4 The following is a consolidated list of the kernel parameters as
5 implemented by the __setup(), core_param() and module_param() macros
6 and sorted into English Dictionary order (defined as ignoring all
7 punctuation and sorting digits before letters in a case insensitive
8 manner), and with descriptions where known.
10 The kernel parses parameters from the kernel command line up to "--";
11 if it doesn't recognize a parameter and it doesn't contain a '.', the
12 parameter gets passed to init: parameters with '=' go into init's
13 environment, others are passed as command line arguments to init.
14 Everything after "--" is passed as an argument to init.
16 Module parameters can be specified in two ways: via the kernel command
17 line with a module name prefix, or via modprobe, e.g.:
19 (kernel command line) usbcore.blinkenlights=1
20 (modprobe command line) modprobe usbcore blinkenlights=1
22 Parameters for modules which are built into the kernel need to be
23 specified on the kernel command line. modprobe looks through the
24 kernel command line (/proc/cmdline) and collects module parameters
25 when it loads a module, so the kernel command line can be used for
28 Hyphens (dashes) and underscores are equivalent in parameter names, so
29 log_buf_len=1M print-fatal-signals=1
30 can also be entered as
31 log-buf-len=1M print_fatal_signals=1
33 Double-quotes can be used to protect spaces in values, e.g.:
34 param="spaces in here"
36 This document may not be entirely up to date and comprehensive. The command
37 "modinfo -p ${modulename}" shows a current list of all parameters of a loadable
38 module. Loadable modules, after being loaded into the running kernel, also
39 reveal their parameters in /sys/module/${modulename}/parameters/. Some of these
40 parameters may be changed at runtime by the command
41 "echo -n ${value} > /sys/module/${modulename}/parameters/${parm}".
43 The parameters listed below are only valid if certain kernel build options were
44 enabled and if respective hardware is present. The text in square brackets at
45 the beginning of each description states the restrictions within which a
46 parameter is applicable:
48 ACPI ACPI support is enabled.
49 AGP AGP (Accelerated Graphics Port) is enabled.
50 ALSA ALSA sound support is enabled.
51 APIC APIC support is enabled.
52 APM Advanced Power Management support is enabled.
53 ARM ARM architecture is enabled.
54 AVR32 AVR32 architecture is enabled.
55 AX25 Appropriate AX.25 support is enabled.
56 BLACKFIN Blackfin architecture is enabled.
57 CLK Common clock infrastructure is enabled.
58 CMA Contiguous Memory Area support is enabled.
59 DRM Direct Rendering Management support is enabled.
60 DYNAMIC_DEBUG Build in debug messages and enable them at runtime
61 EDD BIOS Enhanced Disk Drive Services (EDD) is enabled
62 EFI EFI Partitioning (GPT) is enabled
63 EIDE EIDE/ATAPI support is enabled.
64 EVM Extended Verification Module
65 FB The frame buffer device is enabled.
66 FTRACE Function tracing enabled.
67 GCOV GCOV profiling is enabled.
68 HW Appropriate hardware is enabled.
69 IA-64 IA-64 architecture is enabled.
70 IMA Integrity measurement architecture is enabled.
71 IOSCHED More than one I/O scheduler is enabled.
72 IP_PNP IP DHCP, BOOTP, or RARP is enabled.
73 IPV6 IPv6 support is enabled.
74 ISAPNP ISA PnP code is enabled.
75 ISDN Appropriate ISDN support is enabled.
76 JOY Appropriate joystick support is enabled.
77 KGDB Kernel debugger support is enabled.
78 KVM Kernel Virtual Machine support is enabled.
79 LIBATA Libata driver is enabled
80 LP Printer support is enabled.
81 LOOP Loopback device support is enabled.
82 M68k M68k architecture is enabled.
83 These options have more detailed description inside of
84 Documentation/m68k/kernel-options.txt.
85 MDA MDA console support is enabled.
86 MIPS MIPS architecture is enabled.
87 MOUSE Appropriate mouse support is enabled.
88 MSI Message Signaled Interrupts (PCI).
89 MTD MTD (Memory Technology Device) support is enabled.
90 NET Appropriate network support is enabled.
91 NUMA NUMA support is enabled.
92 NFS Appropriate NFS support is enabled.
93 OSS OSS sound support is enabled.
94 PV_OPS A paravirtualized kernel is enabled.
95 PARIDE The ParIDE (parallel port IDE) subsystem is enabled.
96 PARISC The PA-RISC architecture is enabled.
97 PCI PCI bus support is enabled.
98 PCIE PCI Express support is enabled.
99 PCMCIA The PCMCIA subsystem is enabled.
100 PNP Plug & Play support is enabled.
101 PPC PowerPC architecture is enabled.
102 PPT Parallel port support is enabled.
103 PS2 Appropriate PS/2 support is enabled.
104 RAM RAM disk support is enabled.
105 S390 S390 architecture is enabled.
106 SCSI Appropriate SCSI support is enabled.
107 A lot of drivers have their options described inside
108 the Documentation/scsi/ sub-directory.
109 SECURITY Different security models are enabled.
110 SELINUX SELinux support is enabled.
111 APPARMOR AppArmor support is enabled.
112 SERIAL Serial support is enabled.
113 SH SuperH architecture is enabled.
114 SMP The kernel is an SMP kernel.
115 SPARC Sparc architecture is enabled.
116 SWSUSP Software suspend (hibernation) is enabled.
117 SUSPEND System suspend states are enabled.
118 TPM TPM drivers are enabled.
119 TS Appropriate touchscreen support is enabled.
120 UMS USB Mass Storage support is enabled.
121 USB USB support is enabled.
122 USBHID USB Human Interface Device support is enabled.
123 V4L Video For Linux support is enabled.
124 VMMIO Driver for memory mapped virtio devices is enabled.
125 VGA The VGA console has been enabled.
126 VT Virtual terminal support is enabled.
127 WDT Watchdog support is enabled.
128 XT IBM PC/XT MFM hard disk support is enabled.
129 X86-32 X86-32, aka i386 architecture is enabled.
130 X86-64 X86-64 architecture is enabled.
131 More X86-64 boot options can be found in
132 Documentation/x86/x86_64/boot-options.txt .
133 X86 Either 32-bit or 64-bit x86 (same as X86-32+X86-64)
134 X86_UV SGI UV support is enabled.
135 XEN Xen support is enabled
137 In addition, the following text indicates that the option:
139 BUGS= Relates to possible processor bugs on the said processor.
140 KNL Is a kernel start-up parameter.
141 BOOT Is a boot loader parameter.
143 Parameters denoted with BOOT are actually interpreted by the boot
144 loader, and have no meaning to the kernel directly.
145 Do not modify the syntax of boot loader parameters without extreme
146 need or coordination with <Documentation/x86/boot.txt>.
148 There are also arch-specific kernel-parameters not documented here.
149 See for example <Documentation/x86/x86_64/boot-options.txt>.
151 Note that ALL kernel parameters listed below are CASE SENSITIVE, and that
152 a trailing = on the name of any parameter states that that parameter will
153 be entered as an environment variable, whereas its absence indicates that
154 it will appear as a kernel argument readable via /proc/cmdline by programs
155 running once the system is up.
157 The number of kernel parameters is not limited, but the length of the
158 complete command line (parameters including spaces etc.) is limited to
159 a fixed number of characters. This limit depends on the architecture
160 and is between 256 and 4096 characters. It is defined in the file
161 ./include/asm/setup.h as COMMAND_LINE_SIZE.
163 Finally, the [KMG] suffix is commonly described after a number of kernel
164 parameter values. These 'K', 'M', and 'G' letters represent the _binary_
165 multipliers 'Kilo', 'Mega', and 'Giga', equalling 2^10, 2^20, and 2^30
166 bytes respectively. Such letter suffixes can also be entirely omitted.
169 acpi= [HW,ACPI,X86,ARM64]
170 Advanced Configuration and Power Interface
171 Format: { force | on | off | strict | noirq | rsdt |
173 force -- enable ACPI if default was off
174 on -- enable ACPI but allow fallback to DT [arm64]
175 off -- disable ACPI if default was on
176 noirq -- do not use ACPI for IRQ routing
177 strict -- Be less tolerant of platforms that are not
178 strictly ACPI specification compliant.
179 rsdt -- prefer RSDT over (default) XSDT
180 copy_dsdt -- copy DSDT to memory
181 For ARM64, ONLY "acpi=off", "acpi=on" or "acpi=force"
184 See also Documentation/power/runtime_pm.txt, pci=noacpi
186 acpi_apic_instance= [ACPI, IOAPIC]
188 2: use 2nd APIC table, if available
189 1,0: use 1st APIC table
192 acpi_backlight= [HW,ACPI]
193 acpi_backlight=vendor
195 If set to vendor, prefer vendor specific driver
196 (e.g. thinkpad_acpi, sony_acpi, etc.) instead
197 of the ACPI video.ko driver.
199 acpi_force_32bit_fadt_addr
200 force FADT to use 32 bit addresses rather than the
201 64 bit X_* addresses. Some firmware have broken 64
202 bit addresses for force ACPI ignore these and use
203 the older legacy 32 bit addresses.
205 acpica_no_return_repair [HW, ACPI]
206 Disable AML predefined validation mechanism
207 This mechanism can repair the evaluation result to make
208 the return objects more ACPI specification compliant.
209 This option is useful for developers to identify the
210 root cause of an AML interpreter issue when the issue
211 has something to do with the repair mechanism.
213 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG]
214 acpi.debug_level= [HW,ACPI,ACPI_DEBUG]
216 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
217 debug output. Bits in debug_layer correspond to a
218 _COMPONENT in an ACPI source file, e.g.,
219 #define _COMPONENT ACPI_PCI_COMPONENT
220 Bits in debug_level correspond to a level in
221 ACPI_DEBUG_PRINT statements, e.g.,
222 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
223 The debug_level mask defaults to "info". See
224 Documentation/acpi/debug.txt for more information about
225 debug layers and levels.
227 Enable processor driver info messages:
228 acpi.debug_layer=0x20000000
229 Enable PCI/PCI interrupt routing info messages:
230 acpi.debug_layer=0x400000
231 Enable AML "Debug" output, i.e., stores to the Debug
232 object while interpreting AML:
233 acpi.debug_layer=0xffffffff acpi.debug_level=0x2
234 Enable all messages related to ACPI hardware:
235 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
237 Some values produce so much output that the system is
238 unusable. The "log_buf_len" parameter may be useful
239 if you need to capture more output.
241 acpi_enforce_resources= [ACPI]
242 { strict | lax | no }
243 Check for resource conflicts between native drivers
244 and ACPI OperationRegions (SystemIO and SystemMemory
245 only). IO ports and memory declared in ACPI might be
246 used by the ACPI subsystem in arbitrary AML code and
247 can interfere with legacy drivers.
248 strict (default): access to resources claimed by ACPI
249 is denied; legacy drivers trying to access reserved
250 resources will fail to bind to device using them.
251 lax: access to resources claimed by ACPI is allowed;
252 legacy drivers trying to access reserved resources
253 will bind successfully but a warning message is logged.
254 no: ACPI OperationRegions are not marked as reserved,
255 no further checks are performed.
257 acpi_force_table_verification [HW,ACPI]
258 Enable table checksum verification during early stage.
259 By default, this is disabled due to x86 early mapping
262 acpi_irq_balance [HW,ACPI]
263 ACPI will balance active IRQs
266 acpi_irq_nobalance [HW,ACPI]
267 ACPI will not move active IRQs (default)
270 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA
271 Format: <irq>,<irq>...
273 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for
275 Format: <irq>,<irq>...
277 acpi_no_auto_serialize [HW,ACPI]
278 Disable auto-serialization of AML methods
279 AML control methods that contain the opcodes to create
280 named objects will be marked as "Serialized" by the
281 auto-serialization feature.
282 This feature is enabled by default.
283 This option allows to turn off the feature.
285 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump
288 acpi_no_static_ssdt [HW,ACPI]
289 Disable installation of static SSDTs at early boot time
290 By default, SSDTs contained in the RSDT/XSDT will be
291 installed automatically and they will appear under
292 /sys/firmware/acpi/tables.
293 This option turns off this feature.
294 Note that specifying this option does not affect
295 dynamic table installation which will install SSDT
296 tables to /sys/firmware/acpi/tables/dynamic.
298 acpi_rsdp= [ACPI,EFI,KEXEC]
299 Pass the RSDP address to the kernel, mostly used
300 on machines running EFI runtime service to boot the
301 second kernel for kdump.
303 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
304 Format: To spoof as Windows 98: ="Microsoft Windows"
306 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead
307 of 2 which is mandated by ACPI 6) as the supported ACPI
308 specification revision (when using this switch, it may
309 be necessary to carry out a cold reboot _twice_ in a
310 row to make it take effect on the platform firmware).
312 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
313 acpi_osi="string1" # add string1
314 acpi_osi="!string2" # remove string2
315 acpi_osi=!* # remove all strings
316 acpi_osi=! # disable all built-in OS vendor
318 acpi_osi= # disable all strings
320 'acpi_osi=!' can be used in combination with single or
321 multiple 'acpi_osi="string1"' to support specific OS
322 vendor string(s). Note that such command can only
323 affect the default state of the OS vendor strings, thus
324 it cannot affect the default state of the feature group
325 strings and the current state of the OS vendor strings,
326 specifying it multiple times through kernel command line
327 is meaningless. This command is useful when one do not
328 care about the state of the feature group strings which
329 should be controlled by the OSPM.
331 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
332 to 'acpi_osi="Windows 2000" acpi_osi=!', they all
333 can make '_OSI("Windows 2000")' TRUE.
335 'acpi_osi=' cannot be used in combination with other
336 'acpi_osi=' command lines, the _OSI method will not
337 exist in the ACPI namespace. NOTE that such command can
338 only affect the _OSI support state, thus specifying it
339 multiple times through kernel command line is also
342 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
345 'acpi_osi=!*' can be used in combination with single or
346 multiple 'acpi_osi="string1"' to support specific
347 string(s). Note that such command can affect the
348 current state of both the OS vendor strings and the
349 feature group strings, thus specifying it multiple times
350 through kernel command line is meaningful. But it may
351 still not able to affect the final state of a string if
352 there are quirks related to this string. This command
353 is useful when one want to control the state of the
354 feature group strings to debug BIOS issues related to
357 1. 'acpi_osi="Module Device" acpi_osi=!*' can make
358 '_OSI("Module Device")' FALSE.
359 2. 'acpi_osi=!* acpi_osi="Module Device"' can make
360 '_OSI("Module Device")' TRUE.
361 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
363 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
365 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
366 they all will make '_OSI("Windows 2000")' TRUE.
369 Override the pmtimer bug detection: force the kernel
370 to assume that this machine's pmtimer latches its value
371 and always returns good values.
373 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
374 Format: { level | edge | high | low }
376 acpi_skip_timer_override [HW,ACPI]
377 Recognize and ignore IRQ0/pin2 Interrupt Override.
378 For broken nForce2 BIOS resulting in XT-PIC timer.
380 acpi_sleep= [HW,ACPI] Sleep options
381 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
382 old_ordering, nonvs, sci_force_enable }
383 See Documentation/power/video.txt for information on
385 s3_beep is for debugging; it makes the PC's speaker beep
386 as soon as the kernel's real-mode entry point is called.
387 s4_nohwsig prevents ACPI hardware signature from being
388 used during resume from hibernation.
389 old_ordering causes the ACPI 1.0 ordering of the _PTS
390 control method, with respect to putting devices into
391 low power states, to be enforced (the ACPI 2.0 ordering
392 of _PTS is used by default).
393 nonvs prevents the kernel from saving/restoring the
394 ACPI NVS memory during suspend/hibernation and resume.
395 sci_force_enable causes the kernel to set SCI_EN directly
396 on resume from S1/S3 (which is against the ACPI spec,
397 but some broken systems don't work without it).
399 acpi_use_timer_override [HW,ACPI]
400 Use timer override. For some broken Nvidia NF5 boards
401 that require a timer override, but don't have HPET
403 add_efi_memmap [EFI; X86] Include EFI memory map in
404 kernel's map of available physical RAM.
407 { off | try_unsupported }
408 off: disable AGP support
409 try_unsupported: try to drive unsupported chipsets
410 (may crash computer or cause data corruption)
413 See Documentation/sound/alsa/alsa-parameters.txt
416 Allow the default userspace alignment fault handler
417 behaviour to be specified. Bit 0 enables warnings,
418 bit 1 enables fixups, and bit 2 sends a segfault.
420 align_va_addr= [X86-64]
421 Align virtual addresses by clearing slice [14:12] when
422 allocating a VMA at process creation time. This option
423 gives you up to 3% performance improvement on AMD F15h
424 machines (where it is enabled by default) for a
425 CPU-intensive style benchmark, and it can vary highly in
426 a microbenchmark depending on workload and compiler.
428 32: only for 32-bit processes
429 64: only for 64-bit processes
430 on: enable for both 32- and 64-bit processes
431 off: disable for both 32- and 64-bit processes
433 alloc_snapshot [FTRACE]
434 Allocate the ftrace snapshot buffer on boot up when the
435 main buffer is allocated. This is handy if debugging
436 and you need to use tracing_snapshot() on boot up, and
437 do not want to use tracing_snapshot_alloc() as it needs
438 to be done where GFP_KERNEL allocations are allowed.
440 amd_iommu= [HW,X86-64]
441 Pass parameters to the AMD IOMMU driver in the system.
443 fullflush - enable flushing of IO/TLB entries when
444 they are unmapped. Otherwise they are
445 flushed before they will be reused, which
447 off - do not initialize any AMD IOMMU found in
449 force_isolation - Force device isolation for all
450 devices. The IOMMU driver is not
451 allowed anymore to lift isolation
452 requirements as needed. This option
453 does not override iommu=pt
455 amd_iommu_dump= [HW,X86-64]
456 Enable AMD IOMMU driver option to dump the ACPI table
457 for AMD IOMMU. With this option enabled, AMD IOMMU
458 driver will print ACPI tables for AMD IOMMU during
459 IOMMU initialization.
461 amijoy.map= [HW,JOY] Amiga joystick support
462 Map of devices attached to JOY0DAT and JOY1DAT
464 See also Documentation/input/joystick.txt
466 analog.map= [HW,JOY] Analog joystick and gamepad support
467 Specifies type or capabilities of an analog joystick
468 connected to one of 16 gameports
469 Format: <type1>,<type2>,..<type16>
472 Power management functions (SPARCstation-4/5 + deriv.)
474 Disable APC CPU standby support. SPARCstation-Fox does
475 not play well with APC CPU idle - disable it if you have
476 APC and your system crashes randomly.
478 apic= [APIC,X86-32] Advanced Programmable Interrupt Controller
479 Change the output verbosity whilst booting
480 Format: { quiet (default) | verbose | debug }
481 Change the amount of debugging information output
482 when initialising the APIC and IO-APIC components.
