4 The following is a consolidated list of the kernel parameters as implemented
5 (mostly) by the __setup() macro and sorted into English Dictionary order
6 (defined as ignoring all punctuation and sorting digits before letters in a
7 case insensitive manner), and with descriptions where known.
9 Module parameters for loadable modules are specified only as the
10 parameter name with optional '=' and value as appropriate, such as:
12 modprobe usbcore blinkenlights=1
14 Module parameters for modules that are built into the kernel image
15 are specified on the kernel command line with the module name plus
16 '.' plus parameter name, with '=' and value if appropriate, such as:
18 usbcore.blinkenlights=1
20 Hyphens (dashes) and underscores are equivalent in parameter names, so
21 log_buf_len=1M print-fatal-signals=1
22 can also be entered as
23 log-buf-len=1M print_fatal_signals=1
26 This document may not be entirely up to date and comprehensive. The command
27 "modinfo -p ${modulename}" shows a current list of all parameters of a loadable
28 module. Loadable modules, after being loaded into the running kernel, also
29 reveal their parameters in /sys/module/${modulename}/parameters/. Some of these
30 parameters may be changed at runtime by the command
31 "echo -n ${value} > /sys/module/${modulename}/parameters/${parm}".
33 The parameters listed below are only valid if certain kernel build options were
34 enabled and if respective hardware is present. The text in square brackets at
35 the beginning of each description states the restrictions within which a
36 parameter is applicable:
38 ACPI ACPI support is enabled.
39 AGP AGP (Accelerated Graphics Port) is enabled.
40 ALSA ALSA sound support is enabled.
41 APIC APIC support is enabled.
42 APM Advanced Power Management support is enabled.
43 ARM ARM architecture is enabled.
44 AVR32 AVR32 architecture is enabled.
45 AX25 Appropriate AX.25 support is enabled.
46 BLACKFIN Blackfin architecture is enabled.
47 CLK Common clock infrastructure is enabled.
48 CMA Contiguous Memory Area support is enabled.
49 DRM Direct Rendering Management support is enabled.
50 DYNAMIC_DEBUG Build in debug messages and enable them at runtime
51 EDD BIOS Enhanced Disk Drive Services (EDD) is enabled
52 EFI EFI Partitioning (GPT) is enabled
53 EIDE EIDE/ATAPI support is enabled.
54 EVM Extended Verification Module
55 FB The frame buffer device is enabled.
56 FTRACE Function tracing enabled.
57 GCOV GCOV profiling is enabled.
58 HW Appropriate hardware is enabled.
59 IA-64 IA-64 architecture is enabled.
60 IMA Integrity measurement architecture is enabled.
61 IOSCHED More than one I/O scheduler is enabled.
62 IP_PNP IP DHCP, BOOTP, or RARP is enabled.
63 IPV6 IPv6 support is enabled.
64 ISAPNP ISA PnP code is enabled.
65 ISDN Appropriate ISDN support is enabled.
66 JOY Appropriate joystick support is enabled.
67 KGDB Kernel debugger support is enabled.
68 KVM Kernel Virtual Machine support is enabled.
69 LIBATA Libata driver is enabled
70 LP Printer support is enabled.
71 LOOP Loopback device support is enabled.
72 M68k M68k architecture is enabled.
73 These options have more detailed description inside of
74 Documentation/m68k/kernel-options.txt.
75 MDA MDA console support is enabled.
76 MIPS MIPS architecture is enabled.
77 MOUSE Appropriate mouse support is enabled.
78 MSI Message Signaled Interrupts (PCI).
79 MTD MTD (Memory Technology Device) support is enabled.
80 NET Appropriate network support is enabled.
81 NUMA NUMA support is enabled.
82 NFS Appropriate NFS support is enabled.
83 OSS OSS sound support is enabled.
84 PV_OPS A paravirtualized kernel is enabled.
85 PARIDE The ParIDE (parallel port IDE) subsystem is enabled.
86 PARISC The PA-RISC architecture is enabled.
87 PCI PCI bus support is enabled.
88 PCIE PCI Express support is enabled.
89 PCMCIA The PCMCIA subsystem is enabled.
90 PNP Plug & Play support is enabled.
91 PPC PowerPC architecture is enabled.
92 PPT Parallel port support is enabled.
93 PS2 Appropriate PS/2 support is enabled.
94 RAM RAM disk support is enabled.
95 S390 S390 architecture is enabled.
96 SCSI Appropriate SCSI support is enabled.
97 A lot of drivers have their options described inside
98 the Documentation/scsi/ sub-directory.
99 SECURITY Different security models are enabled.
100 SELINUX SELinux support is enabled.
101 APPARMOR AppArmor support is enabled.
102 SERIAL Serial support is enabled.
103 SH SuperH architecture is enabled.
104 SMP The kernel is an SMP kernel.
105 SPARC Sparc architecture is enabled.
106 SWSUSP Software suspend (hibernation) is enabled.
107 SUSPEND System suspend states are enabled.
108 TPM TPM drivers are enabled.
109 TS Appropriate touchscreen support is enabled.
110 UMS USB Mass Storage support is enabled.
111 USB USB support is enabled.
112 USBHID USB Human Interface Device support is enabled.
113 V4L Video For Linux support is enabled.
114 VMMIO Driver for memory mapped virtio devices is enabled.
115 VGA The VGA console has been enabled.
116 VT Virtual terminal support is enabled.
117 WDT Watchdog support is enabled.
118 XT IBM PC/XT MFM hard disk support is enabled.
119 X86-32 X86-32, aka i386 architecture is enabled.
120 X86-64 X86-64 architecture is enabled.
121 More X86-64 boot options can be found in
122 Documentation/x86/x86_64/boot-options.txt .
123 X86 Either 32-bit or 64-bit x86 (same as X86-32+X86-64)
124 XEN Xen support is enabled
126 In addition, the following text indicates that the option:
128 BUGS= Relates to possible processor bugs on the said processor.
129 KNL Is a kernel start-up parameter.
130 BOOT Is a boot loader parameter.
132 Parameters denoted with BOOT are actually interpreted by the boot
133 loader, and have no meaning to the kernel directly.
134 Do not modify the syntax of boot loader parameters without extreme
135 need or coordination with <Documentation/x86/boot.txt>.
137 There are also arch-specific kernel-parameters not documented here.
138 See for example <Documentation/x86/x86_64/boot-options.txt>.
140 Note that ALL kernel parameters listed below are CASE SENSITIVE, and that
141 a trailing = on the name of any parameter states that that parameter will
142 be entered as an environment variable, whereas its absence indicates that
143 it will appear as a kernel argument readable via /proc/cmdline by programs
144 running once the system is up.
146 The number of kernel parameters is not limited, but the length of the
147 complete command line (parameters including spaces etc.) is limited to
148 a fixed number of characters. This limit depends on the architecture
149 and is between 256 and 4096 characters. It is defined in the file
150 ./include/asm/setup.h as COMMAND_LINE_SIZE.
152 Finally, the [KMG] suffix is commonly described after a number of kernel
153 parameter values. These 'K', 'M', and 'G' letters represent the _binary_
154 multipliers 'Kilo', 'Mega', and 'Giga', equalling 2^10, 2^20, and 2^30
155 bytes respectively. Such letter suffixes can also be entirely omitted.
159 Advanced Configuration and Power Interface
160 Format: { force | off | strict | noirq | rsdt }
161 force -- enable ACPI if default was off
162 off -- disable ACPI if default was on
163 noirq -- do not use ACPI for IRQ routing
164 strict -- Be less tolerant of platforms that are not
165 strictly ACPI specification compliant.
166 rsdt -- prefer RSDT over (default) XSDT
167 copy_dsdt -- copy DSDT to memory
169 See also Documentation/power/runtime_pm.txt, pci=noacpi
171 acpi_rsdp= [ACPI,EFI,KEXEC]
172 Pass the RSDP address to the kernel, mostly used
173 on machines running EFI runtime service to boot the
174 second kernel for kdump.
176 acpi_apic_instance= [ACPI, IOAPIC]
178 2: use 2nd APIC table, if available
179 1,0: use 1st APIC table
182 acpi_backlight= [HW,ACPI]
183 acpi_backlight=vendor
185 If set to vendor, prefer vendor specific driver
186 (e.g. thinkpad_acpi, sony_acpi, etc.) instead
187 of the ACPI video.ko driver.
189 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG]
190 acpi.debug_level= [HW,ACPI,ACPI_DEBUG]
192 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
193 debug output. Bits in debug_layer correspond to a
194 _COMPONENT in an ACPI source file, e.g.,
195 #define _COMPONENT ACPI_PCI_COMPONENT
196 Bits in debug_level correspond to a level in
197 ACPI_DEBUG_PRINT statements, e.g.,
198 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
199 The debug_level mask defaults to "info". See
200 Documentation/acpi/debug.txt for more information about
201 debug layers and levels.
203 Enable processor driver info messages:
204 acpi.debug_layer=0x20000000
205 Enable PCI/PCI interrupt routing info messages:
206 acpi.debug_layer=0x400000
207 Enable AML "Debug" output, i.e., stores to the Debug
208 object while interpreting AML:
209 acpi.debug_layer=0xffffffff acpi.debug_level=0x2
210 Enable all messages related to ACPI hardware:
211 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
213 Some values produce so much output that the system is
214 unusable. The "log_buf_len" parameter may be useful
215 if you need to capture more output.
217 acpi_irq_balance [HW,ACPI]
218 ACPI will balance active IRQs
221 acpi_irq_nobalance [HW,ACPI]
222 ACPI will not move active IRQs (default)
225 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA
226 Format: <irq>,<irq>...
228 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for
230 Format: <irq>,<irq>...
232 acpi_no_auto_serialize [HW,ACPI]
233 Disable auto-serialization of AML methods
234 AML control methods that contain the opcodes to create
235 named objects will be marked as "Serialized" by the
236 auto-serialization feature.
237 This feature is enabled by default.
238 This option allows to turn off the feature.
240 acpi_no_static_ssdt [HW,ACPI]
241 Disable installation of static SSDTs at early boot time
242 By default, SSDTs contained in the RSDT/XSDT will be
243 installed automatically and they will appear under
244 /sys/firmware/acpi/tables.
245 This option turns off this feature.
246 Note that specifying this option does not affect
247 dynamic table installation which will install SSDT
248 tables to /sys/firmware/acpi/tables/dynamic.
250 acpica_no_return_repair [HW, ACPI]
251 Disable AML predefined validation mechanism
252 This mechanism can repair the evaluation result to make
253 the return objects more ACPI specification compliant.
254 This option is useful for developers to identify the
255 root cause of an AML interpreter issue when the issue
256 has something to do with the repair mechanism.
258 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
259 Format: To spoof as Windows 98: ="Microsoft Windows"
261 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
262 acpi_osi="string1" # add string1
263 acpi_osi="!string2" # remove string2
264 acpi_osi=!* # remove all strings
265 acpi_osi=! # disable all built-in OS vendor
267 acpi_osi= # disable all strings
269 'acpi_osi=!' can be used in combination with single or
270 multiple 'acpi_osi="string1"' to support specific OS
271 vendor string(s). Note that such command can only
272 affect the default state of the OS vendor strings, thus
273 it cannot affect the default state of the feature group
274 strings and the current state of the OS vendor strings,
275 specifying it multiple times through kernel command line
276 is meaningless. This command is useful when one do not
277 care about the state of the feature group strings which
278 should be controlled by the OSPM.
280 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
281 to 'acpi_osi="Windows 2000" acpi_osi=!', they all
282 can make '_OSI("Windows 2000")' TRUE.
284 'acpi_osi=' cannot be used in combination with other
285 'acpi_osi=' command lines, the _OSI method will not
286 exist in the ACPI namespace. NOTE that such command can
287 only affect the _OSI support state, thus specifying it
288 multiple times through kernel command line is also
291 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
294 'acpi_osi=!*' can be used in combination with single or
295 multiple 'acpi_osi="string1"' to support specific
296 string(s). Note that such command can affect the
297 current state of both the OS vendor strings and the
298 feature group strings, thus specifying it multiple times
299 through kernel command line is meaningful. But it may
300 still not able to affect the final state of a string if
301 there are quirks related to this string. This command
302 is useful when one want to control the state of the
303 feature group strings to debug BIOS issues related to
306 1. 'acpi_osi="Module Device" acpi_osi=!*' can make
307 '_OSI("Module Device")' FALSE.
308 2. 'acpi_osi=!* acpi_osi="Module Device"' can make
309 '_OSI("Module Device")' TRUE.
310 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
312 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
314 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
315 they all will make '_OSI("Windows 2000")' TRUE.
318 Override the pmtimer bug detection: force the kernel
319 to assume that this machine's pmtimer latches its value
320 and always returns good values.
322 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
323 Format: { level | edge | high | low }
325 acpi_skip_timer_override [HW,ACPI]
326 Recognize and ignore IRQ0/pin2 Interrupt Override.
327 For broken nForce2 BIOS resulting in XT-PIC timer.
329 acpi_sleep= [HW,ACPI] Sleep options
330 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
331 old_ordering, nonvs, sci_force_enable }
332 See Documentation/power/video.txt for information on
334 s3_beep is for debugging; it makes the PC's speaker beep
335 as soon as the kernel's real-mode entry point is called.
336 s4_nohwsig prevents ACPI hardware signature from being
337 used during resume from hibernation.
338 old_ordering causes the ACPI 1.0 ordering of the _PTS
339 control method, with respect to putting devices into
340 low power states, to be enforced (the ACPI 2.0 ordering
341 of _PTS is used by default).
342 nonvs prevents the kernel from saving/restoring the
343 ACPI NVS memory during suspend/hibernation and resume.
344 sci_force_enable causes the kernel to set SCI_EN directly
345 on resume from S1/S3 (which is against the ACPI spec,
346 but some broken systems don't work without it).
348 acpi_use_timer_override [HW,ACPI]
349 Use timer override. For some broken Nvidia NF5 boards
350 that require a timer override, but don't have HPET
352 acpi_enforce_resources= [ACPI]
353 { strict | lax | no }
354 Check for resource conflicts between native drivers
355 and ACPI OperationRegions (SystemIO and SystemMemory
356 only). IO ports and memory declared in ACPI might be
357 used by the ACPI subsystem in arbitrary AML code and
358 can interfere with legacy drivers.
359 strict (default): access to resources claimed by ACPI
360 is denied; legacy drivers trying to access reserved
361 resources will fail to bind to device using them.
362 lax: access to resources claimed by ACPI is allowed;
363 legacy drivers trying to access reserved resources
364 will bind successfully but a warning message is logged.
365 no: ACPI OperationRegions are not marked as reserved,
366 no further checks are performed.
368 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump
371 add_efi_memmap [EFI; X86] Include EFI memory map in
372 kernel's map of available physical RAM.
375 { off | try_unsupported }
376 off: disable AGP support
377 try_unsupported: try to drive unsupported chipsets
378 (may crash computer or cause data corruption)
381 See Documentation/sound/alsa/alsa-parameters.txt
384 Allow the default userspace alignment fault handler
385 behaviour to be specified. Bit 0 enables warnings,
386 bit 1 enables fixups, and bit 2 sends a segfault.
