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_auto_ssdt [HW,ACPI] Disable automatic loading of SSDT
242 acpica_no_return_repair [HW, ACPI]
243 Disable AML predefined validation mechanism
244 This mechanism can repair the evaluation result to make
245 the return objects more ACPI specification compliant.
246 This option is useful for developers to identify the
247 root cause of an AML interpreter issue when the issue
248 has something to do with the repair mechanism.
250 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
251 Format: To spoof as Windows 98: ="Microsoft Windows"
253 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
254 acpi_osi="string1" # add string1
255 acpi_osi="!string2" # remove string2
256 acpi_osi=!* # remove all strings
257 acpi_osi=! # disable all built-in OS vendor
259 acpi_osi= # disable all strings
261 'acpi_osi=!' can be used in combination with single or
262 multiple 'acpi_osi="string1"' to support specific OS
263 vendor string(s). Note that such command can only
264 affect the default state of the OS vendor strings, thus
265 it cannot affect the default state of the feature group
266 strings and the current state of the OS vendor strings,
267 specifying it multiple times through kernel command line
268 is meaningless. This command is useful when one do not
269 care about the state of the feature group strings which
270 should be controlled by the OSPM.
272 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
273 to 'acpi_osi="Windows 2000" acpi_osi=!', they all
274 can make '_OSI("Windows 2000")' TRUE.
276 'acpi_osi=' cannot be used in combination with other
277 'acpi_osi=' command lines, the _OSI method will not
278 exist in the ACPI namespace. NOTE that such command can
279 only affect the _OSI support state, thus specifying it
280 multiple times through kernel command line is also
283 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
286 'acpi_osi=!*' can be used in combination with single or
287 multiple 'acpi_osi="string1"' to support specific
288 string(s). Note that such command can affect the
289 current state of both the OS vendor strings and the
290 feature group strings, thus specifying it multiple times
291 through kernel command line is meaningful. But it may
292 still not able to affect the final state of a string if
293 there are quirks related to this string. This command
294 is useful when one want to control the state of the
295 feature group strings to debug BIOS issues related to
298 1. 'acpi_osi="Module Device" acpi_osi=!*' can make
299 '_OSI("Module Device")' FALSE.
300 2. 'acpi_osi=!* acpi_osi="Module Device"' can make
301 '_OSI("Module Device")' TRUE.
302 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
304 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
306 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
307 they all will make '_OSI("Windows 2000")' TRUE.
310 Override the pmtimer bug detection: force the kernel
311 to assume that this machine's pmtimer latches its value
312 and always returns good values.
314 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
315 Format: { level | edge | high | low }
317 acpi_skip_timer_override [HW,ACPI]
318 Recognize and ignore IRQ0/pin2 Interrupt Override.
319 For broken nForce2 BIOS resulting in XT-PIC timer.
321 acpi_sleep= [HW,ACPI] Sleep options
322 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
323 old_ordering, nonvs, sci_force_enable }
324 See Documentation/power/video.txt for information on
326 s3_beep is for debugging; it makes the PC's speaker beep
327 as soon as the kernel's real-mode entry point is called.
328 s4_nohwsig prevents ACPI hardware signature from being
329 used during resume from hibernation.
330 old_ordering causes the ACPI 1.0 ordering of the _PTS
331 control method, with respect to putting devices into
332 low power states, to be enforced (the ACPI 2.0 ordering
333 of _PTS is used by default).
334 nonvs prevents the kernel from saving/restoring the
335 ACPI NVS memory during suspend/hibernation and resume.
336 sci_force_enable causes the kernel to set SCI_EN directly
337 on resume from S1/S3 (which is against the ACPI spec,
338 but some broken systems don't work without it).
340 acpi_use_timer_override [HW,ACPI]
341 Use timer override. For some broken Nvidia NF5 boards
342 that require a timer override, but don't have HPET
344 acpi_enforce_resources= [ACPI]
345 { strict | lax | no }
346 Check for resource conflicts between native drivers
347 and ACPI OperationRegions (SystemIO and SystemMemory
348 only). IO ports and memory declared in ACPI might be
349 used by the ACPI subsystem in arbitrary AML code and
350 can interfere with legacy drivers.
351 strict (default): access to resources claimed by ACPI
352 is denied; legacy drivers trying to access reserved
353 resources will fail to bind to device using them.