484 apic_extnmi= [APIC,X86] External NMI delivery setting
485 Format: { bsp (default) | all | none }
486 bsp: External NMI is delivered only to CPU 0
487 all: External NMIs are broadcast to all CPUs as a
489 none: External NMI is masked for all CPUs. This is
490 useful so that a dump capture kernel won't be
494 See Documentation/networking/ipv6.txt.
496 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
497 Limit apic dumping. The parameter defines the maximal
498 number of local apics being dumped. Also it is possible
499 to set it to "all" by meaning -- no limit here.
500 Format: { 1 (default) | 2 | ... | all }.
501 The parameter valid if only apic=debug or
502 apic=verbose is specified.
503 Example: apic=debug show_lapic=all
505 apm= [APM] Advanced Power Management
506 See header of arch/x86/kernel/apm_32.c.
508 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
509 Format: <io>,<irq>,<nodeID>
513 atarimouse= [HW,MOUSE] Atari Mouse
515 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
516 EzKey and similar keyboards
518 atkbd.reset= [HW] Reset keyboard during initialization
520 atkbd.set= [HW] Select keyboard code set
521 Format: <int> (2 = AT (default), 3 = PS/2)
523 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
526 atkbd.softraw= [HW] Choose between synthetic and real raw mode
527 Format: <bool> (0 = real, 1 = synthetic (default))
529 atkbd.softrepeat= [HW]
530 Use software keyboard repeat
532 audit= [KNL] Enable the audit sub-system
533 Format: { "0" | "1" } (0 = disabled, 1 = enabled)
534 0 - kernel audit is disabled and can not be enabled
535 until the next reboot
536 unset - kernel audit is initialized but disabled and
537 will be fully enabled by the userspace auditd.
538 1 - kernel audit is initialized and partially enabled,
539 storing at most audit_backlog_limit messages in
540 RAM until it is fully enabled by the userspace
544 audit_backlog_limit= [KNL] Set the audit queue size limit.
545 Format: <int> (must be >=0)
548 bau= [X86_UV] Enable the BAU on SGI UV. The default
549 behavior is to disable the BAU (i.e. bau=0).
550 Format: { "0" | "1" }
553 unset - Disable the BAU.
555 baycom_epp= [HW,AX25]
558 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
560 See header of drivers/net/hamradio/baycom_par.c.
562 baycom_ser_fdx= [HW,AX25]
563 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
564 Format: <io>,<irq>,<mode>[,<baud>]
565 See header of drivers/net/hamradio/baycom_ser_fdx.c.
567 baycom_ser_hdx= [HW,AX25]
568 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
569 Format: <io>,<irq>,<mode>
570 See header of drivers/net/hamradio/baycom_ser_hdx.c.
572 blkdevparts= Manual partition parsing of block device(s) for
573 embedded devices based on command line input.
574 See Documentation/block/cmdline-partition.txt
576 boot_delay= Milliseconds to delay each printk during boot.
577 Values larger than 10 seconds (10000) are changed to
581 bootmem_debug [KNL] Enable bootmem allocator debug messages.
583 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
584 bttv.radio= Most important insmod options are available as
586 bttv.pll= See Documentation/video4linux/bttv/Insmod-options
589 bulk_remove=off [PPC] This parameter disables the use of the pSeries
590 firmware feature for flushing multiple hpte entries
593 c101= [NET] Moxa C101 synchronous serial card
595 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
596 Sometimes CPU hardware bugs make them report the cache
597 size incorrectly. The kernel will attempt work arounds
598 to fix known problems, but for some CPUs it is not
599 possible to determine what the correct size should be.
600 This option provides an override for these situations.
602 ca_keys= [KEYS] This parameter identifies a specific key(s) on
603 the system trusted keyring to be used for certificate
605 format: { id:<keyid> | builtin }
607 cca= [MIPS] Override the kernel pages' cache coherency
608 algorithm. Accepted values range from 0 to 7
609 inclusive. See arch/mips/include/asm/pgtable-bits.h
610 for platform specific values (SB1, Loongson3 and
613 ccw_timeout_log [S390]
614 See Documentation/s390/CommonIO for details.
616 cgroup_disable= [KNL] Disable a particular controller
617 Format: {name of the controller(s) to disable}
618 The effects of cgroup_disable=foo are:
619 - foo isn't auto-mounted if you mount all cgroups in
621 - foo isn't visible as an individually mountable
623 {Currently only "memory" controller deal with this and
624 cut the overhead, others just disable the usage. So
625 only cgroup_disable=memory is actually worthy}
627 cgroup_no_v1= [KNL] Disable one, multiple, all cgroup controllers in v1
628 Format: { controller[,controller...] | "all" }
629 Like cgroup_disable, but only applies to cgroup v1;
630 the blacklisted controllers remain available in cgroup2.
632 cgroup.memory= [KNL] Pass options to the cgroup memory controller.
634 nosocket -- Disable socket memory accounting.
635 nokmem -- Disable kernel memory accounting.
637 checkreqprot [SELINUX] Set initial checkreqprot flag value.
638 Format: { "0" | "1" }
639 See security/selinux/Kconfig help text.
640 0 -- check protection applied by kernel (includes
641 any implied execute protection).
642 1 -- check protection requested by application.
643 Default value is set via a kernel config option.
644 Value can be changed at runtime via
645 /selinux/checkreqprot.
648 See Documentation/s390/CommonIO for details.
651 Prevents the clock framework from automatically gating
652 clocks that have not been explicitly enabled by a Linux
653 device driver but are enabled in hardware at reset or
654 by the bootloader/firmware. Note that this does not
655 force such clocks to be always-on nor does it reserve
656 those clocks in any way. This parameter is useful for
657 debug and development, but should not be needed on a
658 platform with proper driver support. For more
659 information, see Documentation/clk.txt.
661 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
663 Forces specified clocksource (if available) to be used
664 when calculating gettimeofday(). If specified
665 clocksource is not available, it defaults to PIT.
666 Format: { pit | tsc | cyclone | pmtmr }
668 clocksource= Override the default clocksource
670 Override the default clocksource and use the clocksource
671 with the name specified.
672 Some clocksource names to choose from, depending on
674 [all] jiffies (this is the base, fallback clocksource)
676 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
677 pxa_timer,timer3,32k_counter,timer0_1
679 [X86-32] pit,hpet,tsc;
680 scx200_hrt on Geode; cyclone on IBM x440
688 clearcpuid=BITNUM [X86]
689 Disable CPUID feature X for the kernel. See
690 arch/x86/include/asm/cpufeatures.h for the valid bit
691 numbers. Note the Linux specific bits are not necessarily
692 stable over kernel options, but the vendor specific
694 Also note that user programs calling CPUID directly
695 or using the feature without checking anything
696 will still see it. This just prevents it from
697 being used by the kernel or shown in /proc/cpuinfo.
698 Also note the kernel might malfunction if you disable
701 cma=nn[MG]@[start[MG][-end[MG]]]
703 Sets the size of kernel global memory area for
704 contiguous memory allocations and optionally the
705 placement constraint by the physical address range of
706 memory allocations. A value of 0 disables CMA
707 altogether. For more information, see
708 include/linux/dma-contiguous.h
710 cmo_free_hint= [PPC] Format: { yes | no }
711 Specify whether pages are marked as being inactive
712 when they are freed. This is used in CMO environments
713 to determine OS memory pressure for page stealing by
717 coherent_pool=nn[KMG] [ARM,KNL]
718 Sets the size of memory pool for coherent, atomic dma
719 allocations, by default set to 256K.
721 code_bytes [X86] How many bytes of object code to print
726 com20020= [HW,NET] ARCnet - COM20020 chipset
728 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
730 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
734 ARCnet - COM90xx chipset (memory-mapped buffers)
735 Format: <io>[,<irq>[,<memstart>]]
737 condev= [HW,S390] console device
740 console= [KNL] Output console device and options.
742 tty<n> Use the virtual console device <n>.
746 Use the specified serial port. The options are of
747 the form "bbbbpnf", where "bbbb" is the baud rate,
748 "p" is parity ("n", "o", or "e"), "n" is number of
749 bits, and "f" is flow control ("r" for RTS or
750 omit it). Default is "9600n8".
752 See Documentation/serial-console.txt for more
754 Documentation/networking/netconsole.txt for an
757 uart[8250],io,<addr>[,options]
758 uart[8250],mmio,<addr>[,options]
759 uart[8250],mmio16,<addr>[,options]
760 uart[8250],mmio32,<addr>[,options]
761 uart[8250],0x<addr>[,options]
762 Start an early, polled-mode console on the 8250/16550
763 UART at the specified I/O port or MMIO address,
764 switching to the matching ttyS device later.
765 MMIO inter-register address stride is either 8-bit
766 (mmio), 16-bit (mmio16), or 32-bit (mmio32).
767 If none of [io|mmio|mmio16|mmio32], <addr> is assumed
768 to be equivalent to 'mmio'. 'options' are specified in
769 the same format described for ttyS above; if unspecified,
770 the h/w is not re-initialized.
772 hvc<n> Use the hypervisor console device <n>. This is for
773 both Xen and PowerPC hypervisors.
775 If the device connected to the port is not a TTY but a braille
776 device, prepend "brl," before the device type, for instance
778 For now, only VisioBraille is supported.
780 consoleblank= [KNL] The console blank (screen saver) timeout in
781 seconds. Defaults to 10*60 = 10mins. A value of 0
782 disables the blank timer.
785 [KNL] Change the default value for
786 /proc/<pid>/coredump_filter.
787 See also Documentation/filesystems/proc.txt.
789 cpuidle.off=1 [CPU_IDLE]
790 disable the cpuidle sub-system
793 [X86] Delay for N microsec between assert and de-assert
794 of APIC INIT to start processors. This delay occurs
795 on every CPU online, such as boot, and resume from suspend.
798 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
800 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
802 crashkernel=size[KMG][@offset[KMG]]
803 [KNL] Using kexec, Linux can switch to a 'crash kernel'
804 upon panic. This parameter reserves the physical
805 memory region [offset, offset + size] for that kernel
806 image. If '@offset' is omitted, then a suitable offset
807 is selected automatically. Check
808 Documentation/kdump/kdump.txt for further details.
810 crashkernel=range1:size1[,range2:size2,...][@offset]
811 [KNL] Same as above, but depends on the memory
812 in the running system. The syntax of range is
813 start-[end] where start and end are both
814 a memory unit (amount[KMG]). See also
815 Documentation/kdump/kdump.txt for an example.
817 crashkernel=size[KMG],high
818 [KNL, x86_64] range could be above 4G. Allow kernel
819 to allocate physical memory region from top, so could
820 be above 4G if system have more than 4G ram installed.
821 Otherwise memory region will be allocated below 4G, if
823 It will be ignored if crashkernel=X is specified.
824 crashkernel=size[KMG],low
825 [KNL, x86_64] range under 4G. When crashkernel=X,high
826 is passed, kernel could allocate physical memory region
827 above 4G, that cause second kernel crash on system
828 that require some amount of low memory, e.g. swiotlb
829 requires at least 64M+32K low memory, also enough extra
830 low memory is needed to make sure DMA buffers for 32-bit
831 devices won't run out. Kernel would try to allocate at
832 at least 256M below 4G automatically.
833 This one let user to specify own low range under 4G
834 for second kernel instead.
835 0: to disable low allocation.
836 It will be ignored when crashkernel=X,high is not used
837 or memory reserved is below 4G.
842 cs89x0_media= [HW,NET]
843 Format: { rj45 | aui | bnc }
846 See header of drivers/s390/block/dasd_devmap.c.
848 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
849 (one device per port)
850 Format: <port#>,<type>
851 See also Documentation/input/joystick-parport.txt
853 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
854 time. See Documentation/dynamic-debug-howto.txt for
855 details. Deprecated, see dyndbg.
857 debug [KNL] Enable kernel debugging (events log level).
860 [KNL] verbose self-tests
862 Print debugging info while doing the locking API
864 We default to 0 (no extra messages), setting it to
865 1 will print _a lot_ more information - normally
866 only useful to kernel developers.
868 debug_objects [KNL] Enable object debugging
871 [KNL] Disable object debugging
873 debug_guardpage_minorder=
874 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
875 parameter allows control of the order of pages that will
876 be intentionally kept free (and hence protected) by the
877 buddy allocator. Bigger value increase the probability
878 of catching random memory corruption, but reduce the
879 amount of memory for normal system use. The maximum
880 possible value is MAX_ORDER/2. Setting this parameter
881 to 1 or 2 should be enough to identify most random
882 memory corruption problems caused by bugs in kernel or
883 driver code when a CPU writes to (or reads from) a
884 random memory location. Note that there exists a class
885 of memory corruptions problems caused by buggy H/W or
886 F/W or by drivers badly programing DMA (basically when
887 memory is written at bus level and the CPU MMU is
888 bypassed) which are not detectable by
889 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
890 tracking down these problems.
893 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
894 parameter enables the feature at boot time. In
895 default, it is disabled. We can avoid allocating huge
896 chunk of memory for debug pagealloc if we don't enable
897 it at boot time and the system will work mostly same
898 with the kernel built without CONFIG_DEBUG_PAGEALLOC.
899 on: enable the feature
901 debugpat [X86] Enable PAT debugging
903 decnet.addr= [HW,NET]
904 Format: <area>[,<node>]
905 See also Documentation/networking/decnet.txt.
908 [same as hugepagesz=] The size of the default
909 HugeTLB page size. This is the size represented by
910 the legacy /proc/ hugepages APIs, used for SHM, and
911 default size when mounting hugetlbfs filesystems.
912 Defaults to the default architecture's huge page size
916 Set number of hash buckets for dentry cache.
919 See Documentation/networking/ipv6.txt.
921 disable_cpu_apicid= [X86,APIC,SMP]
923 The number of initial APIC ID for the
924 corresponding CPU to be disabled at boot,
925 mostly used for the kdump 2nd kernel to
926 disable BSP to wake up multiple CPUs without
927 causing system reset or hang due to sending
930 disable_ddw [PPC/PSERIES]
931 Disable Dynamic DMA Window support. Use this if
932 to workaround buggy firmware.
935 See Documentation/networking/ipv6.txt.
937 disable_mtrr_cleanup [X86]
938 The kernel tries to adjust MTRR layout from continuous
939 to discrete, to make X server driver able to add WB
940 entry later. This parameter disables that.
942 disable_mtrr_trim [X86, Intel and AMD only]
943 By default the kernel will trim any uncacheable
944 memory out of your available memory pool based on
945 MTRR settings. This parameter disables that behavior,
946 possibly causing your machine to run very slowly.
948 disable_timer_pin_1 [X86]
949 Disable PIN 1 of APIC timer
950 Can be useful to work around chipset bugs.
952 dis_ucode_ldr [X86] Disable the microcode loader.
954 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
955 this option disables the debugging code at boot.
957 dma_debug_entries=<number>
958 This option allows to tune the number of preallocated
959 entries for DMA-API debugging code. One entry is
960 required per DMA-API allocation. Use this if the
961 DMA-API debugging code disables itself because the
962 architectural default is too low.
964 dma_debug_driver=<driver_name>
965 With this option the DMA-API debugging driver
966 filter feature can be enabled at boot time. Just
967 pass the driver to filter for as the parameter.
968 The filter can be disabled or changed to another
969 driver later using sysfs.
971 drm_kms_helper.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
972 Broken monitors, graphic adapters, KVMs and EDIDless
973 panels may send no or incorrect EDID data sets.
974 This parameter allows to specify an EDID data sets
975 in the /lib/firmware directory that are used instead.
976 Generic built-in EDID data sets are used, if one of
977 edid/1024x768.bin, edid/1280x1024.bin,
978 edid/1680x1050.bin, or edid/1920x1080.bin is given
979 and no file with the same name exists. Details and
980 instructions how to build your own EDID data are
981 available in Documentation/EDID/HOWTO.txt. An EDID
982 data set will only be used for a particular connector,
983 if its name and a colon are prepended to the EDID
984 name. Each connector may use a unique EDID data
985 set by separating the files with a comma. An EDID
986 data set with no connector name will be used for
987 any connectors not explicitly specified.
991 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
992 module.dyndbg[="val"]
993 Enable debug messages at boot time. See
994 Documentation/dynamic-debug-howto.txt for details.
996 nompx [X86] Disables Intel Memory Protection Extensions.
997 See Documentation/x86/intel_mpx.txt for more
998 information about the feature.
1000 nopku [X86] Disable Memory Protection Keys CPU feature found
1004 on enable eager fpu restore
1005 off disable eager fpu restore
1006 auto selects the default scheme, which automatically
1007 enables eagerfpu restore for xsaveopt.
1009 module.async_probe [KNL]
1010 Enable asynchronous probe on this module.
1012 early_ioremap_debug [KNL]
1013 Enable debug messages in early_ioremap support. This
1014 is useful for tracking down temporary early mappings
1015 which are not unmapped.
1017 earlycon= [KNL] Output early console device and options.
1019 When used with no options, the early console is
1020 determined by the stdout-path property in device
1024 Start an early, polled-mode console on a cadence serial
1025 port at the specified address. The cadence serial port
1026 must already be setup and configured. Options are not
1029 uart[8250],io,<addr>[,options]
1030 uart[8250],mmio,<addr>[,options]
1031 uart[8250],mmio32,<addr>[,options]
1032 uart[8250],mmio32be,<addr>[,options]
1033 uart[8250],0x<addr>[,options]
1034 Start an early, polled-mode console on the 8250/16550
1035 UART at the specified I/O port or MMIO address.
1036 MMIO inter-register address stride is either 8-bit
1037 (mmio) or 32-bit (mmio32 or mmio32be).
1038 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
1039 to be equivalent to 'mmio'. 'options' are specified
1040 in the same format described for "console=ttyS<n>"; if
1041 unspecified, the h/w is not initialized.
1045 Start an early, polled-mode console on a pl011 serial
1046 port at the specified address. The pl011 serial port
1047 must already be setup and configured. Options are not
1048 yet supported. If 'mmio32' is specified, then only
1049 the driver will use only 32-bit accessors to read/write
1050 the device registers.