388 align_va_addr= [X86-64]
389 Align virtual addresses by clearing slice [14:12] when
390 allocating a VMA at process creation time. This option
391 gives you up to 3% performance improvement on AMD F15h
392 machines (where it is enabled by default) for a
393 CPU-intensive style benchmark, and it can vary highly in
394 a microbenchmark depending on workload and compiler.
396 32: only for 32-bit processes
397 64: only for 64-bit processes
398 on: enable for both 32- and 64-bit processes
399 off: disable for both 32- and 64-bit processes
401 alloc_snapshot [FTRACE]
402 Allocate the ftrace snapshot buffer on boot up when the
403 main buffer is allocated. This is handy if debugging
404 and you need to use tracing_snapshot() on boot up, and
405 do not want to use tracing_snapshot_alloc() as it needs
406 to be done where GFP_KERNEL allocations are allowed.
408 amd_iommu= [HW,X86-64]
409 Pass parameters to the AMD IOMMU driver in the system.
411 fullflush - enable flushing of IO/TLB entries when
412 they are unmapped. Otherwise they are
413 flushed before they will be reused, which
415 off - do not initialize any AMD IOMMU found in
417 force_isolation - Force device isolation for all
418 devices. The IOMMU driver is not
419 allowed anymore to lift isolation
420 requirements as needed. This option
421 does not override iommu=pt
423 amd_iommu_dump= [HW,X86-64]
424 Enable AMD IOMMU driver option to dump the ACPI table
425 for AMD IOMMU. With this option enabled, AMD IOMMU
426 driver will print ACPI tables for AMD IOMMU during
427 IOMMU initialization.
429 amijoy.map= [HW,JOY] Amiga joystick support
430 Map of devices attached to JOY0DAT and JOY1DAT
432 See also Documentation/input/joystick.txt
434 analog.map= [HW,JOY] Analog joystick and gamepad support
435 Specifies type or capabilities of an analog joystick
436 connected to one of 16 gameports
437 Format: <type1>,<type2>,..<type16>
440 Power management functions (SPARCstation-4/5 + deriv.)
442 Disable APC CPU standby support. SPARCstation-Fox does
443 not play well with APC CPU idle - disable it if you have
444 APC and your system crashes randomly.
446 apic= [APIC,X86-32] Advanced Programmable Interrupt Controller
447 Change the output verbosity whilst booting
448 Format: { quiet (default) | verbose | debug }
449 Change the amount of debugging information output
450 when initialising the APIC and IO-APIC components.
453 See Documentation/networking/ipv6.txt.
455 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
456 Limit apic dumping. The parameter defines the maximal
457 number of local apics being dumped. Also it is possible
458 to set it to "all" by meaning -- no limit here.
459 Format: { 1 (default) | 2 | ... | all }.
460 The parameter valid if only apic=debug or
461 apic=verbose is specified.
462 Example: apic=debug show_lapic=all
464 apm= [APM] Advanced Power Management
465 See header of arch/x86/kernel/apm_32.c.
467 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
468 Format: <io>,<irq>,<nodeID>
472 atarimouse= [HW,MOUSE] Atari Mouse
474 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
475 EzKey and similar keyboards
477 atkbd.reset= [HW] Reset keyboard during initialization
479 atkbd.set= [HW] Select keyboard code set
480 Format: <int> (2 = AT (default), 3 = PS/2)
482 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
485 atkbd.softraw= [HW] Choose between synthetic and real raw mode
486 Format: <bool> (0 = real, 1 = synthetic (default))
488 atkbd.softrepeat= [HW]
489 Use software keyboard repeat
491 audit= [KNL] Enable the audit sub-system
492 Format: { "0" | "1" } (0 = disabled, 1 = enabled)
493 0 - kernel audit is disabled and can not be enabled
494 until the next reboot
495 unset - kernel audit is initialized but disabled and
496 will be fully enabled by the userspace auditd.
497 1 - kernel audit is initialized and partially enabled,
498 storing at most audit_backlog_limit messages in
499 RAM until it is fully enabled by the userspace
503 audit_backlog_limit= [KNL] Set the audit queue size limit.
504 Format: <int> (must be >=0)
507 baycom_epp= [HW,AX25]
510 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
512 See header of drivers/net/hamradio/baycom_par.c.
514 baycom_ser_fdx= [HW,AX25]
515 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
516 Format: <io>,<irq>,<mode>[,<baud>]
517 See header of drivers/net/hamradio/baycom_ser_fdx.c.
519 baycom_ser_hdx= [HW,AX25]
520 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
521 Format: <io>,<irq>,<mode>
522 See header of drivers/net/hamradio/baycom_ser_hdx.c.
524 blkdevparts= Manual partition parsing of block device(s) for
525 embedded devices based on command line input.
526 See Documentation/block/cmdline-partition.txt
528 boot_delay= Milliseconds to delay each printk during boot.
529 Values larger than 10 seconds (10000) are changed to
533 bootmem_debug [KNL] Enable bootmem allocator debug messages.
535 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
536 bttv.radio= Most important insmod options are available as
538 bttv.pll= See Documentation/video4linux/bttv/Insmod-options
541 bulk_remove=off [PPC] This parameter disables the use of the pSeries
542 firmware feature for flushing multiple hpte entries
545 c101= [NET] Moxa C101 synchronous serial card
547 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
548 Sometimes CPU hardware bugs make them report the cache
549 size incorrectly. The kernel will attempt work arounds
550 to fix known problems, but for some CPUs it is not
551 possible to determine what the correct size should be.
552 This option provides an override for these situations.
554 ccw_timeout_log [S390]
555 See Documentation/s390/CommonIO for details.
557 cgroup_disable= [KNL] Disable a particular controller
558 Format: {name of the controller(s) to disable}
559 The effects of cgroup_disable=foo are:
560 - foo isn't auto-mounted if you mount all cgroups in
562 - foo isn't visible as an individually mountable
564 {Currently only "memory" controller deal with this and
565 cut the overhead, others just disable the usage. So
566 only cgroup_disable=memory is actually worthy}
568 checkreqprot [SELINUX] Set initial checkreqprot flag value.
569 Format: { "0" | "1" }
570 See security/selinux/Kconfig help text.
571 0 -- check protection applied by kernel (includes
572 any implied execute protection).
573 1 -- check protection requested by application.
574 Default value is set via a kernel config option.
575 Value can be changed at runtime via
576 /selinux/checkreqprot.
579 See Documentation/s390/CommonIO for details.
582 Keep all clocks already enabled by bootloader on,
583 even if no driver has claimed them. This is useful
584 for debug and development, but should not be
585 needed on a platform with proper driver support.
586 For more information, see Documentation/clk.txt.
588 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
590 Forces specified clocksource (if available) to be used
591 when calculating gettimeofday(). If specified
592 clocksource is not available, it defaults to PIT.
593 Format: { pit | tsc | cyclone | pmtmr }
595 clocksource= Override the default clocksource
597 Override the default clocksource and use the clocksource
598 with the name specified.
599 Some clocksource names to choose from, depending on
601 [all] jiffies (this is the base, fallback clocksource)
603 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
604 pxa_timer,timer3,32k_counter,timer0_1
606 [X86-32] pit,hpet,tsc;
607 scx200_hrt on Geode; cyclone on IBM x440
615 clearcpuid=BITNUM [X86]
616 Disable CPUID feature X for the kernel. See
617 arch/x86/include/asm/cpufeature.h for the valid bit
618 numbers. Note the Linux specific bits are not necessarily
619 stable over kernel options, but the vendor specific
621 Also note that user programs calling CPUID directly
622 or using the feature without checking anything
623 will still see it. This just prevents it from
624 being used by the kernel or shown in /proc/cpuinfo.
625 Also note the kernel might malfunction if you disable
629 Sets the size of kernel global memory area for contiguous
630 memory allocations. For more information, see
631 include/linux/dma-contiguous.h
633 cmo_free_hint= [PPC] Format: { yes | no }
634 Specify whether pages are marked as being inactive
635 when they are freed. This is used in CMO environments
636 to determine OS memory pressure for page stealing by
640 coherent_pool=nn[KMG] [ARM,KNL]
641 Sets the size of memory pool for coherent, atomic dma
642 allocations, by default set to 256K.
644 code_bytes [X86] How many bytes of object code to print
649 com20020= [HW,NET] ARCnet - COM20020 chipset
651 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
653 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
657 ARCnet - COM90xx chipset (memory-mapped buffers)
658 Format: <io>[,<irq>[,<memstart>]]
660 condev= [HW,S390] console device
663 console= [KNL] Output console device and options.
665 tty<n> Use the virtual console device <n>.
669 Use the specified serial port. The options are of
670 the form "bbbbpnf", where "bbbb" is the baud rate,
671 "p" is parity ("n", "o", or "e"), "n" is number of
672 bits, and "f" is flow control ("r" for RTS or
673 omit it). Default is "9600n8".
675 See Documentation/serial-console.txt for more
677 Documentation/networking/netconsole.txt for an
680 uart[8250],io,<addr>[,options]
681 uart[8250],mmio,<addr>[,options]
682 Start an early, polled-mode console on the 8250/16550
683 UART at the specified I/O port or MMIO address,
684 switching to the matching ttyS device later. The
685 options are the same as for ttyS, above.
686 hvc<n> Use the hypervisor console device <n>. This is for
687 both Xen and PowerPC hypervisors.
689 If the device connected to the port is not a TTY but a braille
690 device, prepend "brl," before the device type, for instance
692 For now, only VisioBraille is supported.
694 consoleblank= [KNL] The console blank (screen saver) timeout in
695 seconds. Defaults to 10*60 = 10mins. A value of 0
696 disables the blank timer.
699 [KNL] Change the default value for
700 /proc/<pid>/coredump_filter.
701 See also Documentation/filesystems/proc.txt.
703 cpuidle.off=1 [CPU_IDLE]
704 disable the cpuidle sub-system
706 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
708 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
710 crashkernel=size[KMG][@offset[KMG]]
711 [KNL] Using kexec, Linux can switch to a 'crash kernel'
712 upon panic. This parameter reserves the physical
713 memory region [offset, offset + size] for that kernel
714 image. If '@offset' is omitted, then a suitable offset
715 is selected automatically. Check
716 Documentation/kdump/kdump.txt for further details.
718 crashkernel=range1:size1[,range2:size2,...][@offset]
719 [KNL] Same as above, but depends on the memory
720 in the running system. The syntax of range is
721 start-[end] where start and end are both
722 a memory unit (amount[KMG]). See also
723 Documentation/kdump/kdump.txt for an example.
725 crashkernel=size[KMG],high
726 [KNL, x86_64] range could be above 4G. Allow kernel
727 to allocate physical memory region from top, so could
728 be above 4G if system have more than 4G ram installed.
729 Otherwise memory region will be allocated below 4G, if
731 It will be ignored if crashkernel=X is specified.
732 crashkernel=size[KMG],low
733 [KNL, x86_64] range under 4G. When crashkernel=X,high
734 is passed, kernel could allocate physical memory region
735 above 4G, that cause second kernel crash on system
736 that require some amount of low memory, e.g. swiotlb
737 requires at least 64M+32K low memory. Kernel would
738 try to allocate 72M below 4G automatically.
739 This one let user to specify own low range under 4G
740 for second kernel instead.
741 0: to disable low allocation.
742 It will be ignored when crashkernel=X,high is not used
743 or memory reserved is below 4G.
748 cs89x0_media= [HW,NET]
749 Format: { rj45 | aui | bnc }
752 See header of drivers/s390/block/dasd_devmap.c.
754 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
755 (one device per port)
756 Format: <port#>,<type>
757 See also Documentation/input/joystick-parport.txt
759 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
760 time. See Documentation/dynamic-debug-howto.txt for
761 details. Deprecated, see dyndbg.
763 debug [KNL] Enable kernel debugging (events log level).
766 [KNL] verbose self-tests
768 Print debugging info while doing the locking API
770 We default to 0 (no extra messages), setting it to
771 1 will print _a lot_ more information - normally
772 only useful to kernel developers.
774 debug_objects [KNL] Enable object debugging
777 [KNL] Disable object debugging
779 debug_guardpage_minorder=
780 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
781 parameter allows control of the order of pages that will
782 be intentionally kept free (and hence protected) by the
783 buddy allocator. Bigger value increase the probability
784 of catching random memory corruption, but reduce the
785 amount of memory for normal system use. The maximum
786 possible value is MAX_ORDER/2. Setting this parameter
787 to 1 or 2 should be enough to identify most random
788 memory corruption problems caused by bugs in kernel or
789 driver code when a CPU writes to (or reads from) a
790 random memory location. Note that there exists a class
791 of memory corruptions problems caused by buggy H/W or
792 F/W or by drivers badly programing DMA (basically when
793 memory is written at bus level and the CPU MMU is
794 bypassed) which are not detectable by
795 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
796 tracking down these problems.
798 debugpat [X86] Enable PAT debugging
800 decnet.addr= [HW,NET]
801 Format: <area>[,<node>]
802 See also Documentation/networking/decnet.txt.
805 [same as hugepagesz=] The size of the default
806 HugeTLB page size. This is the size represented by
807 the legacy /proc/ hugepages APIs, used for SHM, and
808 default size when mounting hugetlbfs filesystems.
809 Defaults to the default architecture's huge page size
813 Set number of hash buckets for dentry cache.
816 See Documentation/networking/ipv6.txt.
818 disable_cpu_apicid= [X86,APIC,SMP]
820 The number of initial APIC ID for the
821 corresponding CPU to be disabled at boot,
822 mostly used for the kdump 2nd kernel to
823 disable BSP to wake up multiple CPUs without
824 causing system reset or hang due to sending
827 disable_ddw [PPC/PSERIES]
828 Disable Dynamic DMA Window support. Use this if
829 to workaround buggy firmware.
832 See Documentation/networking/ipv6.txt.
834 disable_mtrr_cleanup [X86]
835 The kernel tries to adjust MTRR layout from continuous
836 to discrete, to make X server driver able to add WB
837 entry later. This parameter disables that.
839 disable_mtrr_trim [X86, Intel and AMD only]
840 By default the kernel will trim any uncacheable
841 memory out of your available memory pool based on
842 MTRR settings. This parameter disables that behavior,
843 possibly causing your machine to run very slowly.
845 disable_timer_pin_1 [X86]
846 Disable PIN 1 of APIC timer
847 Can be useful to work around chipset bugs.
849 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
850 this option disables the debugging code at boot.
852 dma_debug_entries=<number>
853 This option allows to tune the number of preallocated
854 entries for DMA-API debugging code. One entry is
855 required per DMA-API allocation. Use this if the
856 DMA-API debugging code disables itself because the
857 architectural default is too low.
859 dma_debug_driver=<driver_name>
860 With this option the DMA-API debugging driver
861 filter feature can be enabled at boot time. Just
862 pass the driver to filter for as the parameter.
863 The filter can be disabled or changed to another
864 driver later using sysfs.
