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 | off | strict | noirq | rsdt |
173 force -- enable ACPI if default was off
174 off -- disable ACPI if default was on
175 noirq -- do not use ACPI for IRQ routing
176 strict -- Be less tolerant of platforms that are not
177 strictly ACPI specification compliant.
178 rsdt -- prefer RSDT over (default) XSDT
179 copy_dsdt -- copy DSDT to memory
180 For ARM64, ONLY "acpi=off" or "acpi=force" are available
182 See also Documentation/power/runtime_pm.txt, pci=noacpi
184 acpi_apic_instance= [ACPI, IOAPIC]
186 2: use 2nd APIC table, if available
187 1,0: use 1st APIC table
190 acpi_backlight= [HW,ACPI]
191 acpi_backlight=vendor
193 If set to vendor, prefer vendor specific driver
194 (e.g. thinkpad_acpi, sony_acpi, etc.) instead
195 of the ACPI video.ko driver.
197 acpi_force_32bit_fadt_addr
198 force FADT to use 32 bit addresses rather than the
199 64 bit X_* addresses. Some firmware have broken 64
200 bit addresses for force ACPI ignore these and use
201 the older legacy 32 bit addresses.
203 acpica_no_return_repair [HW, ACPI]
204 Disable AML predefined validation mechanism
205 This mechanism can repair the evaluation result to make
206 the return objects more ACPI specification compliant.
207 This option is useful for developers to identify the
208 root cause of an AML interpreter issue when the issue
209 has something to do with the repair mechanism.
211 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG]
212 acpi.debug_level= [HW,ACPI,ACPI_DEBUG]
214 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
215 debug output. Bits in debug_layer correspond to a
216 _COMPONENT in an ACPI source file, e.g.,
217 #define _COMPONENT ACPI_PCI_COMPONENT
218 Bits in debug_level correspond to a level in
219 ACPI_DEBUG_PRINT statements, e.g.,
220 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
221 The debug_level mask defaults to "info". See
222 Documentation/acpi/debug.txt for more information about
223 debug layers and levels.
225 Enable processor driver info messages:
226 acpi.debug_layer=0x20000000
227 Enable PCI/PCI interrupt routing info messages:
228 acpi.debug_layer=0x400000
229 Enable AML "Debug" output, i.e., stores to the Debug
230 object while interpreting AML:
231 acpi.debug_layer=0xffffffff acpi.debug_level=0x2
232 Enable all messages related to ACPI hardware:
233 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
235 Some values produce so much output that the system is
236 unusable. The "log_buf_len" parameter may be useful
237 if you need to capture more output.
239 acpi_enforce_resources= [ACPI]
240 { strict | lax | no }
241 Check for resource conflicts between native drivers
242 and ACPI OperationRegions (SystemIO and SystemMemory
243 only). IO ports and memory declared in ACPI might be
244 used by the ACPI subsystem in arbitrary AML code and
245 can interfere with legacy drivers.
246 strict (default): access to resources claimed by ACPI
247 is denied; legacy drivers trying to access reserved
248 resources will fail to bind to device using them.
249 lax: access to resources claimed by ACPI is allowed;
250 legacy drivers trying to access reserved resources
251 will bind successfully but a warning message is logged.
252 no: ACPI OperationRegions are not marked as reserved,
253 no further checks are performed.
255 acpi_force_table_verification [HW,ACPI]
256 Enable table checksum verification during early stage.
257 By default, this is disabled due to x86 early mapping
260 acpi_irq_balance [HW,ACPI]
261 ACPI will balance active IRQs
264 acpi_irq_nobalance [HW,ACPI]
265 ACPI will not move active IRQs (default)
268 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA
269 Format: <irq>,<irq>...
271 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for
273 Format: <irq>,<irq>...
275 acpi_no_auto_serialize [HW,ACPI]
276 Disable auto-serialization of AML methods
277 AML control methods that contain the opcodes to create
278 named objects will be marked as "Serialized" by the
279 auto-serialization feature.
280 This feature is enabled by default.
281 This option allows to turn off the feature.
283 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump
286 acpi_no_static_ssdt [HW,ACPI]
287 Disable installation of static SSDTs at early boot time
288 By default, SSDTs contained in the RSDT/XSDT will be
289 installed automatically and they will appear under
290 /sys/firmware/acpi/tables.
291 This option turns off this feature.
292 Note that specifying this option does not affect
293 dynamic table installation which will install SSDT
294 tables to /sys/firmware/acpi/tables/dynamic.
296 acpi_rsdp= [ACPI,EFI,KEXEC]
297 Pass the RSDP address to the kernel, mostly used
298 on machines running EFI runtime service to boot the
299 second kernel for kdump.
301 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
302 Format: To spoof as Windows 98: ="Microsoft Windows"
304 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead
305 of 2 which is mandated by ACPI 6) as the supported ACPI
306 specification revision (when using this switch, it may
307 be necessary to carry out a cold reboot _twice_ in a
308 row to make it take effect on the platform firmware).
310 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
311 acpi_osi="string1" # add string1
312 acpi_osi="!string2" # remove string2
313 acpi_osi=!* # remove all strings
314 acpi_osi=! # disable all built-in OS vendor
316 acpi_osi= # disable all strings
318 'acpi_osi=!' can be used in combination with single or
319 multiple 'acpi_osi="string1"' to support specific OS
320 vendor string(s). Note that such command can only
321 affect the default state of the OS vendor strings, thus
322 it cannot affect the default state of the feature group
323 strings and the current state of the OS vendor strings,
324 specifying it multiple times through kernel command line
325 is meaningless. This command is useful when one do not
326 care about the state of the feature group strings which
327 should be controlled by the OSPM.
329 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
330 to 'acpi_osi="Windows 2000" acpi_osi=!', they all
331 can make '_OSI("Windows 2000")' TRUE.
333 'acpi_osi=' cannot be used in combination with other
334 'acpi_osi=' command lines, the _OSI method will not
335 exist in the ACPI namespace. NOTE that such command can
336 only affect the _OSI support state, thus specifying it
337 multiple times through kernel command line is also
340 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
343 'acpi_osi=!*' can be used in combination with single or
344 multiple 'acpi_osi="string1"' to support specific
345 string(s). Note that such command can affect the
346 current state of both the OS vendor strings and the
347 feature group strings, thus specifying it multiple times
348 through kernel command line is meaningful. But it may
349 still not able to affect the final state of a string if
350 there are quirks related to this string. This command
351 is useful when one want to control the state of the
352 feature group strings to debug BIOS issues related to
355 1. 'acpi_osi="Module Device" acpi_osi=!*' can make
356 '_OSI("Module Device")' FALSE.
357 2. 'acpi_osi=!* acpi_osi="Module Device"' can make
358 '_OSI("Module Device")' TRUE.
359 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
361 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
363 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
364 they all will make '_OSI("Windows 2000")' TRUE.
367 Override the pmtimer bug detection: force the kernel
368 to assume that this machine's pmtimer latches its value
369 and always returns good values.
371 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
372 Format: { level | edge | high | low }
374 acpi_skip_timer_override [HW,ACPI]
375 Recognize and ignore IRQ0/pin2 Interrupt Override.
376 For broken nForce2 BIOS resulting in XT-PIC timer.
378 acpi_sleep= [HW,ACPI] Sleep options
379 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
380 old_ordering, nonvs, sci_force_enable }
381 See Documentation/power/video.txt for information on
383 s3_beep is for debugging; it makes the PC's speaker beep
384 as soon as the kernel's real-mode entry point is called.
385 s4_nohwsig prevents ACPI hardware signature from being
386 used during resume from hibernation.
387 old_ordering causes the ACPI 1.0 ordering of the _PTS
388 control method, with respect to putting devices into
389 low power states, to be enforced (the ACPI 2.0 ordering
390 of _PTS is used by default).
391 nonvs prevents the kernel from saving/restoring the
392 ACPI NVS memory during suspend/hibernation and resume.
393 sci_force_enable causes the kernel to set SCI_EN directly
394 on resume from S1/S3 (which is against the ACPI spec,
395 but some broken systems don't work without it).
397 acpi_use_timer_override [HW,ACPI]
398 Use timer override. For some broken Nvidia NF5 boards
399 that require a timer override, but don't have HPET
401 add_efi_memmap [EFI; X86] Include EFI memory map in
402 kernel's map of available physical RAM.
405 { off | try_unsupported }
406 off: disable AGP support
407 try_unsupported: try to drive unsupported chipsets
408 (may crash computer or cause data corruption)
411 See Documentation/sound/alsa/alsa-parameters.txt
414 Allow the default userspace alignment fault handler
415 behaviour to be specified. Bit 0 enables warnings,
416 bit 1 enables fixups, and bit 2 sends a segfault.
418 align_va_addr= [X86-64]
419 Align virtual addresses by clearing slice [14:12] when
420 allocating a VMA at process creation time. This option
421 gives you up to 3% performance improvement on AMD F15h
422 machines (where it is enabled by default) for a
423 CPU-intensive style benchmark, and it can vary highly in
424 a microbenchmark depending on workload and compiler.
426 32: only for 32-bit processes
427 64: only for 64-bit processes
428 on: enable for both 32- and 64-bit processes
429 off: disable for both 32- and 64-bit processes
431 alloc_snapshot [FTRACE]
432 Allocate the ftrace snapshot buffer on boot up when the
433 main buffer is allocated. This is handy if debugging
434 and you need to use tracing_snapshot() on boot up, and
435 do not want to use tracing_snapshot_alloc() as it needs
436 to be done where GFP_KERNEL allocations are allowed.
438 amd_iommu= [HW,X86-64]
439 Pass parameters to the AMD IOMMU driver in the system.
441 fullflush - enable flushing of IO/TLB entries when
442 they are unmapped. Otherwise they are
443 flushed before they will be reused, which
445 off - do not initialize any AMD IOMMU found in
447 force_isolation - Force device isolation for all
448 devices. The IOMMU driver is not
449 allowed anymore to lift isolation
450 requirements as needed. This option
451 does not override iommu=pt
453 amd_iommu_dump= [HW,X86-64]
454 Enable AMD IOMMU driver option to dump the ACPI table
455 for AMD IOMMU. With this option enabled, AMD IOMMU
456 driver will print ACPI tables for AMD IOMMU during
457 IOMMU initialization.
459 amijoy.map= [HW,JOY] Amiga joystick support
460 Map of devices attached to JOY0DAT and JOY1DAT
462 See also Documentation/input/joystick.txt
464 analog.map= [HW,JOY] Analog joystick and gamepad support
465 Specifies type or capabilities of an analog joystick
466 connected to one of 16 gameports
467 Format: <type1>,<type2>,..<type16>
470 Power management functions (SPARCstation-4/5 + deriv.)
472 Disable APC CPU standby support. SPARCstation-Fox does
473 not play well with APC CPU idle - disable it if you have
474 APC and your system crashes randomly.
476 apic= [APIC,X86-32] Advanced Programmable Interrupt Controller
477 Change the output verbosity whilst booting
478 Format: { quiet (default) | verbose | debug }
479 Change the amount of debugging information output
480 when initialising the APIC and IO-APIC components.
482 apic_extnmi= [APIC,X86] External NMI delivery setting
483 Format: { bsp (default) | all | none }
484 bsp: External NMI is delivered only to CPU 0
485 all: External NMIs are broadcast to all CPUs as a
487 none: External NMI is masked for all CPUs. This is
488 useful so that a dump capture kernel won't be
492 See Documentation/networking/ipv6.txt.
494 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
495 Limit apic dumping. The parameter defines the maximal
496 number of local apics being dumped. Also it is possible
497 to set it to "all" by meaning -- no limit here.
498 Format: { 1 (default) | 2 | ... | all }.
499 The parameter valid if only apic=debug or
500 apic=verbose is specified.
501 Example: apic=debug show_lapic=all
503 apm= [APM] Advanced Power Management
504 See header of arch/x86/kernel/apm_32.c.
506 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
507 Format: <io>,<irq>,<nodeID>
511 atarimouse= [HW,MOUSE] Atari Mouse
513 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
514 EzKey and similar keyboards
516 atkbd.reset= [HW] Reset keyboard during initialization
518 atkbd.set= [HW] Select keyboard code set
519 Format: <int> (2 = AT (default), 3 = PS/2)
521 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
524 atkbd.softraw= [HW] Choose between synthetic and real raw mode
525 Format: <bool> (0 = real, 1 = synthetic (default))
527 atkbd.softrepeat= [HW]
528 Use software keyboard repeat
530 audit= [KNL] Enable the audit sub-system
531 Format: { "0" | "1" } (0 = disabled, 1 = enabled)
532 0 - kernel audit is disabled and can not be enabled
533 until the next reboot
534 unset - kernel audit is initialized but disabled and
535 will be fully enabled by the userspace auditd.
536 1 - kernel audit is initialized and partially enabled,
537 storing at most audit_backlog_limit messages in
538 RAM until it is fully enabled by the userspace
542 audit_backlog_limit= [KNL] Set the audit queue size limit.
543 Format: <int> (must be >=0)
546 bau= [X86_UV] Enable the BAU on SGI UV. The default
547 behavior is to disable the BAU (i.e. bau=0).
548 Format: { "0" | "1" }
551 unset - Disable the BAU.
553 baycom_epp= [HW,AX25]
556 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
558 See header of drivers/net/hamradio/baycom_par.c.
560 baycom_ser_fdx= [HW,AX25]
561 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
562 Format: <io>,<irq>,<mode>[,<baud>]
563 See header of drivers/net/hamradio/baycom_ser_fdx.c.
565 baycom_ser_hdx= [HW,AX25]
566 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
567 Format: <io>,<irq>,<mode>
568 See header of drivers/net/hamradio/baycom_ser_hdx.c.
570 blkdevparts= Manual partition parsing of block device(s) for
571 embedded devices based on command line input.
572 See Documentation/block/cmdline-partition.txt
574 boot_delay= Milliseconds to delay each printk during boot.
575 Values larger than 10 seconds (10000) are changed to
579 bootmem_debug [KNL] Enable bootmem allocator debug messages.
581 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
582 bttv.radio= Most important insmod options are available as
584 bttv.pll= See Documentation/video4linux/bttv/Insmod-options
587 bulk_remove=off [PPC] This parameter disables the use of the pSeries
588 firmware feature for flushing multiple hpte entries
591 c101= [NET] Moxa C101 synchronous serial card
593 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
594 Sometimes CPU hardware bugs make them report the cache
595 size incorrectly. The kernel will attempt work arounds
596 to fix known problems, but for some CPUs it is not
597 possible to determine what the correct size should be.
598 This option provides an override for these situations.
600 ca_keys= [KEYS] This parameter identifies a specific key(s) on
601 the system trusted keyring to be used for certificate
603 format: { id:<keyid> | builtin }
605 cca= [MIPS] Override the kernel pages' cache coherency
606 algorithm. Accepted values range from 0 to 7
607 inclusive. See arch/mips/include/asm/pgtable-bits.h
608 for platform specific values (SB1, Loongson3 and
611 ccw_timeout_log [S390]
612 See Documentation/s390/CommonIO for details.
614 cgroup_disable= [KNL] Disable a particular controller
615 Format: {name of the controller(s) to disable}
616 The effects of cgroup_disable=foo are:
617 - foo isn't auto-mounted if you mount all cgroups in
619 - foo isn't visible as an individually mountable
621 {Currently only "memory" controller deal with this and
622 cut the overhead, others just disable the usage. So
623 only cgroup_disable=memory is actually worthy}
625 cgroup_no_v1= [KNL] Disable one, multiple, all cgroup controllers in v1
626 Format: { controller[,controller...] | "all" }
627 Like cgroup_disable, but only applies to cgroup v1;
628 the blacklisted controllers remain available in cgroup2.
630 cgroup.memory= [KNL] Pass options to the cgroup memory controller.
632 nosocket -- Disable socket memory accounting.
633 nokmem -- Disable kernel memory accounting.
635 checkreqprot [SELINUX] Set initial checkreqprot flag value.
636 Format: { "0" | "1" }
637 See security/selinux/Kconfig help text.
638 0 -- check protection applied by kernel (includes
639 any implied execute protection).
640 1 -- check protection requested by application.
641 Default value is set via a kernel config option.
642 Value can be changed at runtime via
643 /selinux/checkreqprot.
646 See Documentation/s390/CommonIO for details.
649 Prevents the clock framework from automatically gating
650 clocks that have not been explicitly enabled by a Linux
651 device driver but are enabled in hardware at reset or
652 by the bootloader/firmware. Note that this does not
653 force such clocks to be always-on nor does it reserve
654 those clocks in any way. This parameter is useful for
655 debug and development, but should not be needed on a
656 platform with proper driver support. For more
657 information, see Documentation/clk.txt.
659 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
661 Forces specified clocksource (if available) to be used
662 when calculating gettimeofday(). If specified
663 clocksource is not available, it defaults to PIT.
664 Format: { pit | tsc | cyclone | pmtmr }
666 clocksource= Override the default clocksource
668 Override the default clocksource and use the clocksource
669 with the name specified.
670 Some clocksource names to choose from, depending on
672 [all] jiffies (this is the base, fallback clocksource)
674 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
675 pxa_timer,timer3,32k_counter,timer0_1
677 [X86-32] pit,hpet,tsc;
678 scx200_hrt on Geode; cyclone on IBM x440
686 clearcpuid=BITNUM [X86]
687 Disable CPUID feature X for the kernel. See
688 arch/x86/include/asm/cpufeatures.h for the valid bit
689 numbers. Note the Linux specific bits are not necessarily
690 stable over kernel options, but the vendor specific
692 Also note that user programs calling CPUID directly
693 or using the feature without checking anything
694 will still see it. This just prevents it from
695 being used by the kernel or shown in /proc/cpuinfo.