1053 Start an early, polled-mode console on an msm serial
1054 port at the specified address. The serial port
1055 must already be setup and configured. Options are not
1058 msm_serial_dm,<addr>
1059 Start an early, polled-mode console on an msm serial
1060 dm port at the specified address. The serial port
1061 must already be setup and configured. Options are not
1064 smh Use ARM semihosting calls for early console.
1072 Use early console provided by serial driver available
1073 on Samsung SoCs, requires selecting proper type and
1074 a correct base address of the selected UART port. The
1075 serial port must already be setup and configured.
1076 Options are not yet supported.
1080 Use early console provided by Freescale LP UART driver
1081 found on Freescale Vybrid and QorIQ LS1021A processors.
1082 A valid base address must be provided, and the serial
1083 port must already be setup and configured.
1085 armada3700_uart,<addr>
1086 Start an early, polled-mode console on the
1087 Armada 3700 serial port at the specified
1088 address. The serial port must already be setup
1089 and configured. Options are not yet supported.
1091 earlyprintk= [X86,SH,BLACKFIN,ARM,M68k]
1095 earlyprintk=serial[,ttySn[,baudrate]]
1096 earlyprintk=serial[,0x...[,baudrate]]
1097 earlyprintk=ttySn[,baudrate]
1098 earlyprintk=dbgp[debugController#]
1099 earlyprintk=pciserial,bus:device.function[,baudrate]
1101 earlyprintk is useful when the kernel crashes before
1102 the normal console is initialized. It is not enabled by
1103 default because it has some cosmetic problems.
1105 Append ",keep" to not disable it when the real console
1108 Only one of vga, efi, serial, or usb debug port can
1111 Currently only ttyS0 and ttyS1 may be specified by
1112 name. Other I/O ports may be explicitly specified
1113 on some architectures (x86 and arm at least) by
1114 replacing ttySn with an I/O port address, like this:
1115 earlyprintk=serial,0x1008,115200
1116 You can find the port for a given device in
1117 /proc/tty/driver/serial:
1118 2: uart:ST16650V2 port:00001008 irq:18 ...
1120 Interaction with the standard serial driver is not
1123 The VGA and EFI output is eventually overwritten by
1126 The xen output can only be used by Xen PV guests.
1128 edac_report= [HW,EDAC] Control how to report EDAC event
1129 Format: {"on" | "off" | "force"}
1130 on: enable EDAC to report H/W event. May be overridden
1131 by other higher priority error reporting module.
1132 off: disable H/W event reporting through EDAC.
1133 force: enforce the use of EDAC to report H/W event.
1136 ekgdboc= [X86,KGDB] Allow early kernel console debugging
1139 This is designed to be used in conjunction with
1140 the boot argument: earlyprintk=vga
1143 Format: {"off" | "on" | "skip[mbr]"}
1146 Format: { "old_map", "nochunk", "noruntime", "debug" }
1147 old_map [X86-64]: switch to the old ioremap-based EFI
1148 runtime services mapping. 32-bit still uses this one by
1150 nochunk: disable reading files in "chunks" in the EFI
1151 boot stub, as chunking can cause problems with some
1152 firmware implementations.
1153 noruntime : disable EFI runtime services support
1154 debug: enable misc debug output
1156 efi_no_storage_paranoia [EFI; X86]
1157 Using this parameter you can use more than 50% of
1158 your efi variable storage. Use this parameter only if
1159 you are really sure that your UEFI does sane gc and
1160 fulfills the spec otherwise your board may brick.
1162 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1163 Add arbitrary attribute to specific memory range by
1164 updating original EFI memory map.
1165 Region of memory which aa attribute is added to is
1167 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1168 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1169 attribute is added to range 0x100000000-0x180000000 and
1170 0x10a0000000-0x1120000000.
1172 Using this parameter you can do debugging of EFI memmap
1173 related feature. For example, you can do debugging of
1174 Address Range Mirroring feature even if your box
1177 eisa_irq_edge= [PARISC,HW]
1178 See header of drivers/parisc/eisa.c.
1181 See comment before function elanfreq_setup() in
1182 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1185 Format: {"cfq" | "deadline" | "noop"}
1186 See Documentation/block/cfq-iosched.txt and
1187 Documentation/block/deadline-iosched.txt for details.
1189 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1190 Specifies physical address of start of kernel core
1191 image elf header and optionally the size. Generally
1192 kexec loader will pass this option to capture kernel.
1193 See Documentation/kdump/kdump.txt for details.
1195 enable_mtrr_cleanup [X86]
1196 The kernel tries to adjust MTRR layout from continuous
1197 to discrete, to make X server driver able to add WB
1198 entry later. This parameter enables that.
1200 enable_timer_pin_1 [X86]
1201 Enable PIN 1 of APIC timer
1202 Can be useful to work around chipset bugs
1203 (in particular on some ATI chipsets).
1204 The kernel tries to set a reasonable default.
1206 enforcing [SELINUX] Set initial enforcing status.
1208 See security/selinux/Kconfig help text.
1209 0 -- permissive (log only, no denials).
1210 1 -- enforcing (deny and log).
1212 Value can be changed at runtime via /selinux/enforce.
1215 Disable Error Record Serialization Table (ERST)
1218 ether= [HW,NET] Ethernet cards parameters
1219 This option is obsoleted by the "netdev=" option, which
1220 has equivalent usage. See its documentation for details.
1224 Permit 'security.evm' to be updated regardless of
1225 current integrity status.
1229 fail_make_request=[KNL]
1230 General fault injection mechanism.
1231 Format: <interval>,<probability>,<space>,<times>
1232 See also Documentation/fault-injection/.
1235 See Documentation/blockdev/floppy.txt.
1237 force_pal_cache_flush
1238 [IA-64] Avoid check_sal_cache_flush which may hang on
1239 buggy SAL_CACHE_FLUSH implementations. Using this
1240 parameter will force ia64_sal_cache_flush to call
1241 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1244 Forcefully enable Physical Address Extension (PAE).
1245 Many Pentium M systems disable PAE but may have a
1246 functionally usable PAE implementation.
1247 Warning: use of this parameter will taint the kernel
1248 and may cause unknown problems.
1251 [FTRACE] will set and start the specified tracer
1252 as early as possible in order to facilitate early
1255 ftrace_dump_on_oops[=orig_cpu]
1256 [FTRACE] will dump the trace buffers on oops.
1257 If no parameter is passed, ftrace will dump
1258 buffers of all CPUs, but if you pass orig_cpu, it will
1259 dump only the buffer of the CPU that triggered the
1262 ftrace_filter=[function-list]
1263 [FTRACE] Limit the functions traced by the function
1264 tracer at boot up. function-list is a comma separated
1265 list of functions. This list can be changed at run
1266 time by the set_ftrace_filter file in the debugfs
1269 ftrace_notrace=[function-list]
1270 [FTRACE] Do not trace the functions specified in
1271 function-list. This list can be changed at run time
1272 by the set_ftrace_notrace file in the debugfs
1275 ftrace_graph_filter=[function-list]
1276 [FTRACE] Limit the top level callers functions traced
1277 by the function graph tracer at boot up.
1278 function-list is a comma separated list of functions
1279 that can be changed at run time by the
1280 set_graph_function file in the debugfs tracing directory.
1282 ftrace_graph_notrace=[function-list]
1283 [FTRACE] Do not trace from the functions specified in
1284 function-list. This list is a comma separated list of
1285 functions that can be changed at run time by the
1286 set_graph_notrace file in the debugfs tracing directory.
1289 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1290 support via parallel port (up to 5 devices per port)
1291 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1292 See also Documentation/input/joystick-parport.txt
1296 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART
1300 gcov_persist= [GCOV] When non-zero (default), profiling data for
1301 kernel modules is saved and remains accessible via
1302 debugfs, even when the module is unloaded/reloaded.
1303 When zero, profiling data is discarded and associated
1304 debugfs files are removed at module unload time.
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 hardlockup_all_cpu_backtrace=
1330 [KNL] Should the hard-lockup detector generate
1331 backtraces on all cpus.
1334 hashdist= [KNL,NUMA] Large hashes allocated during boot
1335 are distributed across NUMA nodes. Defaults on
1336 for 64-bit NUMA, off otherwise.
1337 Format: 0 | 1 (for off | on)
1339 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1341 hd= [EIDE] (E)IDE hard drive subsystem geometry
1342 Format: <cyl>,<head>,<sect>
1345 Disable Hardware Error Source Table (HEST) support;
1346 corresponding firmware-first mode error processing
1347 logic will be disabled.
1349 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1350 size of <nn>. This works even on boxes that have no
1351 highmem otherwise. This also works to reduce highmem
1352 size on bigger boxes.
1354 highres= [KNL] Enable/disable high resolution timer mode.
1355 Valid parameters: "on", "off"
1359 See Documentation/isdn/README.HiSax.
1363 hpet= [X86-32,HPET] option to control HPET usage
1364 Format: { enable (default) | disable | force |
1366 disable: disable HPET and use PIT instead
1367 force: allow force enabled of undocumented chips (ICH4,
1369 verbose: show contents of HPET registers during setup
1371 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1372 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1374 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
1375 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
1376 On x86-64 and powerpc, this option can be specified
1377 multiple times interleaved with hugepages= to reserve
1378 huge pages of different sizes. Valid pages sizes on
1379 x86-64 are 2M (when the CPU supports "pse") and 1G
1380 (when the CPU supports the "pdpe1gb" cpuinfo flag).
1382 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1383 terminal devices. Valid values: 0..8
1384 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1385 If specified, z/VM IUCV HVC accepts connections
1386 from listed z/VM user IDs only.
1388 hwthread_map= [METAG] Comma-separated list of Linux cpu id to
1389 hardware thread id mappings.
1390 Format: <cpu>:<hwthread>
1393 Do not unregister boot console at start. This is only
1394 useful for debugging when something happens in the window
1395 between unregistering the boot console and initializing
1398 i2c_bus= [HW] Override the default board specific I2C bus speed
1399 or register an additional I2C bus that is not
1400 registered from board initialization code.
1404 i8042.debug [HW] Toggle i8042 debug mode
1405 i8042.unmask_kbd_data
1406 [HW] Enable printing of interrupt data from the KBD port
1407 (disabled by default, and as a pre-condition
1408 requires that i8042.debug=1 be enabled)
1409 i8042.direct [HW] Put keyboard port into non-translated mode
1410 i8042.dumbkbd [HW] Pretend that controller can only read data from
1411 keyboard and cannot control its state
1412 (Don't attempt to blink the leds)
1413 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1414 i8042.nokbd [HW] Don't check/create keyboard port
1415 i8042.noloop [HW] Disable the AUX Loopback command while probing
1417 i8042.nomux [HW] Don't check presence of an active multiplexing
1419 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1421 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1422 i8042.reset [HW] Reset the controller during init and cleanup
1423 i8042.unlock [HW] Unlock (ignore) the keylock
1424 i8042.kbdreset [HW] Reset device connected to KBD port
1428 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
1429 indicates that the driver is running on unsupported
1431 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
1432 does not match list of supported models.
1434 [HW] Report power status in /proc/i8k
1435 (disabled by default)
1436 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
1439 i915.invert_brightness=
1440 [DRM] Invert the sense of the variable that is used to
1441 set the brightness of the panel backlight. Normally a
1442 brightness value of 0 indicates backlight switched off,
1443 and the maximum of the brightness value sets the backlight
1444 to maximum brightness. If this parameter is set to 0
1445 (default) and the machine requires it, or this parameter
1446 is set to 1, a brightness value of 0 sets the backlight
1447 to maximum brightness, and the maximum of the brightness
1448 value switches the backlight off.
1449 -1 -- never invert brightness
1450 0 -- machine default
1451 1 -- force brightness inversion
1454 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1456 ide-core.nodma= [HW] (E)IDE subsystem
1457 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1458 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1459 .cdrom .chs .ignore_cable are additional options
1460 See Documentation/ide/ide.txt.
1462 ide-generic.probe-mask= [HW] (E)IDE subsystem
1464 Probe mask for legacy ISA IDE ports. Depending on
1465 platform up to 6 ports are supported, enabled by
1466 setting corresponding bits in the mask to 1. The
1467 default value is 0x0, which has a special meaning.
1468 On systems that have PCI, it triggers scanning the
1469 PCI bus for the first and the second port, which
1470 are then probed. On systems without PCI the value
1471 of 0x0 enables probing the two first ports as if it
1474 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1475 Claim all unknown PCI IDE storage controllers.
1478 Format: idle=poll, idle=halt, idle=nomwait
1479 Poll forces a polling idle loop that can slightly
1480 improve the performance of waking up a idle CPU, but
1481 will use a lot of power and make the system run hot.
1483 idle=halt: Halt is forced to be used for CPU idle.
1484 In such case C2/C3 won't be used again.
1485 idle=nomwait: Disable mwait for CPU C-states
1487 ieee754= [MIPS] Select IEEE Std 754 conformance mode
1488 Format: { strict | legacy | 2008 | relaxed }
1491 Choose which programs will be accepted for execution
1492 based on the IEEE 754 NaN encoding(s) supported by
1493 the FPU and the NaN encoding requested with the value
1494 of an ELF file header flag individually set by each
1495 binary. Hardware implementations are permitted to
1496 support either or both of the legacy and the 2008 NaN
1499 Available settings are as follows:
1500 strict accept binaries that request a NaN encoding
1501 supported by the FPU
1502 legacy only accept legacy-NaN binaries, if supported
1504 2008 only accept 2008-NaN binaries, if supported
1506 relaxed accept any binaries regardless of whether
1507 supported by the FPU
1509 The FPU emulator is always able to support both NaN
1510 encodings, so if no FPU hardware is present or it has
1511 been disabled with 'nofpu', then the settings of
1512 'legacy' and '2008' strap the emulator accordingly,
1513 'relaxed' straps the emulator for both legacy-NaN and
1514 2008-NaN, whereas 'strict' enables legacy-NaN only on
1515 legacy processors and both NaN encodings on MIPS32 or
1518 The setting for ABS.fmt/NEG.fmt instruction execution
1519 mode generally follows that for the NaN encoding,
1520 except where unsupported by hardware.
1522 ignore_loglevel [KNL]
1523 Ignore loglevel setting - this will print /all/
1524 kernel messages to the console. Useful for debugging.
1525 We also add it as printk module parameter, so users
1526 could change it dynamically, usually by
1527 /sys/module/printk/parameters/ignore_loglevel.
1530 Ignore RLIMIT_DATA setting for data mappings,
1531 print warning at first misuse. Can be changed via
1532 /sys/module/kernel/parameters/ignore_rlimit_data.
1534 ihash_entries= [KNL]
1535 Set number of hash buckets for inode cache.
1537 ima_appraise= [IMA] appraise integrity measurements
1538 Format: { "off" | "enforce" | "fix" | "log" }
1541 ima_appraise_tcb [IMA]
1542 The builtin appraise policy appraises all files
1546 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1550 The list of supported hash algorithms is defined
1551 in crypto/hash_info.h.
1554 The builtin measurement policy to load during IMA
1555 setup. Specyfing "tcb" as the value, measures all
1556 programs exec'd, files mmap'd for exec, and all files
1557 opened with the read mode bit set by either the
1558 effective uid (euid=0) or uid=0.
1561 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
1562 Load a policy which meets the needs of the Trusted
1563 Computing Base. This means IMA will measure all
1564 programs exec'd, files mmap'd for exec, and all files
1565 opened for read by uid=0.
1568 Select one of defined IMA measurements template formats.
1569 Formats: { "ima" | "ima-ng" | "ima-sig" }
1573 [IMA] Define a custom template format.
1574 Format: { "field1|...|fieldN" }
1576 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1577 Format: <min_file_size>
1578 Set the minimal file size for using asynchronous hash.
1579 If left unspecified, ahash usage is disabled.
1581 ahash performance varies for different data sizes on
1582 different crypto accelerators. This option can be used
1583 to achieve the best performance for a particular HW.
1585 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1587 Set hashing buffer size. Default: 4k.
1589 ahash performance varies for different chunk sizes on
1590 different crypto accelerators. This option can be used
1591 to achieve best performance for particular HW.
1595 Run specified binary instead of /sbin/init as init
1598 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1599 for working out where the kernel is dying during
1602 initcall_blacklist= [KNL] Do not execute a comma-separated list of
1603 initcall functions. Useful for debugging built-in
1604 modules and initcalls.
1606 initrd= [BOOT] Specify the location of the initial ramdisk
1608 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1611 int_pln_enable [x86] Enable power limit notification interrupt
1613 integrity_audit=[IMA]
1614 Format: { "0" | "1" }
1615 0 -- basic integrity auditing messages. (Default)
1616 1 -- additional integrity auditing messages.
1618 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
1620 Enable intel iommu driver.
1622 Disable intel iommu driver.
1623 igfx_off [Default Off]
1624 By default, gfx is mapped as normal device. If a gfx
1625 device has a dedicated DMAR unit, the DMAR unit is
1626 bypassed by not enabling DMAR with this option. In
1627 this case, gfx device will use physical address for
1630 With this option iommu will not optimize to look
1631 for io virtual address below 32-bit forcing dual
1632 address cycle on pci bus for cards supporting greater
1633 than 32-bit addressing. The default is to look
1634 for translation below 32-bit and if not available
1635 then look in the higher range.
1636 strict [Default Off]
1637 With this option on every unmap_single operation will
1638 result in a hardware IOTLB flush operation as opposed
1639 to batching them for performance.
1640 sp_off [Default Off]
1641 By default, super page will be supported if Intel IOMMU
1642 has the capability. With this option, super page will
1644 ecs_off [Default Off]
1645 By default, extended context tables will be supported if
1646 the hardware advertises that it has support both for the
1647 extended tables themselves, and also PASID support. With
1648 this option set, extended tables will not be used even
1649 on hardware which claims to support them.
1651 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
1652 0 disables intel_idle and fall back on acpi_idle.
1653 1 to 6 specify maximum depth of C-state.
1657 Do not enable intel_pstate as the default
1658 scaling driver for the supported processors
1660 Enable intel_pstate on systems that prohibit it by default
1661 in favor of acpi-cpufreq. Forcing the intel_pstate driver
1662 instead of acpi-cpufreq may disable platform features, such
1663 as thermal controls and power capping, that rely on ACPI
1664 P-States information being indicated to OSPM and therefore
1665 should be used with caution. This option does not work with
1666 processors that aren't supported by the intel_pstate driver
1667 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
1669 Do not enable hardware P state control (HWP)
1672 Only load intel_pstate on systems which support
1673 hardware P state control (HWP) if available.