354 lax: access to resources claimed by ACPI is allowed;
355 legacy drivers trying to access reserved resources
356 will bind successfully but a warning message is logged.
357 no: ACPI OperationRegions are not marked as reserved,
358 no further checks are performed.
360 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump
363 add_efi_memmap [EFI; X86] Include EFI memory map in
364 kernel's map of available physical RAM.
367 { off | try_unsupported }
368 off: disable AGP support
369 try_unsupported: try to drive unsupported chipsets
370 (may crash computer or cause data corruption)
373 See Documentation/sound/alsa/alsa-parameters.txt
376 Allow the default userspace alignment fault handler
377 behaviour to be specified. Bit 0 enables warnings,
378 bit 1 enables fixups, and bit 2 sends a segfault.
380 align_va_addr= [X86-64]
381 Align virtual addresses by clearing slice [14:12] when
382 allocating a VMA at process creation time. This option
383 gives you up to 3% performance improvement on AMD F15h
384 machines (where it is enabled by default) for a
385 CPU-intensive style benchmark, and it can vary highly in
386 a microbenchmark depending on workload and compiler.
388 32: only for 32-bit processes
389 64: only for 64-bit processes
390 on: enable for both 32- and 64-bit processes
391 off: disable for both 32- and 64-bit processes
393 alloc_snapshot [FTRACE]
394 Allocate the ftrace snapshot buffer on boot up when the
395 main buffer is allocated. This is handy if debugging
396 and you need to use tracing_snapshot() on boot up, and
397 do not want to use tracing_snapshot_alloc() as it needs
398 to be done where GFP_KERNEL allocations are allowed.
400 amd_iommu= [HW,X86-64]
401 Pass parameters to the AMD IOMMU driver in the system.
403 fullflush - enable flushing of IO/TLB entries when
404 they are unmapped. Otherwise they are
405 flushed before they will be reused, which
407 off - do not initialize any AMD IOMMU found in
409 force_isolation - Force device isolation for all
410 devices. The IOMMU driver is not
411 allowed anymore to lift isolation
412 requirements as needed. This option
413 does not override iommu=pt
415 amd_iommu_dump= [HW,X86-64]
416 Enable AMD IOMMU driver option to dump the ACPI table
417 for AMD IOMMU. With this option enabled, AMD IOMMU
418 driver will print ACPI tables for AMD IOMMU during
419 IOMMU initialization.
421 amijoy.map= [HW,JOY] Amiga joystick support
422 Map of devices attached to JOY0DAT and JOY1DAT
424 See also Documentation/input/joystick.txt
426 analog.map= [HW,JOY] Analog joystick and gamepad support
427 Specifies type or capabilities of an analog joystick
428 connected to one of 16 gameports
429 Format: <type1>,<type2>,..<type16>
432 Power management functions (SPARCstation-4/5 + deriv.)
434 Disable APC CPU standby support. SPARCstation-Fox does
435 not play well with APC CPU idle - disable it if you have
436 APC and your system crashes randomly.
438 apic= [APIC,X86-32] Advanced Programmable Interrupt Controller
439 Change the output verbosity whilst booting
440 Format: { quiet (default) | verbose | debug }
441 Change the amount of debugging information output
442 when initialising the APIC and IO-APIC components.
445 See Documentation/networking/ipv6.txt.
447 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
448 Limit apic dumping. The parameter defines the maximal
449 number of local apics being dumped. Also it is possible
450 to set it to "all" by meaning -- no limit here.
451 Format: { 1 (default) | 2 | ... | all }.
452 The parameter valid if only apic=debug or
453 apic=verbose is specified.
454 Example: apic=debug show_lapic=all
456 apm= [APM] Advanced Power Management
457 See header of arch/x86/kernel/apm_32.c.
459 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
460 Format: <io>,<irq>,<nodeID>
464 atarimouse= [HW,MOUSE] Atari Mouse
466 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
467 EzKey and similar keyboards
469 atkbd.reset= [HW] Reset keyboard during initialization
471 atkbd.set= [HW] Select keyboard code set
472 Format: <int> (2 = AT (default), 3 = PS/2)
474 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
477 atkbd.softraw= [HW] Choose between synthetic and real raw mode
478 Format: <bool> (0 = real, 1 = synthetic (default))
480 atkbd.softrepeat= [HW]
481 Use software keyboard repeat
483 audit= [KNL] Enable the audit sub-system
484 Format: { "0" | "1" } (0 = disabled, 1 = enabled)
485 0 - kernel audit is disabled and can not be enabled
486 until the next reboot
487 unset - kernel audit is initialized but disabled and
488 will be fully enabled by the userspace auditd.
