspend time delay (in seconds) used
4688 for newly-detected USB devices (default 2). This
4689 is the time required before an idle device will be
4690 autosuspended. Devices for which the delay is set
4691 to a negative value won't be autosuspended at all.
4693 usbcore.usbfs_snoop=
4694 [USB] Set to log all usbfs traffic (default 0 = off).
4696 usbcore.usbfs_snoop_max=
4697 [USB] Maximum number of bytes to snoop in each URB
4700 usbcore.blinkenlights=
4701 [USB] Set to cycle leds on hubs (default 0 = off).
4703 usbcore.old_scheme_first=
4704 [USB] Start with the old device initialization
4705 scheme, applies only to low and full-speed devices
4708 usbcore.usbfs_memory_mb=
4709 [USB] Memory limit (in MB) for buffers allocated by
4710 usbfs (default = 16, 0 = max = 2047).
4712 usbcore.use_both_schemes=
4713 [USB] Try the other device initialization scheme
4714 if the first one fails (default 1 = enabled).
4716 usbcore.initial_descriptor_timeout=
4717 [USB] Specifies timeout for the initial 64-byte
4718 USB_REQ_GET_DESCRIPTOR request in milliseconds
4719 (default 5000 = 5.0 seconds).
4721 usbcore.nousb [USB] Disable the USB subsystem
4724 [USB] A list of quirk entries to augment the built-in
4725 usb core quirk list. List entries are separated by
4726 commas. Each entry has the form
4727 VendorID:ProductID:Flags. The IDs are 4-digit hex
4728 numbers and Flags is a set of letters. Each letter
4729 will change the built-in quirk; setting it if it is
4730 clear and clearing it if it is set. The letters have
4731 the following meanings:
4732 a = USB_QUIRK_STRING_FETCH_255 (string
4733 descriptors must not be fetched using
4735 b = USB_QUIRK_RESET_RESUME (device can't resume
4736 correctly so reset it instead);
4737 c = USB_QUIRK_NO_SET_INTF (device can't handle
4738 Set-Interface requests);
4739 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't
4740 handle its Configuration or Interface
4742 e = USB_QUIRK_RESET (device can't be reset
4743 (e.g morph devices), don't use reset);
4744 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has
4745 more interface descriptions than the
4746 bNumInterfaces count, and can't handle
4747 talking to these interfaces);
4748 g = USB_QUIRK_DELAY_INIT (device needs a pause
4749 during initialization, after we read
4750 the device descriptor);
4751 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For
4752 high speed and super speed interrupt
4753 endpoints, the USB 2.0 and USB 3.0 spec
4754 require the interval in microframes (1
4755 microframe = 125 microseconds) to be
4756 calculated as interval = 2 ^
4758 Devices with this quirk report their
4759 bInterval as the result of this
4760 calculation instead of the exponent
4761 variable used in the calculation);
4762 i = USB_QUIRK_DEVICE_QUALIFIER (device can't
4763 handle device_qualifier descriptor
4765 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device
4766 generates spurious wakeup, ignore
4767 remote wakeup capability);
4768 k = USB_QUIRK_NO_LPM (device can't handle Link
4770 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL
4771 (Device reports its bInterval as linear
4772 frames instead of the USB 2.0
4774 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs
4775 to be disconnected before suspend to
4776 prevent spurious wakeup);
4777 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a
4778 pause after every control message);
4779 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra
4780 delay after resetting its port);
4781 Example: quirks=0781:5580:bk,0a5c:5834:gij
4784 [USBHID] The interval which mice are to be polled at.
4787 [USBHID] The interval which joysticks are to be polled at.
4790 [USBHID] The interval which keyboards are to be polled at.
4792 usb-storage.delay_use=
4793 [UMS] The delay in seconds before a new device is
4794 scanned for Logical Units (default 1).
4797 [UMS] A list of quirks entries to supplement or
4798 override the built-in unusual_devs list. List
4799 entries are separated by commas. Each entry has
4800 the form VID:PID:Flags where VID and PID are Vendor
4801 and Product ID values (4-digit hex numbers) and
4802 Flags is a set of characters, each corresponding
4803 to a common usb-storage quirk flag as follows:
4804 a = SANE_SENSE (collect more than 18 bytes
4806 b = BAD_SENSE (don't collect more than 18
4807 bytes of sense data);
4808 c = FIX_CAPACITY (decrease the reported
4809 device capacity by one sector);
4810 d = NO_READ_DISC_INFO (don't use
4811 READ_DISC_INFO command);
4812 e = NO_READ_CAPACITY_16 (don't use
4813 READ_CAPACITY_16 command);
4814 f = NO_REPORT_OPCODES (don't use report opcodes
4816 g = MAX_SECTORS_240 (don't transfer more than
4817 240 sectors at a time, uas only);
4818 h = CAPACITY_HEURISTICS (decrease the
4819 reported device capacity by one
4820 sector if the number is odd);
4821 i = IGNORE_DEVICE (don't bind to this
4823 j = NO_REPORT_LUNS (don't use report luns
4825 l = NOT_LOCKABLE (don't try to lock and
4826 unlock ejectable media);
4827 m = MAX_SECTORS_64 (don't transfer more
4828 than 64 sectors = 32 KB at a time);
4829 n = INITIAL_READ10 (force a retry of the
4830 initial READ(10) command);
4831 o = CAPACITY_OK (accept the capacity
4832 reported by the device);
4833 p = WRITE_CACHE (the device cache is ON
4835 r = IGNORE_RESIDUE (the device reports
4836 bogus residue values);
4837 s = SINGLE_LUN (the device has only one
4839 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
4840 commands, uas only);
4841 u = IGNORE_UAS (don't bind to the uas driver);
4842 w = NO_WP_DETECT (don't test whether the
4843 medium is write-protected).