1675 Enforce ACPI _PPC performance limits. If the Fixed ACPI
1676 Description Table, specifies preferred power management
1677 profile as "Enterprise Server" or "Performance Server",
1678 then this feature is turned on by default.
1680 intremap= [X86-64, Intel-IOMMU]
1681 on enable Interrupt Remapping (default)
1682 off disable Interrupt Remapping
1683 nosid disable Source ID checking
1685 BIOS x2APIC opt-out request will be ignored
1686 nopost disable Interrupt Posting
1688 iomem= Disable strict checking of access to MMIO memory
1689 strict regions from userspace.
1704 nobypass [PPC/POWERNV]
1705 Disable IOMMU bypass, using IOMMU for PCI devices.
1708 io7= [HW] IO7 for Marvel based alpha systems
1709 See comment before marvel_specify_io7 in
1710 arch/alpha/kernel/core_marvel.c.
1712 io_delay= [X86] I/O delay method
1714 Standard port 0x80 based delay
1716 Alternate port 0xed based delay (needed on some systems)
1718 Simple two microseconds delay
1723 See Documentation/filesystems/nfs/nfsroot.txt.
1725 irqaffinity= [SMP] Set the default irq affinity mask
1727 <cpu number>,...,<cpu number>
1729 <cpu number>-<cpu number>
1730 (must be a positive range in ascending order)
1732 <cpu number>,...,<cpu number>-<cpu number>
1735 When an interrupt is not handled search all handlers
1736 for it. Intended to get systems with badly broken
1740 When an interrupt is not handled search all handlers
1741 for it. Also check all handlers each timer
1742 interrupt. Intended to get systems with badly broken
1746 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
1748 isolcpus= [KNL,SMP] Isolate CPUs from the general scheduler.
1750 <cpu number>,...,<cpu number>
1752 <cpu number>-<cpu number>
1753 (must be a positive range in ascending order)
1755 <cpu number>,...,<cpu number>-<cpu number>
1757 This option can be used to specify one or more CPUs
1758 to isolate from the general SMP balancing and scheduling
1759 algorithms. You can move a process onto or off an
1760 "isolated" CPU via the CPU affinity syscalls or cpuset.
1761 <cpu number> begins at 0 and the maximum value is
1762 "number of CPUs in system - 1".
1764 This option is the preferred way to isolate CPUs. The
1765 alternative -- manually setting the CPU mask of all
1766 tasks in the system -- can cause problems and
1767 suboptimal load balancer performance.
1771 ivrs_ioapic [HW,X86_64]
1772 Provide an override to the IOAPIC-ID<->DEVICE-ID
1773 mapping provided in the IVRS ACPI table. For
1774 example, to map IOAPIC-ID decimal 10 to
1775 PCI device 00:14.0 write the parameter as:
1776 ivrs_ioapic[10]=00:14.0
1778 ivrs_hpet [HW,X86_64]
1779 Provide an override to the HPET-ID<->DEVICE-ID
1780 mapping provided in the IVRS ACPI table. For
1781 example, to map HPET-ID decimal 0 to
1782 PCI device 00:14.0 write the parameter as:
1783 ivrs_hpet[0]=00:14.0
1785 js= [HW,JOY] Analog joystick
1786 See Documentation/input/joystick.txt.
1789 Enable/disable kernel and module base offset ASLR
1790 (Address Space Layout Randomization) if built into
1791 the kernel. When CONFIG_HIBERNATION is selected,
1792 kASLR is disabled by default. When kASLR is enabled,
1793 hibernation will be disabled.
1797 kernelcore= [KNL,X86,IA-64,PPC]
1798 Format: nn[KMGTPE] | "mirror"
1800 specifies the amount of memory usable by the kernel
1801 for non-movable allocations. The requested amount is
1802 spread evenly throughout all nodes in the system. The
1803 remaining memory in each node is used for Movable
1804 pages. In the event, a node is too small to have both
1805 kernelcore and Movable pages, kernelcore pages will
1806 take priority and other nodes will have a larger number
1807 of Movable pages. The Movable zone is used for the
1808 allocation of pages that may be reclaimed or moved
1809 by the page migration subsystem. This means that
1810 HugeTLB pages may not be allocated from this zone.
1811 Note that allocations like PTEs-from-HighMem still
1812 use the HighMem zone if it exists, and the Normal
1813 zone if it does not.
1815 Instead of specifying the amount of memory (nn[KMGTPE]),
1816 you can specify "mirror" option. In case "mirror"
1817 option is specified, mirrored (reliable) memory is used
1818 for non-movable allocations and remaining memory is used
1819 for Movable pages. nn[KMGTPE] and "mirror" are exclusive,
1820 so you can NOT specify nn[KMGTPE] and "mirror" at the same
1823 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
1824 Format: <Controller#>[,poll interval]
1825 The controller # is the number of the ehci usb debug
1826 port as it is probed via PCI. The poll interval is
1827 optional and is the number seconds in between
1828 each poll cycle to the debug port in case you need
1829 the functionality for interrupting the kernel with
1830 gdb or control-c on the dbgp connection. When
1831 not using this parameter you use sysrq-g to break into
1832 the kernel debugger.
1834 kgdboc= [KGDB,HW] kgdb over consoles.
1835 Requires a tty driver that supports console polling,
1836 or a supported polling keyboard driver (non-usb).
1837 Serial only format: <serial_device>[,baud]
1838 keyboard only format: kbd
1839 keyboard and serial format: kbd,<serial_device>[,baud]
1840 Optional Kernel mode setting:
1841 kms, kbd format: kms,kbd
1842 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
1844 kgdbwait [KGDB] Stop kernel execution and enter the
1845 kernel debugger at the earliest opportunity.
1847 kmac= [MIPS] korina ethernet MAC address.
1848 Configure the RouterBoard 532 series on-chip
1849 Ethernet adapter MAC address.
1851 kmemleak= [KNL] Boot-time kmemleak enable/disable
1852 Valid arguments: on, off
1854 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
1857 kmemcheck= [X86] Boot-time kmemcheck enable/disable/one-shot mode
1858 Valid arguments: 0, 1, 2
1859 kmemcheck=0 (disabled)
1860 kmemcheck=1 (enabled)
1861 kmemcheck=2 (one-shot mode)
1862 Default: 2 (one-shot mode)
1864 kstack=N [X86] Print N words from the kernel stack
1867 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
1868 Default is 0 (don't ignore, but inject #GP)
1870 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit
1874 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
1875 Default is 1 (enabled)
1877 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
1879 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
1881 kvm-intel.ept= [KVM,Intel] Disable extended page tables
1882 (virtualized MMU) support on capable Intel chips.
1883 Default is 1 (enabled)
1885 kvm-intel.emulate_invalid_guest_state=
1886 [KVM,Intel] Enable emulation of invalid guest states
1887 Default is 0 (disabled)
1889 kvm-intel.flexpriority=
1890 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
1891 Default is 1 (enabled)
1894 [KVM,Intel] Enable VMX nesting (nVMX).
1895 Default is 0 (disabled)
1897 kvm-intel.unrestricted_guest=
1898 [KVM,Intel] Disable unrestricted guest feature
1899 (virtualized real and unpaged mode) on capable
1900 Intel chips. Default is 1 (enabled)
1902 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
1903 feature (tagged TLBs) on capable Intel chips.
1904 Default is 1 (enabled)
1910 lapic [X86-32,APIC] Enable the local APIC even if BIOS
1913 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline
1914 value for LAPIC timer one-shot implementation. Default
1915 back to the programmable timer unit in the LAPIC.
1917 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
1920 libata.dma= [LIBATA] DMA control
1921 libata.dma=0 Disable all PATA and SATA DMA
1922 libata.dma=1 PATA and SATA Disk DMA only
1923 libata.dma=2 ATAPI (CDROM) DMA only
1924 libata.dma=4 Compact Flash DMA only
1925 Combinations also work, so libata.dma=3 enables DMA
1926 for disks and CDROMs, but not CFs.
1928 libata.ignore_hpa= [LIBATA] Ignore HPA limit
1929 libata.ignore_hpa=0 keep BIOS limits (default)
1930 libata.ignore_hpa=1 ignore limits, using full disk
1932 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
1936 libata.force= [LIBATA] Force configurations. The format is comma
1937 separated list of "[ID:]VAL" where ID is
1938 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
1939 matching port, link or device. Basically, it matches
1940 the ATA ID string printed on console by libata. If
1941 the whole ID part is omitted, the last PORT and DEVICE
1942 values are used. If ID hasn't been specified yet, the
1943 configuration applies to all ports, links and devices.
1945 If only DEVICE is omitted, the parameter applies to
1946 the port and all links and devices behind it. DEVICE
1947 number of 0 either selects the first device or the
1948 first fan-out link behind PMP device. It does not
1949 select the host link. DEVICE number of 15 selects the
1950 host link and device attached to it.
1952 The VAL specifies the configuration to force. As long
1953 as there's no ambiguity shortcut notation is allowed.
1954 For example, both 1.5 and 1.5G would work for 1.5Gbps.
1955 The following configurations can be forced.
1957 * Cable type: 40c, 80c, short40c, unk, ign or sata.
1958 Any ID with matching PORT is used.
1960 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
1962 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
1963 udma[/][16,25,33,44,66,100,133] notation is also
1966 * [no]ncq: Turn on or off NCQ.
1968 * [no]ncqtrim: Turn off queued DSM TRIM.
1970 * nohrst, nosrst, norst: suppress hard, soft
1973 * rstonce: only attempt one reset during
1974 hot-unplug link recovery
1976 * dump_id: dump IDENTIFY data.
1978 * atapi_dmadir: Enable ATAPI DMADIR bridge support
1980 * disable: Disable this device.
1982 If there are multiple matching configurations changing
1983 the same attribute, the last one is used.
1985 memblock=debug [KNL] Enable memblock debug messages.
1987 load_ramdisk= [RAM] List of ramdisks to load from floppy
1988 See Documentation/blockdev/ramdisk.txt.
1990 lockd.nlm_grace_period=P [NFS] Assign grace period.
1993 lockd.nlm_tcpport=N [NFS] Assign TCP port.
1996 lockd.nlm_timeout=T [NFS] Assign timeout value.
1999 lockd.nlm_udpport=M [NFS] Assign UDP port.
2002 locktorture.nreaders_stress= [KNL]
2003 Set the number of locking read-acquisition kthreads.
2004 Defaults to being automatically set based on the
2005 number of online CPUs.
2007 locktorture.nwriters_stress= [KNL]
2008 Set the number of locking write-acquisition kthreads.
2010 locktorture.onoff_holdoff= [KNL]
2011 Set time (s) after boot for CPU-hotplug testing.
2013 locktorture.onoff_interval= [KNL]
2014 Set time (s) between CPU-hotplug operations, or
2015 zero to disable CPU-hotplug testing.
2017 locktorture.shuffle_interval= [KNL]
2018 Set task-shuffle interval (jiffies). Shuffling
2019 tasks allows some CPUs to go into dyntick-idle
2020 mode during the locktorture test.
2022 locktorture.shutdown_secs= [KNL]
2023 Set time (s) after boot system shutdown. This
2024 is useful for hands-off automated testing.
2026 locktorture.stat_interval= [KNL]
2027 Time (s) between statistics printk()s.
2029 locktorture.stutter= [KNL]
2030 Time (s) to stutter testing, for example,
2031 specifying five seconds causes the test to run for
2032 five seconds, wait for five seconds, and so on.
2033 This tests the locking primitive's ability to
2034 transition abruptly to and from idle.
2036 locktorture.torture_runnable= [BOOT]
2037 Start locktorture running at boot time.
2039 locktorture.torture_type= [KNL]
2040 Specify the locking implementation to test.
2042 locktorture.verbose= [KNL]
2043 Enable additional printk() statements.
2045 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
2048 loglevel= All Kernel Messages with a loglevel smaller than the
2049 console loglevel will be printed to the console. It can
2050 also be changed with klogd or other programs. The
2051 loglevels are defined as follows:
2053 0 (KERN_EMERG) system is unusable
2054 1 (KERN_ALERT) action must be taken immediately
2055 2 (KERN_CRIT) critical conditions
2056 3 (KERN_ERR) error conditions
2057 4 (KERN_WARNING) warning conditions
2058 5 (KERN_NOTICE) normal but significant condition
2059 6 (KERN_INFO) informational
2060 7 (KERN_DEBUG) debug-level messages
2062 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
2063 in bytes. n must be a power of two and greater
2064 than the minimal size. The minimal size is defined
2065 by LOG_BUF_SHIFT kernel config parameter. There is
2066 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
2067 that allows to increase the default size depending on
2068 the number of CPUs. See init/Kconfig for more details.
2070 logo.nologo [FB] Disables display of the built-in Linux logo.
2071 This may be used to provide more screen space for
2072 kernel log messages and is useful when debugging
2073 kernel boot problems.
2075 lp=0 [LP] Specify parallel ports to use, e.g,
2076 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
2077 lp=reset first parallel port). 'lp=0' disables the
2078 lp=auto printer driver. 'lp=reset' (which can be
2079 specified in addition to the ports) causes
2080 attached printers to be reset. Using
2081 lp=port1,port2,... specifies the parallel ports
2082 to associate lp devices with, starting with
2083 lp0. A port specification may be 'none' to skip
2084 that lp device, or a parport name such as
2085 'parport0'. Specifying 'lp=auto' instead of a
2086 port specification list means that device IDs
2087 from each port should be examined, to see if
2088 an IEEE 1284-compliant printer is attached; if
2089 so, the driver will manage that printer.
2090 See also header of drivers/char/lp.c.
2093 Sets loops_per_jiffy to given constant, thus avoiding
2094 time-consuming boot-time autodetection (up to 250 ms per
2095 CPU). 0 enables autodetection (default). To determine
2096 the correct value for your kernel, boot with normal
2097 autodetection and see what value is printed. Note that
2098 on SMP systems the preset will be applied to all CPUs,
2099 which is likely to cause problems if your CPUs need
2100 significantly divergent settings. An incorrect value
2101 will cause delays in the kernel to be wrong, leading to
2102 unpredictable I/O errors and other breakage. Although
2103 unlikely, in the extreme case this might damage your
2107 Format: <io>,<irq>,<dma>
2109 machvec= [IA-64] Force the use of a particular machine-vector
2110 (machvec) in a generic kernel.
2111 Example: machvec=hpzx1_swiotlb
2113 machtype= [Loongson] Share the same kernel image file between different
2115 Example: machtype=lemote-yeeloong-2f-7inch
2117 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
2118 than or equal to this physical address is ignored.
2120 maxcpus= [SMP] Maximum number of processors that an SMP kernel
2121 should make use of. maxcpus=n : n >= 0 limits the
2122 kernel to using 'n' processors. n=0 is a special case,
2123 it is equivalent to "nosmp", which also disables
2126 max_loop= [LOOP] The number of loop block devices that get
2127 (loop.max_loop) unconditionally pre-created at init time. The default
2128 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
2129 of statically allocating a predefined number, loop
2130 devices can be requested on-demand with the
2131 /dev/loop-control interface.
2133 mce [X86-32] Machine Check Exception
2135 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.txt
2137 md= [HW] RAID subsystems devices and level
2138 See Documentation/md.txt.
2141 Format: <first>,<last>
2142 Specifies range of consoles to be captured by the MDA.
2144 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
2145 Amount of memory to be used when the kernel is not able
2146 to see the whole system memory or for test.
2147 [X86] Work as limiting max address. Use together
2148 with memmap= to avoid physical address space collisions.
2149 Without memmap= PCI devices could be placed at addresses
2150 belonging to unused RAM.
2152 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
2156 [KNL,SH] Allow user to override the default size for
2157 per-device physically contiguous DMA buffers.
2159 memmap=exactmap [KNL,X86] Enable setting of an exact
2160 E820 memory map, as specified by the user.
2161 Such memmap=exactmap lines can be constructed based on
2162 BIOS output or other requirements. See the memmap=nn@ss
2165 memmap=nn[KMG]@ss[KMG]
2166 [KNL] Force usage of a specific region of memory.
2167 Region of memory to be used is from ss to ss+nn.
2169 memmap=nn[KMG]#ss[KMG]
2170 [KNL,ACPI] Mark specific memory as ACPI data.
2171 Region of memory to be marked is from ss to ss+nn.
2173 memmap=nn[KMG]$ss[KMG]
2174 [KNL,ACPI] Mark specific memory as reserved.
2175 Region of memory to be reserved is from ss to ss+nn.
2176 Example: Exclude memory from 0x18690000-0x1869ffff
2177 memmap=64K$0x18690000
2179 memmap=0x10000$0x18690000
2181 memmap=nn[KMG]!ss[KMG]
2182 [KNL,X86] Mark specific memory as protected.
2183 Region of memory to be used, from ss to ss+nn.
2184 The memory region may be marked as e820 type 12 (0xc)
2185 and is NVDIMM or ADR memory.
2187 memory_corruption_check=0/1 [X86]
2188 Some BIOSes seem to corrupt the first 64k of
2189 memory when doing things like suspend/resume.
2190 Setting this option will scan the memory
2191 looking for corruption. Enabling this will
2192 both detect corruption and prevent the kernel
2193 from using the memory being corrupted.
2194 However, its intended as a diagnostic tool; if
2195 repeatable BIOS-originated corruption always
2196 affects the same memory, you can use memmap=
2197 to prevent the kernel from using that memory.
2199 memory_corruption_check_size=size [X86]
2200 By default it checks for corruption in the low
2201 64k, making this memory unavailable for normal
2202 use. Use this parameter to scan for
2203 corruption in more or less memory.
2205 memory_corruption_check_period=seconds [X86]
2206 By default it checks for corruption every 60
2207 seconds. Use this parameter to check at some
2208 other rate. 0 disables periodic checking.
2210 memtest= [KNL,X86,ARM] Enable memtest
2212 default : 0 <disable>
2213 Specifies the number of memtest passes to be
2214 performed. Each pass selects another test
2215 pattern from a given set of patterns. Memtest
2216 fills the memory with this pattern, validates
2217 memory contents and reserves bad memory
2218 regions that are detected.
2220 meye.*= [HW] Set MotionEye Camera parameters
2221 See Documentation/video4linux/meye.txt.
2223 mfgpt_irq= [IA-32] Specify the IRQ to use for the
2224 Multi-Function General Purpose Timers on AMD Geode
2227 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
2228 the BIOS has incorrectly applied a workaround. TinyBIOS
2229 version 0.98 is known to be affected, 0.99 fixes the
2230 problem by letting the user disable the workaround.