696 Also note the kernel might malfunction if you disable
699 cma=nn[MG]@[start[MG][-end[MG]]]
701 Sets the size of kernel global memory area for
702 contiguous memory allocations and optionally the
703 placement constraint by the physical address range of
704 memory allocations. A value of 0 disables CMA
705 altogether. For more information, see
706 include/linux/dma-contiguous.h
708 cmo_free_hint= [PPC] Format: { yes | no }
709 Specify whether pages are marked as being inactive
710 when they are freed. This is used in CMO environments
711 to determine OS memory pressure for page stealing by
715 coherent_pool=nn[KMG] [ARM,KNL]
716 Sets the size of memory pool for coherent, atomic dma
717 allocations, by default set to 256K.
719 code_bytes [X86] How many bytes of object code to print
724 com20020= [HW,NET] ARCnet - COM20020 chipset
726 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
728 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
732 ARCnet - COM90xx chipset (memory-mapped buffers)
733 Format: <io>[,<irq>[,<memstart>]]
735 condev= [HW,S390] console device
738 console= [KNL] Output console device and options.
740 tty<n> Use the virtual console device <n>.
744 Use the specified serial port. The options are of
745 the form "bbbbpnf", where "bbbb" is the baud rate,
746 "p" is parity ("n", "o", or "e"), "n" is number of
747 bits, and "f" is flow control ("r" for RTS or
748 omit it). Default is "9600n8".
750 See Documentation/serial-console.txt for more
752 Documentation/networking/netconsole.txt for an
755 uart[8250],io,<addr>[,options]
756 uart[8250],mmio,<addr>[,options]
757 uart[8250],mmio16,<addr>[,options]
758 uart[8250],mmio32,<addr>[,options]
759 uart[8250],0x<addr>[,options]
760 Start an early, polled-mode console on the 8250/16550
761 UART at the specified I/O port or MMIO address,
762 switching to the matching ttyS device later.
763 MMIO inter-register address stride is either 8-bit
764 (mmio), 16-bit (mmio16), or 32-bit (mmio32).
765 If none of [io|mmio|mmio16|mmio32], <addr> is assumed
766 to be equivalent to 'mmio'. 'options' are specified in
767 the same format described for ttyS above; if unspecified,
768 the h/w is not re-initialized.
770 hvc<n> Use the hypervisor console device <n>. This is for
771 both Xen and PowerPC hypervisors.
773 If the device connected to the port is not a TTY but a braille
774 device, prepend "brl," before the device type, for instance
776 For now, only VisioBraille is supported.
778 consoleblank= [KNL] The console blank (screen saver) timeout in
779 seconds. Defaults to 10*60 = 10mins. A value of 0
780 disables the blank timer.
783 [KNL] Change the default value for
784 /proc/<pid>/coredump_filter.
785 See also Documentation/filesystems/proc.txt.
787 cpuidle.off=1 [CPU_IDLE]
788 disable the cpuidle sub-system
791 [X86] Delay for N microsec between assert and de-assert
792 of APIC INIT to start processors. This delay occurs
793 on every CPU online, such as boot, and resume from suspend.
796 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
798 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
800 crashkernel=size[KMG][@offset[KMG]]
801 [KNL] Using kexec, Linux can switch to a 'crash kernel'
802 upon panic. This parameter reserves the physical
803 memory region [offset, offset + size] for that kernel
804 image. If '@offset' is omitted, then a suitable offset
805 is selected automatically. Check
806 Documentation/kdump/kdump.txt for further details.
808 crashkernel=range1:size1[,range2:size2,...][@offset]
809 [KNL] Same as above, but depends on the memory
810 in the running system. The syntax of range is
811 start-[end] where start and end are both
812 a memory unit (amount[KMG]). See also
813 Documentation/kdump/kdump.txt for an example.
815 crashkernel=size[KMG],high
816 [KNL, x86_64] range could be above 4G. Allow kernel
817 to allocate physical memory region from top, so could
818 be above 4G if system have more than 4G ram installed.
819 Otherwise memory region will be allocated below 4G, if
821 It will be ignored if crashkernel=X is specified.
822 crashkernel=size[KMG],low
823 [KNL, x86_64] range under 4G. When crashkernel=X,high
824 is passed, kernel could allocate physical memory region
825 above 4G, that cause second kernel crash on system
826 that require some amount of low memory, e.g. swiotlb
827 requires at least 64M+32K low memory, also enough extra
828 low memory is needed to make sure DMA buffers for 32-bit
829 devices won't run out. Kernel would try to allocate at
830 at least 256M below 4G automatically.
831 This one let user to specify own low range under 4G
832 for second kernel instead.
833 0: to disable low allocation.
834 It will be ignored when crashkernel=X,high is not used
835 or memory reserved is below 4G.
840 cs89x0_media= [HW,NET]
841 Format: { rj45 | aui | bnc }
844 See header of drivers/s390/block/dasd_devmap.c.
846 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
847 (one device per port)
848 Format: <port#>,<type>
849 See also Documentation/input/joystick-parport.txt
851 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
852 time. See Documentation/dynamic-debug-howto.txt for
853 details. Deprecated, see dyndbg.
855 debug [KNL] Enable kernel debugging (events log level).
858 [KNL] verbose self-tests
860 Print debugging info while doing the locking API
862 We default to 0 (no extra messages), setting it to
863 1 will print _a lot_ more information - normally
864 only useful to kernel developers.
866 debug_objects [KNL] Enable object debugging
869 [KNL] Disable object debugging
871 debug_guardpage_minorder=
872 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
873 parameter allows control of the order of pages that will
874 be intentionally kept free (and hence protected) by the
875 buddy allocator. Bigger value increase the probability
876 of catching random memory corruption, but reduce the
877 amount of memory for normal system use. The maximum
878 possible value is MAX_ORDER/2. Setting this parameter
879 to 1 or 2 should be enough to identify most random
880 memory corruption problems caused by bugs in kernel or
881 driver code when a CPU writes to (or reads from) a
882 random memory location. Note that there exists a class
883 of memory corruptions problems caused by buggy H/W or
884 F/W or by drivers badly programing DMA (basically when
885 memory is written at bus level and the CPU MMU is
886 bypassed) which are not detectable by
887 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
888 tracking down these problems.
891 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
892 parameter enables the feature at boot time. In
893 default, it is disabled. We can avoid allocating huge
894 chunk of memory for debug pagealloc if we don't enable
895 it at boot time and the system will work mostly same
896 with the kernel built without CONFIG_DEBUG_PAGEALLOC.
897 on: enable the feature
899 debugpat [X86] Enable PAT debugging
901 decnet.addr= [HW,NET]
902 Format: <area>[,<node>]
903 See also Documentation/networking/decnet.txt.
906 [same as hugepagesz=] The size of the default
907 HugeTLB page size. This is the size represented by
908 the legacy /proc/ hugepages APIs, used for SHM, and
909 default size when mounting hugetlbfs filesystems.
910 Defaults to the default architecture's huge page size
914 Set number of hash buckets for dentry cache.
917 See Documentation/networking/ipv6.txt.
919 disable_cpu_apicid= [X86,APIC,SMP]
921 The number of initial APIC ID for the
922 corresponding CPU to be disabled at boot,
923 mostly used for the kdump 2nd kernel to
924 disable BSP to wake up multiple CPUs without
925 causing system reset or hang due to sending
928 disable_ddw [PPC/PSERIES]
929 Disable Dynamic DMA Window support. Use this if
930 to workaround buggy firmware.
933 See Documentation/networking/ipv6.txt.
935 disable_mtrr_cleanup [X86]
936 The kernel tries to adjust MTRR layout from continuous
937 to discrete, to make X server driver able to add WB
938 entry later. This parameter disables that.
940 disable_mtrr_trim [X86, Intel and AMD only]
941 By default the kernel will trim any uncacheable
942 memory out of your available memory pool based on
943 MTRR settings. This parameter disables that behavior,
944 possibly causing your machine to run very slowly.
946 disable_timer_pin_1 [X86]
947 Disable PIN 1 of APIC timer
948 Can be useful to work around chipset bugs.
950 dis_ucode_ldr [X86] Disable the microcode loader.
952 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
953 this option disables the debugging code at boot.
955 dma_debug_entries=<number>
956 This option allows to tune the number of preallocated
957 entries for DMA-API debugging code. One entry is
958 required per DMA-API allocation. Use this if the
959 DMA-API debugging code disables itself because the
960 architectural default is too low.
962 dma_debug_driver=<driver_name>
963 With this option the DMA-API debugging driver
964 filter feature can be enabled at boot time. Just
965 pass the driver to filter for as the parameter.
966 The filter can be disabled or changed to another
967 driver later using sysfs.
969 drm_kms_helper.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
970 Broken monitors, graphic adapters, KVMs and EDIDless
971 panels may send no or incorrect EDID data sets.
972 This parameter allows to specify an EDID data sets
973 in the /lib/firmware directory that are used instead.
974 Generic built-in EDID data sets are used, if one of
975 edid/1024x768.bin, edid/1280x1024.bin,
976 edid/1680x1050.bin, or edid/1920x1080.bin is given
977 and no file with the same name exists. Details and
978 instructions how to build your own EDID data are
979 available in Documentation/EDID/HOWTO.txt. An EDID
980 data set will only be used for a particular connector,
981 if its name and a colon are prepended to the EDID
982 name. Each connector may use a unique EDID data
983 set by separating the files with a comma. An EDID
984 data set with no connector name will be used for
985 any connectors not explicitly specified.
989 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
990 module.dyndbg[="val"]
991 Enable debug messages at boot time. See
992 Documentation/dynamic-debug-howto.txt for details.
994 nompx [X86] Disables Intel Memory Protection Extensions.
995 See Documentation/x86/intel_mpx.txt for more
996 information about the feature.
998 nopku [X86] Disable Memory Protection Keys CPU feature found
1002 on enable eager fpu restore
1003 off disable eager fpu restore
1004 auto selects the default scheme, which automatically
1005 enables eagerfpu restore for xsaveopt.
1007 module.async_probe [KNL]
1008 Enable asynchronous probe on this module.
1010 early_ioremap_debug [KNL]
1011 Enable debug messages in early_ioremap support. This
1012 is useful for tracking down temporary early mappings
1013 which are not unmapped.
1015 earlycon= [KNL] Output early console device and options.
1017 When used with no options, the early console is
1018 determined by the stdout-path property in device
1022 Start an early, polled-mode console on a cadence serial
1023 port at the specified address. The cadence serial port
1024 must already be setup and configured. Options are not
1027 uart[8250],io,<addr>[,options]
1028 uart[8250],mmio,<addr>[,options]
1029 uart[8250],mmio32,<addr>[,options]
1030 uart[8250],mmio32be,<addr>[,options]
1031 uart[8250],0x<addr>[,options]
1032 Start an early, polled-mode console on the 8250/16550
1033 UART at the specified I/O port or MMIO address.
1034 MMIO inter-register address stride is either 8-bit
1035 (mmio) or 32-bit (mmio32 or mmio32be).
1036 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
1037 to be equivalent to 'mmio'. 'options' are specified
1038 in the same format described for "console=ttyS<n>"; if
1039 unspecified, the h/w is not initialized.
1043 Start an early, polled-mode console on a pl011 serial
1044 port at the specified address. The pl011 serial port
1045 must already be setup and configured. Options are not
1046 yet supported. If 'mmio32' is specified, then only
1047 the driver will use only 32-bit accessors to read/write
1048 the device registers.
1051 Start an early, polled-mode console on an msm serial
1052 port at the specified address. The serial port
1053 must already be setup and configured. Options are not
1056 msm_serial_dm,<addr>
1057 Start an early, polled-mode console on an msm serial
1058 dm port at the specified address. The serial port
1059 must already be setup and configured. Options are not
1062 smh Use ARM semihosting calls for early console.
1070 Use early console provided by serial driver available
1071 on Samsung SoCs, requires selecting proper type and
1072 a correct base address of the selected UART port. The
1073 serial port must already be setup and configured.
1074 Options are not yet supported.
1078 Use early console provided by Freescale LP UART driver
1079 found on Freescale Vybrid and QorIQ LS1021A processors.
1080 A valid base address must be provided, and the serial
1081 port must already be setup and configured.
1083 armada3700_uart,<addr>
1084 Start an early, polled-mode console on the
1085 Armada 3700 serial port at the specified
1086 address. The serial port must already be setup
1087 and configured. Options are not yet supported.
1089 earlyprintk= [X86,SH,BLACKFIN,ARM,M68k]
1093 earlyprintk=serial[,ttySn[,baudrate]]
1094 earlyprintk=serial[,0x...[,baudrate]]
1095 earlyprintk=ttySn[,baudrate]
1096 earlyprintk=dbgp[debugController#]
1097 earlyprintk=pciserial,bus:device.function[,baudrate]
1099 earlyprintk is useful when the kernel crashes before
1100 the normal console is initialized. It is not enabled by
1101 default because it has some cosmetic problems.
1103 Append ",keep" to not disable it when the real console
1106 Only one of vga, efi, serial, or usb debug port can
1109 Currently only ttyS0 and ttyS1 may be specified by
1110 name. Other I/O ports may be explicitly specified
1111 on some architectures (x86 and arm at least) by
1112 replacing ttySn with an I/O port address, like this:
1113 earlyprintk=serial,0x1008,115200
1114 You can find the port for a given device in
1115 /proc/tty/driver/serial:
1116 2: uart:ST16650V2 port:00001008 irq:18 ...
1118 Interaction with the standard serial driver is not
1121 The VGA and EFI output is eventually overwritten by
1124 The xen output can only be used by Xen PV guests.
1126 edac_report= [HW,EDAC] Control how to report EDAC event
1127 Format: {"on" | "off" | "force"}
1128 on: enable EDAC to report H/W event. May be overridden
1129 by other higher priority error reporting module.
1130 off: disable H/W event reporting through EDAC.
1131 force: enforce the use of EDAC to report H/W event.
1134 ekgdboc= [X86,KGDB] Allow early kernel console debugging
1137 This is designed to be used in conjunction with
1138 the boot argument: earlyprintk=vga
1141 Format: {"off" | "on" | "skip[mbr]"}
1144 Format: { "old_map", "nochunk", "noruntime", "debug" }
1145 old_map [X86-64]: switch to the old ioremap-based EFI
1146 runtime services mapping. 32-bit still uses this one by
1148 nochunk: disable reading files in "chunks" in the EFI
1149 boot stub, as chunking can cause problems with some
1150 firmware implementations.
1151 noruntime : disable EFI runtime services support
1152 debug: enable misc debug output
1154 efi_no_storage_paranoia [EFI; X86]
1155 Using this parameter you can use more than 50% of
1156 your efi variable storage. Use this parameter only if
1157 you are really sure that your UEFI does sane gc and
1158 fulfills the spec otherwise your board may brick.
1160 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1161 Add arbitrary attribute to specific memory range by
1162 updating original EFI memory map.
1163 Region of memory which aa attribute is added to is
1165 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1166 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1167 attribute is added to range 0x100000000-0x180000000 and
1168 0x10a0000000-0x1120000000.
1170 Using this parameter you can do debugging of EFI memmap
1171 related feature. For example, you can do debugging of
1172 Address Range Mirroring feature even if your box
1175 eisa_irq_edge= [PARISC,HW]
1176 See header of drivers/parisc/eisa.c.
1179 See comment before function elanfreq_setup() in
1180 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1183 Format: {"cfq" | "deadline" | "noop"}
1184 See Documentation/block/cfq-iosched.txt and
1185 Documentation/block/deadline-iosched.txt for details.
1187 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1188 Specifies physical address of start of kernel core
1189 image elf header and optionally the size. Generally
1190 kexec loader will pass this option to capture kernel.
1191 See Documentation/kdump/kdump.txt for details.
1193 enable_mtrr_cleanup [X86]
1194 The kernel tries to adjust MTRR layout from continuous
1195 to discrete, to make X server driver able to add WB
1196 entry later. This parameter enables that.
1198 enable_timer_pin_1 [X86]
1199 Enable PIN 1 of APIC timer
1200 Can be useful to work around chipset bugs
1201 (in particular on some ATI chipsets).
1202 The kernel tries to set a reasonable default.
1204 enforcing [SELINUX] Set initial enforcing status.
1206 See security/selinux/Kconfig help text.
1207 0 -- permissive (log only, no denials).
1208 1 -- enforcing (deny and log).
1210 Value can be changed at runtime via /selinux/enforce.
1213 Disable Error Record Serialization Table (ERST)
1216 ether= [HW,NET] Ethernet cards parameters
1217 This option is obsoleted by the "netdev=" option, which
1218 has equivalent usage. See its documentation for details.
1222 Permit 'security.evm' to be updated regardless of
1223 current integrity status.
1227 fail_make_request=[KNL]
1228 General fault injection mechanism.
1229 Format: <interval>,<probability>,<space>,<times>
1230 See also Documentation/fault-injection/.
1233 See Documentation/blockdev/floppy.txt.
1235 force_pal_cache_flush
1236 [IA-64] Avoid check_sal_cache_flush which may hang on
1237 buggy SAL_CACHE_FLUSH implementations. Using this
1238 parameter will force ia64_sal_cache_flush to call
1239 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1242 Forcefully enable Physical Address Extension (PAE).
1243 Many Pentium M systems disable PAE but may have a
1244 functionally usable PAE implementation.