4844 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE
4845 even if the device claims no cache)
4846 Example: quirks=0419:aaf5:rl,0421:0433:rc
4848 user_debug= [KNL,ARM]
4850 See arch/arm/Kconfig.debug help text.
4851 1 - undefined instruction events
4853 4 - invalid data aborts
4856 Example: user_debug=31
4859 [X86] Flags controlling user PTE allocations.
4861 nohigh = do not allocate PTE pages in
4862 HIGHMEM regardless of setting
4866 On X86_32, this is an alias for vdso32=. Otherwise:
4868 vdso=1: enable VDSO (the default)
4869 vdso=0: disable VDSO mapping
4871 vdso32= [X86] Control the 32-bit vDSO
4872 vdso32=1: enable 32-bit VDSO
4873 vdso32=0 or vdso32=2: disable 32-bit VDSO
4875 See the help text for CONFIG_COMPAT_VDSO for more
4876 details. If CONFIG_COMPAT_VDSO is set, the default is
4877 vdso32=0; otherwise, the default is vdso32=1.
4879 For compatibility with older kernels, vdso32=2 is an
4882 Try vdso32=0 if you encounter an error that says:
4883 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
4886 vector=percpu: enable percpu vector domain
4888 video= [FB] Frame buffer configuration
4889 See Documentation/fb/modedb.txt.
4891 video.brightness_switch_enabled= [0,1]
4892 If set to 1, on receiving an ACPI notify event
4893 generated by hotkey, video driver will adjust brightness
4894 level and then send out the event to user space through
4895 the allocated input device; If set to 0, video driver
4896 will only send out the event without touching backlight
4901 [VMMIO] Memory mapped virtio (platform) device.
4903 <size>@<baseaddr>:<irq>[:<id>]
4905 <size> := size (can use standard suffixes
4907 <baseaddr> := physical base address
4908 <irq> := interrupt number (as passed to
4910 <id> := (optional) platform device id
4912 virtio_mmio.device=1K@0x100b0000:48:7
4914 Can be used multiple times for multiple devices.
4916 vga= [BOOT,X86-32] Select a particular video mode
4917 See Documentation/x86/boot.txt and
4918 Documentation/svga.txt.
4919 Use vga=ask for menu.
4920 This is actually a boot loader parameter; the value is
4921 passed to the kernel using a special protocol.
4923 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y.
4924 May slow down system boot speed, especially when
4925 enabled on systems with a large amount of memory.
4926 All options are enabled by default, and this
4927 interface is meant to allow for selectively
4928 enabling or disabling specific virtual memory
4931 Available options are:
4932 P Enable page structure init time poisoning
4933 - Disable all of the above options
4935 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
4936 size of <nn>. This can be used to increase the
4937 minimum size (128MB on x86). It can also be used to
4938 decrease the size and leave more room for directly
4941 vmcp_cma=nn[MG] [KNL,S390]
4942 Sets the memory size reserved for contiguous memory
4943 allocations for the vmcp device driver.
4945 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
4948 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
4951 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
4955 Controls the behavior of vsyscalls (i.e. calls to
4956 fixed addresses of 0xffffffffff600x00 from legacy
4957 code). Most statically-linked binaries and older
4958 versions of glibc use these calls. Because these
4959 functions are at fixed addresses, they make nice
4960 targets for exploits that can control RIP.
4962 emulate [default] Vsyscalls turn into traps and are
4963 emulated reasonably safely.
4965 native Vsyscalls are native syscall instructions.
4966 This is a little bit faster than trapping
4967 and makes a few dynamic recompilers work
4968 better than they would in emulation mode.
4969 It also makes exploits much easier to write.
4971 none Vsyscalls don't work at all. This makes
4972 them quite hard to use for exploits but
4973 might break your system.
4975 vt.color= [VT] Default text color.
4976 Format: 0xYX, X = foreground, Y = background.
4977 Default: 0x07 = light gray on black.
4979 vt.cur_default= [VT] Default cursor shape.
4980 Format: 0xCCBBAA, where AA, BB, and CC are the same as
4981 the parameters of the <Esc>[?A;B;Cc escape sequence;
4982 see VGA-softcursor.txt. Default: 2 = underline.