866 drm_kms_helper.edid_firmware=[<connector>:]<file>
867 Broken monitors, graphic adapters and KVMs may
868 send no or incorrect EDID data sets. This parameter
869 allows to specify an EDID data set in the
870 /lib/firmware directory that is used instead.
871 Generic built-in EDID data sets are used, if one of
872 edid/1024x768.bin, edid/1280x1024.bin,
873 edid/1680x1050.bin, or edid/1920x1080.bin is given
874 and no file with the same name exists. Details and
875 instructions how to build your own EDID data are
876 available in Documentation/EDID/HOWTO.txt. An EDID
877 data set will only be used for a particular connector,
878 if its name and a colon are prepended to the EDID
883 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
884 module.dyndbg[="val"]
885 Enable debug messages at boot time. See
886 Documentation/dynamic-debug-howto.txt for details.
888 early_ioremap_debug [KNL]
889 Enable debug messages in early_ioremap support. This
890 is useful for tracking down temporary early mappings
891 which are not unmapped.
893 earlycon= [KNL] Output early console device and options.
894 uart[8250],io,<addr>[,options]
895 uart[8250],mmio,<addr>[,options]
896 uart[8250],mmio32,<addr>[,options]
897 Start an early, polled-mode console on the 8250/16550
898 UART at the specified I/O port or MMIO address.
899 MMIO inter-register address stride is either 8-bit
900 (mmio) or 32-bit (mmio32).
901 The options are the same as for ttyS, above.
903 earlyprintk= [X86,SH,BLACKFIN,ARM]
907 earlyprintk=serial[,ttySn[,baudrate]]
908 earlyprintk=serial[,0x...[,baudrate]]
909 earlyprintk=ttySn[,baudrate]
910 earlyprintk=dbgp[debugController#]
912 earlyprintk is useful when the kernel crashes before
913 the normal console is initialized. It is not enabled by
914 default because it has some cosmetic problems.
916 Append ",keep" to not disable it when the real console
919 Only one of vga, efi, serial, or usb debug port can
922 Currently only ttyS0 and ttyS1 may be specified by
923 name. Other I/O ports may be explicitly specified
924 on some architectures (x86 and arm at least) by
925 replacing ttySn with an I/O port address, like this:
926 earlyprintk=serial,0x1008,115200
927 You can find the port for a given device in
928 /proc/tty/driver/serial:
929 2: uart:ST16650V2 port:00001008 irq:18 ...
931 Interaction with the standard serial driver is not
934 The VGA and EFI output is eventually overwritten by
937 The xen output can only be used by Xen PV guests.
939 edac_report= [HW,EDAC] Control how to report EDAC event
940 Format: {"on" | "off" | "force"}
941 on: enable EDAC to report H/W event. May be overridden
942 by other higher priority error reporting module.
943 off: disable H/W event reporting through EDAC.
944 force: enforce the use of EDAC to report H/W event.
947 ekgdboc= [X86,KGDB] Allow early kernel console debugging
950 This is designed to be used in conjunction with
951 the boot argument: earlyprintk=vga
954 Format: {"off" | "on" | "skip[mbr]"}
957 Format: { "old_map" }
958 old_map [X86-64]: switch to the old ioremap-based EFI
959 runtime services mapping. 32-bit still uses this one by
962 efi_no_storage_paranoia [EFI; X86]
963 Using this parameter you can use more than 50% of
964 your efi variable storage. Use this parameter only if
965 you are really sure that your UEFI does sane gc and
966 fulfills the spec otherwise your board may brick.
968 eisa_irq_edge= [PARISC,HW]
969 See header of drivers/parisc/eisa.c.
972 See comment before function elanfreq_setup() in
973 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
976 Format: {"cfq" | "deadline" | "noop"}
977 See Documentation/block/cfq-iosched.txt and
978 Documentation/block/deadline-iosched.txt for details.
980 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
981 Specifies physical address of start of kernel core
982 image elf header and optionally the size. Generally
983 kexec loader will pass this option to capture kernel.
984 See Documentation/kdump/kdump.txt for details.
986 enable_mtrr_cleanup [X86]
987 The kernel tries to adjust MTRR layout from continuous
988 to discrete, to make X server driver able to add WB
989 entry later. This parameter enables that.
991 enable_timer_pin_1 [X86]
992 Enable PIN 1 of APIC timer
993 Can be useful to work around chipset bugs
994 (in particular on some ATI chipsets).
995 The kernel tries to set a reasonable default.
997 enforcing [SELINUX] Set initial enforcing status.
999 See security/selinux/Kconfig help text.
1000 0 -- permissive (log only, no denials).
1001 1 -- enforcing (deny and log).
1003 Value can be changed at runtime via /selinux/enforce.
1006 Disable Error Record Serialization Table (ERST)
1009 ether= [HW,NET] Ethernet cards parameters
1010 This option is obsoleted by the "netdev=" option, which
1011 has equivalent usage. See its documentation for details.
1015 Permit 'security.evm' to be updated regardless of
1016 current integrity status.
1020 fail_make_request=[KNL]
1021 General fault injection mechanism.
1022 Format: <interval>,<probability>,<space>,<times>
1023 See also Documentation/fault-injection/.
1026 See Documentation/blockdev/floppy.txt.
1028 force_pal_cache_flush
1029 [IA-64] Avoid check_sal_cache_flush which may hang on
1030 buggy SAL_CACHE_FLUSH implementations. Using this
1031 parameter will force ia64_sal_cache_flush to call
1032 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1035 Forcefully enable Physical Address Extension (PAE).
1036 Many Pentium M systems disable PAE but may have a
1037 functionally usable PAE implementation.
1038 Warning: use of this parameter will taint the kernel
1039 and may cause unknown problems.
1042 [FTRACE] will set and start the specified tracer
1043 as early as possible in order to facilitate early
1046 ftrace_dump_on_oops[=orig_cpu]
1047 [FTRACE] will dump the trace buffers on oops.
1048 If no parameter is passed, ftrace will dump
1049 buffers of all CPUs, but if you pass orig_cpu, it will
1050 dump only the buffer of the CPU that triggered the
1053 ftrace_filter=[function-list]
1054 [FTRACE] Limit the functions traced by the function
1055 tracer at boot up. function-list is a comma separated
1056 list of functions. This list can be changed at run
1057 time by the set_ftrace_filter file in the debugfs
1060 ftrace_notrace=[function-list]
1061 [FTRACE] Do not trace the functions specified in
1062 function-list. This list can be changed at run time
1063 by the set_ftrace_notrace file in the debugfs
1066 ftrace_graph_filter=[function-list]
1067 [FTRACE] Limit the top level callers functions traced
1068 by the function graph tracer at boot up.
1069 function-list is a comma separated list of functions
1070 that can be changed at run time by the
1071 set_graph_function file in the debugfs tracing directory.
1074 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1075 support via parallel port (up to 5 devices per port)
1076 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1077 See also Documentation/input/joystick-parport.txt
1081 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART
1085 gcov_persist= [GCOV] When non-zero (default), profiling data for
1086 kernel modules is saved and remains accessible via
1087 debugfs, even when the module is unloaded/reloaded.
1088 When zero, profiling data is discarded and associated
1089 debugfs files are removed at module unload time.
1091 gpt [EFI] Forces disk with valid GPT signature but
1092 invalid Protective MBR to be treated as GPT. If the
1093 primary GPT is corrupted, it enables the backup/alternate
1094 GPT to be used instead.
1096 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1097 the "Enable 0" bit of the configuration register.
1100 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1101 the "Enable 0" bit of the configuration register.
1104 grcan.select= [HW] Select which physical interface to use.
1107 grcan.txsize= [HW] Sets the size of the tx buffer.
1108 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1110 grcan.rxsize= [HW] Sets the size of the rx buffer.
1111 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1114 hashdist= [KNL,NUMA] Large hashes allocated during boot
1115 are distributed across NUMA nodes. Defaults on
1116 for 64-bit NUMA, off otherwise.
1117 Format: 0 | 1 (for off | on)
1119 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1121 hd= [EIDE] (E)IDE hard drive subsystem geometry
1122 Format: <cyl>,<head>,<sect>
1125 Disable Hardware Error Source Table (HEST) support;
1126 corresponding firmware-first mode error processing
1127 logic will be disabled.
1129 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1130 size of <nn>. This works even on boxes that have no
1131 highmem otherwise. This also works to reduce highmem
1132 size on bigger boxes.
1134 highres= [KNL] Enable/disable high resolution timer mode.
1135 Valid parameters: "on", "off"
1139 See Documentation/isdn/README.HiSax.
1143 hpet= [X86-32,HPET] option to control HPET usage
1144 Format: { enable (default) | disable | force |
1146 disable: disable HPET and use PIT instead
1147 force: allow force enabled of undocumented chips (ICH4,
1149 verbose: show contents of HPET registers during setup
1151 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1152 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1154 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
1155 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
1156 On x86-64 and powerpc, this option can be specified
1157 multiple times interleaved with hugepages= to reserve
1158 huge pages of different sizes. Valid pages sizes on
1159 x86-64 are 2M (when the CPU supports "pse") and 1G
1160 (when the CPU supports the "pdpe1gb" cpuinfo flag)
1161 Note that 1GB pages can only be allocated at boot time
1162 using hugepages= and not freed afterwards.
1164 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1165 terminal devices. Valid values: 0..8
1166 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1167 If specified, z/VM IUCV HVC accepts connections
1168 from listed z/VM user IDs only.
1170 hwthread_map= [METAG] Comma-separated list of Linux cpu id to
1171 hardware thread id mappings.
1172 Format: <cpu>:<hwthread>
1175 Do not unregister boot console at start. This is only
1176 useful for debugging when something happens in the window
1177 between unregistering the boot console and initializing
1180 i2c_bus= [HW] Override the default board specific I2C bus speed
1181 or register an additional I2C bus that is not
1182 registered from board initialization code.
1186 i8042.debug [HW] Toggle i8042 debug mode
1187 i8042.direct [HW] Put keyboard port into non-translated mode
1188 i8042.dumbkbd [HW] Pretend that controller can only read data from
1189 keyboard and cannot control its state
1190 (Don't attempt to blink the leds)
1191 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1192 i8042.nokbd [HW] Don't check/create keyboard port
1193 i8042.noloop [HW] Disable the AUX Loopback command while probing
1195 i8042.nomux [HW] Don't check presence of an active multiplexing
1197 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1199 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1200 i8042.reset [HW] Reset the controller during init and cleanup
1201 i8042.unlock [HW] Unlock (ignore) the keylock
1205 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
1206 indicates that the driver is running on unsupported
1208 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
1209 does not match list of supported models.
1211 [HW] Report power status in /proc/i8k
1212 (disabled by default)
1213 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
1216 i915.invert_brightness=
1217 [DRM] Invert the sense of the variable that is used to
1218 set the brightness of the panel backlight. Normally a
1219 brightness value of 0 indicates backlight switched off,
1220 and the maximum of the brightness value sets the backlight
1221 to maximum brightness. If this parameter is set to 0
1222 (default) and the machine requires it, or this parameter
1223 is set to 1, a brightness value of 0 sets the backlight
1224 to maximum brightness, and the maximum of the brightness
1225 value switches the backlight off.
1226 -1 -- never invert brightness
1227 0 -- machine default
1228 1 -- force brightness inversion
1231 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1233 ide-core.nodma= [HW] (E)IDE subsystem
1234 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1235 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1236 .cdrom .chs .ignore_cable are additional options
1237 See Documentation/ide/ide.txt.
1239 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1240 Claim all unknown PCI IDE storage controllers.
1243 Format: idle=poll, idle=halt, idle=nomwait
1244 Poll forces a polling idle loop that can slightly
1245 improve the performance of waking up a idle CPU, but
1246 will use a lot of power and make the system run hot.
1248 idle=halt: Halt is forced to be used for CPU idle.
1249 In such case C2/C3 won't be used again.
1250 idle=nomwait: Disable mwait for CPU C-states
1252 ignore_loglevel [KNL]
1253 Ignore loglevel setting - this will print /all/
1254 kernel messages to the console. Useful for debugging.
1255 We also add it as printk module parameter, so users
1256 could change it dynamically, usually by
1257 /sys/module/printk/parameters/ignore_loglevel.
1259 ihash_entries= [KNL]
1260 Set number of hash buckets for inode cache.
1262 ima_appraise= [IMA] appraise integrity measurements
1263 Format: { "off" | "enforce" | "fix" }
1266 ima_appraise_tcb [IMA]
1267 The builtin appraise policy appraises all files
1271 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1275 The list of supported hash algorithms is defined
1276 in crypto/hash_info.h.
1279 Load a policy which meets the needs of the Trusted
1280 Computing Base. This means IMA will measure all
1281 programs exec'd, files mmap'd for exec, and all files
1282 opened for read by uid=0.
1285 Select one of defined IMA measurements template formats.
1286 Formats: { "ima" | "ima-ng" }
1291 Run specified binary instead of /sbin/init as init
1294 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1295 for working out where the kernel is dying during
1298 initrd= [BOOT] Specify the location of the initial ramdisk
1300 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1303 int_pln_enable [x86] Enable power limit notification interrupt
1305 integrity_audit=[IMA]
1306 Format: { "0" | "1" }
1307 0 -- basic integrity auditing messages. (Default)
1308 1 -- additional integrity auditing messages.
1310 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
1312 Enable intel iommu driver.
1314 Disable intel iommu driver.
1315 igfx_off [Default Off]
1316 By default, gfx is mapped as normal device. If a gfx
1317 device has a dedicated DMAR unit, the DMAR unit is
1318 bypassed by not enabling DMAR with this option. In
1319 this case, gfx device will use physical address for
1322 With this option iommu will not optimize to look
1323 for io virtual address below 32-bit forcing dual
1324 address cycle on pci bus for cards supporting greater
1325 than 32-bit addressing. The default is to look
1326 for translation below 32-bit and if not available
1327 then look in the higher range.
1328 strict [Default Off]
1329 With this option on every unmap_single operation will
1330 result in a hardware IOTLB flush operation as opposed
1331 to batching them for performance.
1332 sp_off [Default Off]
1333 By default, super page will be supported if Intel IOMMU
1334 has the capability. With this option, super page will
1337 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
1338 0 disables intel_idle and fall back on acpi_idle.
1339 1 to 6 specify maximum depth of C-state.
1343 Do not enable intel_pstate as the default
1344 scaling driver for the supported processors
1346 intremap= [X86-64, Intel-IOMMU]
1347 on enable Interrupt Remapping (default)
1348 off disable Interrupt Remapping
1349 nosid disable Source ID checking
1351 BIOS x2APIC opt-out request will be ignored
1353 iomem= Disable strict checking of access to MMIO memory
1354 strict regions from userspace.
1371 io7= [HW] IO7 for Marvel based alpha systems
1372 See comment before marvel_specify_io7 in
1373 arch/alpha/kernel/core_marvel.c.