1245 Warning: use of this parameter will taint the kernel
1246 and may cause unknown problems.
1249 [FTRACE] will set and start the specified tracer
1250 as early as possible in order to facilitate early
1253 ftrace_dump_on_oops[=orig_cpu]
1254 [FTRACE] will dump the trace buffers on oops.
1255 If no parameter is passed, ftrace will dump
1256 buffers of all CPUs, but if you pass orig_cpu, it will
1257 dump only the buffer of the CPU that triggered the
1260 ftrace_filter=[function-list]
1261 [FTRACE] Limit the functions traced by the function
1262 tracer at boot up. function-list is a comma separated
1263 list of functions. This list can be changed at run
1264 time by the set_ftrace_filter file in the debugfs
1267 ftrace_notrace=[function-list]
1268 [FTRACE] Do not trace the functions specified in
1269 function-list. This list can be changed at run time
1270 by the set_ftrace_notrace file in the debugfs
1273 ftrace_graph_filter=[function-list]
1274 [FTRACE] Limit the top level callers functions traced
1275 by the function graph tracer at boot up.
1276 function-list is a comma separated list of functions
1277 that can be changed at run time by the
1278 set_graph_function file in the debugfs tracing directory.
1280 ftrace_graph_notrace=[function-list]
1281 [FTRACE] Do not trace from the functions specified in
1282 function-list. This list is a comma separated list of
1283 functions that can be changed at run time by the
1284 set_graph_notrace file in the debugfs tracing directory.
1287 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1288 support via parallel port (up to 5 devices per port)
1289 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1290 See also Documentation/input/joystick-parport.txt
1294 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART
1298 gcov_persist= [GCOV] When non-zero (default), profiling data for
1299 kernel modules is saved and remains accessible via
1300 debugfs, even when the module is unloaded/reloaded.
1301 When zero, profiling data is discarded and associated
1302 debugfs files are removed at module unload time.
1304 gpt [EFI] Forces disk with valid GPT signature but
1305 invalid Protective MBR to be treated as GPT. If the
1306 primary GPT is corrupted, it enables the backup/alternate
1307 GPT to be used instead.
1309 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1310 the "Enable 0" bit of the configuration register.
1313 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1314 the "Enable 0" bit of the configuration register.
1317 grcan.select= [HW] Select which physical interface to use.
1320 grcan.txsize= [HW] Sets the size of the tx buffer.
1321 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1323 grcan.rxsize= [HW] Sets the size of the rx buffer.
1324 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1327 hardlockup_all_cpu_backtrace=
1328 [KNL] Should the hard-lockup detector generate
1329 backtraces on all cpus.
1332 hashdist= [KNL,NUMA] Large hashes allocated during boot
1333 are distributed across NUMA nodes. Defaults on
1334 for 64-bit NUMA, off otherwise.
1335 Format: 0 | 1 (for off | on)
1337 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1339 hd= [EIDE] (E)IDE hard drive subsystem geometry
1340 Format: <cyl>,<head>,<sect>
1343 Disable Hardware Error Source Table (HEST) support;
1344 corresponding firmware-first mode error processing
1345 logic will be disabled.
1347 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1348 size of <nn>. This works even on boxes that have no
1349 highmem otherwise. This also works to reduce highmem
1350 size on bigger boxes.
1352 highres= [KNL] Enable/disable high resolution timer mode.
1353 Valid parameters: "on", "off"
1357 See Documentation/isdn/README.HiSax.
1361 hpet= [X86-32,HPET] option to control HPET usage
1362 Format: { enable (default) | disable | force |
1364 disable: disable HPET and use PIT instead
1365 force: allow force enabled of undocumented chips (ICH4,
1367 verbose: show contents of HPET registers during setup
1369 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1370 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1372 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
1373 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
1374 On x86-64 and powerpc, this option can be specified
1375 multiple times interleaved with hugepages= to reserve
1376 huge pages of different sizes. Valid pages sizes on
1377 x86-64 are 2M (when the CPU supports "pse") and 1G
1378 (when the CPU supports the "pdpe1gb" cpuinfo flag).
1380 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1381 terminal devices. Valid values: 0..8
1382 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1383 If specified, z/VM IUCV HVC accepts connections
1384 from listed z/VM user IDs only.
1386 hwthread_map= [METAG] Comma-separated list of Linux cpu id to
1387 hardware thread id mappings.
1388 Format: <cpu>:<hwthread>
1391 Do not unregister boot console at start. This is only
1392 useful for debugging when something happens in the window
1393 between unregistering the boot console and initializing
1396 i2c_bus= [HW] Override the default board specific I2C bus speed
1397 or register an additional I2C bus that is not
1398 registered from board initialization code.
1402 i8042.debug [HW] Toggle i8042 debug mode
1403 i8042.unmask_kbd_data
1404 [HW] Enable printing of interrupt data from the KBD port
1405 (disabled by default, and as a pre-condition
1406 requires that i8042.debug=1 be enabled)
1407 i8042.direct [HW] Put keyboard port into non-translated mode
1408 i8042.dumbkbd [HW] Pretend that controller can only read data from
1409 keyboard and cannot control its state
1410 (Don't attempt to blink the leds)
1411 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1412 i8042.nokbd [HW] Don't check/create keyboard port
1413 i8042.noloop [HW] Disable the AUX Loopback command while probing
1415 i8042.nomux [HW] Don't check presence of an active multiplexing
1417 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1419 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1420 i8042.reset [HW] Reset the controller during init and cleanup
1421 i8042.unlock [HW] Unlock (ignore) the keylock
1422 i8042.kbdreset [HW] Reset device connected to KBD port
1426 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
1427 indicates that the driver is running on unsupported
1429 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
1430 does not match list of supported models.
1432 [HW] Report power status in /proc/i8k
1433 (disabled by default)
1434 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
1437 i915.invert_brightness=
1438 [DRM] Invert the sense of the variable that is used to
1439 set the brightness of the panel backlight. Normally a
1440 brightness value of 0 indicates backlight switched off,
1441 and the maximum of the brightness value sets the backlight
1442 to maximum brightness. If this parameter is set to 0
1443 (default) and the machine requires it, or this parameter
1444 is set to 1, a brightness value of 0 sets the backlight
1445 to maximum brightness, and the maximum of the brightness
1446 value switches the backlight off.
1447 -1 -- never invert brightness
1448 0 -- machine default
1449 1 -- force brightness inversion
1452 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1454 ide-core.nodma= [HW] (E)IDE subsystem
1455 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1456 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1457 .cdrom .chs .ignore_cable are additional options
1458 See Documentation/ide/ide.txt.
1460 ide-generic.probe-mask= [HW] (E)IDE subsystem
1462 Probe mask for legacy ISA IDE ports. Depending on
1463 platform up to 6 ports are supported, enabled by
1464 setting corresponding bits in the mask to 1. The
1465 default value is 0x0, which has a special meaning.
1466 On systems that have PCI, it triggers scanning the
1467 PCI bus for the first and the second port, which
1468 are then probed. On systems without PCI the value
1469 of 0x0 enables probing the two first ports as if it
1472 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1473 Claim all unknown PCI IDE storage controllers.
1476 Format: idle=poll, idle=halt, idle=nomwait
1477 Poll forces a polling idle loop that can slightly
1478 improve the performance of waking up a idle CPU, but
1479 will use a lot of power and make the system run hot.
1481 idle=halt: Halt is forced to be used for CPU idle.
1482 In such case C2/C3 won't be used again.
1483 idle=nomwait: Disable mwait for CPU C-states
1485 ieee754= [MIPS] Select IEEE Std 754 conformance mode
1486 Format: { strict | legacy | 2008 | relaxed }
1489 Choose which programs will be accepted for execution
1490 based on the IEEE 754 NaN encoding(s) supported by
1491 the FPU and the NaN encoding requested with the value
1492 of an ELF file header flag individually set by each
1493 binary. Hardware implementations are permitted to
1494 support either or both of the legacy and the 2008 NaN
1497 Available settings are as follows:
1498 strict accept binaries that request a NaN encoding
1499 supported by the FPU
1500 legacy only accept legacy-NaN binaries, if supported
1502 2008 only accept 2008-NaN binaries, if supported
1504 relaxed accept any binaries regardless of whether
1505 supported by the FPU
1507 The FPU emulator is always able to support both NaN
1508 encodings, so if no FPU hardware is present or it has
1509 been disabled with 'nofpu', then the settings of
1510 'legacy' and '2008' strap the emulator accordingly,
1511 'relaxed' straps the emulator for both legacy-NaN and
1512 2008-NaN, whereas 'strict' enables legacy-NaN only on
1513 legacy processors and both NaN encodings on MIPS32 or
1516 The setting for ABS.fmt/NEG.fmt instruction execution
1517 mode generally follows that for the NaN encoding,
1518 except where unsupported by hardware.
1520 ignore_loglevel [KNL]
1521 Ignore loglevel setting - this will print /all/
1522 kernel messages to the console. Useful for debugging.
1523 We also add it as printk module parameter, so users
1524 could change it dynamically, usually by
1525 /sys/module/printk/parameters/ignore_loglevel.
1528 Ignore RLIMIT_DATA setting for data mappings,
1529 print warning at first misuse. Can be changed via
1530 /sys/module/kernel/parameters/ignore_rlimit_data.
1532 ihash_entries= [KNL]
1533 Set number of hash buckets for inode cache.
1535 ima_appraise= [IMA] appraise integrity measurements
1536 Format: { "off" | "enforce" | "fix" | "log" }
1539 ima_appraise_tcb [IMA]
1540 The builtin appraise policy appraises all files
1544 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1548 The list of supported hash algorithms is defined
1549 in crypto/hash_info.h.
1552 The builtin measurement policy to load during IMA
1553 setup. Specyfing "tcb" as the value, measures all
1554 programs exec'd, files mmap'd for exec, and all files
1555 opened with the read mode bit set by either the
1556 effective uid (euid=0) or uid=0.
1559 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
1560 Load a policy which meets the needs of the Trusted
1561 Computing Base. This means IMA will measure all
1562 programs exec'd, files mmap'd for exec, and all files
1563 opened for read by uid=0.
1566 Select one of defined IMA measurements template formats.
1567 Formats: { "ima" | "ima-ng" | "ima-sig" }
1571 [IMA] Define a custom template format.
1572 Format: { "field1|...|fieldN" }
1574 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1575 Format: <min_file_size>
1576 Set the minimal file size for using asynchronous hash.
1577 If left unspecified, ahash usage is disabled.
1579 ahash performance varies for different data sizes on
1580 different crypto accelerators. This option can be used
1581 to achieve the best performance for a particular HW.
1583 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1585 Set hashing buffer size. Default: 4k.
1587 ahash performance varies for different chunk sizes on
1588 different crypto accelerators. This option can be used
1589 to achieve best performance for particular HW.
1593 Run specified binary instead of /sbin/init as init
1596 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1597 for working out where the kernel is dying during
1600 initcall_blacklist= [KNL] Do not execute a comma-separated list of
1601 initcall functions. Useful for debugging built-in
1602 modules and initcalls.
1604 initrd= [BOOT] Specify the location of the initial ramdisk
1606 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1609 int_pln_enable [x86] Enable power limit notification interrupt
1611 integrity_audit=[IMA]
1612 Format: { "0" | "1" }
1613 0 -- basic integrity auditing messages. (Default)
1614 1 -- additional integrity auditing messages.
1616 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
1618 Enable intel iommu driver.
1620 Disable intel iommu driver.
1621 igfx_off [Default Off]
1622 By default, gfx is mapped as normal device. If a gfx
1623 device has a dedicated DMAR unit, the DMAR unit is
1624 bypassed by not enabling DMAR with this option. In
1625 this case, gfx device will use physical address for
1628 With this option iommu will not optimize to look
1629 for io virtual address below 32-bit forcing dual
1630 address cycle on pci bus for cards supporting greater
1631 than 32-bit addressing. The default is to look
1632 for translation below 32-bit and if not available
1633 then look in the higher range.
1634 strict [Default Off]
1635 With this option on every unmap_single operation will
1636 result in a hardware IOTLB flush operation as opposed
1637 to batching them for performance.
1638 sp_off [Default Off]
1639 By default, super page will be supported if Intel IOMMU
1640 has the capability. With this option, super page will
1642 ecs_off [Default Off]
1643 By default, extended context tables will be supported if
1644 the hardware advertises that it has support both for the
1645 extended tables themselves, and also PASID support. With
1646 this option set, extended tables will not be used even
1647 on hardware which claims to support them.
1649 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
1650 0 disables intel_idle and fall back on acpi_idle.
1651 1 to 6 specify maximum depth of C-state.
1655 Do not enable intel_pstate as the default
1656 scaling driver for the supported processors
1658 Enable intel_pstate on systems that prohibit it by default
1659 in favor of acpi-cpufreq. Forcing the intel_pstate driver
1660 instead of acpi-cpufreq may disable platform features, such
1661 as thermal controls and power capping, that rely on ACPI
1662 P-States information being indicated to OSPM and therefore
1663 should be used with caution. This option does not work with
1664 processors that aren't supported by the intel_pstate driver
1665 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
1667 Do not enable hardware P state control (HWP)
1670 Only load intel_pstate on systems which support
1671 hardware P state control (HWP) if available.
1673 intremap= [X86-64, Intel-IOMMU]
1674 on enable Interrupt Remapping (default)
1675 off disable Interrupt Remapping
1676 nosid disable Source ID checking
1678 BIOS x2APIC opt-out request will be ignored
1679 nopost disable Interrupt Posting
1681 iomem= Disable strict checking of access to MMIO memory
1682 strict regions from userspace.
1697 nobypass [PPC/POWERNV]
1698 Disable IOMMU bypass, using IOMMU for PCI devices.
1701 io7= [HW] IO7 for Marvel based alpha systems
1702 See comment before marvel_specify_io7 in
1703 arch/alpha/kernel/core_marvel.c.
1705 io_delay= [X86] I/O delay method
1707 Standard port 0x80 based delay
1709 Alternate port 0xed based delay (needed on some systems)
1711 Simple two microseconds delay
1716 See Documentation/filesystems/nfs/nfsroot.txt.
1718 irqaffinity= [SMP] Set the default irq affinity mask
1720 <cpu number>,...,<cpu number>
1722 <cpu number>-<cpu number>
1723 (must be a positive range in ascending order)
1725 <cpu number>,...,<cpu number>-<cpu number>
1728 When an interrupt is not handled search all handlers
1729 for it. Intended to get systems with badly broken
1733 When an interrupt is not handled search all handlers
1734 for it. Also check all handlers each timer
1735 interrupt. Intended to get systems with badly broken
1739 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
1741 isolcpus= [KNL,SMP] Isolate CPUs from the general scheduler.
1743 <cpu number>,...,<cpu number>
1745 <cpu number>-<cpu number>
1746 (must be a positive range in ascending order)
1748 <cpu number>,...,<cpu number>-<cpu number>
1750 This option can be used to specify one or more CPUs
1751 to isolate from the general SMP balancing and scheduling
1752 algorithms. You can move a process onto or off an
1753 "isolated" CPU via the CPU affinity syscalls or cpuset.
1754 <cpu number> begins at 0 and the maximum value is
1755 "number of CPUs in system - 1".
1757 This option is the preferred way to isolate CPUs. The
1758 alternative -- manually setting the CPU mask of all
1759 tasks in the system -- can cause problems and
1760 suboptimal load balancer performance.
1764 ivrs_ioapic [HW,X86_64]
1765 Provide an override to the IOAPIC-ID<->DEVICE-ID
1766 mapping provided in the IVRS ACPI table. For
1767 example, to map IOAPIC-ID decimal 10 to
1768 PCI device 00:14.0 write the parameter as:
1769 ivrs_ioapic[10]=00:14.0
1771 ivrs_hpet [HW,X86_64]
1772 Provide an override to the HPET-ID<->DEVICE-ID
1773 mapping provided in the IVRS ACPI table. For
1774 example, to map HPET-ID decimal 0 to
1775 PCI device 00:14.0 write the parameter as:
1776 ivrs_hpet[0]=00:14.0
1778 js= [HW,JOY] Analog joystick
1779 See Documentation/input/joystick.txt.
1782 Enable/disable kernel and module base offset ASLR
1783 (Address Space Layout Randomization) if built into
1784 the kernel. When CONFIG_HIBERNATION is selected,
1785 kASLR is disabled by default. When kASLR is enabled,
1786 hibernation will be disabled.
1790 kernelcore= [KNL,X86,IA-64,PPC]
1791 Format: nn[KMGTPE] | "mirror"
1793 specifies the amount of memory usable by the kernel
1794 for non-movable allocations. The requested amount is
1795 spread evenly throughout all nodes in the system. The
1796 remaining memory in each node is used for Movable
1797 pages. In the event, a node is too small to have both
1798 kernelcore and Movable pages, kernelcore pages will
1799 take priority and other nodes will have a larger number
1800 of Movable pages. The Movable zone is used for the
1801 allocation of pages that may be reclaimed or moved
1802 by the page migration subsystem. This means that
1803 HugeTLB pages may not be allocated from this zone.
1804 Note that allocations like PTEs-from-HighMem still
1805 use the HighMem zone if it exists, and the Normal
1806 zone if it does not.