2234 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
2235 physical address is ignored.
2237 mini2440= [ARM,HW,KNL]
2238 Format:[0..2][b][c][t]
2240 MINI2440 configuration specification:
2241 0 - The attached screen is the 3.5" TFT
2242 1 - The attached screen is the 7" TFT
2243 2 - The VGA Shield is attached (1024x768)
2244 Leaving out the screen size parameter will not load
2245 the TFT driver, and the framebuffer will be left
2247 b - Enable backlight. The TFT backlight pin will be
2248 linked to the kernel VESA blanking code and a GPIO
2249 LED. This parameter is not necessary when using the
2251 c - Enable the s3c camera interface.
2252 t - Reserved for enabling touchscreen support. The
2253 touchscreen support is not enabled in the mainstream
2254 kernel as of 2.6.30, a preliminary port can be found
2255 in the "bleeding edge" mini2440 support kernel at
2256 http://repo.or.cz/w/linux-2.6/mini2440.git
2259 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
2260 parameter allows control of the logging verbosity for
2261 the additional memory initialisation checks. A value
2262 of 0 disables mminit logging and a level of 4 will
2263 log everything. Information is printed at KERN_DEBUG
2264 so loglevel=8 may also need to be specified.
2267 [KNL] When CONFIG_MODULE_SIG is set, this means that
2268 modules without (valid) signatures will fail to load.
2269 Note that if CONFIG_MODULE_SIG_FORCE is set, that
2270 is always true, so this option does nothing.
2273 [MOUSE] Maximum time between finger touching and
2274 leaving touchpad surface for touch to be considered
2275 a tap and be reported as a left button click (for
2276 touchpads working in absolute mode only).
2278 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
2279 reporting absolute coordinates, such as tablets
2280 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
2281 reporting absolute coordinates, such as tablets
2283 movablecore=nn[KMG] [KNL,X86,IA-64,PPC] This parameter
2284 is similar to kernelcore except it specifies the
2285 amount of memory used for migratable allocations.
2286 If both kernelcore and movablecore is specified,
2287 then kernelcore will be at *least* the specified
2288 value but may be more. If movablecore on its own
2289 is specified, the administrator must be careful
2290 that the amount of memory usable for all allocations
2293 movable_node [KNL,X86] Boot-time switch to enable the effects
2294 of CONFIG_MOVABLE_NODE=y. See mm/Kconfig for details.
2296 MTD_Partition= [MTD]
2297 Format: <name>,<region-number>,<size>,<offset>
2299 MTD_Region= [MTD] Format:
2300 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
2303 See drivers/mtd/cmdlinepart.c.
2305 multitce=off [PPC] This parameter disables the use of the pSeries
2306 firmware feature for updating multiple TCE entries
2309 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
2311 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
2313 boundary - index of last SLC block on Flex-OneNAND.
2314 The remaining blocks are configured as MLC blocks.
2315 lock - Configure if Flex-OneNAND boundary should be locked.
2316 Once locked, the boundary cannot be changed.
2317 1 indicates lock status, 0 indicates unlock status.
2320 ARM/S3C2412 JIVE boot control
2322 See arch/arm/mach-s3c2412/mach-jive.c
2324 mtouchusb.raw_coordinates=
2325 [HW] Make the MicroTouch USB driver use raw coordinates
2326 ('y', default) or cooked coordinates ('n')
2328 mtrr_chunk_size=nn[KMG] [X86]
2329 used for mtrr cleanup. It is largest continuous chunk
2330 that could hold holes aka. UC entries.
2332 mtrr_gran_size=nn[KMG] [X86]
2333 Used for mtrr cleanup. It is granularity of mtrr block.
2335 Large value could prevent small alignment from
2338 mtrr_spare_reg_nr=n [X86]
2340 Range: 0,7 : spare reg number
2342 Used for mtrr cleanup. It is spare mtrr entries number.
2343 Set to 2 or more if your graphical card needs more.
2345 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
2347 netdev= [NET] Network devices parameters
2348 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
2349 Note that mem_start is often overloaded to mean
2350 something different and driver-specific.
2351 This usage is only documented in each driver source
2355 [NETFILTER] Enable connection tracking flow accounting
2356 0 to disable accounting
2357 1 to enable accounting
2360 nfsaddrs= [NFS] Deprecated. Use ip= instead.
2361 See Documentation/filesystems/nfs/nfsroot.txt.
2363 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
2364 See Documentation/filesystems/nfs/nfsroot.txt.
2366 nfsrootdebug [NFS] enable nfsroot debugging messages.
2367 See Documentation/filesystems/nfs/nfsroot.txt.
2369 nfs.callback_tcpport=
2370 [NFS] set the TCP port on which the NFSv4 callback
2371 channel should listen.
2374 [NFS] sets the pathname to the program which is used
2375 to update the NFS client cache entries.
2377 nfs.cache_getent_timeout=
2378 [NFS] sets the timeout after which an attempt to
2379 update a cache entry is deemed to have failed.
2381 nfs.idmap_cache_timeout=
2382 [NFS] set the maximum lifetime for idmapper cache
2386 [NFS] enable 64-bit inode numbers.
2387 If zero, the NFS client will fake up a 32-bit inode
2388 number for the readdir() and stat() syscalls instead
2389 of returning the full 64-bit number.
2390 The default is to return 64-bit inode numbers.
2392 nfs.max_session_slots=
2393 [NFSv4.1] Sets the maximum number of session slots
2394 the client will attempt to negotiate with the server.
2395 This limits the number of simultaneous RPC requests
2396 that the client can send to the NFSv4.1 server.
2397 Note that there is little point in setting this
2398 value higher than the max_tcp_slot_table_limit.
2400 nfs.nfs4_disable_idmapping=
2401 [NFSv4] When set to the default of '1', this option
2402 ensures that both the RPC level authentication
2403 scheme and the NFS level operations agree to use
2404 numeric uids/gids if the mount is using the
2405 'sec=sys' security flavour. In effect it is
2406 disabling idmapping, which can make migration from
2407 legacy NFSv2/v3 systems to NFSv4 easier.
2408 Servers that do not support this mode of operation
2409 will be autodetected by the client, and it will fall
2410 back to using the idmapper.
2411 To turn off this behaviour, set the value to '0'.
2413 [NFS4] Specify an additional fixed unique ident-
2414 ification string that NFSv4 clients can insert into
2415 their nfs_client_id4 string. This is typically a
2416 UUID that is generated at system install time.
2418 nfs.send_implementation_id =
2419 [NFSv4.1] Send client implementation identification
2420 information in exchange_id requests.
2421 If zero, no implementation identification information
2423 The default is to send the implementation identification
2426 nfs.recover_lost_locks =
2427 [NFSv4] Attempt to recover locks that were lost due
2428 to a lease timeout on the server. Please note that
2429 doing this risks data corruption, since there are
2430 no guarantees that the file will remain unchanged
2431 after the locks are lost.
2432 If you want to enable the kernel legacy behaviour of
2433 attempting to recover these locks, then set this
2435 The default parameter value of '0' causes the kernel
2436 not to attempt recovery of lost locks.
2438 nfs4.layoutstats_timer =
2439 [NFSv4.2] Change the rate at which the kernel sends
2440 layoutstats to the pNFS metadata server.
2442 Setting this to value to 0 causes the kernel to use
2443 whatever value is the default set by the layout
2444 driver. A non-zero value sets the minimum interval
2445 in seconds between layoutstats transmissions.
2447 nfsd.nfs4_disable_idmapping=
2448 [NFSv4] When set to the default of '1', the NFSv4
2449 server will return only numeric uids and gids to
2450 clients using auth_sys, and will accept numeric uids
2451 and gids from such clients. This is intended to ease
2452 migration from NFSv2/v3.
2454 objlayoutdriver.osd_login_prog=
2455 [NFS] [OBJLAYOUT] sets the pathname to the program which
2456 is used to automatically discover and login into new
2457 osd-targets. Please see:
2458 Documentation/filesystems/pnfs.txt for more explanations
2460 nmi_debug= [KNL,AVR32,SH] Specify one or more actions to take
2461 when a NMI is triggered.
2462 Format: [state][,regs][,debounce][,die]
2464 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
2465 Format: [panic,][nopanic,][num]
2467 0 - turn hardlockup detector in nmi_watchdog off
2468 1 - turn hardlockup detector in nmi_watchdog on
2469 When panic is specified, panic when an NMI watchdog
2470 timeout occurs (or 'nopanic' to override the opposite
2471 default). To disable both hard and soft lockup detectors,
2472 please see 'nowatchdog'.
2473 This is useful when you use a panic=... timeout and
2474 need the box quickly up again.
2476 netpoll.carrier_timeout=
2477 [NET] Specifies amount of time (in seconds) that
2478 netpoll should wait for a carrier. By default netpoll
2481 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
2482 emulation library even if a 387 maths coprocessor
2486 [HW] Never suspend the console
2487 Disable suspending of consoles during suspend and
2488 hibernate operations. Once disabled, debugging
2489 messages can reach various consoles while the rest
2490 of the system is being put to sleep (ie, while
2491 debugging driver suspend/resume hooks). This may
2492 not work reliably with all consoles, but is known
2493 to work with serial and VGA consoles.
2494 To facilitate more flexible debugging, we also add
2495 console_suspend, a printk module parameter to control
2496 it. Users could use console_suspend (usually
2497 /sys/module/printk/parameters/console_suspend) to
2498 turn on/off it dynamically.
2500 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
2501 caches in the slab allocator. Saves per-node memory,
2502 but will impact performance.
2506 noapic [SMP,APIC] Tells the kernel to not make use of any
2507 IOAPICs that may be present in the system.
2509 noautogroup Disable scheduler automatic task group creation.
2511 nobats [PPC] Do not use BATs for mapping kernel lowmem
2512 on "Classic" PPC cores.
2516 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
2518 nodelayacct [KNL] Disable per-task delay accounting
2520 nodisconnect [HW,SCSI,M68K] Disables SCSI disconnects.
2522 nodsp [SH] Disable hardware DSP at boot time.
2524 noefi Disable EFI runtime services support.
2529 On X86-32 available only on PAE configured kernels.
2530 noexec=on: enable non-executable mappings (default)
2531 noexec=off: disable non-executable mappings
2534 Disable SMAP (Supervisor Mode Access Prevention)
2535 even if it is supported by processor.
2538 Disable SMEP (Supervisor Mode Execution Prevention)
2539 even if it is supported by processor.
2542 This affects only 32-bit executables.
2543 noexec32=on: enable non-executable mappings (default)
2544 read doesn't imply executable mappings
2545 noexec32=off: disable non-executable mappings
2546 read implies executable mappings
2548 nofpu [MIPS,SH] Disable hardware FPU at boot time.
2550 nofxsr [BUGS=X86-32] Disables x86 floating point extended
2551 register save and restore. The kernel will only save
2552 legacy floating-point registers on task switch.
2554 nohugeiomap [KNL,x86] Disable kernel huge I/O mappings.
2556 noxsave [BUGS=X86] Disables x86 extended register state save
2557 and restore using xsave. The kernel will fallback to
2558 enabling legacy floating-point and sse state.
2560 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
2561 register states. The kernel will fall back to use
2562 xsave to save the states. By using this parameter,
2563 performance of saving the states is degraded because
2564 xsave doesn't support modified optimization while
2565 xsaveopt supports it on xsaveopt enabled systems.
2567 noxsaves [X86] Disables xsaves and xrstors used in saving and
2568 restoring x86 extended register state in compacted
2569 form of xsave area. The kernel will fall back to use
2570 xsaveopt and xrstor to save and restore the states
2571 in standard form of xsave area. By using this
2572 parameter, xsave area per process might occupy more
2573 memory on xsaves enabled systems.
2575 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
2576 wfi(ARM) instruction doesn't work correctly and not to
2577 use it. This is also useful when using JTAG debugger.
2579 no_file_caps Tells the kernel not to honor file capabilities. The
2580 only way then for a file to be executed with privilege
2581 is to be setuid root or executed by root.
2583 nohalt [IA-64] Tells the kernel not to use the power saving
2584 function PAL_HALT_LIGHT when idle. This increases
2585 power-consumption. On the positive side, it reduces
2586 interrupt wake-up latency, which may improve performance
2587 in certain environments such as networked servers or
2590 nohibernate [HIBERNATION] Disable hibernation and resume.
2592 nohz= [KNL] Boottime enable/disable dynamic ticks
2593 Valid arguments: on, off
2596 nohz_full= [KNL,BOOT]
2597 In kernels built with CONFIG_NO_HZ_FULL=y, set
2598 the specified list of CPUs whose tick will be stopped
2599 whenever possible. The boot CPU will be forced outside
2600 the range to maintain the timekeeping.
2601 The CPUs in this range must also be included in the
2604 noiotrap [SH] Disables trapped I/O port accesses.
2606 noirqdebug [X86-32] Disables the code which attempts to detect and
2607 disable unhandled interrupt sources.
2609 no_timer_check [X86,APIC] Disables the code which tests for
2610 broken timer IRQ sources.
2612 noisapnp [ISAPNP] Disables ISA PnP code.
2614 noinitrd [RAM] Tells the kernel not to load any configured
2617 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
2619 [Deprecated - use intremap=off]
2623 noinvpcid [X86] Disable the INVPCID cpu feature.
2625 nojitter [IA-64] Disables jitter checking for ITC timers.
2627 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
2629 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
2632 no-steal-acc [X86,KVM] Disable paravirtualized steal time accounting.
2633 steal time is computed, but won't influence scheduler
2636 nolapic [X86-32,APIC] Do not enable or use the local APIC.
2638 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
2640 noltlbs [PPC] Do not use large page/tlb entries for kernel
2641 lowmem mapping on PPC40x and PPC8xx
2643 nomca [IA-64] Disable machine check abort handling
2645 nomce [X86-32] Disable Machine Check Exception
2647 nomfgpt [X86-32] Disable Multi-Function General Purpose
2648 Timer usage (for AMD Geode machines).
2650 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
2651 shutdown the other cpus. Instead use the REBOOT_VECTOR
2654 nomodule Disable module load
2656 nopat [X86] Disable PAT (page attribute table extension of
2657 pagetables) support.
2659 norandmaps Don't use address space randomization. Equivalent to
2660 echo 0 > /proc/sys/kernel/randomize_va_space
2662 noreplace-paravirt [X86,IA-64,PV_OPS] Don't patch paravirt_ops
2664 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
2665 with UP alternatives
2667 nordrand [X86] Disable kernel use of the RDRAND and
2668 RDSEED instructions even if they are supported
2669 by the processor. RDRAND and RDSEED are still
2670 available to user space applications.
2672 noresume [SWSUSP] Disables resume and restores original swap
2675 no-scroll [VGA] Disables scrollback.
2676 This is required for the Braillex ib80-piezo Braille
2677 reader made by F.H. Papenmeier (Germany).
2681 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
2683 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
2684 and disable the IO APIC. legacy for "maxcpus=0".
2686 nosoftlockup [KNL] Disable the soft-lockup detector.
2688 nosync [HW,M68K] Disables sync negotiation for all devices.
2690 notsc [BUGS=X86-32] Disable Time Stamp Counter
2692 nowatchdog [KNL] Disable both lockup detectors, i.e.
2693 soft-lockup and NMI watchdog (hard-lockup).
2697 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
2699 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
2700 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
2701 Some features depend on CPU0. Known dependencies are:
2702 1. Resume from suspend/hibernate depends on CPU0.
2703 Suspend/hibernate will fail if CPU0 is offline and you
2704 need to online CPU0 before suspend/hibernate.
2705 2. PIC interrupts also depend on CPU0. CPU0 can't be
2706 removed if a PIC interrupt is detected.
2707 It's said poweroff/reboot may depend on CPU0 on some
2708 machines although I haven't seen such issues so far
2709 after CPU0 is offline on a few tested machines.
2710 If the dependencies are under your control, you can
2711 turn on cpu0_hotplug.
2713 nptcg= [IA-64] Override max number of concurrent global TLB
2714 purges which is reported from either PAL_VM_SUMMARY or
2717 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
2718 could support. nr_cpus=n : n >= 1 limits the kernel to
2719 supporting 'n' processors. Later in runtime you can not
2720 use hotplug cpu feature to put more cpu back to online.
2721 just like you compile the kernel NR_CPUS=n
2723 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
2725 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing.
2726 Allowed values are enable and disable
2728 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
2729 one of ['zone', 'node', 'default'] can be specified
2730 This can be set from sysctl after boot.
2731 See Documentation/sysctl/vm.txt for details.
2733 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
2734 See Documentation/debugging-via-ohci1394.txt for more
2737 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
2738 Rather than timing out after 20 ms if an EC
2739 command is not properly ACKed, override the length
2740 of the timeout. We have interrupts disabled while
2741 waiting for the ACK, so if this is set too high
2742 interrupts *may* be lost!
2744 omap_mux= [OMAP] Override bootloader pin multiplexing.
2745 Format: <mux_mode0.mode_name=value>...
2746 For example, to override I2C bus2:
2747 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
2749 oprofile.timer= [HW]
2750 Use timer interrupt instead of performance counters
2752 oprofile.cpu_type= Force an oprofile cpu type
2753 This might be useful if you have an older oprofile
2754 userland or if you want common events.
2755 Format: { arch_perfmon }
2756 arch_perfmon: [X86] Force use of architectural
2757 perfmon on Intel CPUs instead of the
2758 CPU specific event set.
2759 timer: [X86] Force use of architectural NMI
2760 timer mode (see also oprofile.timer
2761 for generic hr timer mode)
2762 [s390] Force legacy basic mode sampling
2763 (report cpu_type "timer")
2765 oops=panic Always panic on oopses. Default is to just kill the
2766 process, but there is a small probability of
2767 deadlocking the machine.
2768 This will also cause panics on machine check exceptions.
2769 Useful together with panic=30 to trigger a reboot.
2772 See Documentation/sound/oss/oss-parameters.txt
2774 page_owner= [KNL] Boot-time page_owner enabling option.
2775 Storage of the information about who allocated
2776 each page is disabled in default. With this switch,
2778 on: enable the feature
2780 page_poison= [KNL] Boot-time parameter changing the state of
2781 poisoning on the buddy allocator.