1375 io_delay= [X86] I/O delay method
1377 Standard port 0x80 based delay
1379 Alternate port 0xed based delay (needed on some systems)
1381 Simple two microseconds delay
1386 See Documentation/filesystems/nfs/nfsroot.txt.
1388 ip2= [HW] Set IO/IRQ pairs for up to 4 IntelliPort boards
1389 See comment before ip2_setup() in
1390 drivers/char/ip2/ip2base.c.
1393 When an interrupt is not handled search all handlers
1394 for it. Intended to get systems with badly broken
1398 When an interrupt is not handled search all handlers
1399 for it. Also check all handlers each timer
1400 interrupt. Intended to get systems with badly broken
1404 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
1406 isolcpus= [KNL,SMP] Isolate CPUs from the general scheduler.
1408 <cpu number>,...,<cpu number>
1410 <cpu number>-<cpu number>
1411 (must be a positive range in ascending order)
1413 <cpu number>,...,<cpu number>-<cpu number>
1415 This option can be used to specify one or more CPUs
1416 to isolate from the general SMP balancing and scheduling
1417 algorithms. You can move a process onto or off an
1418 "isolated" CPU via the CPU affinity syscalls or cpuset.
1419 <cpu number> begins at 0 and the maximum value is
1420 "number of CPUs in system - 1".
1422 This option is the preferred way to isolate CPUs. The
1423 alternative -- manually setting the CPU mask of all
1424 tasks in the system -- can cause problems and
1425 suboptimal load balancer performance.
1429 ivrs_ioapic [HW,X86_64]
1430 Provide an override to the IOAPIC-ID<->DEVICE-ID
1431 mapping provided in the IVRS ACPI table. For
1432 example, to map IOAPIC-ID decimal 10 to
1433 PCI device 00:14.0 write the parameter as:
1434 ivrs_ioapic[10]=00:14.0
1436 ivrs_hpet [HW,X86_64]
1437 Provide an override to the HPET-ID<->DEVICE-ID
1438 mapping provided in the IVRS ACPI table. For
1439 example, to map HPET-ID decimal 0 to
1440 PCI device 00:14.0 write the parameter as:
1441 ivrs_hpet[0]=00:14.0
1443 js= [HW,JOY] Analog joystick
1444 See Documentation/input/joystick.txt.
1448 kernelcore=nn[KMG] [KNL,X86,IA-64,PPC] This parameter
1449 specifies the amount of memory usable by the kernel
1450 for non-movable allocations. The requested amount is
1451 spread evenly throughout all nodes in the system. The
1452 remaining memory in each node is used for Movable
1453 pages. In the event, a node is too small to have both
1454 kernelcore and Movable pages, kernelcore pages will
1455 take priority and other nodes will have a larger number
1456 of Movable pages. The Movable zone is used for the
1457 allocation of pages that may be reclaimed or moved
1458 by the page migration subsystem. This means that
1459 HugeTLB pages may not be allocated from this zone.
1460 Note that allocations like PTEs-from-HighMem still
1461 use the HighMem zone if it exists, and the Normal
1462 zone if it does not.
1464 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
1465 Format: <Controller#>[,poll interval]
1466 The controller # is the number of the ehci usb debug
1467 port as it is probed via PCI. The poll interval is
1468 optional and is the number seconds in between
1469 each poll cycle to the debug port in case you need
1470 the functionality for interrupting the kernel with
1471 gdb or control-c on the dbgp connection. When
1472 not using this parameter you use sysrq-g to break into
1473 the kernel debugger.
1475 kgdboc= [KGDB,HW] kgdb over consoles.
1476 Requires a tty driver that supports console polling,
1477 or a supported polling keyboard driver (non-usb).
1478 Serial only format: <serial_device>[,baud]
1479 keyboard only format: kbd
1480 keyboard and serial format: kbd,<serial_device>[,baud]
1481 Optional Kernel mode setting:
1482 kms, kbd format: kms,kbd
1483 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
1485 kgdbwait [KGDB] Stop kernel execution and enter the
1486 kernel debugger at the earliest opportunity.
1488 kmac= [MIPS] korina ethernet MAC address.
1489 Configure the RouterBoard 532 series on-chip
1490 Ethernet adapter MAC address.
1492 kmemleak= [KNL] Boot-time kmemleak enable/disable
1493 Valid arguments: on, off
1496 kmemcheck= [X86] Boot-time kmemcheck enable/disable/one-shot mode
1497 Valid arguments: 0, 1, 2
1498 kmemcheck=0 (disabled)
1499 kmemcheck=1 (enabled)
1500 kmemcheck=2 (one-shot mode)
1501 Default: 2 (one-shot mode)
1503 kstack=N [X86] Print N words from the kernel stack
1506 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
1507 Default is 0 (don't ignore, but inject #GP)
1509 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit
1513 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
1514 Default is 1 (enabled)
1516 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
1518 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
1520 kvm-intel.ept= [KVM,Intel] Disable extended page tables
1521 (virtualized MMU) support on capable Intel chips.
1522 Default is 1 (enabled)
1524 kvm-intel.emulate_invalid_guest_state=
1525 [KVM,Intel] Enable emulation of invalid guest states
1526 Default is 0 (disabled)
1528 kvm-intel.flexpriority=
1529 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
1530 Default is 1 (enabled)
1533 [KVM,Intel] Enable VMX nesting (nVMX).
1534 Default is 0 (disabled)
1536 kvm-intel.unrestricted_guest=
1537 [KVM,Intel] Disable unrestricted guest feature
1538 (virtualized real and unpaged mode) on capable
1539 Intel chips. Default is 1 (enabled)
1541 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
1542 feature (tagged TLBs) on capable Intel chips.
1543 Default is 1 (enabled)
1549 lapic [X86-32,APIC] Enable the local APIC even if BIOS
1552 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline
1553 value for LAPIC timer one-shot implementation. Default
1554 back to the programmable timer unit in the LAPIC.
1556 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
1559 libata.dma= [LIBATA] DMA control
1560 libata.dma=0 Disable all PATA and SATA DMA
1561 libata.dma=1 PATA and SATA Disk DMA only
1562 libata.dma=2 ATAPI (CDROM) DMA only
1563 libata.dma=4 Compact Flash DMA only
1564 Combinations also work, so libata.dma=3 enables DMA
1565 for disks and CDROMs, but not CFs.
1567 libata.ignore_hpa= [LIBATA] Ignore HPA limit
1568 libata.ignore_hpa=0 keep BIOS limits (default)
1569 libata.ignore_hpa=1 ignore limits, using full disk
1571 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
1575 libata.force= [LIBATA] Force configurations. The format is comma
1576 separated list of "[ID:]VAL" where ID is
1577 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
1578 matching port, link or device. Basically, it matches
1579 the ATA ID string printed on console by libata. If
1580 the whole ID part is omitted, the last PORT and DEVICE
1581 values are used. If ID hasn't been specified yet, the
1582 configuration applies to all ports, links and devices.
1584 If only DEVICE is omitted, the parameter applies to
1585 the port and all links and devices behind it. DEVICE
1586 number of 0 either selects the first device or the
1587 first fan-out link behind PMP device. It does not
1588 select the host link. DEVICE number of 15 selects the
1589 host link and device attached to it.
1591 The VAL specifies the configuration to force. As long
1592 as there's no ambiguity shortcut notation is allowed.
1593 For example, both 1.5 and 1.5G would work for 1.5Gbps.
1594 The following configurations can be forced.
1596 * Cable type: 40c, 80c, short40c, unk, ign or sata.
1597 Any ID with matching PORT is used.
1599 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
1601 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
1602 udma[/][16,25,33,44,66,100,133] notation is also
1605 * [no]ncq: Turn on or off NCQ.
1607 * nohrst, nosrst, norst: suppress hard, soft
1610 * rstonce: only attempt one reset during
1611 hot-unplug link recovery
1613 * dump_id: dump IDENTIFY data.
1615 * atapi_dmadir: Enable ATAPI DMADIR bridge support
1617 * disable: Disable this device.
1619 If there are multiple matching configurations changing
1620 the same attribute, the last one is used.
1622 memblock=debug [KNL] Enable memblock debug messages.
1624 load_ramdisk= [RAM] List of ramdisks to load from floppy
1625 See Documentation/blockdev/ramdisk.txt.
1627 lockd.nlm_grace_period=P [NFS] Assign grace period.
1630 lockd.nlm_tcpport=N [NFS] Assign TCP port.
1633 lockd.nlm_timeout=T [NFS] Assign timeout value.
1636 lockd.nlm_udpport=M [NFS] Assign UDP port.
1639 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
1642 loglevel= All Kernel Messages with a loglevel smaller than the
1643 console loglevel will be printed to the console. It can
1644 also be changed with klogd or other programs. The
1645 loglevels are defined as follows:
1647 0 (KERN_EMERG) system is unusable
1648 1 (KERN_ALERT) action must be taken immediately
1649 2 (KERN_CRIT) critical conditions
1650 3 (KERN_ERR) error conditions
1651 4 (KERN_WARNING) warning conditions
1652 5 (KERN_NOTICE) normal but significant condition
1653 6 (KERN_INFO) informational
1654 7 (KERN_DEBUG) debug-level messages
1656 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
1657 in bytes. n must be a power of two. The default
1658 size is set in the kernel config file.
1660 logo.nologo [FB] Disables display of the built-in Linux logo.
1661 This may be used to provide more screen space for
1662 kernel log messages and is useful when debugging
1663 kernel boot problems.
1665 lp=0 [LP] Specify parallel ports to use, e.g,
1666 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
1667 lp=reset first parallel port). 'lp=0' disables the
1668 lp=auto printer driver. 'lp=reset' (which can be
1669 specified in addition to the ports) causes
1670 attached printers to be reset. Using
1671 lp=port1,port2,... specifies the parallel ports
1672 to associate lp devices with, starting with
1673 lp0. A port specification may be 'none' to skip
1674 that lp device, or a parport name such as
1675 'parport0'. Specifying 'lp=auto' instead of a
1676 port specification list means that device IDs
1677 from each port should be examined, to see if
1678 an IEEE 1284-compliant printer is attached; if
1679 so, the driver will manage that printer.
1680 See also header of drivers/char/lp.c.
1683 Sets loops_per_jiffy to given constant, thus avoiding
1684 time-consuming boot-time autodetection (up to 250 ms per
1685 CPU). 0 enables autodetection (default). To determine
1686 the correct value for your kernel, boot with normal
1687 autodetection and see what value is printed. Note that
1688 on SMP systems the preset will be applied to all CPUs,
1689 which is likely to cause problems if your CPUs need
1690 significantly divergent settings. An incorrect value
1691 will cause delays in the kernel to be wrong, leading to
1692 unpredictable I/O errors and other breakage. Although
1693 unlikely, in the extreme case this might damage your
1697 Format: <io>,<irq>,<dma>
1699 machvec= [IA-64] Force the use of a particular machine-vector
1700 (machvec) in a generic kernel.
1701 Example: machvec=hpzx1_swiotlb
1703 machtype= [Loongson] Share the same kernel image file between different
1705 Example: machtype=lemote-yeeloong-2f-7inch
1707 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
1708 than or equal to this physical address is ignored.
1710 maxcpus= [SMP] Maximum number of processors that an SMP kernel
1711 should make use of. maxcpus=n : n >= 0 limits the
1712 kernel to using 'n' processors. n=0 is a special case,
1713 it is equivalent to "nosmp", which also disables
1716 max_loop= [LOOP] The number of loop block devices that get
1717 (loop.max_loop) unconditionally pre-created at init time. The default
1718 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
1719 of statically allocating a predefined number, loop
1720 devices can be requested on-demand with the
1721 /dev/loop-control interface.
1723 mce [X86-32] Machine Check Exception
1725 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.txt
1727 md= [HW] RAID subsystems devices and level
1728 See Documentation/md.txt.
1731 Format: <first>,<last>
1732 Specifies range of consoles to be captured by the MDA.
1734 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
1735 Amount of memory to be used when the kernel is not able
1736 to see the whole system memory or for test.
1737 [X86] Work as limiting max address. Use together
1738 with memmap= to avoid physical address space collisions.
1739 Without memmap= PCI devices could be placed at addresses
1740 belonging to unused RAM.
1742 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
1746 [KNL,SH] Allow user to override the default size for
1747 per-device physically contiguous DMA buffers.
1749 memmap=exactmap [KNL,X86] Enable setting of an exact
1750 E820 memory map, as specified by the user.
1751 Such memmap=exactmap lines can be constructed based on
1752 BIOS output or other requirements. See the memmap=nn@ss
1755 memmap=nn[KMG]@ss[KMG]
1756 [KNL] Force usage of a specific region of memory.
1757 Region of memory to be used is from ss to ss+nn.
1759 memmap=nn[KMG]#ss[KMG]
1760 [KNL,ACPI] Mark specific memory as ACPI data.
1761 Region of memory to be marked is from ss to ss+nn.
1763 memmap=nn[KMG]$ss[KMG]
1764 [KNL,ACPI] Mark specific memory as reserved.
1765 Region of memory to be reserved is from ss to ss+nn.
1766 Example: Exclude memory from 0x18690000-0x1869ffff
1767 memmap=64K$0x18690000
1769 memmap=0x10000$0x18690000
1771 memory_corruption_check=0/1 [X86]
1772 Some BIOSes seem to corrupt the first 64k of
1773 memory when doing things like suspend/resume.
1774 Setting this option will scan the memory
1775 looking for corruption. Enabling this will
1776 both detect corruption and prevent the kernel
1777 from using the memory being corrupted.
1778 However, its intended as a diagnostic tool; if
1779 repeatable BIOS-originated corruption always
1780 affects the same memory, you can use memmap=
1781 to prevent the kernel from using that memory.
1783 memory_corruption_check_size=size [X86]
1784 By default it checks for corruption in the low
1785 64k, making this memory unavailable for normal
1786 use. Use this parameter to scan for
1787 corruption in more or less memory.
1789 memory_corruption_check_period=seconds [X86]
1790 By default it checks for corruption every 60
1791 seconds. Use this parameter to check at some
1792 other rate. 0 disables periodic checking.
1794 memtest= [KNL,X86] Enable memtest
1796 default : 0 <disable>
1797 Specifies the number of memtest passes to be
1798 performed. Each pass selects another test
1799 pattern from a given set of patterns. Memtest
1800 fills the memory with this pattern, validates
1801 memory contents and reserves bad memory
1802 regions that are detected.
1804 meye.*= [HW] Set MotionEye Camera parameters
1805 See Documentation/video4linux/meye.txt.
1807 mfgpt_irq= [IA-32] Specify the IRQ to use for the
1808 Multi-Function General Purpose Timers on AMD Geode
1811 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
1812 the BIOS has incorrectly applied a workaround. TinyBIOS
1813 version 0.98 is known to be affected, 0.99 fixes the
1814 problem by letting the user disable the workaround.
1818 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
1819 physical address is ignored.