1808 Instead of specifying the amount of memory (nn[KMGTPE]),
1809 you can specify "mirror" option. In case "mirror"
1810 option is specified, mirrored (reliable) memory is used
1811 for non-movable allocations and remaining memory is used
1812 for Movable pages. nn[KMGTPE] and "mirror" are exclusive,
1813 so you can NOT specify nn[KMGTPE] and "mirror" at the same
1816 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
1817 Format: <Controller#>[,poll interval]
1818 The controller # is the number of the ehci usb debug
1819 port as it is probed via PCI. The poll interval is
1820 optional and is the number seconds in between
1821 each poll cycle to the debug port in case you need
1822 the functionality for interrupting the kernel with
1823 gdb or control-c on the dbgp connection. When
1824 not using this parameter you use sysrq-g to break into
1825 the kernel debugger.
1827 kgdboc= [KGDB,HW] kgdb over consoles.
1828 Requires a tty driver that supports console polling,
1829 or a supported polling keyboard driver (non-usb).
1830 Serial only format: <serial_device>[,baud]
1831 keyboard only format: kbd
1832 keyboard and serial format: kbd,<serial_device>[,baud]
1833 Optional Kernel mode setting:
1834 kms, kbd format: kms,kbd
1835 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
1837 kgdbwait [KGDB] Stop kernel execution and enter the
1838 kernel debugger at the earliest opportunity.
1840 kmac= [MIPS] korina ethernet MAC address.
1841 Configure the RouterBoard 532 series on-chip
1842 Ethernet adapter MAC address.
1844 kmemleak= [KNL] Boot-time kmemleak enable/disable
1845 Valid arguments: on, off
1847 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
1850 kmemcheck= [X86] Boot-time kmemcheck enable/disable/one-shot mode
1851 Valid arguments: 0, 1, 2
1852 kmemcheck=0 (disabled)
1853 kmemcheck=1 (enabled)
1854 kmemcheck=2 (one-shot mode)
1855 Default: 2 (one-shot mode)
1857 kstack=N [X86] Print N words from the kernel stack
1860 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
1861 Default is 0 (don't ignore, but inject #GP)
1863 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit
1867 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
1868 Default is 1 (enabled)
1870 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
1872 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
1874 kvm-intel.ept= [KVM,Intel] Disable extended page tables
1875 (virtualized MMU) support on capable Intel chips.
1876 Default is 1 (enabled)
1878 kvm-intel.emulate_invalid_guest_state=
1879 [KVM,Intel] Enable emulation of invalid guest states
1880 Default is 0 (disabled)
1882 kvm-intel.flexpriority=
1883 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
1884 Default is 1 (enabled)
1887 [KVM,Intel] Enable VMX nesting (nVMX).
1888 Default is 0 (disabled)
1890 kvm-intel.unrestricted_guest=
1891 [KVM,Intel] Disable unrestricted guest feature
1892 (virtualized real and unpaged mode) on capable
1893 Intel chips. Default is 1 (enabled)
1895 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
1896 feature (tagged TLBs) on capable Intel chips.
1897 Default is 1 (enabled)
1903 lapic [X86-32,APIC] Enable the local APIC even if BIOS
1906 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline
1907 value for LAPIC timer one-shot implementation. Default
1908 back to the programmable timer unit in the LAPIC.
1910 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
1913 libata.dma= [LIBATA] DMA control
1914 libata.dma=0 Disable all PATA and SATA DMA
1915 libata.dma=1 PATA and SATA Disk DMA only
1916 libata.dma=2 ATAPI (CDROM) DMA only
1917 libata.dma=4 Compact Flash DMA only
1918 Combinations also work, so libata.dma=3 enables DMA
1919 for disks and CDROMs, but not CFs.
1921 libata.ignore_hpa= [LIBATA] Ignore HPA limit
1922 libata.ignore_hpa=0 keep BIOS limits (default)
1923 libata.ignore_hpa=1 ignore limits, using full disk
1925 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
1929 libata.force= [LIBATA] Force configurations. The format is comma
1930 separated list of "[ID:]VAL" where ID is
1931 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
1932 matching port, link or device. Basically, it matches
1933 the ATA ID string printed on console by libata. If
1934 the whole ID part is omitted, the last PORT and DEVICE
1935 values are used. If ID hasn't been specified yet, the
1936 configuration applies to all ports, links and devices.
1938 If only DEVICE is omitted, the parameter applies to
1939 the port and all links and devices behind it. DEVICE
1940 number of 0 either selects the first device or the
1941 first fan-out link behind PMP device. It does not
1942 select the host link. DEVICE number of 15 selects the
1943 host link and device attached to it.
1945 The VAL specifies the configuration to force. As long
1946 as there's no ambiguity shortcut notation is allowed.
1947 For example, both 1.5 and 1.5G would work for 1.5Gbps.
1948 The following configurations can be forced.
1950 * Cable type: 40c, 80c, short40c, unk, ign or sata.
1951 Any ID with matching PORT is used.
1953 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
1955 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
1956 udma[/][16,25,33,44,66,100,133] notation is also
1959 * [no]ncq: Turn on or off NCQ.
1961 * [no]ncqtrim: Turn off queued DSM TRIM.
1963 * nohrst, nosrst, norst: suppress hard, soft
1966 * rstonce: only attempt one reset during
1967 hot-unplug link recovery
1969 * dump_id: dump IDENTIFY data.
1971 * atapi_dmadir: Enable ATAPI DMADIR bridge support
1973 * disable: Disable this device.
1975 If there are multiple matching configurations changing
1976 the same attribute, the last one is used.
1978 memblock=debug [KNL] Enable memblock debug messages.
1980 load_ramdisk= [RAM] List of ramdisks to load from floppy
1981 See Documentation/blockdev/ramdisk.txt.
1983 lockd.nlm_grace_period=P [NFS] Assign grace period.
1986 lockd.nlm_tcpport=N [NFS] Assign TCP port.
1989 lockd.nlm_timeout=T [NFS] Assign timeout value.
1992 lockd.nlm_udpport=M [NFS] Assign UDP port.
1995 locktorture.nreaders_stress= [KNL]
1996 Set the number of locking read-acquisition kthreads.
1997 Defaults to being automatically set based on the
1998 number of online CPUs.
2000 locktorture.nwriters_stress= [KNL]
2001 Set the number of locking write-acquisition kthreads.
2003 locktorture.onoff_holdoff= [KNL]
2004 Set time (s) after boot for CPU-hotplug testing.
2006 locktorture.onoff_interval= [KNL]
2007 Set time (s) between CPU-hotplug operations, or
2008 zero to disable CPU-hotplug testing.
2010 locktorture.shuffle_interval= [KNL]
2011 Set task-shuffle interval (jiffies). Shuffling
2012 tasks allows some CPUs to go into dyntick-idle
2013 mode during the locktorture test.
2015 locktorture.shutdown_secs= [KNL]
2016 Set time (s) after boot system shutdown. This
2017 is useful for hands-off automated testing.
2019 locktorture.stat_interval= [KNL]
2020 Time (s) between statistics printk()s.
2022 locktorture.stutter= [KNL]
2023 Time (s) to stutter testing, for example,
2024 specifying five seconds causes the test to run for
2025 five seconds, wait for five seconds, and so on.
2026 This tests the locking primitive's ability to
2027 transition abruptly to and from idle.
2029 locktorture.torture_runnable= [BOOT]
2030 Start locktorture running at boot time.
2032 locktorture.torture_type= [KNL]
2033 Specify the locking implementation to test.
2035 locktorture.verbose= [KNL]
2036 Enable additional printk() statements.
2038 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
2041 loglevel= All Kernel Messages with a loglevel smaller than the
2042 console loglevel will be printed to the console. It can
2043 also be changed with klogd or other programs. The
2044 loglevels are defined as follows:
2046 0 (KERN_EMERG) system is unusable
2047 1 (KERN_ALERT) action must be taken immediately
2048 2 (KERN_CRIT) critical conditions
2049 3 (KERN_ERR) error conditions
2050 4 (KERN_WARNING) warning conditions
2051 5 (KERN_NOTICE) normal but significant condition
2052 6 (KERN_INFO) informational
2053 7 (KERN_DEBUG) debug-level messages
2055 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
2056 in bytes. n must be a power of two and greater
2057 than the minimal size. The minimal size is defined
2058 by LOG_BUF_SHIFT kernel config parameter. There is
2059 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
2060 that allows to increase the default size depending on
2061 the number of CPUs. See init/Kconfig for more details.
2063 logo.nologo [FB] Disables display of the built-in Linux logo.
2064 This may be used to provide more screen space for
2065 kernel log messages and is useful when debugging
2066 kernel boot problems.
2068 lp=0 [LP] Specify parallel ports to use, e.g,
2069 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
2070 lp=reset first parallel port). 'lp=0' disables the
2071 lp=auto printer driver. 'lp=reset' (which can be
2072 specified in addition to the ports) causes
2073 attached printers to be reset. Using
2074 lp=port1,port2,... specifies the parallel ports
2075 to associate lp devices with, starting with
2076 lp0. A port specification may be 'none' to skip
2077 that lp device, or a parport name such as
2078 'parport0'. Specifying 'lp=auto' instead of a
2079 port specification list means that device IDs
2080 from each port should be examined, to see if
2081 an IEEE 1284-compliant printer is attached; if
2082 so, the driver will manage that printer.
2083 See also header of drivers/char/lp.c.
2086 Sets loops_per_jiffy to given constant, thus avoiding
2087 time-consuming boot-time autodetection (up to 250 ms per
2088 CPU). 0 enables autodetection (default). To determine
2089 the correct value for your kernel, boot with normal
2090 autodetection and see what value is printed. Note that
2091 on SMP systems the preset will be applied to all CPUs,
2092 which is likely to cause problems if your CPUs need
2093 significantly divergent settings. An incorrect value
2094 will cause delays in the kernel to be wrong, leading to
2095 unpredictable I/O errors and other breakage. Although
2096 unlikely, in the extreme case this might damage your
2100 Format: <io>,<irq>,<dma>
2102 machvec= [IA-64] Force the use of a particular machine-vector
2103 (machvec) in a generic kernel.
2104 Example: machvec=hpzx1_swiotlb
2106 machtype= [Loongson] Share the same kernel image file between different
2108 Example: machtype=lemote-yeeloong-2f-7inch
2110 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
2111 than or equal to this physical address is ignored.
2113 maxcpus= [SMP] Maximum number of processors that an SMP kernel
2114 should make use of. maxcpus=n : n >= 0 limits the
2115 kernel to using 'n' processors. n=0 is a special case,
2116 it is equivalent to "nosmp", which also disables
2119 max_loop= [LOOP] The number of loop block devices that get
2120 (loop.max_loop) unconditionally pre-created at init time. The default
2121 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
2122 of statically allocating a predefined number, loop
2123 devices can be requested on-demand with the
2124 /dev/loop-control interface.
2126 mce [X86-32] Machine Check Exception
2128 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.txt
2130 md= [HW] RAID subsystems devices and level
2131 See Documentation/md.txt.
2134 Format: <first>,<last>
2135 Specifies range of consoles to be captured by the MDA.
2137 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
2138 Amount of memory to be used when the kernel is not able
2139 to see the whole system memory or for test.
2140 [X86] Work as limiting max address. Use together
2141 with memmap= to avoid physical address space collisions.
2142 Without memmap= PCI devices could be placed at addresses
2143 belonging to unused RAM.
2145 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
2149 [KNL,SH] Allow user to override the default size for
2150 per-device physically contiguous DMA buffers.
2152 memmap=exactmap [KNL,X86] Enable setting of an exact
2153 E820 memory map, as specified by the user.
2154 Such memmap=exactmap lines can be constructed based on
2155 BIOS output or other requirements. See the memmap=nn@ss
2158 memmap=nn[KMG]@ss[KMG]
2159 [KNL] Force usage of a specific region of memory.
2160 Region of memory to be used is from ss to ss+nn.
2162 memmap=nn[KMG]#ss[KMG]
2163 [KNL,ACPI] Mark specific memory as ACPI data.
2164 Region of memory to be marked is from ss to ss+nn.
2166 memmap=nn[KMG]$ss[KMG]
2167 [KNL,ACPI] Mark specific memory as reserved.
2168 Region of memory to be reserved is from ss to ss+nn.
2169 Example: Exclude memory from 0x18690000-0x1869ffff
2170 memmap=64K$0x18690000
2172 memmap=0x10000$0x18690000
2174 memmap=nn[KMG]!ss[KMG]
2175 [KNL,X86] Mark specific memory as protected.
2176 Region of memory to be used, from ss to ss+nn.
2177 The memory region may be marked as e820 type 12 (0xc)
2178 and is NVDIMM or ADR memory.
2180 memory_corruption_check=0/1 [X86]
2181 Some BIOSes seem to corrupt the first 64k of
2182 memory when doing things like suspend/resume.
2183 Setting this option will scan the memory
2184 looking for corruption. Enabling this will
2185 both detect corruption and prevent the kernel
2186 from using the memory being corrupted.
2187 However, its intended as a diagnostic tool; if
2188 repeatable BIOS-originated corruption always
2189 affects the same memory, you can use memmap=
2190 to prevent the kernel from using that memory.
2192 memory_corruption_check_size=size [X86]
2193 By default it checks for corruption in the low
2194 64k, making this memory unavailable for normal
2195 use. Use this parameter to scan for
2196 corruption in more or less memory.
2198 memory_corruption_check_period=seconds [X86]
2199 By default it checks for corruption every 60
2200 seconds. Use this parameter to check at some
2201 other rate. 0 disables periodic checking.
2203 memtest= [KNL,X86,ARM] Enable memtest
2205 default : 0 <disable>
2206 Specifies the number of memtest passes to be
2207 performed. Each pass selects another test
2208 pattern from a given set of patterns. Memtest
2209 fills the memory with this pattern, validates
2210 memory contents and reserves bad memory
2211 regions that are detected.
2213 meye.*= [HW] Set MotionEye Camera parameters
2214 See Documentation/video4linux/meye.txt.
2216 mfgpt_irq= [IA-32] Specify the IRQ to use for the
2217 Multi-Function General Purpose Timers on AMD Geode
2220 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
2221 the BIOS has incorrectly applied a workaround. TinyBIOS
2222 version 0.98 is known to be affected, 0.99 fixes the
2223 problem by letting the user disable the workaround.
2227 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
2228 physical address is ignored.
2230 mini2440= [ARM,HW,KNL]
2231 Format:[0..2][b][c][t]
2233 MINI2440 configuration specification:
2234 0 - The attached screen is the 3.5" TFT
2235 1 - The attached screen is the 7" TFT
2236 2 - The VGA Shield is attached (1024x768)
2237 Leaving out the screen size parameter will not load
2238 the TFT driver, and the framebuffer will be left
2240 b - Enable backlight. The TFT backlight pin will be
2241 linked to the kernel VESA blanking code and a GPIO
2242 LED. This parameter is not necessary when using the
2244 c - Enable the s3c camera interface.
2245 t - Reserved for enabling touchscreen support. The
2246 touchscreen support is not enabled in the mainstream
2247 kernel as of 2.6.30, a preliminary port can be found
2248 in the "bleeding edge" mini2440 support kernel at
2249 http://repo.or.cz/w/linux-2.6/mini2440.git
2252 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
2253 parameter allows control of the logging verbosity for
2254 the additional memory initialisation checks. A value
2255 of 0 disables mminit logging and a level of 4 will
2256 log everything. Information is printed at KERN_DEBUG
2257 so loglevel=8 may also need to be specified.
2260 [KNL] When CONFIG_MODULE_SIG is set, this means that
2261 modules without (valid) signatures will fail to load.
2262 Note that if CONFIG_MODULE_SIG_FORCE is set, that
2263 is always true, so this option does nothing.
2266 [MOUSE] Maximum time between finger touching and
2267 leaving touchpad surface for touch to be considered
2268 a tap and be reported as a left button click (for
2269 touchpads working in absolute mode only).
2271 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
2272 reporting absolute coordinates, such as tablets
2273 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
2274 reporting absolute coordinates, such as tablets
2276 movablecore=nn[KMG] [KNL,X86,IA-64,PPC] This parameter
2277 is similar to kernelcore except it specifies the
2278 amount of memory used for migratable allocations.
2279 If both kernelcore and movablecore is specified,
2280 then kernelcore will be at *least* the specified
2281 value but may be more. If movablecore on its own
2282 is specified, the administrator must be careful
2283 that the amount of memory usable for all allocations
2286 movable_node [KNL,X86] Boot-time switch to enable the effects
2287 of CONFIG_MOVABLE_NODE=y. See mm/Kconfig for details.
2289 MTD_Partition= [MTD]
2290 Format: <name>,<region-number>,<size>,<offset>
2292 MTD_Region= [MTD] Format:
2293 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
2296 See drivers/mtd/cmdlinepart.c.
2298 multitce=off [PPC] This parameter disables the use of the pSeries
2299 firmware feature for updating multiple TCE entries
2302 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
2304 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
2306 boundary - index of last SLC block on Flex-OneNAND.
2307 The remaining blocks are configured as MLC blocks.
2308 lock - Configure if Flex-OneNAND boundary should be locked.
2309 Once locked, the boundary cannot be changed.
2310 1 indicates lock status, 0 indicates unlock status.
2313 ARM/S3C2412 JIVE boot control
2315 See arch/arm/mach-s3c2412/mach-jive.c
2317 mtouchusb.raw_coordinates=
2318 [HW] Make the MicroTouch USB driver use raw coordinates
2319 ('y', default) or cooked coordinates ('n')
2321 mtrr_chunk_size=nn[KMG] [X86]
2322 used for mtrr cleanup. It is largest continuous chunk
2323 that could hold holes aka. UC entries.