2782 off: turn off poisoning
2783 on: turn on poisoning
2785 panic= [KNL] Kernel behaviour on panic: delay <timeout>
2786 timeout > 0: seconds before rebooting
2787 timeout = 0: wait forever
2788 timeout < 0: reboot immediately
2791 panic_on_warn panic() instead of WARN(). Useful to cause kdump
2794 crash_kexec_post_notifiers
2795 Run kdump after running panic-notifiers and dumping
2796 kmsg. This only for the users who doubt kdump always
2797 succeeds in any situation.
2798 Note that this also increases risks of kdump failure,
2799 because some panic notifiers can make the crashed
2800 kernel more unstable.
2802 parkbd.port= [HW] Parallel port number the keyboard adapter is
2803 connected to, default is 0.
2805 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
2806 0 for XT, 1 for AT (default is AT).
2809 parport= [HW,PPT] Specify parallel ports. 0 disables.
2810 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
2811 Use 'auto' to force the driver to use any
2812 IRQ/DMA settings detected (the default is to
2813 ignore detected IRQ/DMA settings because of
2814 possible conflicts). You can specify the base
2815 address, IRQ, and DMA settings; IRQ and DMA
2816 should be numbers, or 'auto' (for using detected
2817 settings on that particular port), or 'nofifo'
2818 (to avoid using a FIFO even if it is detected).
2819 Parallel ports are assigned in the order they
2820 are specified on the command line, starting
2823 parport_init_mode= [HW,PPT]
2824 Configure VIA parallel port to operate in
2825 a specific mode. This is necessary on Pegasos
2826 computer where firmware has no options for setting
2827 up parallel port mode and sets it to spp.
2828 Currently this function knows 686a and 8231 chips.
2829 Format: [spp|ps2|epp|ecp|ecpepp]
2832 Halt all CPUs after the first oops has been printed for
2833 the specified number of seconds. This is to be used if
2834 your oopses keep scrolling off the screen.
2839 See header of drivers/block/paride/pcd.c.
2840 See also Documentation/blockdev/paride.txt.
2842 pci=option[,option...] [PCI] various PCI subsystem options:
2843 earlydump [X86] dump PCI config space before the kernel
2845 off [X86] don't probe for the PCI bus
2846 bios [X86-32] force use of PCI BIOS, don't access
2847 the hardware directly. Use this if your machine
2848 has a non-standard PCI host bridge.
2849 nobios [X86-32] disallow use of PCI BIOS, only direct
2850 hardware access methods are allowed. Use this
2851 if you experience crashes upon bootup and you
2852 suspect they are caused by the BIOS.
2853 conf1 [X86] Force use of PCI Configuration Access
2854 Mechanism 1 (config address in IO port 0xCF8,
2855 data in IO port 0xCFC, both 32-bit).
2856 conf2 [X86] Force use of PCI Configuration Access
2857 Mechanism 2 (IO port 0xCF8 is an 8-bit port for
2858 the function, IO port 0xCFA, also 8-bit, sets
2859 bus number. The config space is then accessed
2860 through ports 0xC000-0xCFFF).
2861 See http://wiki.osdev.org/PCI for more info
2862 on the configuration access mechanisms.
2863 noaer [PCIE] If the PCIEAER kernel config parameter is
2864 enabled, this kernel boot option can be used to
2865 disable the use of PCIE advanced error reporting.
2866 nodomains [PCI] Disable support for multiple PCI
2867 root domains (aka PCI segments, in ACPI-speak).
2868 nommconf [X86] Disable use of MMCONFIG for PCI
2870 check_enable_amd_mmconf [X86] check for and enable
2871 properly configured MMIO access to PCI
2872 config space on AMD family 10h CPU
2873 nomsi [MSI] If the PCI_MSI kernel config parameter is
2874 enabled, this kernel boot option can be used to
2875 disable the use of MSI interrupts system-wide.
2876 noioapicquirk [APIC] Disable all boot interrupt quirks.
2877 Safety option to keep boot IRQs enabled. This
2878 should never be necessary.
2879 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
2880 primary IO-APIC for bridges that cannot disable
2881 boot IRQs. This fixes a source of spurious IRQs
2882 when the system masks IRQs.
2883 noioapicreroute [APIC] Disable workaround that uses the
2884 boot IRQ equivalent of an IRQ that connects to
2885 a chipset where boot IRQs cannot be disabled.
2886 The opposite of ioapicreroute.
2887 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
2888 routing table. These calls are known to be buggy
2889 on several machines and they hang the machine
2890 when used, but on other computers it's the only
2891 way to get the interrupt routing table. Try
2892 this option if the kernel is unable to allocate
2893 IRQs or discover secondary PCI buses on your
2895 rom [X86] Assign address space to expansion ROMs.
2896 Use with caution as certain devices share
2897 address decoders between ROMs and other
2899 norom [X86] Do not assign address space to
2900 expansion ROMs that do not already have
2901 BIOS assigned address ranges.
2902 nobar [X86] Do not assign address space to the
2903 BARs that weren't assigned by the BIOS.
2904 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
2905 assigned automatically to PCI devices. You can
2906 make the kernel exclude IRQs of your ISA cards
2908 pirqaddr=0xAAAAA [X86] Specify the physical address
2909 of the PIRQ table (normally generated
2910 by the BIOS) if it is outside the
2911 F0000h-100000h range.
2912 lastbus=N [X86] Scan all buses thru bus #N. Can be
2913 useful if the kernel is unable to find your
2914 secondary buses and you want to tell it
2915 explicitly which ones they are.
2916 assign-busses [X86] Always assign all PCI bus
2917 numbers ourselves, overriding
2918 whatever the firmware may have done.
2919 usepirqmask [X86] Honor the possible IRQ mask stored
2920 in the BIOS $PIR table. This is needed on
2921 some systems with broken BIOSes, notably
2922 some HP Pavilion N5400 and Omnibook XE3
2923 notebooks. This will have no effect if ACPI
2924 IRQ routing is enabled.
2925 noacpi [X86] Do not use ACPI for IRQ routing
2926 or for PCI scanning.
2927 use_crs [X86] Use PCI host bridge window information
2928 from ACPI. On BIOSes from 2008 or later, this
2929 is enabled by default. If you need to use this,
2930 please report a bug.
2931 nocrs [X86] Ignore PCI host bridge windows from ACPI.
2932 If you need to use this, please report a bug.
2933 routeirq Do IRQ routing for all PCI devices.
2934 This is normally done in pci_enable_device(),
2935 so this option is a temporary workaround
2936 for broken drivers that don't call it.
2937 skip_isa_align [X86] do not align io start addr, so can
2938 handle more pci cards
2939 firmware [ARM] Do not re-enumerate the bus but instead
2940 just use the configuration from the
2941 bootloader. This is currently used on
2942 IXP2000 systems where the bus has to be
2943 configured a certain way for adjunct CPUs.
2944 noearly [X86] Don't do any early type 1 scanning.
2945 This might help on some broken boards which
2946 machine check when some devices' config space
2947 is read. But various workarounds are disabled
2948 and some IOMMU drivers will not work.
2949 bfsort Sort PCI devices into breadth-first order.
2950 This sorting is done to get a device
2951 order compatible with older (<= 2.4) kernels.
2952 nobfsort Don't sort PCI devices into breadth-first order.
2953 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
2954 tuning and use the BIOS-configured MPS defaults.
2955 pcie_bus_safe Set every device's MPS to the largest value
2956 supported by all devices below the root complex.
2957 pcie_bus_perf Set device MPS to the largest allowable MPS
2958 based on its parent bus. Also set MRRS (Max
2959 Read Request Size) to the largest supported
2960 value (no larger than the MPS that the device
2961 or bus can support) for best performance.
2962 pcie_bus_peer2peer Set every device's MPS to 128B, which
2963 every device is guaranteed to support. This
2964 configuration allows peer-to-peer DMA between
2965 any pair of devices, possibly at the cost of
2966 reduced performance. This also guarantees
2967 that hot-added devices will work.
2968 cbiosize=nn[KMG] The fixed amount of bus space which is
2969 reserved for the CardBus bridge's IO window.
2970 The default value is 256 bytes.
2971 cbmemsize=nn[KMG] The fixed amount of bus space which is
2972 reserved for the CardBus bridge's memory
2973 window. The default value is 64 megabytes.
2976 [<order of align>@][<domain>:]<bus>:<slot>.<func>[; ...]
2977 Specifies alignment and device to reassign
2978 aligned memory resources.
2979 If <order of align> is not specified,
2980 PAGE_SIZE is used as alignment.
2981 PCI-PCI bridge can be specified, if resource
2982 windows need to be expanded.
2983 ecrc= Enable/disable PCIe ECRC (transaction layer
2984 end-to-end CRC checking).
2985 bios: Use BIOS/firmware settings. This is the
2989 hpiosize=nn[KMG] The fixed amount of bus space which is
2990 reserved for hotplug bridge's IO window.
2991 Default size is 256 bytes.
2992 hpmemsize=nn[KMG] The fixed amount of bus space which is
2993 reserved for hotplug bridge's memory window.
2994 Default size is 2 megabytes.
2995 realloc= Enable/disable reallocating PCI bridge resources
2996 if allocations done by BIOS are too small to
2997 accommodate resources required by all child
2999 off: Turn realloc off
3001 realloc same as realloc=on
3002 noari do not use PCIe ARI.
3003 pcie_scan_all Scan all possible PCIe devices. Otherwise we
3004 only look for one device below a PCIe downstream
3007 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
3010 force Enable ASPM even on devices that claim not to support it.
3011 WARNING: Forcing ASPM on may cause system lockups.
3013 pcie_hp= [PCIE] PCI Express Hotplug driver options:
3014 nomsi Do not use MSI for PCI Express Native Hotplug (this
3015 makes all PCIe ports use INTx for hotplug services).
3017 pcie_ports= [PCIE] PCIe ports handling:
3018 auto Ask the BIOS whether or not to use native PCIe services
3019 associated with PCIe ports (PME, hot-plug, AER). Use
3020 them only if that is allowed by the BIOS.
3021 native Use native PCIe services associated with PCIe ports
3023 compat Treat PCIe ports as PCI-to-PCI bridges, disable the PCIe
3026 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
3027 nomsi Do not use MSI for native PCIe PME signaling (this makes
3028 all PCIe root ports use INTx for all services).
3030 pcmv= [HW,PCMCIA] BadgePAD 4
3034 Keep all power-domains already enabled by bootloader on,
3035 even if no driver has claimed them. This is useful
3036 for debug and development, but should not be
3037 needed on a platform with proper driver support.
3040 See Documentation/blockdev/paride.txt.
3042 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
3045 See arch/parisc/kernel/pdc_chassis.c
3047 percpu_alloc= Select which percpu first chunk allocator to use.
3048 Currently supported values are "embed" and "page".
3049 Archs may support subset or none of the selections.
3050 See comments in mm/percpu.c for details on each
3051 allocator. This parameter is primarily for debugging
3052 and performance comparison.
3055 See Documentation/blockdev/paride.txt.
3058 See Documentation/blockdev/paride.txt.
3060 pirq= [SMP,APIC] Manual mp-table setup
3061 See Documentation/x86/i386/IO-APIC.txt.
3063 plip= [PPT,NET] Parallel port network link
3064 Format: { parport<nr> | timid | 0 }
3065 See also Documentation/parport.txt.
3067 pmtmr= [X86] Manual setup of pmtmr I/O Port.
3068 Override pmtimer IOPort with a hex value.
3072 Enable PNP debug messages (depends on the
3073 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
3074 via /sys/module/pnp/parameters/debug. We always show
3075 current resource usage; turning this on also shows
3076 possible settings and some assignment information.
3082 { on | off | curr | res | no-curr | no-res }
3085 [ISAPNP] Exclude IRQs for the autoconfiguration
3088 [ISAPNP] Exclude DMAs for the autoconfiguration
3090 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
3091 Ranges are in pairs (I/O port base and size).
3094 [ISAPNP] Exclude memory regions for the
3096 Ranges are in pairs (memory base and size).
3098 ports= [IP_VS_FTP] IPVS ftp helper module
3100 Up to 8 (IP_VS_APP_MAX_PORTS) ports
3102 Format: <port>,<port>....
3104 ppc_strict_facility_enable
3105 [PPC] This option catches any kernel floating point,
3106 Altivec, VSX and SPE outside of regions specifically
3107 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
3108 There is some performance impact when enabling this.
3110 print-fatal-signals=
3111 [KNL] debug: print fatal signals
3113 If enabled, warn about various signal handling
3114 related application anomalies: too many signals,
3115 too many POSIX.1 timers, fatal signals causing a
3118 If you hit the warning due to signal overflow,
3119 you might want to try "ulimit -i unlimited".
3123 printk.always_kmsg_dump=
3124 Trigger kmsg_dump for cases other than kernel oops or
3126 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3129 printk.time= Show timing data prefixed to each printk message line
3130 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3132 processor.max_cstate= [HW,ACPI]
3133 Limit processor to maximum C-state
3134 max_cstate=9 overrides any DMI blacklist limit.
3136 processor.nocst [HW,ACPI]
3137 Ignore the _CST method to determine C-states,
3138 instead using the legacy FADT method
3140 profile= [KNL] Enable kernel profiling via /proc/profile
3141 Format: [schedule,]<number>
3142 Param: "schedule" - profile schedule points.
3143 Param: <number> - step/bucket size as a power of 2 for
3144 statistical time based profiling.
3145 Param: "sleep" - profile D-state sleeping (millisecs).
3146 Requires CONFIG_SCHEDSTATS
3147 Param: "kvm" - profile VM exits.
3149 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
3151 See Documentation/blockdev/ramdisk.txt.
3153 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
3154 probe for; one of (bare|imps|exps|lifebook|any).
3155 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
3157 psmouse.resetafter= [HW,MOUSE]
3158 Try to reset the device after so many bad packets
3161 [HW,MOUSE] Set desired mouse resolution, in dpi.
3162 psmouse.smartscroll=
3163 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
3164 0 = disabled, 1 = enabled (default).
3166 pstore.backend= Specify the name of the pstore backend to use
3169 See Documentation/blockdev/paride.txt.
3172 [KNL] Number of legacy pty's. Overwrites compiled-in
3175 quiet [KNL] Disable most log messages
3180 See Documentation/md.txt.
3182 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
3183 See Documentation/blockdev/ramdisk.txt.
3186 In kernels built with CONFIG_RCU_NOCB_CPU=y, set
3187 the specified list of CPUs to be no-callback CPUs.
3188 Invocation of these CPUs' RCU callbacks will
3189 be offloaded to "rcuox/N" kthreads created for
3190 that purpose, where "x" is "b" for RCU-bh, "p"
3191 for RCU-preempt, and "s" for RCU-sched, and "N"
3192 is the CPU number. This reduces OS jitter on the
3193 offloaded CPUs, which can be useful for HPC and
3194 real-time workloads. It can also improve energy
3195 efficiency for asymmetric multiprocessors.
3198 Rather than requiring that offloaded CPUs
3199 (specified by rcu_nocbs= above) explicitly
3200 awaken the corresponding "rcuoN" kthreads,
3201 make these kthreads poll for callbacks.
3202 This improves the real-time response for the
3203 offloaded CPUs by relieving them of the need to
3204 wake up the corresponding kthread, but degrades
3205 energy efficiency by requiring that the kthreads
3206 periodically wake up to do the polling.
3208 rcutree.blimit= [KNL]
3209 Set maximum number of finished RCU callbacks to
3210 process in one batch.
3212 rcutree.dump_tree= [KNL]
3213 Dump the structure of the rcu_node combining tree
3214 out at early boot. This is used for diagnostic
3215 purposes, to verify correct tree setup.
3217 rcutree.gp_cleanup_delay= [KNL]
3218 Set the number of jiffies to delay each step of
3219 RCU grace-period cleanup. This only has effect
3220 when CONFIG_RCU_TORTURE_TEST_SLOW_CLEANUP is set.
3222 rcutree.gp_init_delay= [KNL]
3223 Set the number of jiffies to delay each step of
3224 RCU grace-period initialization. This only has
3225 effect when CONFIG_RCU_TORTURE_TEST_SLOW_INIT
3228 rcutree.gp_preinit_delay= [KNL]
3229 Set the number of jiffies to delay each step of
3230 RCU grace-period pre-initialization, that is,
3231 the propagation of recent CPU-hotplug changes up
3232 the rcu_node combining tree. This only has effect
3233 when CONFIG_RCU_TORTURE_TEST_SLOW_PREINIT is set.
3235 rcutree.rcu_fanout_exact= [KNL]
3236 Disable autobalancing of the rcu_node combining
3237 tree. This is used by rcutorture, and might
3238 possibly be useful for architectures having high
3239 cache-to-cache transfer latencies.
3241 rcutree.rcu_fanout_leaf= [KNL]
3242 Change the number of CPUs assigned to each
3243 leaf rcu_node structure. Useful for very
3244 large systems, which will choose the value 64,
3245 and for NUMA systems with large remote-access
3246 latencies, which will choose a value aligned
3247 with the appropriate hardware boundaries.
3249 rcutree.jiffies_till_sched_qs= [KNL]
3250 Set required age in jiffies for a
3251 given grace period before RCU starts
3252 soliciting quiescent-state help from
3253 rcu_note_context_switch().
3255 rcutree.jiffies_till_first_fqs= [KNL]
3256 Set delay from grace-period initialization to
3257 first attempt to force quiescent states.
3258 Units are jiffies, minimum value is zero,
3259 and maximum value is HZ.
3261 rcutree.jiffies_till_next_fqs= [KNL]
3262 Set delay between subsequent attempts to force
3263 quiescent states. Units are jiffies, minimum
3264 value is one, and maximum value is HZ.
3266 rcutree.kthread_prio= [KNL,BOOT]
3267 Set the SCHED_FIFO priority of the RCU per-CPU
3268 kthreads (rcuc/N). This value is also used for
3269 the priority of the RCU boost threads (rcub/N)
3270 and for the RCU grace-period kthreads (rcu_bh,
3271 rcu_preempt, and rcu_sched). If RCU_BOOST is
3272 set, valid values are 1-99 and the default is 1
3273 (the least-favored priority). Otherwise, when
3274 RCU_BOOST is not set, valid values are 0-99 and
3275 the default is zero (non-realtime operation).
3277 rcutree.rcu_nocb_leader_stride= [KNL]
3278 Set the number of NOCB kthread groups, which
3279 defaults to the square root of the number of
3280 CPUs. Larger numbers reduces the wakeup overhead
3281 on the per-CPU grace-period kthreads, but increases
3282 that same overhead on each group's leader.