4984 vt.default_blu= [VT]
4985 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
4986 Change the default blue palette of the console.
4987 This is a 16-member array composed of values
4990 vt.default_grn= [VT]
4991 Format: <green0>,<green1>,<green2>,...,<green15>
4992 Change the default green palette of the console.
4993 This is a 16-member array composed of values
4996 vt.default_red= [VT]
4997 Format: <red0>,<red1>,<red2>,...,<red15>
4998 Change the default red palette of the console.
4999 This is a 16-member array composed of values
5005 Set system-wide default UTF-8 mode for all tty's.
5006 Default is 1, i.e. UTF-8 mode is enabled for all
5007 newly opened terminals.
5009 vt.global_cursor_default=
5012 Set system-wide default for whether a cursor
5013 is shown on new VTs. Default is -1,
5014 i.e. cursors will be created by default unless
5015 overridden by individual drivers. 0 will hide
5016 cursors, 1 will display them.
5018 vt.italic= [VT] Default color for italic text; 0-15.
5021 vt.underline= [VT] Default color for underlined text; 0-15.
5024 watchdog timers [HW,WDT] For information on watchdog timers,
5025 see Documentation/watchdog/watchdog-parameters.txt
5026 or other driver-specific files in the
5027 Documentation/watchdog/ directory.
5029 workqueue.watchdog_thresh=
5030 If CONFIG_WQ_WATCHDOG is configured, workqueue can
5031 warn stall conditions and dump internal state to
5032 help debugging. 0 disables workqueue stall
5033 detection; otherwise, it's the stall threshold
5034 duration in seconds. The default value is 30 and
5035 it can be updated at runtime by writing to the
5036 corresponding sysfs file.
5038 workqueue.disable_numa
5039 By default, all work items queued to unbound
5040 workqueues are affine to the NUMA nodes they're
5041 issued on, which results in better behavior in
5042 general. If NUMA affinity needs to be disabled for
5043 whatever reason, this option can be used. Note
5044 that this also can be controlled per-workqueue for
5045 workqueues visible under /sys/bus/workqueue/.
5047 workqueue.power_efficient
5048 Per-cpu workqueues are generally preferred because
5049 they show better performance thanks to cache
5050 locality; unfortunately, per-cpu workqueues tend to
5051 be more power hungry than unbound workqueues.
5053 Enabling this makes the per-cpu workqueues which
5054 were observed to contribute significantly to power
5055 consumption unbound, leading to measurably lower
5056 power usage at the cost of small performance
5059 The default value of this parameter is determined by
5060 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
5062 workqueue.debug_force_rr_cpu
5063 Workqueue used to implicitly guarantee that work
5064 items queued without explicit CPU specified are put
5065 on the local CPU. This guarantee is no longer true
5066 and while local CPU is still preferred work items
5067 may be put on foreign CPUs. This debug option
5068 forces round-robin CPU selection to flush out
5069 usages which depend on the now broken guarantee.
5070 When enabled, memory and cache locality will be
5073 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
5074 default x2apic cluster mode on platforms
5077 x86_intel_mid_timer= [X86-32,APBT]
5078 Choose timer option for x86 Intel MID platform.
5079 Two valid options are apbt timer only and lapic timer
5080 plus one apbt timer for broadcast timer.
5081 x86_intel_mid_timer=apbt_only | lapic_and_apbt
5083 xen_512gb_limit [KNL,X86-64,XEN]
5084 Restricts the kernel running paravirtualized under Xen
5085 to use only up to 512 GB of RAM. The reason to do so is
5086 crash analysis tools and Xen tools for doing domain
5087 save/restore/migration must be enabled to handle larger
5090 xen_emul_unplug= [HW,X86,XEN]
5091 Unplug Xen emulated devices
5092 Format: [unplug0,][unplug1]
5093 ide-disks -- unplug primary master IDE devices
5094 aux-ide-disks -- unplug non-primary-master IDE devices
5095 nics -- unplug network devices
5096 all -- unplug all emulated devices (NICs and IDE disks)
5097 unnecessary -- unplugging emulated devices is
5098 unnecessary even if the host did not respond to
5100 never -- do not unplug even if version check succeeds
5102 xen_nopvspin [X86,XEN]
5103 Disables the ticketlock slowpath using Xen PV
5107 Disables the PV optimizations forcing the HVM guest to
5108 run as generic HVM guest with no PV drivers.
5110 xen_scrub_pages= [XEN]
5111 Boolean option to control scrubbing pages before giving them back
5112 to Xen, for use by other domains. Can be also changed at runtime
5113 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages.
5114 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT.
5116 xirc2ps_cs= [NET,PCMCIA]
5118 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
5120 xhci-hcd.quirks [USB,KNL]
5121 A hex value specifying bitmask with supplemental xhci
5122 host controller quirks. Meaning of each bit can be
5123 consulted in header drivers/usb/host/xhci.h.