2325 mtrr_gran_size=nn[KMG] [X86]
2326 Used for mtrr cleanup. It is granularity of mtrr block.
2328 Large value could prevent small alignment from
2331 mtrr_spare_reg_nr=n [X86]
2333 Range: 0,7 : spare reg number
2335 Used for mtrr cleanup. It is spare mtrr entries number.
2336 Set to 2 or more if your graphical card needs more.
2338 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
2340 netdev= [NET] Network devices parameters
2341 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
2342 Note that mem_start is often overloaded to mean
2343 something different and driver-specific.
2344 This usage is only documented in each driver source
2348 [NETFILTER] Enable connection tracking flow accounting
2349 0 to disable accounting
2350 1 to enable accounting
2353 nfsaddrs= [NFS] Deprecated. Use ip= instead.
2354 See Documentation/filesystems/nfs/nfsroot.txt.
2356 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
2357 See Documentation/filesystems/nfs/nfsroot.txt.
2359 nfsrootdebug [NFS] enable nfsroot debugging messages.
2360 See Documentation/filesystems/nfs/nfsroot.txt.
2362 nfs.callback_tcpport=
2363 [NFS] set the TCP port on which the NFSv4 callback
2364 channel should listen.
2367 [NFS] sets the pathname to the program which is used
2368 to update the NFS client cache entries.
2370 nfs.cache_getent_timeout=
2371 [NFS] sets the timeout after which an attempt to
2372 update a cache entry is deemed to have failed.
2374 nfs.idmap_cache_timeout=
2375 [NFS] set the maximum lifetime for idmapper cache
2379 [NFS] enable 64-bit inode numbers.
2380 If zero, the NFS client will fake up a 32-bit inode
2381 number for the readdir() and stat() syscalls instead
2382 of returning the full 64-bit number.
2383 The default is to return 64-bit inode numbers.
2385 nfs.max_session_slots=
2386 [NFSv4.1] Sets the maximum number of session slots
2387 the client will attempt to negotiate with the server.
2388 This limits the number of simultaneous RPC requests
2389 that the client can send to the NFSv4.1 server.
2390 Note that there is little point in setting this
2391 value higher than the max_tcp_slot_table_limit.
2393 nfs.nfs4_disable_idmapping=
2394 [NFSv4] When set to the default of '1', this option
2395 ensures that both the RPC level authentication
2396 scheme and the NFS level operations agree to use
2397 numeric uids/gids if the mount is using the
2398 'sec=sys' security flavour. In effect it is
2399 disabling idmapping, which can make migration from
2400 legacy NFSv2/v3 systems to NFSv4 easier.
2401 Servers that do not support this mode of operation
2402 will be autodetected by the client, and it will fall
2403 back to using the idmapper.
2404 To turn off this behaviour, set the value to '0'.
2406 [NFS4] Specify an additional fixed unique ident-
2407 ification string that NFSv4 clients can insert into
2408 their nfs_client_id4 string. This is typically a
2409 UUID that is generated at system install time.
2411 nfs.send_implementation_id =
2412 [NFSv4.1] Send client implementation identification
2413 information in exchange_id requests.
2414 If zero, no implementation identification information
2416 The default is to send the implementation identification
2419 nfs.recover_lost_locks =
2420 [NFSv4] Attempt to recover locks that were lost due
2421 to a lease timeout on the server. Please note that
2422 doing this risks data corruption, since there are
2423 no guarantees that the file will remain unchanged
2424 after the locks are lost.
2425 If you want to enable the kernel legacy behaviour of
2426 attempting to recover these locks, then set this
2428 The default parameter value of '0' causes the kernel
2429 not to attempt recovery of lost locks.
2431 nfs4.layoutstats_timer =
2432 [NFSv4.2] Change the rate at which the kernel sends
2433 layoutstats to the pNFS metadata server.
2435 Setting this to value to 0 causes the kernel to use
2436 whatever value is the default set by the layout
2437 driver. A non-zero value sets the minimum interval
2438 in seconds between layoutstats transmissions.
2440 nfsd.nfs4_disable_idmapping=
2441 [NFSv4] When set to the default of '1', the NFSv4
2442 server will return only numeric uids and gids to
2443 clients using auth_sys, and will accept numeric uids
2444 and gids from such clients. This is intended to ease
2445 migration from NFSv2/v3.
2447 objlayoutdriver.osd_login_prog=
2448 [NFS] [OBJLAYOUT] sets the pathname to the program which
2449 is used to automatically discover and login into new
2450 osd-targets. Please see:
2451 Documentation/filesystems/pnfs.txt for more explanations
2453 nmi_debug= [KNL,AVR32,SH] Specify one or more actions to take
2454 when a NMI is triggered.
2455 Format: [state][,regs][,debounce][,die]
2457 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
2458 Format: [panic,][nopanic,][num]
2460 0 - turn hardlockup detector in nmi_watchdog off
2461 1 - turn hardlockup detector in nmi_watchdog on
2462 When panic is specified, panic when an NMI watchdog
2463 timeout occurs (or 'nopanic' to override the opposite
2464 default). To disable both hard and soft lockup detectors,
2465 please see 'nowatchdog'.
2466 This is useful when you use a panic=... timeout and
2467 need the box quickly up again.
2469 netpoll.carrier_timeout=
2470 [NET] Specifies amount of time (in seconds) that
2471 netpoll should wait for a carrier. By default netpoll
2474 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
2475 emulation library even if a 387 maths coprocessor
2479 [HW] Never suspend the console
2480 Disable suspending of consoles during suspend and
2481 hibernate operations. Once disabled, debugging
2482 messages can reach various consoles while the rest
2483 of the system is being put to sleep (ie, while
2484 debugging driver suspend/resume hooks). This may
2485 not work reliably with all consoles, but is known
2486 to work with serial and VGA consoles.
2487 To facilitate more flexible debugging, we also add
2488 console_suspend, a printk module parameter to control
2489 it. Users could use console_suspend (usually
2490 /sys/module/printk/parameters/console_suspend) to
2491 turn on/off it dynamically.
2493 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
2494 caches in the slab allocator. Saves per-node memory,
2495 but will impact performance.
2499 noapic [SMP,APIC] Tells the kernel to not make use of any
2500 IOAPICs that may be present in the system.
2502 noautogroup Disable scheduler automatic task group creation.
2504 nobats [PPC] Do not use BATs for mapping kernel lowmem
2505 on "Classic" PPC cores.
2509 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
2511 nodelayacct [KNL] Disable per-task delay accounting
2513 nodisconnect [HW,SCSI,M68K] Disables SCSI disconnects.
2515 nodsp [SH] Disable hardware DSP at boot time.
2517 noefi Disable EFI runtime services support.
2522 On X86-32 available only on PAE configured kernels.
2523 noexec=on: enable non-executable mappings (default)
2524 noexec=off: disable non-executable mappings
2527 Disable SMAP (Supervisor Mode Access Prevention)
2528 even if it is supported by processor.
2531 Disable SMEP (Supervisor Mode Execution Prevention)
2532 even if it is supported by processor.
2535 This affects only 32-bit executables.
2536 noexec32=on: enable non-executable mappings (default)
2537 read doesn't imply executable mappings
2538 noexec32=off: disable non-executable mappings
2539 read implies executable mappings
2541 nofpu [MIPS,SH] Disable hardware FPU at boot time.
2543 nofxsr [BUGS=X86-32] Disables x86 floating point extended
2544 register save and restore. The kernel will only save
2545 legacy floating-point registers on task switch.
2547 nohugeiomap [KNL,x86] Disable kernel huge I/O mappings.
2549 noxsave [BUGS=X86] Disables x86 extended register state save
2550 and restore using xsave. The kernel will fallback to
2551 enabling legacy floating-point and sse state.
2553 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
2554 register states. The kernel will fall back to use
2555 xsave to save the states. By using this parameter,
2556 performance of saving the states is degraded because
2557 xsave doesn't support modified optimization while
2558 xsaveopt supports it on xsaveopt enabled systems.
2560 noxsaves [X86] Disables xsaves and xrstors used in saving and
2561 restoring x86 extended register state in compacted
2562 form of xsave area. The kernel will fall back to use
2563 xsaveopt and xrstor to save and restore the states
2564 in standard form of xsave area. By using this
2565 parameter, xsave area per process might occupy more
2566 memory on xsaves enabled systems.
2568 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
2569 wfi(ARM) instruction doesn't work correctly and not to
2570 use it. This is also useful when using JTAG debugger.
2572 no_file_caps Tells the kernel not to honor file capabilities. The
2573 only way then for a file to be executed with privilege
2574 is to be setuid root or executed by root.
2576 nohalt [IA-64] Tells the kernel not to use the power saving
2577 function PAL_HALT_LIGHT when idle. This increases
2578 power-consumption. On the positive side, it reduces
2579 interrupt wake-up latency, which may improve performance
2580 in certain environments such as networked servers or
2583 nohibernate [HIBERNATION] Disable hibernation and resume.
2585 nohz= [KNL] Boottime enable/disable dynamic ticks
2586 Valid arguments: on, off
2589 nohz_full= [KNL,BOOT]
2590 In kernels built with CONFIG_NO_HZ_FULL=y, set
2591 the specified list of CPUs whose tick will be stopped
2592 whenever possible. The boot CPU will be forced outside
2593 the range to maintain the timekeeping.
2594 The CPUs in this range must also be included in the
2597 noiotrap [SH] Disables trapped I/O port accesses.
2599 noirqdebug [X86-32] Disables the code which attempts to detect and
2600 disable unhandled interrupt sources.
2602 no_timer_check [X86,APIC] Disables the code which tests for
2603 broken timer IRQ sources.
2605 noisapnp [ISAPNP] Disables ISA PnP code.
2607 noinitrd [RAM] Tells the kernel not to load any configured
2610 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
2612 [Deprecated - use intremap=off]
2616 noinvpcid [X86] Disable the INVPCID cpu feature.
2618 nojitter [IA-64] Disables jitter checking for ITC timers.
2620 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
2622 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
2625 no-steal-acc [X86,KVM] Disable paravirtualized steal time accounting.
2626 steal time is computed, but won't influence scheduler
2629 nolapic [X86-32,APIC] Do not enable or use the local APIC.
2631 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
2633 noltlbs [PPC] Do not use large page/tlb entries for kernel
2634 lowmem mapping on PPC40x and PPC8xx
2636 nomca [IA-64] Disable machine check abort handling
2638 nomce [X86-32] Disable Machine Check Exception
2640 nomfgpt [X86-32] Disable Multi-Function General Purpose
2641 Timer usage (for AMD Geode machines).
2643 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
2644 shutdown the other cpus. Instead use the REBOOT_VECTOR
2647 nomodule Disable module load
2649 nopat [X86] Disable PAT (page attribute table extension of
2650 pagetables) support.
2652 norandmaps Don't use address space randomization. Equivalent to
2653 echo 0 > /proc/sys/kernel/randomize_va_space
2655 noreplace-paravirt [X86,IA-64,PV_OPS] Don't patch paravirt_ops
2657 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
2658 with UP alternatives
2660 nordrand [X86] Disable kernel use of the RDRAND and
2661 RDSEED instructions even if they are supported
2662 by the processor. RDRAND and RDSEED are still
2663 available to user space applications.
2665 noresume [SWSUSP] Disables resume and restores original swap
2668 no-scroll [VGA] Disables scrollback.
2669 This is required for the Braillex ib80-piezo Braille
2670 reader made by F.H. Papenmeier (Germany).
2674 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
2676 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
2677 and disable the IO APIC. legacy for "maxcpus=0".
2679 nosoftlockup [KNL] Disable the soft-lockup detector.
2681 nosync [HW,M68K] Disables sync negotiation for all devices.
2683 notsc [BUGS=X86-32] Disable Time Stamp Counter
2685 nowatchdog [KNL] Disable both lockup detectors, i.e.
2686 soft-lockup and NMI watchdog (hard-lockup).
2690 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
2692 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
2693 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
2694 Some features depend on CPU0. Known dependencies are:
2695 1. Resume from suspend/hibernate depends on CPU0.
2696 Suspend/hibernate will fail if CPU0 is offline and you
2697 need to online CPU0 before suspend/hibernate.
2698 2. PIC interrupts also depend on CPU0. CPU0 can't be
2699 removed if a PIC interrupt is detected.
2700 It's said poweroff/reboot may depend on CPU0 on some
2701 machines although I haven't seen such issues so far
2702 after CPU0 is offline on a few tested machines.
2703 If the dependencies are under your control, you can
2704 turn on cpu0_hotplug.
2706 nptcg= [IA-64] Override max number of concurrent global TLB
2707 purges which is reported from either PAL_VM_SUMMARY or
2710 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
2711 could support. nr_cpus=n : n >= 1 limits the kernel to
2712 supporting 'n' processors. Later in runtime you can not
2713 use hotplug cpu feature to put more cpu back to online.
2714 just like you compile the kernel NR_CPUS=n
2716 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
2718 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing.
2719 Allowed values are enable and disable
2721 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
2722 one of ['zone', 'node', 'default'] can be specified
2723 This can be set from sysctl after boot.
2724 See Documentation/sysctl/vm.txt for details.
2726 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
2727 See Documentation/debugging-via-ohci1394.txt for more
2730 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
2731 Rather than timing out after 20 ms if an EC
2732 command is not properly ACKed, override the length
2733 of the timeout. We have interrupts disabled while
2734 waiting for the ACK, so if this is set too high
2735 interrupts *may* be lost!
2737 omap_mux= [OMAP] Override bootloader pin multiplexing.
2738 Format: <mux_mode0.mode_name=value>...
2739 For example, to override I2C bus2:
2740 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
2742 oprofile.timer= [HW]
2743 Use timer interrupt instead of performance counters
2745 oprofile.cpu_type= Force an oprofile cpu type
2746 This might be useful if you have an older oprofile
2747 userland or if you want common events.
2748 Format: { arch_perfmon }
2749 arch_perfmon: [X86] Force use of architectural
2750 perfmon on Intel CPUs instead of the
2751 CPU specific event set.
2752 timer: [X86] Force use of architectural NMI
2753 timer mode (see also oprofile.timer
2754 for generic hr timer mode)
2755 [s390] Force legacy basic mode sampling
2756 (report cpu_type "timer")
2758 oops=panic Always panic on oopses. Default is to just kill the
2759 process, but there is a small probability of
2760 deadlocking the machine.
2761 This will also cause panics on machine check exceptions.
2762 Useful together with panic=30 to trigger a reboot.
2765 See Documentation/sound/oss/oss-parameters.txt
2767 page_owner= [KNL] Boot-time page_owner enabling option.
2768 Storage of the information about who allocated
2769 each page is disabled in default. With this switch,
2771 on: enable the feature
2773 page_poison= [KNL] Boot-time parameter changing the state of
2774 poisoning on the buddy allocator.
2775 off: turn off poisoning
2776 on: turn on poisoning
2778 panic= [KNL] Kernel behaviour on panic: delay <timeout>
2779 timeout > 0: seconds before rebooting
2780 timeout = 0: wait forever
2781 timeout < 0: reboot immediately
2784 panic_on_warn panic() instead of WARN(). Useful to cause kdump
2787 crash_kexec_post_notifiers
2788 Run kdump after running panic-notifiers and dumping
2789 kmsg. This only for the users who doubt kdump always
2790 succeeds in any situation.
2791 Note that this also increases risks of kdump failure,
2792 because some panic notifiers can make the crashed
2793 kernel more unstable.
2795 parkbd.port= [HW] Parallel port number the keyboard adapter is
2796 connected to, default is 0.
2798 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
2799 0 for XT, 1 for AT (default is AT).
2802 parport= [HW,PPT] Specify parallel ports. 0 disables.
2803 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
2804 Use 'auto' to force the driver to use any
2805 IRQ/DMA settings detected (the default is to
2806 ignore detected IRQ/DMA settings because of
2807 possible conflicts). You can specify the base
2808 address, IRQ, and DMA settings; IRQ and DMA
2809 should be numbers, or 'auto' (for using detected
2810 settings on that particular port), or 'nofifo'
2811 (to avoid using a FIFO even if it is detected).
2812 Parallel ports are assigned in the order they
2813 are specified on the command line, starting
2816 parport_init_mode= [HW,PPT]
2817 Configure VIA parallel port to operate in
2818 a specific mode. This is necessary on Pegasos
2819 computer where firmware has no options for setting
2820 up parallel port mode and sets it to spp.
2821 Currently this function knows 686a and 8231 chips.
2822 Format: [spp|ps2|epp|ecp|ecpepp]
2825 Halt all CPUs after the first oops has been printed for
2826 the specified number of seconds. This is to be used if
2827 your oopses keep scrolling off the screen.
2832 See header of drivers/block/paride/pcd.c.
2833 See also Documentation/blockdev/paride.txt.
2835 pci=option[,option...] [PCI] various PCI subsystem options:
2836 earlydump [X86] dump PCI config space before the kernel
2838 off [X86] don't probe for the PCI bus
2839 bios [X86-32] force use of PCI BIOS, don't access
2840 the hardware directly. Use this if your machine
2841 has a non-standard PCI host bridge.
2842 nobios [X86-32] disallow use of PCI BIOS, only direct
2843 hardware access methods are allowed. Use this
2844 if you experience crashes upon bootup and you
2845 suspect they are caused by the BIOS.
2846 conf1 [X86] Force use of PCI Configuration Access
2847 Mechanism 1 (config address in IO port 0xCF8,
2848 data in IO port 0xCFC, both 32-bit).
2849 conf2 [X86] Force use of PCI Configuration Access
2850 Mechanism 2 (IO port 0xCF8 is an 8-bit port for
2851 the function, IO port 0xCFA, also 8-bit, sets
2852 bus number. The config space is then accessed
2853 through ports 0xC000-0xCFFF).