3284 rcutree.qhimark= [KNL]
3285 Set threshold of queued RCU callbacks beyond which
3286 batch limiting is disabled.
3288 rcutree.qlowmark= [KNL]
3289 Set threshold of queued RCU callbacks below which
3290 batch limiting is re-enabled.
3292 rcutree.rcu_idle_gp_delay= [KNL]
3293 Set wakeup interval for idle CPUs that have
3294 RCU callbacks (RCU_FAST_NO_HZ=y).
3296 rcutree.rcu_idle_lazy_gp_delay= [KNL]
3297 Set wakeup interval for idle CPUs that have
3298 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
3299 Lazy RCU callbacks are those which RCU can
3300 prove do nothing more than free memory.
3302 rcuperf.gp_exp= [KNL]
3303 Measure performance of expedited synchronous
3304 grace-period primitives.
3306 rcuperf.holdoff= [KNL]
3307 Set test-start holdoff period. The purpose of
3308 this parameter is to delay the start of the
3309 test until boot completes in order to avoid
3312 rcuperf.nreaders= [KNL]
3313 Set number of RCU readers. The value -1 selects
3314 N, where N is the number of CPUs. A value
3315 "n" less than -1 selects N-n+1, where N is again
3316 the number of CPUs. For example, -2 selects N
3317 (the number of CPUs), -3 selects N+1, and so on.
3318 A value of "n" less than or equal to -N selects
3321 rcuperf.nwriters= [KNL]
3322 Set number of RCU writers. The values operate
3323 the same as for rcuperf.nreaders.
3324 N, where N is the number of CPUs
3326 rcuperf.perf_runnable= [BOOT]
3327 Start rcuperf running at boot time.
3329 rcuperf.shutdown= [KNL]
3330 Shut the system down after performance tests
3331 complete. This is useful for hands-off automated
3334 rcuperf.perf_type= [KNL]
3335 Specify the RCU implementation to test.
3337 rcuperf.verbose= [KNL]
3338 Enable additional printk() statements.
3340 rcutorture.cbflood_inter_holdoff= [KNL]
3341 Set holdoff time (jiffies) between successive
3342 callback-flood tests.
3344 rcutorture.cbflood_intra_holdoff= [KNL]
3345 Set holdoff time (jiffies) between successive
3346 bursts of callbacks within a given callback-flood
3349 rcutorture.cbflood_n_burst= [KNL]
3350 Set the number of bursts making up a given
3351 callback-flood test. Set this to zero to
3352 disable callback-flood testing.
3354 rcutorture.cbflood_n_per_burst= [KNL]
3355 Set the number of callbacks to be registered
3356 in a given burst of a callback-flood test.
3358 rcutorture.fqs_duration= [KNL]
3359 Set duration of force_quiescent_state bursts
3362 rcutorture.fqs_holdoff= [KNL]
3363 Set holdoff time within force_quiescent_state bursts
3366 rcutorture.fqs_stutter= [KNL]
3367 Set wait time between force_quiescent_state bursts
3370 rcutorture.gp_cond= [KNL]
3371 Use conditional/asynchronous update-side
3372 primitives, if available.
3374 rcutorture.gp_exp= [KNL]
3375 Use expedited update-side primitives, if available.
3377 rcutorture.gp_normal= [KNL]
3378 Use normal (non-expedited) asynchronous
3379 update-side primitives, if available.
3381 rcutorture.gp_sync= [KNL]
3382 Use normal (non-expedited) synchronous
3383 update-side primitives, if available. If all
3384 of rcutorture.gp_cond=, rcutorture.gp_exp=,
3385 rcutorture.gp_normal=, and rcutorture.gp_sync=
3386 are zero, rcutorture acts as if is interpreted
3387 they are all non-zero.
3389 rcutorture.n_barrier_cbs= [KNL]
3390 Set callbacks/threads for rcu_barrier() testing.
3392 rcutorture.nfakewriters= [KNL]
3393 Set number of concurrent RCU writers. These just
3394 stress RCU, they don't participate in the actual
3395 test, hence the "fake".
3397 rcutorture.nreaders= [KNL]
3398 Set number of RCU readers. The value -1 selects
3399 N-1, where N is the number of CPUs. A value
3400 "n" less than -1 selects N-n-2, where N is again
3401 the number of CPUs. For example, -2 selects N
3402 (the number of CPUs), -3 selects N+1, and so on.
3404 rcutorture.object_debug= [KNL]
3405 Enable debug-object double-call_rcu() testing.
3407 rcutorture.onoff_holdoff= [KNL]
3408 Set time (s) after boot for CPU-hotplug testing.
3410 rcutorture.onoff_interval= [KNL]
3411 Set time (s) between CPU-hotplug operations, or
3412 zero to disable CPU-hotplug testing.
3414 rcutorture.shuffle_interval= [KNL]
3415 Set task-shuffle interval (s). Shuffling tasks
3416 allows some CPUs to go into dyntick-idle mode
3417 during the rcutorture test.
3419 rcutorture.shutdown_secs= [KNL]
3420 Set time (s) after boot system shutdown. This
3421 is useful for hands-off automated testing.
3423 rcutorture.stall_cpu= [KNL]
3424 Duration of CPU stall (s) to test RCU CPU stall
3425 warnings, zero to disable.
3427 rcutorture.stall_cpu_holdoff= [KNL]
3428 Time to wait (s) after boot before inducing stall.
3430 rcutorture.stat_interval= [KNL]
3431 Time (s) between statistics printk()s.
3433 rcutorture.stutter= [KNL]
3434 Time (s) to stutter testing, for example, specifying
3435 five seconds causes the test to run for five seconds,
3436 wait for five seconds, and so on. This tests RCU's
3437 ability to transition abruptly to and from idle.
3439 rcutorture.test_boost= [KNL]
3440 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
3441 "Maybe" means test if the RCU implementation
3442 under test support RCU priority boosting.
3444 rcutorture.test_boost_duration= [KNL]
3445 Duration (s) of each individual boost test.
3447 rcutorture.test_boost_interval= [KNL]
3448 Interval (s) between each boost test.
3450 rcutorture.test_no_idle_hz= [KNL]
3451 Test RCU's dyntick-idle handling. See also the
3452 rcutorture.shuffle_interval parameter.
3454 rcutorture.torture_runnable= [BOOT]
3455 Start rcutorture running at boot time.
3457 rcutorture.torture_type= [KNL]
3458 Specify the RCU implementation to test.
3460 rcutorture.verbose= [KNL]
3461 Enable additional printk() statements.
3463 rcupdate.rcu_cpu_stall_suppress= [KNL]
3464 Suppress RCU CPU stall warning messages.
3466 rcupdate.rcu_cpu_stall_timeout= [KNL]
3467 Set timeout for RCU CPU stall warning messages.
3469 rcupdate.rcu_expedited= [KNL]
3470 Use expedited grace-period primitives, for
3471 example, synchronize_rcu_expedited() instead
3472 of synchronize_rcu(). This reduces latency,
3473 but can increase CPU utilization, degrade
3474 real-time latency, and degrade energy efficiency.
3475 No effect on CONFIG_TINY_RCU kernels.
3477 rcupdate.rcu_normal= [KNL]
3478 Use only normal grace-period primitives,
3479 for example, synchronize_rcu() instead of
3480 synchronize_rcu_expedited(). This improves
3481 real-time latency, CPU utilization, and
3482 energy efficiency, but can expose users to
3483 increased grace-period latency. This parameter
3484 overrides rcupdate.rcu_expedited. No effect on
3485 CONFIG_TINY_RCU kernels.
3487 rcupdate.rcu_normal_after_boot= [KNL]
3488 Once boot has completed (that is, after
3489 rcu_end_inkernel_boot() has been invoked), use
3490 only normal grace-period primitives. No effect
3491 on CONFIG_TINY_RCU kernels.
3493 rcupdate.rcu_task_stall_timeout= [KNL]
3494 Set timeout in jiffies for RCU task stall warning
3495 messages. Disable with a value less than or equal
3498 rcupdate.rcu_self_test= [KNL]
3499 Run the RCU early boot self tests
3501 rcupdate.rcu_self_test_bh= [KNL]
3502 Run the RCU bh early boot self tests
3504 rcupdate.rcu_self_test_sched= [KNL]
3505 Run the RCU sched early boot self tests
3509 Run specified binary instead of /init from the ramdisk,
3510 used for early userspace startup. See initrd.
3513 Format (x86 or x86_64):
3514 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
3516 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
3518 Where reboot_mode is one of warm (soft) or cold (hard) or gpio,
3519 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
3520 reboot_force is either force or not specified,
3521 reboot_cpu is s[mp]#### with #### being the processor
3522 to be used for rebooting.
3525 [KNL, SMP] Set scheduler's default relax_domain_level.
3526 See Documentation/cgroups/cpusets.txt.
3528 relative_sleep_states=
3529 [SUSPEND] Use sleep state labeling where the deepest
3530 state available other than hibernation is always "mem".
3531 Format: { "0" | "1" }
3532 0 -- Traditional sleep state labels.
3533 1 -- Relative sleep state labels.
3535 reserve= [KNL,BUGS] Force the kernel to ignore some iomem area
3537 reservetop= [X86-32]
3539 Reserves a hole at the top of the kernel virtual
3544 Set the amount of memory to reserve for BIOS at
3545 the bottom of the address space.
3547 reset_devices [KNL] Force drivers to reset the underlying device
3548 during initialization.
3551 Specify the partition device for software suspend
3553 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
3555 resume_offset= [SWSUSP]
3556 Specify the offset from the beginning of the partition
3557 given by "resume=" at which the swap header is located,
3558 in <PAGE_SIZE> units (needed only for swap files).
3559 See Documentation/power/swsusp-and-swap-files.txt
3561 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
3562 read the resume files
3564 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
3565 Useful for devices that are detected asynchronously
3566 (e.g. USB and MMC devices).
3568 hibernate= [HIBERNATION]
3569 noresume Don't check if there's a hibernation image
3570 present during boot.
3571 nocompress Don't compress/decompress hibernation images.
3572 no Disable hibernation and resume.
3574 retain_initrd [RAM] Keep initrd memory after extraction
3576 rfkill.default_state=
3577 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
3578 etc. communication is blocked by default.
3581 rfkill.master_switch_mode=
3582 0 The "airplane mode" button does nothing.
3583 1 The "airplane mode" button toggles between everything
3584 blocked and the previous configuration.
3585 2 The "airplane mode" button toggles between everything
3586 blocked and everything unblocked.
3588 rhash_entries= [KNL,NET]
3589 Set number of hash buckets for route cache
3591 ro [KNL] Mount root device read-only on boot
3594 on Mark read-only kernel memory as read-only (default).
3595 off Leave read-only kernel memory writable for debugging.
3598 Enable the uart passthrough on the designated usb port
3599 on Rockchip SoCs. When active, the signals of the
3600 debug-uart get routed to the D+ and D- pins of the usb
3601 port and the regular usb controller gets disabled.
3603 root= [KNL] Root filesystem
3604 See name_to_dev_t comment in init/do_mounts.c.
3606 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
3607 mount the root filesystem
3609 rootflags= [KNL] Set root filesystem mount option string
3611 rootfstype= [KNL] Set root filesystem type
3613 rootwait [KNL] Wait (indefinitely) for root device to show up.
3614 Useful for devices that are detected asynchronously
3615 (e.g. USB and MMC devices).
3617 rproc_mem=nn[KMG][@address]
3618 [KNL,ARM,CMA] Remoteproc physical memory block.
3619 Memory area to be used by remote processor image,
3622 rw [KNL] Mount root device read-write on boot
3624 S [KNL] Run init in single mode
3626 s390_iommu= [HW,S390]
3627 Set s390 IOTLB flushing mode
3629 With strict flushing every unmap operation will result in
3630 an IOTLB flush. Default is lazy flushing before reuse,
3634 See drivers/net/irda/sa1100_ir.c.
3636 sbni= [NET] Granch SBNI12 leased line adapter
3638 sched_debug [KNL] Enables verbose scheduler debug messages.
3640 schedstats= [KNL,X86] Enable or disable scheduled statistics.
3641 Allowed values are enable and disable. This feature
3642 incurs a small amount of overhead in the scheduler
3643 but is useful for debugging and performance tuning.
3645 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
3646 xtime_lock contention on larger systems, and/or RCU lock
3647 contention on all systems with CONFIG_MAXSMP set.
3648 Format: { "0" | "1" }
3649 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
3651 Note: increases power consumption, thus should only be
3652 enabled if running jitter sensitive (HPC/RT) workloads.
3654 security= [SECURITY] Choose a security module to enable at boot.
3655 If this boot parameter is not specified, only the first
3656 security module asking for security registration will be
3657 loaded. An invalid security module name will be treated
3658 as if no module has been chosen.
3660 selinux= [SELINUX] Disable or enable SELinux at boot time.
3661 Format: { "0" | "1" }
3662 See security/selinux/Kconfig help text.
3665 Default value is set via kernel config option.
3666 If enabled at boot time, /selinux/disable can be used
3667 later to disable prior to initial policy load.
3669 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
3670 Format: { "0" | "1" }
3671 See security/apparmor/Kconfig help text
3674 Default value is set via kernel config option.
3676 serialnumber [BUGS=X86-32]
3679 Maximal number of shapers.
3681 show_msr= [x86] show boot-time MSR settings
3682 Format: { <integer> }
3683 Show boot-time (BIOS-initialized) MSR settings.
3684 The parameter means the number of CPUs to show,
3685 for example 1 means boot CPU only.
3693 Disable merging of slabs with similar size. May be
3694 necessary if there is some reason to distinguish
3695 allocs to different slabs. Debug options disable
3696 merging on their own.
3697 For more information see Documentation/vm/slub.txt.
3699 slab_max_order= [MM, SLAB]
3700 Determines the maximum allowed order for slabs.
3701 A high setting may cause OOMs due to memory
3702 fragmentation. Defaults to 1 for systems with
3703 more than 32MB of RAM, 0 otherwise.
3705 slub_debug[=options[,slabs]] [MM, SLUB]
3706 Enabling slub_debug allows one to determine the
3707 culprit if slab objects become corrupted. Enabling
3708 slub_debug can create guard zones around objects and
3709 may poison objects when not in use. Also tracks the
3710 last alloc / free. For more information see
3711 Documentation/vm/slub.txt.
3713 slub_max_order= [MM, SLUB]
3714 Determines the maximum allowed order for slabs.
3715 A high setting may cause OOMs due to memory
3716 fragmentation. For more information see
3717 Documentation/vm/slub.txt.
3719 slub_min_objects= [MM, SLUB]
3720 The minimum number of objects per slab. SLUB will
3721 increase the slab order up to slub_max_order to
3722 generate a sufficiently large slab able to contain
3723 the number of objects indicated. The higher the number
3724 of objects the smaller the overhead of tracking slabs
3725 and the less frequently locks need to be acquired.
3726 For more information see Documentation/vm/slub.txt.
3728 slub_min_order= [MM, SLUB]
3729 Determines the minimum page order for slabs. Must be
3730 lower than slub_max_order.
3731 For more information see Documentation/vm/slub.txt.
3733 slub_nomerge [MM, SLUB]
3734 Same with slab_nomerge. This is supported for legacy.
3735 See slab_nomerge for more information.
3738 Format: <io1>[,<io2>[,...,<io8>]]
3740 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
3741 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
3742 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
3743 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
3744 smsc-ircc2.ircc_irq= [HW] IRQ line
3745 smsc-ircc2.ircc_dma= [HW] DMA channel
3746 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
3747 0: Toshiba Satellite 1800 (GP data pin select)
3748 1: Fast pin select (default)
3752 [KNL] Should the soft-lockup detector generate panics.
3755 softlockup_all_cpu_backtrace=
3756 [KNL] Should the soft-lockup detector generate
3757 backtraces on all cpus.
3760 sonypi.*= [HW] Sony Programmable I/O Control Device driver
3761 See Documentation/laptops/sonypi.txt
3763 spia_io_base= [HW,MTD]
3769 Enabled the stack tracer on boot up.
3771 stacktrace_filter=[function-list]
3772 [FTRACE] Limit the functions that the stack tracer
3773 will trace at boot up. function-list is a comma separated
3774 list of functions. This list can be changed at run
3775 time by the stack_trace_filter file in the debugfs
3776 tracing directory. Note, this enables stack tracing
3777 and the stacktrace above is not needed.
3781 Set the STI (builtin display/keyboard on the HP-PARISC
3782 machines) console (graphic card) which should be used
3783 as the initial boot-console.
3784 See also comment in drivers/video/console/sticore.c.
3787 See comment in drivers/video/console/sticore.c.
3790 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
3792 sunrpc.min_resvport=
3793 sunrpc.max_resvport=
3795 SunRPC servers often require that client requests
3796 originate from a privileged port (i.e. a port in the
3797 range 0 < portnr < 1024).
3798 An administrator who wishes to reserve some of these
3799 ports for other uses may adjust the range that the
3800 kernel's sunrpc client considers to be privileged
3801 using these two parameters to set the minimum and
3802 maximum port values.
3806 Control how the NFS server code allocates CPUs to
3807 service thread pools. Depending on how many NICs
3808 you have and where their interrupts are bound, this
3809 option will affect which CPUs will do NFS serving.
3810 Note: this parameter cannot be changed while the
3811 NFS server is running.
3813 auto the server chooses an appropriate mode
3814 automatically using heuristics
3815 global a single global pool contains all CPUs
3816 percpu one pool for each CPU
3817 pernode one pool for each NUMA node (equivalent
3818 to global on non-NUMA machines)
3820 sunrpc.tcp_slot_table_entries=
3821 sunrpc.udp_slot_table_entries=
3823 Sets the upper limit on the number of simultaneous
3824 RPC calls that can be sent from the client to a
3825 server. Increasing these values may allow you to
3826 improve throughput, but will also increase the
3827 amount of memory reserved for use by the client.
3829 suspend.pm_test_delay=
3831 Sets the number of seconds to remain in a suspend test
3832 mode before resuming the system (see
3833 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
3834 is set. Default value is 5.
3837 [KNL] Enable accounting of swap in memory resource
3838 controller if no parameter or 1 is given or disable
3839 it if 0 is given (See Documentation/cgroups/memory.txt)
3841 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
3842 Format: { <int> | force }
3843 <int> -- Number of I/O TLB slabs
3844 force -- force using of bounce buffers even if they
3845 wouldn't be automatically used by the kernel
3849 sysfs.deprecated=0|1 [KNL]
3850 Enable/disable old style sysfs layout for old udev
3851 on older distributions. When this option is enabled
3852 very new udev will not work anymore. When this option
3853 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
3854 in older udev will not work anymore.