1821 mini2440= [ARM,HW,KNL]
1822 Format:[0..2][b][c][t]
1824 MINI2440 configuration specification:
1825 0 - The attached screen is the 3.5" TFT
1826 1 - The attached screen is the 7" TFT
1827 2 - The VGA Shield is attached (1024x768)
1828 Leaving out the screen size parameter will not load
1829 the TFT driver, and the framebuffer will be left
1831 b - Enable backlight. The TFT backlight pin will be
1832 linked to the kernel VESA blanking code and a GPIO
1833 LED. This parameter is not necessary when using the
1835 c - Enable the s3c camera interface.
1836 t - Reserved for enabling touchscreen support. The
1837 touchscreen support is not enabled in the mainstream
1838 kernel as of 2.6.30, a preliminary port can be found
1839 in the "bleeding edge" mini2440 support kernel at
1840 http://repo.or.cz/w/linux-2.6/mini2440.git
1843 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
1844 parameter allows control of the logging verbosity for
1845 the additional memory initialisation checks. A value
1846 of 0 disables mminit logging and a level of 4 will
1847 log everything. Information is printed at KERN_DEBUG
1848 so loglevel=8 may also need to be specified.
1851 [KNL] When CONFIG_MODULE_SIG is set, this means that
1852 modules without (valid) signatures will fail to load.
1853 Note that if CONFIG_MODULE_SIG_FORCE is set, that
1854 is always true, so this option does nothing.
1857 [MOUSE] Maximum time between finger touching and
1858 leaving touchpad surface for touch to be considered
1859 a tap and be reported as a left button click (for
1860 touchpads working in absolute mode only).
1862 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
1863 reporting absolute coordinates, such as tablets
1864 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
1865 reporting absolute coordinates, such as tablets
1867 movablecore=nn[KMG] [KNL,X86,IA-64,PPC] This parameter
1868 is similar to kernelcore except it specifies the
1869 amount of memory used for migratable allocations.
1870 If both kernelcore and movablecore is specified,
1871 then kernelcore will be at *least* the specified
1872 value but may be more. If movablecore on its own
1873 is specified, the administrator must be careful
1874 that the amount of memory usable for all allocations
1877 movable_node [KNL,X86] Boot-time switch to enable the effects
1878 of CONFIG_MOVABLE_NODE=y. See mm/Kconfig for details.
1880 MTD_Partition= [MTD]
1881 Format: <name>,<region-number>,<size>,<offset>
1883 MTD_Region= [MTD] Format:
1884 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
1887 See drivers/mtd/cmdlinepart.c.
1889 multitce=off [PPC] This parameter disables the use of the pSeries
1890 firmware feature for updating multiple TCE entries
1893 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
1895 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
1897 boundary - index of last SLC block on Flex-OneNAND.
1898 The remaining blocks are configured as MLC blocks.
1899 lock - Configure if Flex-OneNAND boundary should be locked.
1900 Once locked, the boundary cannot be changed.
1901 1 indicates lock status, 0 indicates unlock status.
1904 ARM/S3C2412 JIVE boot control
1906 See arch/arm/mach-s3c2412/mach-jive.c
1908 mtouchusb.raw_coordinates=
1909 [HW] Make the MicroTouch USB driver use raw coordinates
1910 ('y', default) or cooked coordinates ('n')
1912 mtrr_chunk_size=nn[KMG] [X86]
1913 used for mtrr cleanup. It is largest continuous chunk
1914 that could hold holes aka. UC entries.
1916 mtrr_gran_size=nn[KMG] [X86]
1917 Used for mtrr cleanup. It is granularity of mtrr block.
1919 Large value could prevent small alignment from
1922 mtrr_spare_reg_nr=n [X86]
1924 Range: 0,7 : spare reg number
1926 Used for mtrr cleanup. It is spare mtrr entries number.
1927 Set to 2 or more if your graphical card needs more.
1929 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
1931 netdev= [NET] Network devices parameters
1932 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
1933 Note that mem_start is often overloaded to mean
1934 something different and driver-specific.
1935 This usage is only documented in each driver source
1939 [NETFILTER] Enable connection tracking flow accounting
1940 0 to disable accounting
1941 1 to enable accounting
1944 nfsaddrs= [NFS] Deprecated. Use ip= instead.
1945 See Documentation/filesystems/nfs/nfsroot.txt.
1947 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
1948 See Documentation/filesystems/nfs/nfsroot.txt.
1950 nfsrootdebug [NFS] enable nfsroot debugging messages.
1951 See Documentation/filesystems/nfs/nfsroot.txt.
1953 nfs.callback_tcpport=
1954 [NFS] set the TCP port on which the NFSv4 callback
1955 channel should listen.
1958 [NFS] sets the pathname to the program which is used
1959 to update the NFS client cache entries.
1961 nfs.cache_getent_timeout=
1962 [NFS] sets the timeout after which an attempt to
1963 update a cache entry is deemed to have failed.
1965 nfs.idmap_cache_timeout=
1966 [NFS] set the maximum lifetime for idmapper cache
1970 [NFS] enable 64-bit inode numbers.
1971 If zero, the NFS client will fake up a 32-bit inode
1972 number for the readdir() and stat() syscalls instead
1973 of returning the full 64-bit number.
1974 The default is to return 64-bit inode numbers.
1976 nfs.max_session_slots=
1977 [NFSv4.1] Sets the maximum number of session slots
1978 the client will attempt to negotiate with the server.
1979 This limits the number of simultaneous RPC requests
1980 that the client can send to the NFSv4.1 server.
1981 Note that there is little point in setting this
1982 value higher than the max_tcp_slot_table_limit.
1984 nfs.nfs4_disable_idmapping=
1985 [NFSv4] When set to the default of '1', this option
1986 ensures that both the RPC level authentication
1987 scheme and the NFS level operations agree to use
1988 numeric uids/gids if the mount is using the
1989 'sec=sys' security flavour. In effect it is
1990 disabling idmapping, which can make migration from
1991 legacy NFSv2/v3 systems to NFSv4 easier.
1992 Servers that do not support this mode of operation
1993 will be autodetected by the client, and it will fall
1994 back to using the idmapper.
1995 To turn off this behaviour, set the value to '0'.
1997 [NFS4] Specify an additional fixed unique ident-
1998 ification string that NFSv4 clients can insert into
1999 their nfs_client_id4 string. This is typically a
2000 UUID that is generated at system install time.
2002 nfs.send_implementation_id =
2003 [NFSv4.1] Send client implementation identification
2004 information in exchange_id requests.
2005 If zero, no implementation identification information
2007 The default is to send the implementation identification
2010 nfs.recover_lost_locks =
2011 [NFSv4] Attempt to recover locks that were lost due
2012 to a lease timeout on the server. Please note that
2013 doing this risks data corruption, since there are
2014 no guarantees that the file will remain unchanged
2015 after the locks are lost.
2016 If you want to enable the kernel legacy behaviour of
2017 attempting to recover these locks, then set this
2019 The default parameter value of '0' causes the kernel
2020 not to attempt recovery of lost locks.
2022 nfsd.nfs4_disable_idmapping=
2023 [NFSv4] When set to the default of '1', the NFSv4
2024 server will return only numeric uids and gids to
2025 clients using auth_sys, and will accept numeric uids
2026 and gids from such clients. This is intended to ease
2027 migration from NFSv2/v3.
2029 objlayoutdriver.osd_login_prog=
2030 [NFS] [OBJLAYOUT] sets the pathname to the program which
2031 is used to automatically discover and login into new
2032 osd-targets. Please see:
2033 Documentation/filesystems/pnfs.txt for more explanations
2035 nmi_debug= [KNL,AVR32,SH] Specify one or more actions to take
2036 when a NMI is triggered.
2037 Format: [state][,regs][,debounce][,die]
2039 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
2040 Format: [panic,][nopanic,][num]
2042 0 - turn nmi_watchdog off
2043 When panic is specified, panic when an NMI watchdog
2044 timeout occurs (or 'nopanic' to override the opposite
2046 This is useful when you use a panic=... timeout and
2047 need the box quickly up again.
2049 netpoll.carrier_timeout=
2050 [NET] Specifies amount of time (in seconds) that
2051 netpoll should wait for a carrier. By default netpoll
2054 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
2055 emulation library even if a 387 maths coprocessor
2059 [HW] Never suspend the console
2060 Disable suspending of consoles during suspend and
2061 hibernate operations. Once disabled, debugging
2062 messages can reach various consoles while the rest
2063 of the system is being put to sleep (ie, while
2064 debugging driver suspend/resume hooks). This may
2065 not work reliably with all consoles, but is known
2066 to work with serial and VGA consoles.
2067 To facilitate more flexible debugging, we also add
2068 console_suspend, a printk module parameter to control
2069 it. Users could use console_suspend (usually
2070 /sys/module/printk/parameters/console_suspend) to
2071 turn on/off it dynamically.
2073 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
2074 caches in the slab allocator. Saves per-node memory,
2075 but will impact performance.
2079 noapic [SMP,APIC] Tells the kernel to not make use of any
2080 IOAPICs that may be present in the system.
2083 Disable kernel and module base offset ASLR (Address
2084 Space Layout Randomization) if built into the kernel.
2086 noautogroup Disable scheduler automatic task group creation.
2088 nobats [PPC] Do not use BATs for mapping kernel lowmem
2089 on "Classic" PPC cores.
2093 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
2095 nodelayacct [KNL] Disable per-task delay accounting
2097 nodisconnect [HW,SCSI,M68K] Disables SCSI disconnects.
2099 nodsp [SH] Disable hardware DSP at boot time.
2101 noefi [X86] Disable EFI runtime services support.
2106 On X86-32 available only on PAE configured kernels.
2107 noexec=on: enable non-executable mappings (default)
2108 noexec=off: disable non-executable mappings
2111 Disable SMAP (Supervisor Mode Access Prevention)
2112 even if it is supported by processor.
2115 Disable SMEP (Supervisor Mode Execution Prevention)
2116 even if it is supported by processor.
2119 This affects only 32-bit executables.
2120 noexec32=on: enable non-executable mappings (default)
2121 read doesn't imply executable mappings
2122 noexec32=off: disable non-executable mappings
2123 read implies executable mappings
2125 nofpu [SH] Disable hardware FPU at boot time.
2127 nofxsr [BUGS=X86-32] Disables x86 floating point extended
2128 register save and restore. The kernel will only save
2129 legacy floating-point registers on task switch.
2131 noxsave [BUGS=X86] Disables x86 extended register state save
2132 and restore using xsave. The kernel will fallback to
2133 enabling legacy floating-point and sse state.
2136 on enable eager fpu restore
2137 off disable eager fpu restore
2138 auto selects the default scheme, which automatically
2139 enables eagerfpu restore for xsaveopt.
2141 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
2142 wfi(ARM) instruction doesn't work correctly and not to
2143 use it. This is also useful when using JTAG debugger.
2145 no_file_caps Tells the kernel not to honor file capabilities. The
2146 only way then for a file to be executed with privilege
2147 is to be setuid root or executed by root.
2149 nohalt [IA-64] Tells the kernel not to use the power saving
2150 function PAL_HALT_LIGHT when idle. This increases
2151 power-consumption. On the positive side, it reduces
2152 interrupt wake-up latency, which may improve performance
2153 in certain environments such as networked servers or
2156 nohz= [KNL] Boottime enable/disable dynamic ticks
2157 Valid arguments: on, off
2160 nohz_full= [KNL,BOOT]
2161 In kernels built with CONFIG_NO_HZ_FULL=y, set
2162 the specified list of CPUs whose tick will be stopped
2163 whenever possible. The boot CPU will be forced outside
2164 the range to maintain the timekeeping.
2165 The CPUs in this range must also be included in the
2168 noiotrap [SH] Disables trapped I/O port accesses.
2170 noirqdebug [X86-32] Disables the code which attempts to detect and
2171 disable unhandled interrupt sources.
2173 no_timer_check [X86,APIC] Disables the code which tests for
2174 broken timer IRQ sources.
2176 noisapnp [ISAPNP] Disables ISA PnP code.
2178 noinitrd [RAM] Tells the kernel not to load any configured
2181 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
2183 [Deprecated - use intremap=off]
2187 nojitter [IA-64] Disables jitter checking for ITC timers.
2189 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
2191 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
2194 no-steal-acc [X86,KVM] Disable paravirtualized steal time accounting.
2195 steal time is computed, but won't influence scheduler
2198 nolapic [X86-32,APIC] Do not enable or use the local APIC.
2200 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
2202 noltlbs [PPC] Do not use large page/tlb entries for kernel
2203 lowmem mapping on PPC40x.
2205 nomca [IA-64] Disable machine check abort handling
2207 nomce [X86-32] Machine Check Exception
2209 nomfgpt [X86-32] Disable Multi-Function General Purpose
2210 Timer usage (for AMD Geode machines).
2212 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
2213 shutdown the other cpus. Instead use the REBOOT_VECTOR
2216 nomodule Disable module load
2218 nopat [X86] Disable PAT (page attribute table extension of
2219 pagetables) support.
2221 norandmaps Don't use address space randomization. Equivalent to
2222 echo 0 > /proc/sys/kernel/randomize_va_space
2224 noreplace-paravirt [X86,IA-64,PV_OPS] Don't patch paravirt_ops
2226 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
2227 with UP alternatives
2229 nordrand [X86] Disable the direct use of the RDRAND
2230 instruction even if it is supported by the
2231 processor. RDRAND is still available to user
2234 noresume [SWSUSP] Disables resume and restores original swap
2237 no-scroll [VGA] Disables scrollback.
2238 This is required for the Braillex ib80-piezo Braille
2239 reader made by F.H. Papenmeier (Germany).
2243 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
2245 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
2246 and disable the IO APIC. legacy for "maxcpus=0".
2248 nosoftlockup [KNL] Disable the soft-lockup detector.
2250 nosync [HW,M68K] Disables sync negotiation for all devices.
2252 notsc [BUGS=X86-32] Disable Time Stamp Counter
2254 nousb [USB] Disable the USB subsystem
2256 nowatchdog [KNL] Disable the lockup detector (NMI watchdog).
2260 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
2262 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
2263 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
2264 Some features depend on CPU0. Known dependencies are:
2265 1. Resume from suspend/hibernate depends on CPU0.
2266 Suspend/hibernate will fail if CPU0 is offline and you
2267 need to online CPU0 before suspend/hibernate.
2268 2. PIC interrupts also depend on CPU0. CPU0 can't be
2269 removed if a PIC interrupt is detected.
2270 It's said poweroff/reboot may depend on CPU0 on some
2271 machines although I haven't seen such issues so far
2272 after CPU0 is offline on a few tested machines.
2273 If the dependencies are under your control, you can
2274 turn on cpu0_hotplug.
2276 nptcg= [IA-64] Override max number of concurrent global TLB
2277 purges which is reported from either PAL_VM_SUMMARY or
2280 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
2281 could support. nr_cpus=n : n >= 1 limits the kernel to
2282 supporting 'n' processors. Later in runtime you can not
2283 use hotplug cpu feature to put more cpu back to online.
2284 just like you compile the kernel NR_CPUS=n
2286 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
2288 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing.