2854 See http://wiki.osdev.org/PCI for more info
2855 on the configuration access mechanisms.
2856 noaer [PCIE] If the PCIEAER kernel config parameter is
2857 enabled, this kernel boot option can be used to
2858 disable the use of PCIE advanced error reporting.
2859 nodomains [PCI] Disable support for multiple PCI
2860 root domains (aka PCI segments, in ACPI-speak).
2861 nommconf [X86] Disable use of MMCONFIG for PCI
2863 check_enable_amd_mmconf [X86] check for and enable
2864 properly configured MMIO access to PCI
2865 config space on AMD family 10h CPU
2866 nomsi [MSI] If the PCI_MSI kernel config parameter is
2867 enabled, this kernel boot option can be used to
2868 disable the use of MSI interrupts system-wide.
2869 noioapicquirk [APIC] Disable all boot interrupt quirks.
2870 Safety option to keep boot IRQs enabled. This
2871 should never be necessary.
2872 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
2873 primary IO-APIC for bridges that cannot disable
2874 boot IRQs. This fixes a source of spurious IRQs
2875 when the system masks IRQs.
2876 noioapicreroute [APIC] Disable workaround that uses the
2877 boot IRQ equivalent of an IRQ that connects to
2878 a chipset where boot IRQs cannot be disabled.
2879 The opposite of ioapicreroute.
2880 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
2881 routing table. These calls are known to be buggy
2882 on several machines and they hang the machine
2883 when used, but on other computers it's the only
2884 way to get the interrupt routing table. Try
2885 this option if the kernel is unable to allocate
2886 IRQs or discover secondary PCI buses on your
2888 rom [X86] Assign address space to expansion ROMs.
2889 Use with caution as certain devices share
2890 address decoders between ROMs and other
2892 norom [X86] Do not assign address space to
2893 expansion ROMs that do not already have
2894 BIOS assigned address ranges.
2895 nobar [X86] Do not assign address space to the
2896 BARs that weren't assigned by the BIOS.
2897 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
2898 assigned automatically to PCI devices. You can
2899 make the kernel exclude IRQs of your ISA cards
2901 pirqaddr=0xAAAAA [X86] Specify the physical address
2902 of the PIRQ table (normally generated
2903 by the BIOS) if it is outside the
2904 F0000h-100000h range.
2905 lastbus=N [X86] Scan all buses thru bus #N. Can be
2906 useful if the kernel is unable to find your
2907 secondary buses and you want to tell it
2908 explicitly which ones they are.
2909 assign-busses [X86] Always assign all PCI bus
2910 numbers ourselves, overriding
2911 whatever the firmware may have done.
2912 usepirqmask [X86] Honor the possible IRQ mask stored
2913 in the BIOS $PIR table. This is needed on
2914 some systems with broken BIOSes, notably
2915 some HP Pavilion N5400 and Omnibook XE3
2916 notebooks. This will have no effect if ACPI
2917 IRQ routing is enabled.
2918 noacpi [X86] Do not use ACPI for IRQ routing
2919 or for PCI scanning.
2920 use_crs [X86] Use PCI host bridge window information
2921 from ACPI. On BIOSes from 2008 or later, this
2922 is enabled by default. If you need to use this,
2923 please report a bug.
2924 nocrs [X86] Ignore PCI host bridge windows from ACPI.
2925 If you need to use this, please report a bug.
2926 routeirq Do IRQ routing for all PCI devices.
2927 This is normally done in pci_enable_device(),
2928 so this option is a temporary workaround
2929 for broken drivers that don't call it.
2930 skip_isa_align [X86] do not align io start addr, so can
2931 handle more pci cards
2932 firmware [ARM] Do not re-enumerate the bus but instead
2933 just use the configuration from the
2934 bootloader. This is currently used on
2935 IXP2000 systems where the bus has to be
2936 configured a certain way for adjunct CPUs.
2937 noearly [X86] Don't do any early type 1 scanning.
2938 This might help on some broken boards which
2939 machine check when some devices' config space
2940 is read. But various workarounds are disabled
2941 and some IOMMU drivers will not work.
2942 bfsort Sort PCI devices into breadth-first order.
2943 This sorting is done to get a device
2944 order compatible with older (<= 2.4) kernels.
2945 nobfsort Don't sort PCI devices into breadth-first order.
2946 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
2947 tuning and use the BIOS-configured MPS defaults.
2948 pcie_bus_safe Set every device's MPS to the largest value
2949 supported by all devices below the root complex.
2950 pcie_bus_perf Set device MPS to the largest allowable MPS
2951 based on its parent bus. Also set MRRS (Max
2952 Read Request Size) to the largest supported
2953 value (no larger than the MPS that the device
2954 or bus can support) for best performance.
2955 pcie_bus_peer2peer Set every device's MPS to 128B, which
2956 every device is guaranteed to support. This
2957 configuration allows peer-to-peer DMA between
2958 any pair of devices, possibly at the cost of
2959 reduced performance. This also guarantees
2960 that hot-added devices will work.
2961 cbiosize=nn[KMG] The fixed amount of bus space which is
2962 reserved for the CardBus bridge's IO window.
2963 The default value is 256 bytes.
2964 cbmemsize=nn[KMG] The fixed amount of bus space which is
2965 reserved for the CardBus bridge's memory
2966 window. The default value is 64 megabytes.
2969 [<order of align>@][<domain>:]<bus>:<slot>.<func>[; ...]
2970 Specifies alignment and device to reassign
2971 aligned memory resources.
2972 If <order of align> is not specified,
2973 PAGE_SIZE is used as alignment.
2974 PCI-PCI bridge can be specified, if resource
2975 windows need to be expanded.
2976 ecrc= Enable/disable PCIe ECRC (transaction layer
2977 end-to-end CRC checking).
2978 bios: Use BIOS/firmware settings. This is the
2982 hpiosize=nn[KMG] The fixed amount of bus space which is
2983 reserved for hotplug bridge's IO window.
2984 Default size is 256 bytes.
2985 hpmemsize=nn[KMG] The fixed amount of bus space which is
2986 reserved for hotplug bridge's memory window.
2987 Default size is 2 megabytes.
2988 realloc= Enable/disable reallocating PCI bridge resources
2989 if allocations done by BIOS are too small to
2990 accommodate resources required by all child
2992 off: Turn realloc off
2994 realloc same as realloc=on
2995 noari do not use PCIe ARI.
2996 pcie_scan_all Scan all possible PCIe devices. Otherwise we
2997 only look for one device below a PCIe downstream
3000 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
3003 force Enable ASPM even on devices that claim not to support it.
3004 WARNING: Forcing ASPM on may cause system lockups.
3006 pcie_hp= [PCIE] PCI Express Hotplug driver options:
3007 nomsi Do not use MSI for PCI Express Native Hotplug (this
3008 makes all PCIe ports use INTx for hotplug services).
3010 pcie_ports= [PCIE] PCIe ports handling:
3011 auto Ask the BIOS whether or not to use native PCIe services
3012 associated with PCIe ports (PME, hot-plug, AER). Use
3013 them only if that is allowed by the BIOS.
3014 native Use native PCIe services associated with PCIe ports
3016 compat Treat PCIe ports as PCI-to-PCI bridges, disable the PCIe
3019 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
3020 nomsi Do not use MSI for native PCIe PME signaling (this makes
3021 all PCIe root ports use INTx for all services).
3023 pcmv= [HW,PCMCIA] BadgePAD 4
3027 Keep all power-domains already enabled by bootloader on,
3028 even if no driver has claimed them. This is useful
3029 for debug and development, but should not be
3030 needed on a platform with proper driver support.
3033 See Documentation/blockdev/paride.txt.
3035 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
3038 See arch/parisc/kernel/pdc_chassis.c
3040 percpu_alloc= Select which percpu first chunk allocator to use.
3041 Currently supported values are "embed" and "page".
3042 Archs may support subset or none of the selections.
3043 See comments in mm/percpu.c for details on each
3044 allocator. This parameter is primarily for debugging
3045 and performance comparison.
3048 See Documentation/blockdev/paride.txt.
3051 See Documentation/blockdev/paride.txt.
3053 pirq= [SMP,APIC] Manual mp-table setup
3054 See Documentation/x86/i386/IO-APIC.txt.
3056 plip= [PPT,NET] Parallel port network link
3057 Format: { parport<nr> | timid | 0 }
3058 See also Documentation/parport.txt.
3060 pmtmr= [X86] Manual setup of pmtmr I/O Port.
3061 Override pmtimer IOPort with a hex value.
3065 Enable PNP debug messages (depends on the
3066 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
3067 via /sys/module/pnp/parameters/debug. We always show
3068 current resource usage; turning this on also shows
3069 possible settings and some assignment information.
3075 { on | off | curr | res | no-curr | no-res }
3078 [ISAPNP] Exclude IRQs for the autoconfiguration
3081 [ISAPNP] Exclude DMAs for the autoconfiguration
3083 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
3084 Ranges are in pairs (I/O port base and size).
3087 [ISAPNP] Exclude memory regions for the
3089 Ranges are in pairs (memory base and size).
3091 ports= [IP_VS_FTP] IPVS ftp helper module
3093 Up to 8 (IP_VS_APP_MAX_PORTS) ports
3095 Format: <port>,<port>....
3097 ppc_strict_facility_enable
3098 [PPC] This option catches any kernel floating point,
3099 Altivec, VSX and SPE outside of regions specifically
3100 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
3101 There is some performance impact when enabling this.
3103 print-fatal-signals=
3104 [KNL] debug: print fatal signals
3106 If enabled, warn about various signal handling
3107 related application anomalies: too many signals,
3108 too many POSIX.1 timers, fatal signals causing a
3111 If you hit the warning due to signal overflow,
3112 you might want to try "ulimit -i unlimited".
3116 printk.always_kmsg_dump=
3117 Trigger kmsg_dump for cases other than kernel oops or
3119 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3122 printk.time= Show timing data prefixed to each printk message line
3123 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3125 processor.max_cstate= [HW,ACPI]
3126 Limit processor to maximum C-state
3127 max_cstate=9 overrides any DMI blacklist limit.
3129 processor.nocst [HW,ACPI]
3130 Ignore the _CST method to determine C-states,
3131 instead using the legacy FADT method
3133 profile= [KNL] Enable kernel profiling via /proc/profile
3134 Format: [schedule,]<number>
3135 Param: "schedule" - profile schedule points.
3136 Param: <number> - step/bucket size as a power of 2 for
3137 statistical time based profiling.
3138 Param: "sleep" - profile D-state sleeping (millisecs).
3139 Requires CONFIG_SCHEDSTATS
3140 Param: "kvm" - profile VM exits.
3142 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
3144 See Documentation/blockdev/ramdisk.txt.
3146 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
3147 probe for; one of (bare|imps|exps|lifebook|any).
3148 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
3150 psmouse.resetafter= [HW,MOUSE]
3151 Try to reset the device after so many bad packets
3154 [HW,MOUSE] Set desired mouse resolution, in dpi.
3155 psmouse.smartscroll=
3156 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
3157 0 = disabled, 1 = enabled (default).
3159 pstore.backend= Specify the name of the pstore backend to use
3162 See Documentation/blockdev/paride.txt.
3165 [KNL] Number of legacy pty's. Overwrites compiled-in
3168 quiet [KNL] Disable most log messages
3173 See Documentation/md.txt.
3175 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
3176 See Documentation/blockdev/ramdisk.txt.
3179 In kernels built with CONFIG_RCU_NOCB_CPU=y, set
3180 the specified list of CPUs to be no-callback CPUs.
3181 Invocation of these CPUs' RCU callbacks will
3182 be offloaded to "rcuox/N" kthreads created for
3183 that purpose, where "x" is "b" for RCU-bh, "p"
3184 for RCU-preempt, and "s" for RCU-sched, and "N"
3185 is the CPU number. This reduces OS jitter on the
3186 offloaded CPUs, which can be useful for HPC and
3187 real-time workloads. It can also improve energy
3188 efficiency for asymmetric multiprocessors.
3191 Rather than requiring that offloaded CPUs
3192 (specified by rcu_nocbs= above) explicitly
3193 awaken the corresponding "rcuoN" kthreads,
3194 make these kthreads poll for callbacks.
3195 This improves the real-time response for the
3196 offloaded CPUs by relieving them of the need to
3197 wake up the corresponding kthread, but degrades
3198 energy efficiency by requiring that the kthreads
3199 periodically wake up to do the polling.
3201 rcutree.blimit= [KNL]
3202 Set maximum number of finished RCU callbacks to
3203 process in one batch.
3205 rcutree.dump_tree= [KNL]
3206 Dump the structure of the rcu_node combining tree
3207 out at early boot. This is used for diagnostic
3208 purposes, to verify correct tree setup.
3210 rcutree.gp_cleanup_delay= [KNL]
3211 Set the number of jiffies to delay each step of
3212 RCU grace-period cleanup. This only has effect
3213 when CONFIG_RCU_TORTURE_TEST_SLOW_CLEANUP is set.
3215 rcutree.gp_init_delay= [KNL]
3216 Set the number of jiffies to delay each step of
3217 RCU grace-period initialization. This only has
3218 effect when CONFIG_RCU_TORTURE_TEST_SLOW_INIT
3221 rcutree.gp_preinit_delay= [KNL]
3222 Set the number of jiffies to delay each step of
3223 RCU grace-period pre-initialization, that is,
3224 the propagation of recent CPU-hotplug changes up
3225 the rcu_node combining tree. This only has effect
3226 when CONFIG_RCU_TORTURE_TEST_SLOW_PREINIT is set.
3228 rcutree.rcu_fanout_exact= [KNL]
3229 Disable autobalancing of the rcu_node combining
3230 tree. This is used by rcutorture, and might
3231 possibly be useful for architectures having high
3232 cache-to-cache transfer latencies.
3234 rcutree.rcu_fanout_leaf= [KNL]
3235 Change the number of CPUs assigned to each
3236 leaf rcu_node structure. Useful for very
3237 large systems, which will choose the value 64,
3238 and for NUMA systems with large remote-access
3239 latencies, which will choose a value aligned
3240 with the appropriate hardware boundaries.
3242 rcutree.jiffies_till_sched_qs= [KNL]
3243 Set required age in jiffies for a
3244 given grace period before RCU starts
3245 soliciting quiescent-state help from
3246 rcu_note_context_switch().
3248 rcutree.jiffies_till_first_fqs= [KNL]
3249 Set delay from grace-period initialization to
3250 first attempt to force quiescent states.
3251 Units are jiffies, minimum value is zero,
3252 and maximum value is HZ.
3254 rcutree.jiffies_till_next_fqs= [KNL]
3255 Set delay between subsequent attempts to force
3256 quiescent states. Units are jiffies, minimum
3257 value is one, and maximum value is HZ.
3259 rcutree.kthread_prio= [KNL,BOOT]
3260 Set the SCHED_FIFO priority of the RCU per-CPU
3261 kthreads (rcuc/N). This value is also used for
3262 the priority of the RCU boost threads (rcub/N)
3263 and for the RCU grace-period kthreads (rcu_bh,
3264 rcu_preempt, and rcu_sched). If RCU_BOOST is
3265 set, valid values are 1-99 and the default is 1
3266 (the least-favored priority). Otherwise, when
3267 RCU_BOOST is not set, valid values are 0-99 and
3268 the default is zero (non-realtime operation).
3270 rcutree.rcu_nocb_leader_stride= [KNL]
3271 Set the number of NOCB kthread groups, which
3272 defaults to the square root of the number of
3273 CPUs. Larger numbers reduces the wakeup overhead
3274 on the per-CPU grace-period kthreads, but increases
3275 that same overhead on each group's leader.
3277 rcutree.qhimark= [KNL]
3278 Set threshold of queued RCU callbacks beyond which
3279 batch limiting is disabled.
3281 rcutree.qlowmark= [KNL]
3282 Set threshold of queued RCU callbacks below which
3283 batch limiting is re-enabled.
3285 rcutree.rcu_idle_gp_delay= [KNL]
3286 Set wakeup interval for idle CPUs that have
3287 RCU callbacks (RCU_FAST_NO_HZ=y).
3289 rcutree.rcu_idle_lazy_gp_delay= [KNL]
3290 Set wakeup interval for idle CPUs that have
3291 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
3292 Lazy RCU callbacks are those which RCU can
3293 prove do nothing more than free memory.
3295 rcuperf.gp_exp= [KNL]
3296 Measure performance of expedited synchronous
3297 grace-period primitives.
3299 rcuperf.holdoff= [KNL]
3300 Set test-start holdoff period. The purpose of
3301 this parameter is to delay the start of the
3302 test until boot completes in order to avoid
3305 rcuperf.nreaders= [KNL]
3306 Set number of RCU readers. The value -1 selects
3307 N, where N is the number of CPUs. A value
3308 "n" less than -1 selects N-n+1, where N is again
3309 the number of CPUs. For example, -2 selects N
3310 (the number of CPUs), -3 selects N+1, and so on.
3311 A value of "n" less than or equal to -N selects
3314 rcuperf.nwriters= [KNL]
3315 Set number of RCU writers. The values operate
3316 the same as for rcuperf.nreaders.
3317 N, where N is the number of CPUs
3319 rcuperf.perf_runnable= [BOOT]
3320 Start rcuperf running at boot time.
3322 rcuperf.shutdown= [KNL]
3323 Shut the system down after performance tests
3324 complete. This is useful for hands-off automated
3327 rcuperf.perf_type= [KNL]
3328 Specify the RCU implementation to test.
3330 rcuperf.verbose= [KNL]
3331 Enable additional printk() statements.
3333 rcutorture.cbflood_inter_holdoff= [KNL]
3334 Set holdoff time (jiffies) between successive
3335 callback-flood tests.