3855 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
3856 the kernel configuration.
3858 sysrq_always_enabled
3860 Ignore sysrq setting - this boot parameter will
3861 neutralize any effect of /proc/sys/kernel/sysrq.
3862 Useful for debugging.
3864 tcpmhash_entries= [KNL,NET]
3865 Set the number of tcp_metrics_hash slots.
3866 Default value is 8192 or 16384 depending on total
3867 ram pages. This is used to specify the TCP metrics
3868 cache size. See Documentation/networking/ip-sysctl.txt
3869 "tcp_no_metrics_save" section for more details.
3873 test_suspend= [SUSPEND][,N]
3874 Specify "mem" (for Suspend-to-RAM) or "standby" (for
3875 standby suspend) or "freeze" (for suspend type freeze)
3876 as the system sleep state during system startup with
3877 the optional capability to repeat N number of times.
3878 The system is woken from this state using a
3879 wakeup-capable RTC alarm.
3881 thash_entries= [KNL,NET]
3882 Set number of hash buckets for TCP connection
3884 thermal.act= [HW,ACPI]
3885 -1: disable all active trip points in all thermal zones
3886 <degrees C>: override all lowest active trip points
3888 thermal.crt= [HW,ACPI]
3889 -1: disable all critical trip points in all thermal zones
3890 <degrees C>: override all critical trip points
3892 thermal.nocrt= [HW,ACPI]
3893 Set to disable actions on ACPI thermal zone
3894 critical and hot trip points.
3896 thermal.off= [HW,ACPI]
3897 1: disable ACPI thermal control
3899 thermal.psv= [HW,ACPI]
3900 -1: disable all passive trip points
3901 <degrees C>: override all passive trip points to this
3904 thermal.tzp= [HW,ACPI]
3905 Specify global default ACPI thermal zone polling rate
3906 <deci-seconds>: poll all this frequency
3907 0: no polling (default)
3910 Force threading of all interrupt handlers except those
3911 marked explicitly IRQF_NO_THREAD.
3914 Enable the Transcendent memory driver if built-in.
3916 tmem.cleancache=0|1 [KNL, XEN]
3917 Default is on (1). Disable the usage of the cleancache
3918 API to send anonymous pages to the hypervisor.
3920 tmem.frontswap=0|1 [KNL, XEN]
3921 Default is on (1). Disable the usage of the frontswap
3922 API to send swap pages to the hypervisor. If disabled
3923 the selfballooning and selfshrinking are force disabled.
3925 tmem.selfballooning=0|1 [KNL, XEN]
3926 Default is on (1). Disable the driving of swap pages
3929 tmem.selfshrinking=0|1 [KNL, XEN]
3930 Default is on (1). Partial swapoff that immediately
3931 transfers pages from Xen hypervisor back to the
3932 kernel based on different criteria.
3936 Specify if the kernel should make use of the cpu
3937 topology information if the hardware supports this.
3938 The scheduler will make use of this information and
3939 e.g. base its process migration decisions on it.
3942 topology_updates= [KNL, PPC, NUMA]
3944 Specify if the kernel should ignore (off)
3945 topology updates sent by the hypervisor to this
3950 tpm_suspend_pcr=[HW,TPM]
3951 Format: integer pcr id
3952 Specify that at suspend time, the tpm driver
3953 should extend the specified pcr with zeros,
3954 as a workaround for some chips which fail to
3955 flush the last written pcr on TPM_SaveState.
3956 This will guarantee that all the other pcrs
3959 trace_buf_size=nn[KMG]
3960 [FTRACE] will set tracing buffer size on each cpu.
3962 trace_event=[event-list]
3963 [FTRACE] Set and start specified trace events in order
3964 to facilitate early boot debugging.
3965 See also Documentation/trace/events.txt
3967 trace_options=[option-list]
3968 [FTRACE] Enable or disable tracer options at boot.
3969 The option-list is a comma delimited list of options
3970 that can be enabled or disabled just as if you were
3971 to echo the option name into
3973 /sys/kernel/debug/tracing/trace_options
3975 For example, to enable stacktrace option (to dump the
3976 stack trace of each event), add to the command line:
3978 trace_options=stacktrace
3980 See also Documentation/trace/ftrace.txt "trace options"
3984 Have the tracepoints sent to printk as well as the
3985 tracing ring buffer. This is useful for early boot up
3986 where the system hangs or reboots and does not give the
3987 option for reading the tracing buffer or performing a
3988 ftrace_dump_on_oops.
3990 To turn off having tracepoints sent to printk,
3991 echo 0 > /proc/sys/kernel/tracepoint_printk
3992 Note, echoing 1 into this file without the
3993 tracepoint_printk kernel cmdline option has no effect.
3997 Having tracepoints sent to printk() and activating high
3998 frequency tracepoints such as irq or sched, can cause
3999 the system to live lock.
4002 [FTRACE] enable this option to disable tracing when a
4003 warning is hit. This turns off "tracing_on". Tracing can
4004 be enabled again by echoing '1' into the "tracing_on"
4005 file located in /sys/kernel/debug/tracing/
4007 This option is useful, as it disables the trace before
4008 the WARNING dump is called, which prevents the trace to
4009 be filled with content caused by the warning output.
4011 This option can also be set at run time via the sysctl
4012 option: kernel/traceoff_on_warning
4014 transparent_hugepage=
4016 Format: [always|madvise|never]
4017 Can be used to control the default behavior of the system
4018 with respect to transparent hugepages.
4019 See Documentation/vm/transhuge.txt for more details.
4021 tsc= Disable clocksource stability checks for TSC.
4023 [x86] reliable: mark tsc clocksource as reliable, this
4024 disables clocksource verification at runtime, as well
4025 as the stability checks done at bootup. Used to enable
4026 high-resolution timer mode on older hardware, and in
4027 virtualized environment.
4028 [x86] noirqtime: Do not use TSC to do irq accounting.
4029 Used to run time disable IRQ_TIME_ACCOUNTING on any
4030 platforms where RDTSC is slow and this accounting
4033 turbografx.map[2|3]= [HW,JOY]
4034 TurboGraFX parallel port interface
4036 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
4037 See also Documentation/input/joystick-parport.txt
4039 udbg-immortal [PPC] When debugging early kernel crashes that
4040 happen after console_init() and before a proper
4041 console driver takes over, this boot options might
4042 help "seeing" what's going on.
4044 uhash_entries= [KNL,NET]
4045 Set number of hash buckets for UDP/UDP-Lite connections
4048 [USB] Ignore overcurrent events (default N).
4049 Some badly-designed motherboards generate lots of
4050 bogus events, for ports that aren't wired to
4051 anything. Set this parameter to avoid log spamming.
4052 Note that genuine overcurrent events won't be
4056 [X86] Cause panic on unknown NMI.
4058 usbcore.authorized_default=
4059 [USB] Default USB device authorization:
4060 (default -1 = authorized except for wireless USB,
4061 0 = not authorized, 1 = authorized)
4063 usbcore.autosuspend=
4064 [USB] The autosuspend time delay (in seconds) used
4065 for newly-detected USB devices (default 2). This
4066 is the time required before an idle device will be
4067 autosuspended. Devices for which the delay is set
4068 to a negative value won't be autosuspended at all.
4070 usbcore.usbfs_snoop=
4071 [USB] Set to log all usbfs traffic (default 0 = off).
4073 usbcore.usbfs_snoop_max=
4074 [USB] Maximum number of bytes to snoop in each URB
4077 usbcore.blinkenlights=
4078 [USB] Set to cycle leds on hubs (default 0 = off).
4080 usbcore.old_scheme_first=
4081 [USB] Start with the old device initialization
4082 scheme (default 0 = off).
4084 usbcore.usbfs_memory_mb=
4085 [USB] Memory limit (in MB) for buffers allocated by
4086 usbfs (default = 16, 0 = max = 2047).
4088 usbcore.use_both_schemes=
4089 [USB] Try the other device initialization scheme
4090 if the first one fails (default 1 = enabled).
4092 usbcore.initial_descriptor_timeout=
4093 [USB] Specifies timeout for the initial 64-byte
4094 USB_REQ_GET_DESCRIPTOR request in milliseconds
4095 (default 5000 = 5.0 seconds).
4097 usbcore.nousb [USB] Disable the USB subsystem
4100 [USBHID] The interval which mice are to be polled at.
4102 usb-storage.delay_use=
4103 [UMS] The delay in seconds before a new device is
4104 scanned for Logical Units (default 1).
4107 [UMS] A list of quirks entries to supplement or
4108 override the built-in unusual_devs list. List
4109 entries are separated by commas. Each entry has
4110 the form VID:PID:Flags where VID and PID are Vendor
4111 and Product ID values (4-digit hex numbers) and
4112 Flags is a set of characters, each corresponding
4113 to a common usb-storage quirk flag as follows:
4114 a = SANE_SENSE (collect more than 18 bytes
4116 b = BAD_SENSE (don't collect more than 18
4117 bytes of sense data);
4118 c = FIX_CAPACITY (decrease the reported
4119 device capacity by one sector);
4120 d = NO_READ_DISC_INFO (don't use
4121 READ_DISC_INFO command);
4122 e = NO_READ_CAPACITY_16 (don't use
4123 READ_CAPACITY_16 command);
4124 f = NO_REPORT_OPCODES (don't use report opcodes
4126 g = MAX_SECTORS_240 (don't transfer more than
4127 240 sectors at a time, uas only);
4128 h = CAPACITY_HEURISTICS (decrease the
4129 reported device capacity by one
4130 sector if the number is odd);
4131 i = IGNORE_DEVICE (don't bind to this
4133 j = NO_REPORT_LUNS (don't use report luns
4135 l = NOT_LOCKABLE (don't try to lock and
4136 unlock ejectable media);
4137 m = MAX_SECTORS_64 (don't transfer more
4138 than 64 sectors = 32 KB at a time);
4139 n = INITIAL_READ10 (force a retry of the
4140 initial READ(10) command);
4141 o = CAPACITY_OK (accept the capacity
4142 reported by the device);
4143 p = WRITE_CACHE (the device cache is ON
4145 r = IGNORE_RESIDUE (the device reports
4146 bogus residue values);
4147 s = SINGLE_LUN (the device has only one
4149 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
4150 commands, uas only);
4151 u = IGNORE_UAS (don't bind to the uas driver);
4152 w = NO_WP_DETECT (don't test whether the
4153 medium is write-protected).
4154 Example: quirks=0419:aaf5:rl,0421:0433:rc
4156 user_debug= [KNL,ARM]
4158 See arch/arm/Kconfig.debug help text.
4159 1 - undefined instruction events
4161 4 - invalid data aborts
4164 Example: user_debug=31
4167 [X86] Flags controlling user PTE allocations.
4169 nohigh = do not allocate PTE pages in
4170 HIGHMEM regardless of setting
4174 On X86_32, this is an alias for vdso32=. Otherwise:
4176 vdso=1: enable VDSO (the default)
4177 vdso=0: disable VDSO mapping
4179 vdso32= [X86] Control the 32-bit vDSO
4180 vdso32=1: enable 32-bit VDSO
4181 vdso32=0 or vdso32=2: disable 32-bit VDSO
4183 See the help text for CONFIG_COMPAT_VDSO for more
4184 details. If CONFIG_COMPAT_VDSO is set, the default is
4185 vdso32=0; otherwise, the default is vdso32=1.
4187 For compatibility with older kernels, vdso32=2 is an
4190 Try vdso32=0 if you encounter an error that says:
4191 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
4194 vector=percpu: enable percpu vector domain
4196 video= [FB] Frame buffer configuration
4197 See Documentation/fb/modedb.txt.
4199 video.brightness_switch_enabled= [0,1]
4200 If set to 1, on receiving an ACPI notify event
4201 generated by hotkey, video driver will adjust brightness
4202 level and then send out the event to user space through
4203 the allocated input device; If set to 0, video driver
4204 will only send out the event without touching backlight
4209 [VMMIO] Memory mapped virtio (platform) device.
4211 <size>@<baseaddr>:<irq>[:<id>]
4213 <size> := size (can use standard suffixes
4215 <baseaddr> := physical base address
4216 <irq> := interrupt number (as passed to
4218 <id> := (optional) platform device id
4220 virtio_mmio.device=1K@0x100b0000:48:7
4222 Can be used multiple times for multiple devices.
4224 vga= [BOOT,X86-32] Select a particular video mode
4225 See Documentation/x86/boot.txt and
4226 Documentation/svga.txt.
4227 Use vga=ask for menu.
4228 This is actually a boot loader parameter; the value is
4229 passed to the kernel using a special protocol.
4231 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
4232 size of <nn>. This can be used to increase the
4233 minimum size (128MB on x86). It can also be used to
4234 decrease the size and leave more room for directly
4237 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
4240 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
4243 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
4247 Controls the behavior of vsyscalls (i.e. calls to
4248 fixed addresses of 0xffffffffff600x00 from legacy
4249 code). Most statically-linked binaries and older
4250 versions of glibc use these calls. Because these
4251 functions are at fixed addresses, they make nice
4252 targets for exploits that can control RIP.
4254 emulate [default] Vsyscalls turn into traps and are
4255 emulated reasonably safely.
4257 native Vsyscalls are native syscall instructions.
4258 This is a little bit faster than trapping
4259 and makes a few dynamic recompilers work
4260 better than they would in emulation mode.
4261 It also makes exploits much easier to write.
4263 none Vsyscalls don't work at all. This makes
4264 them quite hard to use for exploits but
4265 might break your system.
4267 vt.color= [VT] Default text color.
4268 Format: 0xYX, X = foreground, Y = background.
4269 Default: 0x07 = light gray on black.
4271 vt.cur_default= [VT] Default cursor shape.
4272 Format: 0xCCBBAA, where AA, BB, and CC are the same as
4273 the parameters of the <Esc>[?A;B;Cc escape sequence;
4274 see VGA-softcursor.txt. Default: 2 = underline.
4276 vt.default_blu= [VT]
4277 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
4278 Change the default blue palette of the console.
4279 This is a 16-member array composed of values
4282 vt.default_grn= [VT]
4283 Format: <green0>,<green1>,<green2>,...,<green15>
4284 Change the default green palette of the console.
4285 This is a 16-member array composed of values
4288 vt.default_red= [VT]
4289 Format: <red0>,<red1>,<red2>,...,<red15>
4290 Change the default red palette of the console.
4291 This is a 16-member array composed of values
4297 Set system-wide default UTF-8 mode for all tty's.
4298 Default is 1, i.e. UTF-8 mode is enabled for all
4299 newly opened terminals.
4301 vt.global_cursor_default=
4304 Set system-wide default for whether a cursor
4305 is shown on new VTs. Default is -1,
4306 i.e. cursors will be created by default unless
4307 overridden by individual drivers. 0 will hide
4308 cursors, 1 will display them.
4310 vt.italic= [VT] Default color for italic text; 0-15.
4313 vt.underline= [VT] Default color for underlined text; 0-15.
4316 watchdog timers [HW,WDT] For information on watchdog timers,
4317 see Documentation/watchdog/watchdog-parameters.txt
4318 or other driver-specific files in the
4319 Documentation/watchdog/ directory.
4321 workqueue.watchdog_thresh=
4322 If CONFIG_WQ_WATCHDOG is configured, workqueue can
4323 warn stall conditions and dump internal state to
4324 help debugging. 0 disables workqueue stall
4325 detection; otherwise, it's the stall threshold
4326 duration in seconds. The default value is 30 and
4327 it can be updated at runtime by writing to the
4328 corresponding sysfs file.
4330 workqueue.disable_numa
4331 By default, all work items queued to unbound
4332 workqueues are affine to the NUMA nodes they're
4333 issued on, which results in better behavior in
4334 general. If NUMA affinity needs to be disabled for
4335 whatever reason, this option can be used. Note
4336 that this also can be controlled per-workqueue for
4337 workqueues visible under /sys/bus/workqueue/.
4339 workqueue.power_efficient
4340 Per-cpu workqueues are generally preferred because
4341 they show better performance thanks to cache
4342 locality; unfortunately, per-cpu workqueues tend to
4343 be more power hungry than unbound workqueues.
4345 Enabling this makes the per-cpu workqueues which
4346 were observed to contribute significantly to power
4347 consumption unbound, leading to measurably lower
4348 power usage at the cost of small performance
4351 The default value of this parameter is determined by
4352 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
4354 workqueue.debug_force_rr_cpu
4355 Workqueue used to implicitly guarantee that work
4356 items queued without explicit CPU specified are put
4357 on the local CPU. This guarantee is no longer true
4358 and while local CPU is still preferred work items
4359 may be put on foreign CPUs. This debug option
4360 forces round-robin CPU selection to flush out
4361 usages which depend on the now broken guarantee.
4362 When enabled, memory and cache locality will be
4365 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
4366 default x2apic cluster mode on platforms
4369 x86_intel_mid_timer= [X86-32,APBT]
4370 Choose timer option for x86 Intel MID platform.
4371 Two valid options are apbt timer only and lapic timer
4372 plus one apbt timer for broadcast timer.
4373 x86_intel_mid_timer=apbt_only | lapic_and_apbt
4375 xen_512gb_limit [KNL,X86-64,XEN]
4376 Restricts the kernel running paravirtualized under Xen
4377 to use only up to 512 GB of RAM. The reason to do so is
4378 crash analysis tools and Xen tools for doing domain
4379 save/restore/migration must be enabled to handle larger
4382 xen_emul_unplug= [HW,X86,XEN]
4383 Unplug Xen emulated devices
4384 Format: [unplug0,][unplug1]
4385 ide-disks -- unplug primary master IDE devices
4386 aux-ide-disks -- unplug non-primary-master IDE devices
4387 nics -- unplug network devices
4388 all -- unplug all emulated devices (NICs and IDE disks)
4389 unnecessary -- unplugging emulated devices is
4390 unnecessary even if the host did not respond to
4392 never -- do not unplug even if version check succeeds
4394 xen_nopvspin [X86,XEN]
4395 Disables the ticketlock slowpath using Xen PV
4399 Disables the PV optimizations forcing the HVM guest to
4400 run as generic HVM guest with no PV drivers.
4402 xirc2ps_cs= [NET,PCMCIA]
4404 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
4406 ______________________________________________________________________
4410 Add more DRM drivers.
projects / linux-2.6-block.git / blob