2289 Allowed values are enable and disable
2291 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
2292 one of ['zone', 'node', 'default'] can be specified
2293 This can be set from sysctl after boot.
2294 See Documentation/sysctl/vm.txt for details.
2296 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
2297 See Documentation/debugging-via-ohci1394.txt for more
2300 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
2301 Rather than timing out after 20 ms if an EC
2302 command is not properly ACKed, override the length
2303 of the timeout. We have interrupts disabled while
2304 waiting for the ACK, so if this is set too high
2305 interrupts *may* be lost!
2307 omap_mux= [OMAP] Override bootloader pin multiplexing.
2308 Format: <mux_mode0.mode_name=value>...
2309 For example, to override I2C bus2:
2310 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
2312 oprofile.timer= [HW]
2313 Use timer interrupt instead of performance counters
2315 oprofile.cpu_type= Force an oprofile cpu type
2316 This might be useful if you have an older oprofile
2317 userland or if you want common events.
2318 Format: { arch_perfmon }
2319 arch_perfmon: [X86] Force use of architectural
2320 perfmon on Intel CPUs instead of the
2321 CPU specific event set.
2322 timer: [X86] Force use of architectural NMI
2323 timer mode (see also oprofile.timer
2324 for generic hr timer mode)
2325 [s390] Force legacy basic mode sampling
2326 (report cpu_type "timer")
2328 oops=panic Always panic on oopses. Default is to just kill the
2329 process, but there is a small probability of
2330 deadlocking the machine.
2331 This will also cause panics on machine check exceptions.
2332 Useful together with panic=30 to trigger a reboot.
2335 See Documentation/sound/oss/oss-parameters.txt
2337 panic= [KNL] Kernel behaviour on panic: delay <timeout>
2338 timeout > 0: seconds before rebooting
2339 timeout = 0: wait forever
2340 timeout < 0: reboot immediately
2343 parkbd.port= [HW] Parallel port number the keyboard adapter is
2344 connected to, default is 0.
2346 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
2347 0 for XT, 1 for AT (default is AT).
2350 parport= [HW,PPT] Specify parallel ports. 0 disables.
2351 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
2352 Use 'auto' to force the driver to use any
2353 IRQ/DMA settings detected (the default is to
2354 ignore detected IRQ/DMA settings because of
2355 possible conflicts). You can specify the base
2356 address, IRQ, and DMA settings; IRQ and DMA
2357 should be numbers, or 'auto' (for using detected
2358 settings on that particular port), or 'nofifo'
2359 (to avoid using a FIFO even if it is detected).
2360 Parallel ports are assigned in the order they
2361 are specified on the command line, starting
2364 parport_init_mode= [HW,PPT]
2365 Configure VIA parallel port to operate in
2366 a specific mode. This is necessary on Pegasos
2367 computer where firmware has no options for setting
2368 up parallel port mode and sets it to spp.
2369 Currently this function knows 686a and 8231 chips.
2370 Format: [spp|ps2|epp|ecp|ecpepp]
2373 Halt all CPUs after the first oops has been printed for
2374 the specified number of seconds. This is to be used if
2375 your oopses keep scrolling off the screen.
2380 See header of drivers/block/paride/pcd.c.
2381 See also Documentation/blockdev/paride.txt.
2383 pci=option[,option...] [PCI] various PCI subsystem options:
2384 earlydump [X86] dump PCI config space before the kernel
2386 off [X86] don't probe for the PCI bus
2387 bios [X86-32] force use of PCI BIOS, don't access
2388 the hardware directly. Use this if your machine
2389 has a non-standard PCI host bridge.
2390 nobios [X86-32] disallow use of PCI BIOS, only direct
2391 hardware access methods are allowed. Use this
2392 if you experience crashes upon bootup and you
2393 suspect they are caused by the BIOS.
2394 conf1 [X86] Force use of PCI Configuration
2396 conf2 [X86] Force use of PCI Configuration
2398 noaer [PCIE] If the PCIEAER kernel config parameter is
2399 enabled, this kernel boot option can be used to
2400 disable the use of PCIE advanced error reporting.
2401 nodomains [PCI] Disable support for multiple PCI
2402 root domains (aka PCI segments, in ACPI-speak).
2403 nommconf [X86] Disable use of MMCONFIG for PCI
2405 check_enable_amd_mmconf [X86] check for and enable
2406 properly configured MMIO access to PCI
2407 config space on AMD family 10h CPU
2408 nomsi [MSI] If the PCI_MSI kernel config parameter is
2409 enabled, this kernel boot option can be used to
2410 disable the use of MSI interrupts system-wide.
2411 noioapicquirk [APIC] Disable all boot interrupt quirks.
2412 Safety option to keep boot IRQs enabled. This
2413 should never be necessary.
2414 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
2415 primary IO-APIC for bridges that cannot disable
2416 boot IRQs. This fixes a source of spurious IRQs
2417 when the system masks IRQs.
2418 noioapicreroute [APIC] Disable workaround that uses the
2419 boot IRQ equivalent of an IRQ that connects to
2420 a chipset where boot IRQs cannot be disabled.
2421 The opposite of ioapicreroute.
2422 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
2423 routing table. These calls are known to be buggy
2424 on several machines and they hang the machine
2425 when used, but on other computers it's the only
2426 way to get the interrupt routing table. Try
2427 this option if the kernel is unable to allocate
2428 IRQs or discover secondary PCI buses on your
2430 rom [X86] Assign address space to expansion ROMs.
2431 Use with caution as certain devices share
2432 address decoders between ROMs and other
2434 norom [X86] Do not assign address space to
2435 expansion ROMs that do not already have
2436 BIOS assigned address ranges.
2437 nobar [X86] Do not assign address space to the
2438 BARs that weren't assigned by the BIOS.
2439 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
2440 assigned automatically to PCI devices. You can
2441 make the kernel exclude IRQs of your ISA cards
2443 pirqaddr=0xAAAAA [X86] Specify the physical address
2444 of the PIRQ table (normally generated
2445 by the BIOS) if it is outside the
2446 F0000h-100000h range.
2447 lastbus=N [X86] Scan all buses thru bus #N. Can be
2448 useful if the kernel is unable to find your
2449 secondary buses and you want to tell it
2450 explicitly which ones they are.
2451 assign-busses [X86] Always assign all PCI bus
2452 numbers ourselves, overriding
2453 whatever the firmware may have done.
2454 usepirqmask [X86] Honor the possible IRQ mask stored
2455 in the BIOS $PIR table. This is needed on
2456 some systems with broken BIOSes, notably
2457 some HP Pavilion N5400 and Omnibook XE3
2458 notebooks. This will have no effect if ACPI
2459 IRQ routing is enabled.
2460 noacpi [X86] Do not use ACPI for IRQ routing
2461 or for PCI scanning.
2462 use_crs [X86] Use PCI host bridge window information
2463 from ACPI. On BIOSes from 2008 or later, this
2464 is enabled by default. If you need to use this,
2465 please report a bug.
2466 nocrs [X86] Ignore PCI host bridge windows from ACPI.
2467 If you need to use this, please report a bug.
2468 routeirq Do IRQ routing for all PCI devices.
2469 This is normally done in pci_enable_device(),
2470 so this option is a temporary workaround
2471 for broken drivers that don't call it.
2472 skip_isa_align [X86] do not align io start addr, so can
2473 handle more pci cards
2474 firmware [ARM] Do not re-enumerate the bus but instead
2475 just use the configuration from the
2476 bootloader. This is currently used on
2477 IXP2000 systems where the bus has to be
2478 configured a certain way for adjunct CPUs.
2479 noearly [X86] Don't do any early type 1 scanning.
2480 This might help on some broken boards which
2481 machine check when some devices' config space
2482 is read. But various workarounds are disabled
2483 and some IOMMU drivers will not work.
2484 bfsort Sort PCI devices into breadth-first order.
2485 This sorting is done to get a device
2486 order compatible with older (<= 2.4) kernels.
2487 nobfsort Don't sort PCI devices into breadth-first order.
2488 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
2489 tuning and use the BIOS-configured MPS defaults.
2490 pcie_bus_safe Set every device's MPS to the largest value
2491 supported by all devices below the root complex.
2492 pcie_bus_perf Set device MPS to the largest allowable MPS
2493 based on its parent bus. Also set MRRS (Max
2494 Read Request Size) to the largest supported
2495 value (no larger than the MPS that the device
2496 or bus can support) for best performance.
2497 pcie_bus_peer2peer Set every device's MPS to 128B, which
2498 every device is guaranteed to support. This
2499 configuration allows peer-to-peer DMA between
2500 any pair of devices, possibly at the cost of
2501 reduced performance. This also guarantees
2502 that hot-added devices will work.
2503 cbiosize=nn[KMG] The fixed amount of bus space which is
2504 reserved for the CardBus bridge's IO window.
2505 The default value is 256 bytes.
2506 cbmemsize=nn[KMG] The fixed amount of bus space which is
2507 reserved for the CardBus bridge's memory
2508 window. The default value is 64 megabytes.
2511 [<order of align>@][<domain>:]<bus>:<slot>.<func>[; ...]
2512 Specifies alignment and device to reassign
2513 aligned memory resources.
2514 If <order of align> is not specified,
2515 PAGE_SIZE is used as alignment.
2516 PCI-PCI bridge can be specified, if resource
2517 windows need to be expanded.
2518 ecrc= Enable/disable PCIe ECRC (transaction layer
2519 end-to-end CRC checking).
2520 bios: Use BIOS/firmware settings. This is the
2524 hpiosize=nn[KMG] The fixed amount of bus space which is
2525 reserved for hotplug bridge's IO window.
2526 Default size is 256 bytes.
2527 hpmemsize=nn[KMG] The fixed amount of bus space which is
2528 reserved for hotplug bridge's memory window.
2529 Default size is 2 megabytes.
2530 realloc= Enable/disable reallocating PCI bridge resources
2531 if allocations done by BIOS are too small to
2532 accommodate resources required by all child
2534 off: Turn realloc off
2536 realloc same as realloc=on
2537 noari do not use PCIe ARI.
2538 pcie_scan_all Scan all possible PCIe devices. Otherwise we
2539 only look for one device below a PCIe downstream
2542 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
2545 force Enable ASPM even on devices that claim not to support it.
2546 WARNING: Forcing ASPM on may cause system lockups.
2548 pcie_hp= [PCIE] PCI Express Hotplug driver options:
2549 nomsi Do not use MSI for PCI Express Native Hotplug (this
2550 makes all PCIe ports use INTx for hotplug services).
2552 pcie_ports= [PCIE] PCIe ports handling:
2553 auto Ask the BIOS whether or not to use native PCIe services
2554 associated with PCIe ports (PME, hot-plug, AER). Use
2555 them only if that is allowed by the BIOS.
2556 native Use native PCIe services associated with PCIe ports
2558 compat Treat PCIe ports as PCI-to-PCI bridges, disable the PCIe
2561 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
2562 nomsi Do not use MSI for native PCIe PME signaling (this makes
2563 all PCIe root ports use INTx for all services).
2565 pcmv= [HW,PCMCIA] BadgePAD 4
2569 Keep all power-domains already enabled by bootloader on,
2570 even if no driver has claimed them. This is useful
2571 for debug and development, but should not be
2572 needed on a platform with proper driver support.
2575 See Documentation/blockdev/paride.txt.
2577 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
2580 See arch/parisc/kernel/pdc_chassis.c
2582 percpu_alloc= Select which percpu first chunk allocator to use.
2583 Currently supported values are "embed" and "page".
2584 Archs may support subset or none of the selections.
2585 See comments in mm/percpu.c for details on each
2586 allocator. This parameter is primarily for debugging
2587 and performance comparison.
2590 See Documentation/blockdev/paride.txt.
2593 See Documentation/blockdev/paride.txt.
2595 pirq= [SMP,APIC] Manual mp-table setup
2596 See Documentation/x86/i386/IO-APIC.txt.
2598 plip= [PPT,NET] Parallel port network link
2599 Format: { parport<nr> | timid | 0 }
2600 See also Documentation/parport.txt.
2602 pmtmr= [X86] Manual setup of pmtmr I/O Port.
2603 Override pmtimer IOPort with a hex value.
2607 Enable PNP debug messages (depends on the
2608 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
2609 via /sys/module/pnp/parameters/debug. We always show
2610 current resource usage; turning this on also shows
2611 possible settings and some assignment information.
2617 { on | off | curr | res | no-curr | no-res }
2620 [ISAPNP] Exclude IRQs for the autoconfiguration
2623 [ISAPNP] Exclude DMAs for the autoconfiguration
2625 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
2626 Ranges are in pairs (I/O port base and size).
2629 [ISAPNP] Exclude memory regions for the
2631 Ranges are in pairs (memory base and size).
2633 ports= [IP_VS_FTP] IPVS ftp helper module
2635 Up to 8 (IP_VS_APP_MAX_PORTS) ports
2637 Format: <port>,<port>....
2639 print-fatal-signals=
2640 [KNL] debug: print fatal signals
2642 If enabled, warn about various signal handling
2643 related application anomalies: too many signals,
2644 too many POSIX.1 timers, fatal signals causing a
2647 If you hit the warning due to signal overflow,
2648 you might want to try "ulimit -i unlimited".
2652 printk.always_kmsg_dump=
2653 Trigger kmsg_dump for cases other than kernel oops or
2655 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
2658 printk.time= Show timing data prefixed to each printk message line
2659 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
2661 processor.max_cstate= [HW,ACPI]
2662 Limit processor to maximum C-state
2663 max_cstate=9 overrides any DMI blacklist limit.
2665 processor.nocst [HW,ACPI]
2666 Ignore the _CST method to determine C-states,
2667 instead using the legacy FADT method
2669 profile= [KNL] Enable kernel profiling via /proc/profile
2670 Format: [schedule,]<number>
2671 Param: "schedule" - profile schedule points.
2672 Param: <number> - step/bucket size as a power of 2 for
2673 statistical time based profiling.
2674 Param: "sleep" - profile D-state sleeping (millisecs).
2675 Requires CONFIG_SCHEDSTATS
2676 Param: "kvm" - profile VM exits.
2678 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
2680 See Documentation/blockdev/ramdisk.txt.
2682 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
2683 probe for; one of (bare|imps|exps|lifebook|any).
2684 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
2686 psmouse.resetafter= [HW,MOUSE]
2687 Try to reset the device after so many bad packets
2690 [HW,MOUSE] Set desired mouse resolution, in dpi.
2691 psmouse.smartscroll=
2692 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
2693 0 = disabled, 1 = enabled (default).
2695 pstore.backend= Specify the name of the pstore backend to use
2698 See Documentation/blockdev/paride.txt.
2701 [KNL] Number of legacy pty's. Overwrites compiled-in
2704 quiet [KNL] Disable most log messages
2709 See Documentation/md.txt.
2711 ramdisk_blocksize= [RAM]
2712 See Documentation/blockdev/ramdisk.txt.
2714 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
2715 See Documentation/blockdev/ramdisk.txt.
2718 In kernels built with CONFIG_RCU_NOCB_CPU=y, set
2719 the specified list of CPUs to be no-callback CPUs.