3337 rcutorture.cbflood_intra_holdoff= [KNL]
3338 Set holdoff time (jiffies) between successive
3339 bursts of callbacks within a given callback-flood
3342 rcutorture.cbflood_n_burst= [KNL]
3343 Set the number of bursts making up a given
3344 callback-flood test. Set this to zero to
3345 disable callback-flood testing.
3347 rcutorture.cbflood_n_per_burst= [KNL]
3348 Set the number of callbacks to be registered
3349 in a given burst of a callback-flood test.
3351 rcutorture.fqs_duration= [KNL]
3352 Set duration of force_quiescent_state bursts
3355 rcutorture.fqs_holdoff= [KNL]
3356 Set holdoff time within force_quiescent_state bursts
3359 rcutorture.fqs_stutter= [KNL]
3360 Set wait time between force_quiescent_state bursts
3363 rcutorture.gp_cond= [KNL]
3364 Use conditional/asynchronous update-side
3365 primitives, if available.
3367 rcutorture.gp_exp= [KNL]
3368 Use expedited update-side primitives, if available.
3370 rcutorture.gp_normal= [KNL]
3371 Use normal (non-expedited) asynchronous
3372 update-side primitives, if available.
3374 rcutorture.gp_sync= [KNL]
3375 Use normal (non-expedited) synchronous
3376 update-side primitives, if available. If all
3377 of rcutorture.gp_cond=, rcutorture.gp_exp=,
3378 rcutorture.gp_normal=, and rcutorture.gp_sync=
3379 are zero, rcutorture acts as if is interpreted
3380 they are all non-zero.
3382 rcutorture.n_barrier_cbs= [KNL]
3383 Set callbacks/threads for rcu_barrier() testing.
3385 rcutorture.nfakewriters= [KNL]
3386 Set number of concurrent RCU writers. These just
3387 stress RCU, they don't participate in the actual
3388 test, hence the "fake".
3390 rcutorture.nreaders= [KNL]
3391 Set number of RCU readers. The value -1 selects
3392 N-1, where N is the number of CPUs. A value
3393 "n" less than -1 selects N-n-2, where N is again
3394 the number of CPUs. For example, -2 selects N
3395 (the number of CPUs), -3 selects N+1, and so on.
3397 rcutorture.object_debug= [KNL]
3398 Enable debug-object double-call_rcu() testing.
3400 rcutorture.onoff_holdoff= [KNL]
3401 Set time (s) after boot for CPU-hotplug testing.
3403 rcutorture.onoff_interval= [KNL]
3404 Set time (s) between CPU-hotplug operations, or
3405 zero to disable CPU-hotplug testing.
3407 rcutorture.shuffle_interval= [KNL]
3408 Set task-shuffle interval (s). Shuffling tasks
3409 allows some CPUs to go into dyntick-idle mode
3410 during the rcutorture test.
3412 rcutorture.shutdown_secs= [KNL]
3413 Set time (s) after boot system shutdown. This
3414 is useful for hands-off automated testing.
3416 rcutorture.stall_cpu= [KNL]
3417 Duration of CPU stall (s) to test RCU CPU stall
3418 warnings, zero to disable.
3420 rcutorture.stall_cpu_holdoff= [KNL]
3421 Time to wait (s) after boot before inducing stall.
3423 rcutorture.stat_interval= [KNL]
3424 Time (s) between statistics printk()s.
3426 rcutorture.stutter= [KNL]
3427 Time (s) to stutter testing, for example, specifying
3428 five seconds causes the test to run for five seconds,
3429 wait for five seconds, and so on. This tests RCU's
3430 ability to transition abruptly to and from idle.
3432 rcutorture.test_boost= [KNL]
3433 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
3434 "Maybe" means test if the RCU implementation
3435 under test support RCU priority boosting.
3437 rcutorture.test_boost_duration= [KNL]
3438 Duration (s) of each individual boost test.
3440 rcutorture.test_boost_interval= [KNL]
3441 Interval (s) between each boost test.
3443 rcutorture.test_no_idle_hz= [KNL]
3444 Test RCU's dyntick-idle handling. See also the
3445 rcutorture.shuffle_interval parameter.
3447 rcutorture.torture_runnable= [BOOT]
3448 Start rcutorture running at boot time.
3450 rcutorture.torture_type= [KNL]
3451 Specify the RCU implementation to test.
3453 rcutorture.verbose= [KNL]
3454 Enable additional printk() statements.
3456 rcupdate.rcu_cpu_stall_suppress= [KNL]
3457 Suppress RCU CPU stall warning messages.
3459 rcupdate.rcu_cpu_stall_timeout= [KNL]
3460 Set timeout for RCU CPU stall warning messages.
3462 rcupdate.rcu_expedited= [KNL]
3463 Use expedited grace-period primitives, for
3464 example, synchronize_rcu_expedited() instead
3465 of synchronize_rcu(). This reduces latency,
3466 but can increase CPU utilization, degrade
3467 real-time latency, and degrade energy efficiency.
3468 No effect on CONFIG_TINY_RCU kernels.
3470 rcupdate.rcu_normal= [KNL]
3471 Use only normal grace-period primitives,
3472 for example, synchronize_rcu() instead of
3473 synchronize_rcu_expedited(). This improves
3474 real-time latency, CPU utilization, and
3475 energy efficiency, but can expose users to
3476 increased grace-period latency. This parameter
3477 overrides rcupdate.rcu_expedited. No effect on
3478 CONFIG_TINY_RCU kernels.
3480 rcupdate.rcu_normal_after_boot= [KNL]
3481 Once boot has completed (that is, after
3482 rcu_end_inkernel_boot() has been invoked), use
3483 only normal grace-period primitives. No effect
3484 on CONFIG_TINY_RCU kernels.
3486 rcupdate.rcu_task_stall_timeout= [KNL]
3487 Set timeout in jiffies for RCU task stall warning
3488 messages. Disable with a value less than or equal
3491 rcupdate.rcu_self_test= [KNL]
3492 Run the RCU early boot self tests
3494 rcupdate.rcu_self_test_bh= [KNL]
3495 Run the RCU bh early boot self tests
3497 rcupdate.rcu_self_test_sched= [KNL]
3498 Run the RCU sched early boot self tests
3502 Run specified binary instead of /init from the ramdisk,
3503 used for early userspace startup. See initrd.
3506 Format (x86 or x86_64):
3507 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
3509 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
3511 Where reboot_mode is one of warm (soft) or cold (hard) or gpio,
3512 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
3513 reboot_force is either force or not specified,
3514 reboot_cpu is s[mp]#### with #### being the processor
3515 to be used for rebooting.
3518 [KNL, SMP] Set scheduler's default relax_domain_level.
3519 See Documentation/cgroups/cpusets.txt.
3521 relative_sleep_states=
3522 [SUSPEND] Use sleep state labeling where the deepest
3523 state available other than hibernation is always "mem".
3524 Format: { "0" | "1" }
3525 0 -- Traditional sleep state labels.
3526 1 -- Relative sleep state labels.
3528 reserve= [KNL,BUGS] Force the kernel to ignore some iomem area
3530 reservetop= [X86-32]
3532 Reserves a hole at the top of the kernel virtual
3537 Set the amount of memory to reserve for BIOS at
3538 the bottom of the address space.
3540 reset_devices [KNL] Force drivers to reset the underlying device
3541 during initialization.
3544 Specify the partition device for software suspend
3546 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
3548 resume_offset= [SWSUSP]
3549 Specify the offset from the beginning of the partition
3550 given by "resume=" at which the swap header is located,
3551 in <PAGE_SIZE> units (needed only for swap files).
3552 See Documentation/power/swsusp-and-swap-files.txt
3554 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
3555 read the resume files
3557 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
3558 Useful for devices that are detected asynchronously
3559 (e.g. USB and MMC devices).
3561 hibernate= [HIBERNATION]
3562 noresume Don't check if there's a hibernation image
3563 present during boot.
3564 nocompress Don't compress/decompress hibernation images.
3565 no Disable hibernation and resume.
3567 retain_initrd [RAM] Keep initrd memory after extraction
3569 rfkill.default_state=
3570 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
3571 etc. communication is blocked by default.
3574 rfkill.master_switch_mode=
3575 0 The "airplane mode" button does nothing.
3576 1 The "airplane mode" button toggles between everything
3577 blocked and the previous configuration.
3578 2 The "airplane mode" button toggles between everything
3579 blocked and everything unblocked.
3581 rhash_entries= [KNL,NET]
3582 Set number of hash buckets for route cache
3584 ro [KNL] Mount root device read-only on boot
3587 on Mark read-only kernel memory as read-only (default).
3588 off Leave read-only kernel memory writable for debugging.
3591 Enable the uart passthrough on the designated usb port
3592 on Rockchip SoCs. When active, the signals of the
3593 debug-uart get routed to the D+ and D- pins of the usb
3594 port and the regular usb controller gets disabled.
3596 root= [KNL] Root filesystem
3597 See name_to_dev_t comment in init/do_mounts.c.
3599 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
3600 mount the root filesystem
3602 rootflags= [KNL] Set root filesystem mount option string
3604 rootfstype= [KNL] Set root filesystem type
3606 rootwait [KNL] Wait (indefinitely) for root device to show up.
3607 Useful for devices that are detected asynchronously
3608 (e.g. USB and MMC devices).
3610 rproc_mem=nn[KMG][@address]
3611 [KNL,ARM,CMA] Remoteproc physical memory block.
3612 Memory area to be used by remote processor image,
3615 rw [KNL] Mount root device read-write on boot
3617 S [KNL] Run init in single mode
3619 s390_iommu= [HW,S390]
3620 Set s390 IOTLB flushing mode
3622 With strict flushing every unmap operation will result in
3623 an IOTLB flush. Default is lazy flushing before reuse,
3627 See drivers/net/irda/sa1100_ir.c.
3629 sbni= [NET] Granch SBNI12 leased line adapter
3631 sched_debug [KNL] Enables verbose scheduler debug messages.
3633 schedstats= [KNL,X86] Enable or disable scheduled statistics.
3634 Allowed values are enable and disable. This feature
3635 incurs a small amount of overhead in the scheduler
3636 but is useful for debugging and performance tuning.
3638 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
3639 xtime_lock contention on larger systems, and/or RCU lock
3640 contention on all systems with CONFIG_MAXSMP set.
3641 Format: { "0" | "1" }
3642 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
3644 Note: increases power consumption, thus should only be
3645 enabled if running jitter sensitive (HPC/RT) workloads.
3647 security= [SECURITY] Choose a security module to enable at boot.
3648 If this boot parameter is not specified, only the first
3649 security module asking for security registration will be
3650 loaded. An invalid security module name will be treated
3651 as if no module has been chosen.
3653 selinux= [SELINUX] Disable or enable SELinux at boot time.
3654 Format: { "0" | "1" }
3655 See security/selinux/Kconfig help text.
3658 Default value is set via kernel config option.
3659 If enabled at boot time, /selinux/disable can be used
3660 later to disable prior to initial policy load.
3662 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
3663 Format: { "0" | "1" }
3664 See security/apparmor/Kconfig help text
3667 Default value is set via kernel config option.
3669 serialnumber [BUGS=X86-32]
3672 Maximal number of shapers.
3674 show_msr= [x86] show boot-time MSR settings
3675 Format: { <integer> }
3676 Show boot-time (BIOS-initialized) MSR settings.
3677 The parameter means the number of CPUs to show,
3678 for example 1 means boot CPU only.
3686 Disable merging of slabs with similar size. May be
3687 necessary if there is some reason to distinguish
3688 allocs to different slabs. Debug options disable
3689 merging on their own.
3690 For more information see Documentation/vm/slub.txt.
3692 slab_max_order= [MM, SLAB]
3693 Determines the maximum allowed order for slabs.
3694 A high setting may cause OOMs due to memory
3695 fragmentation. Defaults to 1 for systems with
3696 more than 32MB of RAM, 0 otherwise.
3698 slub_debug[=options[,slabs]] [MM, SLUB]
3699 Enabling slub_debug allows one to determine the
3700 culprit if slab objects become corrupted. Enabling
3701 slub_debug can create guard zones around objects and
3702 may poison objects when not in use. Also tracks the
3703 last alloc / free. For more information see
3704 Documentation/vm/slub.txt.
3706 slub_max_order= [MM, SLUB]
3707 Determines the maximum allowed order for slabs.
3708 A high setting may cause OOMs due to memory
3709 fragmentation. For more information see
3710 Documentation/vm/slub.txt.
3712 slub_min_objects= [MM, SLUB]
3713 The minimum number of objects per slab. SLUB will
3714 increase the slab order up to slub_max_order to
3715 generate a sufficiently large slab able to contain
3716 the number of objects indicated. The higher the number
3717 of objects the smaller the overhead of tracking slabs
3718 and the less frequently locks need to be acquired.
3719 For more information see Documentation/vm/slub.txt.
3721 slub_min_order= [MM, SLUB]
3722 Determines the minimum page order for slabs. Must be
3723 lower than slub_max_order.
3724 For more information see Documentation/vm/slub.txt.
3726 slub_nomerge [MM, SLUB]
3727 Same with slab_nomerge. This is supported for legacy.
3728 See slab_nomerge for more information.
3731 Format: <io1>[,<io2>[,...,<io8>]]
3733 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
3734 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
3735 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
3736 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
3737 smsc-ircc2.ircc_irq= [HW] IRQ line
3738 smsc-ircc2.ircc_dma= [HW] DMA channel
3739 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
3740 0: Toshiba Satellite 1800 (GP data pin select)
3741 1: Fast pin select (default)
3745 [KNL] Should the soft-lockup detector generate panics.
3748 softlockup_all_cpu_backtrace=
3749 [KNL] Should the soft-lockup detector generate
3750 backtraces on all cpus.
3753 sonypi.*= [HW] Sony Programmable I/O Control Device driver
3754 See Documentation/laptops/sonypi.txt
3756 spia_io_base= [HW,MTD]
3762 Enabled the stack tracer on boot up.
3764 stacktrace_filter=[function-list]
3765 [FTRACE] Limit the functions that the stack tracer
3766 will trace at boot up. function-list is a comma separated
3767 list of functions. This list can be changed at run
3768 time by the stack_trace_filter file in the debugfs
3769 tracing directory. Note, this enables stack tracing
3770 and the stacktrace above is not needed.
3774 Set the STI (builtin display/keyboard on the HP-PARISC
3775 machines) console (graphic card) which should be used
3776 as the initial boot-console.
3777 See also comment in drivers/video/console/sticore.c.
3780 See comment in drivers/video/console/sticore.c.
3783 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
3785 sunrpc.min_resvport=
3786 sunrpc.max_resvport=
3788 SunRPC servers often require that client requests
3789 originate from a privileged port (i.e. a port in the
3790 range 0 < portnr < 1024).
3791 An administrator who wishes to reserve some of these
3792 ports for other uses may adjust the range that the
3793 kernel's sunrpc client considers to be privileged
3794 using these two parameters to set the minimum and
3795 maximum port values.
3799 Control how the NFS server code allocates CPUs to
3800 service thread pools. Depending on how many NICs
3801 you have and where their interrupts are bound, this
3802 option will affect which CPUs will do NFS serving.
3803 Note: this parameter cannot be changed while the
3804 NFS server is running.
3806 auto the server chooses an appropriate mode
3807 automatically using heuristics
3808 global a single global pool contains all CPUs
3809 percpu one pool for each CPU
3810 pernode one pool for each NUMA node (equivalent
3811 to global on non-NUMA machines)
3813 sunrpc.tcp_slot_table_entries=
3814 sunrpc.udp_slot_table_entries=
3816 Sets the upper limit on the number of simultaneous
3817 RPC calls that can be sent from the client to a
3818 server. Increasing these values may allow you to
3819 improve throughput, but will also increase the
3820 amount of memory reserved for use by the client.
3822 suspend.pm_test_delay=
3824 Sets the number of seconds to remain in a suspend test
3825 mode before resuming the system (see
3826 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
3827 is set. Default value is 5.
3830 [KNL] Enable accounting of swap in memory resource
3831 controller if no parameter or 1 is given or disable
3832 it if 0 is given (See Documentation/cgroups/memory.txt)
3834 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
3835 Format: { <int> | force }
3836 <int> -- Number of I/O TLB slabs
3837 force -- force using of bounce buffers even if they
3838 wouldn't be automatically used by the kernel
3842 sysfs.deprecated=0|1 [KNL]
3843 Enable/disable old style sysfs layout for old udev
3844 on older distributions. When this option is enabled
3845 very new udev will not work anymore. When this option
3846 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
3847 in older udev will not work anymore.
3848 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
3849 the kernel configuration.
3851 sysrq_always_enabled
3853 Ignore sysrq setting - this boot parameter will
3854 neutralize any effect of /proc/sys/kernel/sysrq.
3855 Useful for debugging.
3857 tcpmhash_entries= [KNL,NET]
3858 Set the number of tcp_metrics_hash slots.
3859 Default value is 8192 or 16384 depending on total
3860 ram pages. This is used to specify the TCP metrics
3861 cache size. See Documentation/networking/ip-sysctl.txt
3862 "tcp_no_metrics_save" section for more details.
3866 test_suspend= [SUSPEND][,N]
3867 Specify "mem" (for Suspend-to-RAM) or "standby" (for
3868 standby suspend) or "freeze" (for suspend type freeze)
3869 as the system sleep state during system startup with
3870 the optional capability to repeat N number of times.
3871 The system is woken from this state using a
3872 wakeup-capable RTC alarm.
3874 thash_entries= [KNL,NET]
3875 Set number of hash buckets for TCP connection
3877 thermal.act= [HW,ACPI]
3878 -1: disable all active trip points in all thermal zones
3879 <degrees C>: override all lowest active trip points
3881 thermal.crt= [HW,ACPI]
3882 -1: disable all critical trip points in all thermal zones
3883 <degrees C>: override all critical trip points
3885 thermal.nocrt= [HW,ACPI]
3886 Set to disable actions on ACPI thermal zone
3887 critical and hot trip points.
3889 thermal.off= [HW,ACPI]
3890 1: disable ACPI thermal control
3892 thermal.psv= [HW,ACPI]
3893 -1: disable all passive trip points
3894 <degrees C>: override all passive trip points to this
3897 thermal.tzp= [HW,ACPI]
3898 Specify global default ACPI thermal zone polling rate
3899 <deci-seconds>: poll all this frequency
3900 0: no polling (default)
3903 Force threading of all interrupt handlers except those
3904 marked explicitly IRQF_NO_THREAD.
3907 Enable the Transcendent memory driver if built-in.
3909 tmem.cleancache=0|1 [KNL, XEN]
3910 Default is on (1). Disable the usage of the cleancache
3911 API to send anonymous pages to the hypervisor.
3913 tmem.frontswap=0|1 [KNL, XEN]
3914 Default is on (1). Disable the usage of the frontswap
3915 API to send swap pages to the hypervisor. If disabled
3916 the selfballooning and selfshrinking are force disabled.
3918 tmem.selfballooning=0|1 [KNL, XEN]
3919 Default is on (1). Disable the driving of swap pages
3922 tmem.selfshrinking=0|1 [KNL, XEN]
3923 Default is on (1). Partial swapoff that immediately
3924 transfers pages from Xen hypervisor back to the
3925 kernel based on different criteria.
3929 Specify if the kernel should make use of the cpu
3930 topology information if the hardware supports this.
3931 The scheduler will make use of this information and
3932 e.g. base its process migration decisions on it.
3935 topology_updates= [KNL, PPC, NUMA]
3937 Specify if the kernel should ignore (off)
3938 topology updates sent by the hypervisor to this
3943 tpm_suspend_pcr=[HW,TPM]
3944 Format: integer pcr id
3945 Specify that at suspend time, the tpm driver
3946 should extend the specified pcr with zeros,
3947 as a workaround for some chips which fail to
3948 flush the last written pcr on TPM_SaveState.
3949 This will guarantee that all the other pcrs
3952 trace_buf_size=nn[KMG]
3953 [FTRACE] will set tracing buffer size on each cpu.
3955 trace_event=[event-list]
3956 [FTRACE] Set and start specified trace events in order
3957 to facilitate early boot debugging.
3958 See also Documentation/trace/events.txt
3960 trace_options=[option-list]
3961 [FTRACE] Enable or disable tracer options at boot.
3962 The option-list is a comma delimited list of options
3963 that can be enabled or disabled just as if you were
3964 to echo the option name into
3966 /sys/kernel/debug/tracing/trace_options
3968 For example, to enable stacktrace option (to dump the
3969 stack trace of each event), add to the command line:
3971 trace_options=stacktrace
3973 See also Documentation/trace/ftrace.txt "trace options"
3977 Have the tracepoints sent to printk as well as the
3978 tracing ring buffer. This is useful for early boot up
3979 where the system hangs or reboots and does not give the
3980 option for reading the tracing buffer or performing a
3981 ftrace_dump_on_oops.
3983 To turn off having tracepoints sent to printk,
3984 echo 0 > /proc/sys/kernel/tracepoint_printk
3985 Note, echoing 1 into this file without the
3986 tracepoint_printk kernel cmdline option has no effect.
3990 Having tracepoints sent to printk() and activating high
3991 frequency tracepoints such as irq or sched, can cause
3992 the system to live lock.
3995 [FTRACE] enable this option to disable tracing when a
3996 warning is hit. This turns off "tracing_on". Tracing can
3997 be enabled again by echoing '1' into the "tracing_on"
3998 file located in /sys/kernel/debug/tracing/
4000 This option is useful, as it disables the trace before
4001 the WARNING dump is called, which prevents the trace to
4002 be filled with content caused by the warning output.
4004 This option can also be set at run time via the sysctl
4005 option: kernel/traceoff_on_warning
4007 transparent_hugepage=
4009 Format: [always|madvise|never]
4010 Can be used to control the default behavior of the system
4011 with respect to transparent hugepages.
4012 See Documentation/vm/transhuge.txt for more details.
4014 tsc= Disable clocksource stability checks for TSC.
4016 [x86] reliable: mark tsc clocksource as reliable, this
4017 disables clocksource verification at runtime, as well
4018 as the stability checks done at bootup. Used to enable
4019 high-resolution timer mode on older hardware, and in
4020 virtualized environment.
4021 [x86] noirqtime: Do not use TSC to do irq accounting.
4022 Used to run time disable IRQ_TIME_ACCOUNTING on any
4023 platforms where RDTSC is slow and this accounting
4026 turbografx.map[2|3]= [HW,JOY]
4027 TurboGraFX parallel port interface
4029 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
4030 See also Documentation/input/joystick-parport.txt
4032 udbg-immortal [PPC] When debugging early kernel crashes that
4033 happen after console_init() and before a proper
4034 console driver takes over, this boot options might
4035 help "seeing" what's going on.
4037 uhash_entries= [KNL,NET]
4038 Set number of hash buckets for UDP/UDP-Lite connections
4041 [USB] Ignore overcurrent events (default N).
4042 Some badly-designed motherboards generate lots of
4043 bogus events, for ports that aren't wired to
4044 anything. Set this parameter to avoid log spamming.
4045 Note that genuine overcurrent events won't be
4049 [X86] Cause panic on unknown NMI.
4051 usbcore.authorized_default=
4052 [USB] Default USB device authorization:
4053 (default -1 = authorized except for wireless USB,
4054 0 = not authorized, 1 = authorized)
4056 usbcore.autosuspend=
4057 [USB] The autosuspend time delay (in seconds) used
4058 for newly-detected USB devices (default 2). This
4059 is the time required before an idle device will be
4060 autosuspended. Devices for which the delay is set
4061 to a negative value won't be autosuspended at all.
4063 usbcore.usbfs_snoop=
4064 [USB] Set to log all usbfs traffic (default 0 = off).
4066 usbcore.usbfs_snoop_max=
4067 [USB] Maximum number of bytes to snoop in each URB
4070 usbcore.blinkenlights=
4071 [USB] Set to cycle leds on hubs (default 0 = off).
4073 usbcore.old_scheme_first=
4074 [USB] Start with the old device initialization
4075 scheme (default 0 = off).
4077 usbcore.usbfs_memory_mb=
4078 [USB] Memory limit (in MB) for buffers allocated by
4079 usbfs (default = 16, 0 = max = 2047).
4081 usbcore.use_both_schemes=
4082 [USB] Try the other device initialization scheme
4083 if the first one fails (default 1 = enabled).
4085 usbcore.initial_descriptor_timeout=
4086 [USB] Specifies timeout for the initial 64-byte
4087 USB_REQ_GET_DESCRIPTOR request in milliseconds
4088 (default 5000 = 5.0 seconds).
4090 usbcore.nousb [USB] Disable the USB subsystem
4093 [USBHID] The interval which mice are to be polled at.
4095 usb-storage.delay_use=
4096 [UMS] The delay in seconds before a new device is
4097 scanned for Logical Units (default 1).
4100 [UMS] A list of quirks entries to supplement or
4101 override the built-in unusual_devs list. List
4102 entries are separated by commas. Each entry has
4103 the form VID:PID:Flags where VID and PID are Vendor
4104 and Product ID values (4-digit hex numbers) and
4105 Flags is a set of characters, each corresponding
4106 to a common usb-storage quirk flag as follows:
4107 a = SANE_SENSE (collect more than 18 bytes
4109 b = BAD_SENSE (don't collect more than 18
4110 bytes of sense data);
4111 c = FIX_CAPACITY (decrease the reported
4112 device capacity by one sector);
4113 d = NO_READ_DISC_INFO (don't use
4114 READ_DISC_INFO command);
4115 e = NO_READ_CAPACITY_16 (don't use
4116 READ_CAPACITY_16 command);
4117 f = NO_REPORT_OPCODES (don't use report opcodes
4119 g = MAX_SECTORS_240 (don't transfer more than
4120 240 sectors at a time, uas only);
4121 h = CAPACITY_HEURISTICS (decrease the
4122 reported device capacity by one
4123 sector if the number is odd);
4124 i = IGNORE_DEVICE (don't bind to this
4126 j = NO_REPORT_LUNS (don't use report luns
4128 l = NOT_LOCKABLE (don't try to lock and
4129 unlock ejectable media);
4130 m = MAX_SECTORS_64 (don't transfer more
4131 than 64 sectors = 32 KB at a time);
4132 n = INITIAL_READ10 (force a retry of the
4133 initial READ(10) command);
4134 o = CAPACITY_OK (accept the capacity
4135 reported by the device);
4136 p = WRITE_CACHE (the device cache is ON
4138 r = IGNORE_RESIDUE (the device reports
4139 bogus residue values);
4140 s = SINGLE_LUN (the device has only one
4142 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
4143 commands, uas only);
4144 u = IGNORE_UAS (don't bind to the uas driver);
4145 w = NO_WP_DETECT (don't test whether the
4146 medium is write-protected).
4147 Example: quirks=0419:aaf5:rl,0421:0433:rc
4149 user_debug= [KNL,ARM]
4151 See arch/arm/Kconfig.debug help text.
4152 1 - undefined instruction events
4154 4 - invalid data aborts
4157 Example: user_debug=31
4160 [X86] Flags controlling user PTE allocations.
4162 nohigh = do not allocate PTE pages in
4163 HIGHMEM regardless of setting
4167 On X86_32, this is an alias for vdso32=. Otherwise:
4169 vdso=1: enable VDSO (the default)
4170 vdso=0: disable VDSO mapping
4172 vdso32= [X86] Control the 32-bit vDSO
4173 vdso32=1: enable 32-bit VDSO
4174 vdso32=0 or vdso32=2: disable 32-bit VDSO
4176 See the help text for CONFIG_COMPAT_VDSO for more
4177 details. If CONFIG_COMPAT_VDSO is set, the default is
4178 vdso32=0; otherwise, the default is vdso32=1.
4180 For compatibility with older kernels, vdso32=2 is an
4183 Try vdso32=0 if you encounter an error that says:
4184 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
4187 vector=percpu: enable percpu vector domain
4189 video= [FB] Frame buffer configuration
4190 See Documentation/fb/modedb.txt.
4192 video.brightness_switch_enabled= [0,1]
4193 If set to 1, on receiving an ACPI notify event
4194 generated by hotkey, video driver will adjust brightness
4195 level and then send out the event to user space through
4196 the allocated input device; If set to 0, video driver
4197 will only send out the event without touching backlight
4202 [VMMIO] Memory mapped virtio (platform) device.
4204 <size>@<baseaddr>:<irq>[:<id>]
4206 <size> := size (can use standard suffixes
4208 <baseaddr> := physical base address
4209 <irq> := interrupt number (as passed to
4211 <id> := (optional) platform device id
4213 virtio_mmio.device=1K@0x100b0000:48:7
4215 Can be used multiple times for multiple devices.
4217 vga= [BOOT,X86-32] Select a particular video mode
4218 See Documentation/x86/boot.txt and
4219 Documentation/svga.txt.
4220 Use vga=ask for menu.
4221 This is actually a boot loader parameter; the value is
4222 passed to the kernel using a special protocol.
4224 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
4225 size of <nn>. This can be used to increase the
4226 minimum size (128MB on x86). It can also be used to
4227 decrease the size and leave more room for directly
4230 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
4233 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
4236 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
4240 Controls the behavior of vsyscalls (i.e. calls to
4241 fixed addresses of 0xffffffffff600x00 from legacy
4242 code). Most statically-linked binaries and older
4243 versions of glibc use these calls. Because these
4244 functions are at fixed addresses, they make nice
4245 targets for exploits that can control RIP.
4247 emulate [default] Vsyscalls turn into traps and are
4248 emulated reasonably safely.
4250 native Vsyscalls are native syscall instructions.
4251 This is a little bit faster than trapping
4252 and makes a few dynamic recompilers work
4253 better than they would in emulation mode.
4254 It also makes exploits much easier to write.
4256 none Vsyscalls don't work at all. This makes
4257 them quite hard to use for exploits but
4258 might break your system.
4260 vt.color= [VT] Default text color.
4261 Format: 0xYX, X = foreground, Y = background.
4262 Default: 0x07 = light gray on black.
4264 vt.cur_default= [VT] Default cursor shape.
4265 Format: 0xCCBBAA, where AA, BB, and CC are the same as
4266 the parameters of the <Esc>[?A;B;Cc escape sequence;
4267 see VGA-softcursor.txt. Default: 2 = underline.
4269 vt.default_blu= [VT]
4270 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
4271 Change the default blue palette of the console.
4272 This is a 16-member array composed of values
4275 vt.default_grn= [VT]
4276 Format: <green0>,<green1>,<green2>,...,<green15>
4277 Change the default green palette of the console.
4278 This is a 16-member array composed of values
4281 vt.default_red= [VT]
4282 Format: <red0>,<red1>,<red2>,...,<red15>
4283 Change the default red palette of the console.
4284 This is a 16-member array composed of values
4290 Set system-wide default UTF-8 mode for all tty's.
4291 Default is 1, i.e. UTF-8 mode is enabled for all
4292 newly opened terminals.
4294 vt.global_cursor_default=
4297 Set system-wide default for whether a cursor
4298 is shown on new VTs. Default is -1,
4299 i.e. cursors will be created by default unless
4300 overridden by individual drivers. 0 will hide
4301 cursors, 1 will display them.
4303 vt.italic= [VT] Default color for italic text; 0-15.
4306 vt.underline= [VT] Default color for underlined text; 0-15.
4309 watchdog timers [HW,WDT] For information on watchdog timers,
4310 see Documentation/watchdog/watchdog-parameters.txt
4311 or other driver-specific files in the
4312 Documentation/watchdog/ directory.
4314 workqueue.watchdog_thresh=
4315 If CONFIG_WQ_WATCHDOG is configured, workqueue can
4316 warn stall conditions and dump internal state to
4317 help debugging. 0 disables workqueue stall
4318 detection; otherwise, it's the stall threshold
4319 duration in seconds. The default value is 30 and
4320 it can be updated at runtime by writing to the
4321 corresponding sysfs file.
4323 workqueue.disable_numa
4324 By default, all work items queued to unbound
4325 workqueues are affine to the NUMA nodes they're
4326 issued on, which results in better behavior in
4327 general. If NUMA affinity needs to be disabled for
4328 whatever reason, this option can be used. Note
4329 that this also can be controlled per-workqueue for
4330 workqueues visible under /sys/bus/workqueue/.
4332 workqueue.power_efficient
4333 Per-cpu workqueues are generally preferred because
4334 they show better performance thanks to cache
4335 locality; unfortunately, per-cpu workqueues tend to
4336 be more power hungry than unbound workqueues.
4338 Enabling this makes the per-cpu workqueues which
4339 were observed to contribute significantly to power
4340 consumption unbound, leading to measurably lower
4341 power usage at the cost of small performance
4344 The default value of this parameter is determined by
4345 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
4347 workqueue.debug_force_rr_cpu
4348 Workqueue used to implicitly guarantee that work
4349 items queued without explicit CPU specified are put
4350 on the local CPU. This guarantee is no longer true
4351 and while local CPU is still preferred work items
4352 may be put on foreign CPUs. This debug option
4353 forces round-robin CPU selection to flush out
4354 usages which depend on the now broken guarantee.
4355 When enabled, memory and cache locality will be
4358 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
4359 default x2apic cluster mode on platforms
4362 x86_intel_mid_timer= [X86-32,APBT]
4363 Choose timer option for x86 Intel MID platform.
4364 Two valid options are apbt timer only and lapic timer
4365 plus one apbt timer for broadcast timer.
4366 x86_intel_mid_timer=apbt_only | lapic_and_apbt
4368 xen_512gb_limit [KNL,X86-64,XEN]
4369 Restricts the kernel running paravirtualized under Xen
4370 to use only up to 512 GB of RAM. The reason to do so is
4371 crash analysis tools and Xen tools for doing domain
4372 save/restore/migration must be enabled to handle larger
4375 xen_emul_unplug= [HW,X86,XEN]
4376 Unplug Xen emulated devices
4377 Format: [unplug0,][unplug1]
4378 ide-disks -- unplug primary master IDE devices
4379 aux-ide-disks -- unplug non-primary-master IDE devices
4380 nics -- unplug network devices
4381 all -- unplug all emulated devices (NICs and IDE disks)
4382 unnecessary -- unplugging emulated devices is
4383 unnecessary even if the host did not respond to
4385 never -- do not unplug even if version check succeeds
4387 xen_nopvspin [X86,XEN]
4388 Disables the ticketlock slowpath using Xen PV
4392 Disables the PV optimizations forcing the HVM guest to
4393 run as generic HVM guest with no PV drivers.
4395 xirc2ps_cs= [NET,PCMCIA]
4397 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
4399 ______________________________________________________________________
4403 Add more DRM drivers.
projects / karo-tx-linux.git / blob