2720 Invocation of these CPUs' RCU callbacks will
2721 be offloaded to "rcuox/N" kthreads created for
2722 that purpose, where "x" is "b" for RCU-bh, "p"
2723 for RCU-preempt, and "s" for RCU-sched, and "N"
2724 is the CPU number. This reduces OS jitter on the
2725 offloaded CPUs, which can be useful for HPC and
2726 real-time workloads. It can also improve energy
2727 efficiency for asymmetric multiprocessors.
2730 Rather than requiring that offloaded CPUs
2731 (specified by rcu_nocbs= above) explicitly
2732 awaken the corresponding "rcuoN" kthreads,
2733 make these kthreads poll for callbacks.
2734 This improves the real-time response for the
2735 offloaded CPUs by relieving them of the need to
2736 wake up the corresponding kthread, but degrades
2737 energy efficiency by requiring that the kthreads
2738 periodically wake up to do the polling.
2740 rcutree.blimit= [KNL]
2741 Set maximum number of finished RCU callbacks to
2742 process in one batch.
2744 rcutree.rcu_fanout_leaf= [KNL]
2745 Increase the number of CPUs assigned to each
2746 leaf rcu_node structure. Useful for very large
2749 rcutree.jiffies_till_first_fqs= [KNL]
2750 Set delay from grace-period initialization to
2751 first attempt to force quiescent states.
2752 Units are jiffies, minimum value is zero,
2753 and maximum value is HZ.
2755 rcutree.jiffies_till_next_fqs= [KNL]
2756 Set delay between subsequent attempts to force
2757 quiescent states. Units are jiffies, minimum
2758 value is one, and maximum value is HZ.
2760 rcutree.qhimark= [KNL]
2761 Set threshold of queued RCU callbacks beyond which
2762 batch limiting is disabled.
2764 rcutree.qlowmark= [KNL]
2765 Set threshold of queued RCU callbacks below which
2766 batch limiting is re-enabled.
2768 rcutree.rcu_idle_gp_delay= [KNL]
2769 Set wakeup interval for idle CPUs that have
2770 RCU callbacks (RCU_FAST_NO_HZ=y).
2772 rcutree.rcu_idle_lazy_gp_delay= [KNL]
2773 Set wakeup interval for idle CPUs that have
2774 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
2775 Lazy RCU callbacks are those which RCU can
2776 prove do nothing more than free memory.
2778 rcutorture.fqs_duration= [KNL]
2779 Set duration of force_quiescent_state bursts.
2781 rcutorture.fqs_holdoff= [KNL]
2782 Set holdoff time within force_quiescent_state bursts.
2784 rcutorture.fqs_stutter= [KNL]
2785 Set wait time between force_quiescent_state bursts.
2787 rcutorture.gp_exp= [KNL]
2788 Use expedited update-side primitives.
2790 rcutorture.gp_normal= [KNL]
2791 Use normal (non-expedited) update-side primitives.
2792 If both gp_exp and gp_normal are set, do both.
2793 If neither gp_exp nor gp_normal are set, still
2796 rcutorture.n_barrier_cbs= [KNL]
2797 Set callbacks/threads for rcu_barrier() testing.
2799 rcutorture.nfakewriters= [KNL]
2800 Set number of concurrent RCU writers. These just
2801 stress RCU, they don't participate in the actual
2802 test, hence the "fake".
2804 rcutorture.nreaders= [KNL]
2805 Set number of RCU readers.
2807 rcutorture.object_debug= [KNL]
2808 Enable debug-object double-call_rcu() testing.
2810 rcutorture.onoff_holdoff= [KNL]
2811 Set time (s) after boot for CPU-hotplug testing.
2813 rcutorture.onoff_interval= [KNL]
2814 Set time (s) between CPU-hotplug operations, or
2815 zero to disable CPU-hotplug testing.
2817 rcutorture.rcutorture_runnable= [BOOT]
2818 Start rcutorture running at boot time.
2820 rcutorture.shuffle_interval= [KNL]
2821 Set task-shuffle interval (s). Shuffling tasks
2822 allows some CPUs to go into dyntick-idle mode
2823 during the rcutorture test.
2825 rcutorture.shutdown_secs= [KNL]
2826 Set time (s) after boot system shutdown. This
2827 is useful for hands-off automated testing.
2829 rcutorture.stall_cpu= [KNL]
2830 Duration of CPU stall (s) to test RCU CPU stall
2831 warnings, zero to disable.
2833 rcutorture.stall_cpu_holdoff= [KNL]
2834 Time to wait (s) after boot before inducing stall.
2836 rcutorture.stat_interval= [KNL]
2837 Time (s) between statistics printk()s.
2839 rcutorture.stutter= [KNL]
2840 Time (s) to stutter testing, for example, specifying
2841 five seconds causes the test to run for five seconds,
2842 wait for five seconds, and so on. This tests RCU's
2843 ability to transition abruptly to and from idle.
2845 rcutorture.test_boost= [KNL]
2846 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
2847 "Maybe" means test if the RCU implementation
2848 under test support RCU priority boosting.
2850 rcutorture.test_boost_duration= [KNL]
2851 Duration (s) of each individual boost test.
2853 rcutorture.test_boost_interval= [KNL]
2854 Interval (s) between each boost test.
2856 rcutorture.test_no_idle_hz= [KNL]
2857 Test RCU's dyntick-idle handling. See also the
2858 rcutorture.shuffle_interval parameter.
2860 rcutorture.torture_type= [KNL]
2861 Specify the RCU implementation to test.
2863 rcutorture.verbose= [KNL]
2864 Enable additional printk() statements.
2866 rcupdate.rcu_expedited= [KNL]
2867 Use expedited grace-period primitives, for
2868 example, synchronize_rcu_expedited() instead
2869 of synchronize_rcu(). This reduces latency,
2870 but can increase CPU utilization, degrade
2871 real-time latency, and degrade energy efficiency.
2873 rcupdate.rcu_cpu_stall_suppress= [KNL]
2874 Suppress RCU CPU stall warning messages.
2876 rcupdate.rcu_cpu_stall_timeout= [KNL]
2877 Set timeout for RCU CPU stall warning messages.
2881 Run specified binary instead of /init from the ramdisk,
2882 used for early userspace startup. See initrd.
2885 Format (x86 or x86_64):
2886 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
2888 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
2890 Where reboot_mode is one of warm (soft) or cold (hard) or gpio,
2891 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
2892 reboot_force is either force or not specified,
2893 reboot_cpu is s[mp]#### with #### being the processor
2894 to be used for rebooting.
2897 [KNL, SMP] Set scheduler's default relax_domain_level.
2898 See Documentation/cgroups/cpusets.txt.
2900 reserve= [KNL,BUGS] Force the kernel to ignore some iomem area
2902 reservetop= [X86-32]
2904 Reserves a hole at the top of the kernel virtual
2909 Set the amount of memory to reserve for BIOS at
2910 the bottom of the address space.
2912 reset_devices [KNL] Force drivers to reset the underlying device
2913 during initialization.
2916 Specify the partition device for software suspend
2918 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
2920 resume_offset= [SWSUSP]
2921 Specify the offset from the beginning of the partition
2922 given by "resume=" at which the swap header is located,
2923 in <PAGE_SIZE> units (needed only for swap files).
2924 See Documentation/power/swsusp-and-swap-files.txt
2926 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
2927 read the resume files
2929 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
2930 Useful for devices that are detected asynchronously
2931 (e.g. USB and MMC devices).
2933 hibernate= [HIBERNATION]
2934 noresume Don't check if there's a hibernation image
2935 present during boot.
2936 nocompress Don't compress/decompress hibernation images.
2938 retain_initrd [RAM] Keep initrd memory after extraction
2940 rhash_entries= [KNL,NET]
2941 Set number of hash buckets for route cache
2943 ro [KNL] Mount root device read-only on boot
2945 root= [KNL] Root filesystem
2946 See name_to_dev_t comment in init/do_mounts.c.
2948 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
2949 mount the root filesystem
2951 rootflags= [KNL] Set root filesystem mount option string
2953 rootfstype= [KNL] Set root filesystem type
2955 rootwait [KNL] Wait (indefinitely) for root device to show up.
2956 Useful for devices that are detected asynchronously
2957 (e.g. USB and MMC devices).
2959 rproc_mem=nn[KMG][@address]
2960 [KNL,ARM,CMA] Remoteproc physical memory block.
2961 Memory area to be used by remote processor image,
2964 rw [KNL] Mount root device read-write on boot
2966 S [KNL] Run init in single mode
2969 See drivers/net/irda/sa1100_ir.c.
2971 sbni= [NET] Granch SBNI12 leased line adapter
2973 sched_debug [KNL] Enables verbose scheduler debug messages.
2975 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
2976 xtime_lock contention on larger systems, and/or RCU lock
2977 contention on all systems with CONFIG_MAXSMP set.
2978 Format: { "0" | "1" }
2979 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
2981 Note: increases power consumption, thus should only be
2982 enabled if running jitter sensitive (HPC/RT) workloads.
2984 security= [SECURITY] Choose a security module to enable at boot.
2985 If this boot parameter is not specified, only the first
2986 security module asking for security registration will be
2987 loaded. An invalid security module name will be treated
2988 as if no module has been chosen.
2990 selinux= [SELINUX] Disable or enable SELinux at boot time.
2991 Format: { "0" | "1" }
2992 See security/selinux/Kconfig help text.
2995 Default value is set via kernel config option.
2996 If enabled at boot time, /selinux/disable can be used
2997 later to disable prior to initial policy load.
2999 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
3000 Format: { "0" | "1" }
3001 See security/apparmor/Kconfig help text
3004 Default value is set via kernel config option.
3006 serialnumber [BUGS=X86-32]
3009 Maximal number of shapers.
3011 show_msr= [x86] show boot-time MSR settings
3012 Format: { <integer> }
3013 Show boot-time (BIOS-initialized) MSR settings.
3014 The parameter means the number of CPUs to show,
3015 for example 1 means boot CPU only.
3022 slab_max_order= [MM, SLAB]
3023 Determines the maximum allowed order for slabs.
3024 A high setting may cause OOMs due to memory
3025 fragmentation. Defaults to 1 for systems with
3026 more than 32MB of RAM, 0 otherwise.
3028 slub_debug[=options[,slabs]] [MM, SLUB]
3029 Enabling slub_debug allows one to determine the
3030 culprit if slab objects become corrupted. Enabling
3031 slub_debug can create guard zones around objects and
3032 may poison objects when not in use. Also tracks the
3033 last alloc / free. For more information see
3034 Documentation/vm/slub.txt.
3036 slub_max_order= [MM, SLUB]
3037 Determines the maximum allowed order for slabs.
3038 A high setting may cause OOMs due to memory
3039 fragmentation. For more information see
3040 Documentation/vm/slub.txt.
3042 slub_min_objects= [MM, SLUB]
3043 The minimum number of objects per slab. SLUB will
3044 increase the slab order up to slub_max_order to
3045 generate a sufficiently large slab able to contain
3046 the number of objects indicated. The higher the number
3047 of objects the smaller the overhead of tracking slabs
3048 and the less frequently locks need to be acquired.
3049 For more information see Documentation/vm/slub.txt.
3051 slub_min_order= [MM, SLUB]
3052 Determines the minimum page order for slabs. Must be
3053 lower than slub_max_order.
3054 For more information see Documentation/vm/slub.txt.
3056 slub_nomerge [MM, SLUB]
3057 Disable merging of slabs with similar size. May be
3058 necessary if there is some reason to distinguish
3059 allocs to different slabs. Debug options disable
3060 merging on their own.
3061 For more information see Documentation/vm/slub.txt.
3064 Format: <io1>[,<io2>[,...,<io8>]]
3066 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
3067 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
3068 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
3069 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
3070 smsc-ircc2.ircc_irq= [HW] IRQ line
3071 smsc-ircc2.ircc_dma= [HW] DMA channel
3072 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
3073 0: Toshiba Satellite 1800 (GP data pin select)
3074 1: Fast pin select (default)
3078 [KNL] Should the soft-lockup detector generate panics.
3081 sonypi.*= [HW] Sony Programmable I/O Control Device driver
3082 See Documentation/laptops/sonypi.txt
3084 spia_io_base= [HW,MTD]
3090 Enabled the stack tracer on boot up.
3092 stacktrace_filter=[function-list]
3093 [FTRACE] Limit the functions that the stack tracer
3094 will trace at boot up. function-list is a comma separated
3095 list of functions. This list can be changed at run
3096 time by the stack_trace_filter file in the debugfs
3097 tracing directory. Note, this enables stack tracing
3098 and the stacktrace above is not needed.
3102 Set the STI (builtin display/keyboard on the HP-PARISC
3103 machines) console (graphic card) which should be used
3104 as the initial boot-console.
3105 See also comment in drivers/video/console/sticore.c.
3108 See comment in drivers/video/console/sticore.c.
3111 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
3113 sunrpc.min_resvport=
3114 sunrpc.max_resvport=
3116 SunRPC servers often require that client requests
3117 originate from a privileged port (i.e. a port in the
3118 range 0 < portnr < 1024).
3119 An administrator who wishes to reserve some of these
3120 ports for other uses may adjust the range that the
3121 kernel's sunrpc client considers to be privileged
3122 using these two parameters to set the minimum and
3123 maximum port values.
3127 Control how the NFS server code allocates CPUs to
3128 service thread pools. Depending on how many NICs
3129 you have and where their interrupts are bound, this
3130 option will affect which CPUs will do NFS serving.
3131 Note: this parameter cannot be changed while the
3132 NFS server is running.
3134 auto the server chooses an appropriate mode
3135 automatically using heuristics
3136 global a single global pool contains all CPUs
3137 percpu one pool for each CPU
3138 pernode one pool for each NUMA node (equivalent
3139 to global on non-NUMA machines)
3141 sunrpc.tcp_slot_table_entries=
3142 sunrpc.udp_slot_table_entries=
3144 Sets the upper limit on the number of simultaneous
3145 RPC calls that can be sent from the client to a
3146 server. Increasing these values may allow you to
3147 improve throughput, but will also increase the
3148 amount of memory reserved for use by the client.
3151 [KNL] Enable accounting of swap in memory resource
3152 controller if no parameter or 1 is given or disable
3153 it if 0 is given (See Documentation/cgroups/memory.txt)
3155 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
3156 Format: { <int> | force }
3157 <int> -- Number of I/O TLB slabs
3158 force -- force using of bounce buffers even if they
3159 wouldn't be automatically used by the kernel
3163 sysfs.deprecated=0|1 [KNL]
3164 Enable/disable old style sysfs layout for old udev
3165 on older distributions. When this option is enabled
3166 very new udev will not work anymore. When this option
3167 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
3168 in older udev will not work anymore.
3169 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
3170 the kernel configuration.
3172 sysrq_always_enabled
3174 Ignore sysrq setting - this boot parameter will
3175 neutralize any effect of /proc/sys/kernel/sysrq.
3176 Useful for debugging.
3180 test_suspend= [SUSPEND]
3181 Specify "mem" (for Suspend-to-RAM) or "standby" (for
3182 standby suspend) as the system sleep state to briefly
3183 enter during system startup. The system is woken from
3184 this state using a wakeup-capable RTC alarm.
3186 thash_entries= [KNL,NET]
3187 Set number of hash buckets for TCP connection
3189 thermal.act= [HW,ACPI]
3190 -1: disable all active trip points in all thermal zones
3191 <degrees C>: override all lowest active trip points
3193 thermal.crt= [HW,ACPI]
3194 -1: disable all critical trip points in all thermal zones
3195 <degrees C>: override all critical trip points
3197 thermal.nocrt= [HW,ACPI]
3198 Set to disable actions on ACPI thermal zone
3199 critical and hot trip points.
3201 thermal.off= [HW,ACPI]
3202 1: disable ACPI thermal control
3204 thermal.psv= [HW,ACPI]
3205 -1: disable all passive trip points
3206 <degrees C>: override all passive trip points to this
3209 thermal.tzp= [HW,ACPI]
3210 Specify global default ACPI thermal zone polling rate
3211 <deci-seconds>: poll all this frequency
3212 0: no polling (default)
3215 Force threading of all interrupt handlers except those
3216 marked explicitly IRQF_NO_THREAD.
3219 Enable the Transcendent memory driver if built-in.
3221 tmem.cleancache=0|1 [KNL, XEN]
3222 Default is on (1). Disable the usage of the cleancache
3223 API to send anonymous pages to the hypervisor.
3225 tmem.frontswap=0|1 [KNL, XEN]
3226 Default is on (1). Disable the usage of the frontswap
3227 API to send swap pages to the hypervisor. If disabled
3228 the selfballooning and selfshrinking are force disabled.
3230 tmem.selfballooning=0|1 [KNL, XEN]
3231 Default is on (1). Disable the driving of swap pages
3234 tmem.selfshrinking=0|1 [KNL, XEN]
3235 Default is on (1). Partial swapoff that immediately
3236 transfers pages from Xen hypervisor back to the
3237 kernel based on different criteria.
3241 Specify if the kernel should make use of the cpu
3242 topology information if the hardware supports this.
3243 The scheduler will make use of this information and
3244 e.g. base its process migration decisions on it.
3249 tpm_suspend_pcr=[HW,TPM]
3250 Format: integer pcr id
3251 Specify that at suspend time, the tpm driver
3252 should extend the specified pcr with zeros,
3253 as a workaround for some chips which fail to
3254 flush the last written pcr on TPM_SaveState.
3255 This will guarantee that all the other pcrs
3258 trace_buf_size=nn[KMG]
3259 [FTRACE] will set tracing buffer size.
3261 trace_event=[event-list]
3262 [FTRACE] Set and start specified trace events in order
3263 to facilitate early boot debugging.
3264 See also Documentation/trace/events.txt
3266 trace_options=[option-list]
3267 [FTRACE] Enable or disable tracer options at boot.
3268 The option-list is a comma delimited list of options
3269 that can be enabled or disabled just as if you were
3270 to echo the option name into
3272 /sys/kernel/debug/tracing/trace_options
3274 For example, to enable stacktrace option (to dump the
3275 stack trace of each event), add to the command line:
3277 trace_options=stacktrace
3279 See also Documentation/trace/ftrace.txt "trace options"
3283 [FTRACE] enable this option to disable tracing when a
3284 warning is hit. This turns off "tracing_on". Tracing can
3285 be enabled again by echoing '1' into the "tracing_on"
3286 file located in /sys/kernel/debug/tracing/
3288 This option is useful, as it disables the trace before
3289 the WARNING dump is called, which prevents the trace to
3290 be filled with content caused by the warning output.
3292 This option can also be set at run time via the sysctl
3293 option: kernel/traceoff_on_warning
3295 transparent_hugepage=
3297 Format: [always|madvise|never]
3298 Can be used to control the default behavior of the system
3299 with respect to transparent hugepages.
3300 See Documentation/vm/transhuge.txt for more details.
3302 tsc= Disable clocksource stability checks for TSC.
3304 [x86] reliable: mark tsc clocksource as reliable, this
3305 disables clocksource verification at runtime, as well
3306 as the stability checks done at bootup. Used to enable
3307 high-resolution timer mode on older hardware, and in
3308 virtualized environment.
3309 [x86] noirqtime: Do not use TSC to do irq accounting.
3310 Used to run time disable IRQ_TIME_ACCOUNTING on any
3311 platforms where RDTSC is slow and this accounting
3314 turbografx.map[2|3]= [HW,JOY]
3315 TurboGraFX parallel port interface
3317 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
3318 See also Documentation/input/joystick-parport.txt
3320 udbg-immortal [PPC] When debugging early kernel crashes that
3321 happen after console_init() and before a proper
3322 console driver takes over, this boot options might
3323 help "seeing" what's going on.
3325 uhash_entries= [KNL,NET]
3326 Set number of hash buckets for UDP/UDP-Lite connections
3329 [USB] Ignore overcurrent events (default N).
3330 Some badly-designed motherboards generate lots of
3331 bogus events, for ports that aren't wired to
3332 anything. Set this parameter to avoid log spamming.
3333 Note that genuine overcurrent events won't be
3337 [X86] Cause panic on unknown NMI.
3339 usbcore.authorized_default=
3340 [USB] Default USB device authorization:
3341 (default -1 = authorized except for wireless USB,
3342 0 = not authorized, 1 = authorized)
3344 usbcore.autosuspend=
3345 [USB] The autosuspend time delay (in seconds) used
3346 for newly-detected USB devices (default 2). This
3347 is the time required before an idle device will be
3348 autosuspended. Devices for which the delay is set
3349 to a negative value won't be autosuspended at all.
3351 usbcore.usbfs_snoop=
3352 [USB] Set to log all usbfs traffic (default 0 = off).
3354 usbcore.blinkenlights=
3355 [USB] Set to cycle leds on hubs (default 0 = off).
3357 usbcore.old_scheme_first=
3358 [USB] Start with the old device initialization
3359 scheme (default 0 = off).
3361 usbcore.usbfs_memory_mb=
3362 [USB] Memory limit (in MB) for buffers allocated by
3363 usbfs (default = 16, 0 = max = 2047).
3365 usbcore.use_both_schemes=
3366 [USB] Try the other device initialization scheme
3367 if the first one fails (default 1 = enabled).
3369 usbcore.initial_descriptor_timeout=
3370 [USB] Specifies timeout for the initial 64-byte
3371 USB_REQ_GET_DESCRIPTOR request in milliseconds
3372 (default 5000 = 5.0 seconds).
3375 [USBHID] The interval which mice are to be polled at.
3377 usb-storage.delay_use=
3378 [UMS] The delay in seconds before a new device is
3379 scanned for Logical Units (default 5).
3382 [UMS] A list of quirks entries to supplement or
3383 override the built-in unusual_devs list. List
3384 entries are separated by commas. Each entry has
3385 the form VID:PID:Flags where VID and PID are Vendor
3386 and Product ID values (4-digit hex numbers) and
3387 Flags is a set of characters, each corresponding
3388 to a common usb-storage quirk flag as follows:
3389 a = SANE_SENSE (collect more than 18 bytes
3391 b = BAD_SENSE (don't collect more than 18
3392 bytes of sense data);
3393 c = FIX_CAPACITY (decrease the reported
3394 device capacity by one sector);
3395 d = NO_READ_DISC_INFO (don't use
3396 READ_DISC_INFO command);
3397 e = NO_READ_CAPACITY_16 (don't use
3398 READ_CAPACITY_16 command);
3399 h = CAPACITY_HEURISTICS (decrease the
3400 reported device capacity by one
3401 sector if the number is odd);
3402 i = IGNORE_DEVICE (don't bind to this
3404 l = NOT_LOCKABLE (don't try to lock and
3405 unlock ejectable media);
3406 m = MAX_SECTORS_64 (don't transfer more
3407 than 64 sectors = 32 KB at a time);
3408 n = INITIAL_READ10 (force a retry of the
3409 initial READ(10) command);
3410 o = CAPACITY_OK (accept the capacity
3411 reported by the device);
3412 p = WRITE_CACHE (the device cache is ON
3414 r = IGNORE_RESIDUE (the device reports
3415 bogus residue values);
3416 s = SINGLE_LUN (the device has only one
3418 w = NO_WP_DETECT (don't test whether the
3419 medium is write-protected).
3420 Example: quirks=0419:aaf5:rl,0421:0433:rc
3422 user_debug= [KNL,ARM]
3424 See arch/arm/Kconfig.debug help text.
3425 1 - undefined instruction events
3427 4 - invalid data aborts
3430 Example: user_debug=31
3433 [X86] Flags controlling user PTE allocations.
3435 nohigh = do not allocate PTE pages in
3436 HIGHMEM regardless of setting
3440 On X86_32, this is an alias for vdso32=. Otherwise:
3442 vdso=1: enable VDSO (the default)
3443 vdso=0: disable VDSO mapping
3445 vdso32= [X86] Control the 32-bit vDSO
3446 vdso32=1: enable 32-bit VDSO
3447 vdso32=0 or vdso32=2: disable 32-bit VDSO
3449 See the help text for CONFIG_COMPAT_VDSO for more
3450 details. If CONFIG_COMPAT_VDSO is set, the default is
3451 vdso32=0; otherwise, the default is vdso32=1.
3453 For compatibility with older kernels, vdso32=2 is an
3456 Try vdso32=0 if you encounter an error that says:
3457 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
3460 vector=percpu: enable percpu vector domain
3462 video= [FB] Frame buffer configuration
3463 See Documentation/fb/modedb.txt.
3465 video.brightness_switch_enabled= [0,1]
3466 If set to 1, on receiving an ACPI notify event
3467 generated by hotkey, video driver will adjust brightness
3468 level and then send out the event to user space through
3469 the allocated input device; If set to 0, video driver
3470 will only send out the event without touching backlight
3475 [VMMIO] Memory mapped virtio (platform) device.
3477 <size>@<baseaddr>:<irq>[:<id>]
3479 <size> := size (can use standard suffixes
3481 <baseaddr> := physical base address
3482 <irq> := interrupt number (as passed to
3484 <id> := (optional) platform device id
3486 virtio_mmio.device=1K@0x100b0000:48:7
3488 Can be used multiple times for multiple devices.
3490 vga= [BOOT,X86-32] Select a particular video mode
3491 See Documentation/x86/boot.txt and
3492 Documentation/svga.txt.
3493 Use vga=ask for menu.
3494 This is actually a boot loader parameter; the value is
3495 passed to the kernel using a special protocol.
3497 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
3498 size of <nn>. This can be used to increase the
3499 minimum size (128MB on x86). It can also be used to
3500 decrease the size and leave more room for directly
3503 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
3506 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
3509 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
3513 Controls the behavior of vsyscalls (i.e. calls to
3514 fixed addresses of 0xffffffffff600x00 from legacy
3515 code). Most statically-linked binaries and older
3516 versions of glibc use these calls. Because these
3517 functions are at fixed addresses, they make nice
3518 targets for exploits that can control RIP.
3520 emulate [default] Vsyscalls turn into traps and are
3521 emulated reasonably safely.
3523 native Vsyscalls are native syscall instructions.
3524 This is a little bit faster than trapping
3525 and makes a few dynamic recompilers work
3526 better than they would in emulation mode.
3527 It also makes exploits much easier to write.
3529 none Vsyscalls don't work at all. This makes
3530 them quite hard to use for exploits but
3531 might break your system.
3533 vt.color= [VT] Default text color.
3534 Format: 0xYX, X = foreground, Y = background.
3535 Default: 0x07 = light gray on black.
3537 vt.cur_default= [VT] Default cursor shape.
3538 Format: 0xCCBBAA, where AA, BB, and CC are the same as
3539 the parameters of the <Esc>[?A;B;Cc escape sequence;
3540 see VGA-softcursor.txt. Default: 2 = underline.
3542 vt.default_blu= [VT]
3543 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
3544 Change the default blue palette of the console.
3545 This is a 16-member array composed of values
3548 vt.default_grn= [VT]
3549 Format: <green0>,<green1>,<green2>,...,<green15>
3550 Change the default green palette of the console.
3551 This is a 16-member array composed of values
3554 vt.default_red= [VT]
3555 Format: <red0>,<red1>,<red2>,...,<red15>
3556 Change the default red palette of the console.
3557 This is a 16-member array composed of values
3563 Set system-wide default UTF-8 mode for all tty's.
3564 Default is 1, i.e. UTF-8 mode is enabled for all
3565 newly opened terminals.
3567 vt.global_cursor_default=
3570 Set system-wide default for whether a cursor
3571 is shown on new VTs. Default is -1,
3572 i.e. cursors will be created by default unless
3573 overridden by individual drivers. 0 will hide
3574 cursors, 1 will display them.
3576 vt.italic= [VT] Default color for italic text; 0-15.
3579 vt.underline= [VT] Default color for underlined text; 0-15.
3582 watchdog timers [HW,WDT] For information on watchdog timers,
3583 see Documentation/watchdog/watchdog-parameters.txt
3584 or other driver-specific files in the
3585 Documentation/watchdog/ directory.
3587 workqueue.disable_numa
3588 By default, all work items queued to unbound
3589 workqueues are affine to the NUMA nodes they're
3590 issued on, which results in better behavior in
3591 general. If NUMA affinity needs to be disabled for
3592 whatever reason, this option can be used. Note
3593 that this also can be controlled per-workqueue for
3594 workqueues visible under /sys/bus/workqueue/.
3596 workqueue.power_efficient
3597 Per-cpu workqueues are generally preferred because
3598 they show better performance thanks to cache
3599 locality; unfortunately, per-cpu workqueues tend to
3600 be more power hungry than unbound workqueues.
3602 Enabling this makes the per-cpu workqueues which
3603 were observed to contribute significantly to power
3604 consumption unbound, leading to measurably lower
3605 power usage at the cost of small performance
3608 The default value of this parameter is determined by
3609 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
3611 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
3612 default x2apic cluster mode on platforms
3615 x86_intel_mid_timer= [X86-32,APBT]
3616 Choose timer option for x86 Intel MID platform.
3617 Two valid options are apbt timer only and lapic timer
3618 plus one apbt timer for broadcast timer.
3619 x86_intel_mid_timer=apbt_only | lapic_and_apbt
3621 xen_emul_unplug= [HW,X86,XEN]
3622 Unplug Xen emulated devices
3623 Format: [unplug0,][unplug1]
3624 ide-disks -- unplug primary master IDE devices
3625 aux-ide-disks -- unplug non-primary-master IDE devices
3626 nics -- unplug network devices
3627 all -- unplug all emulated devices (NICs and IDE disks)
3628 unnecessary -- unplugging emulated devices is
3629 unnecessary even if the host did not respond to
3631 never -- do not unplug even if version check succeeds
3633 xen_nopvspin [X86,XEN]
3634 Disables the ticketlock slowpath using Xen PV
3637 xirc2ps_cs= [NET,PCMCIA]
3639 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
3641 ______________________________________________________________________
3645 Add more DRM drivers.