489 1 - kernel audit is initialized and partially enabled,
490 storing at most audit_backlog_limit messages in
491 RAM until it is fully enabled by the userspace
495 audit_backlog_limit= [KNL] Set the audit queue size limit.
496 Format: <int> (must be >=0)
499 baycom_epp= [HW,AX25]
502 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
504 See header of drivers/net/hamradio/baycom_par.c.
506 baycom_ser_fdx= [HW,AX25]
507 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
508 Format: <io>,<irq>,<mode>[,<baud>]
509 See header of drivers/net/hamradio/baycom_ser_fdx.c.
511 baycom_ser_hdx= [HW,AX25]
512 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
513 Format: <io>,<irq>,<mode>
514 See header of drivers/net/hamradio/baycom_ser_hdx.c.
516 blkdevparts= Manual partition parsing of block device(s) for
517 embedded devices based on command line input.
518 See Documentation/block/cmdline-partition.txt
520 boot_delay= Milliseconds to delay each printk during boot.
521 Values larger than 10 seconds (10000) are changed to
525 bootmem_debug [KNL] Enable bootmem allocator debug messages.
527 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
528 bttv.radio= Most important insmod options are available as
530 bttv.pll= See Documentation/video4linux/bttv/Insmod-options
533 bulk_remove=off [PPC] This parameter disables the use of the pSeries
534 firmware feature for flushing multiple hpte entries
537 c101= [NET] Moxa C101 synchronous serial card
539 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
540 Sometimes CPU hardware bugs make them report the cache
541 size incorrectly. The kernel will attempt work arounds
542 to fix known problems, but for some CPUs it is not
543 possible to determine what the correct size should be.
544 This option provides an override for these situations.
546 ccw_timeout_log [S390]
547 See Documentation/s390/CommonIO for details.
549 cgroup_disable= [KNL] Disable a particular controller
550 Format: {name of the controller(s) to disable}
551 The effects of cgroup_disable=foo are:
552 - foo isn't auto-mounted if you mount all cgroups in
554 - foo isn't visible as an individually mountable
556 {Currently only "memory" controller deal with this and
557 cut the overhead, others just disable the usage. So
558 only cgroup_disable=memory is actually worthy}
560 checkreqprot [SELINUX] Set initial checkreqprot flag value.
561 Format: { "0" | "1" }
562 See security/selinux/Kconfig help text.
563 0 -- check protection applied by kernel (includes
564 any implied execute protection).
565 1 -- check protection requested by application.
566 Default value is set via a kernel config option.
567 Value can be changed at runtime via
568 /selinux/checkreqprot.
571 See Documentation/s390/CommonIO for details.
574 Keep all clocks already enabled by bootloader on,
575 even if no driver has claimed them. This is useful
576 for debug and development, but should not be
577 needed on a platform with proper driver support.
578 For more information, see Documentation/clk.txt.
580 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
582 Forces specified clocksource (if available) to be used
583 when calculating gettimeofday(). If specified
584 clocksource is not available, it defaults to PIT.
585 Format: { pit | tsc | cyclone | pmtmr }
587 clocksource= Override the default clocksource
589 Override the default clocksource and use the clocksource
590 with the name specified.
591 Some clocksource names to choose from, depending on
593 [all] jiffies (this is the base, fallback clocksource)
595 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
596 pxa_timer,timer3,32k_counter,timer0_1
598 [X86-32] pit,hpet,tsc;
599 scx200_hrt on Geode; cyclone on IBM x440
607 clearcpuid=BITNUM [X86]
608 Disable CPUID feature X for the kernel. See
609 arch/x86/include/asm/cpufeature.h for the valid bit
610 numbers. Note the Linux specific bits are not necessarily
611 stable over kernel options, but the vendor specific
613 Also note that user programs calling CPUID directly
614 or using the feature without checking anything
615 will still see it. This just prevents it from
616 being used by the kernel or shown in /proc/cpuinfo.
617 Also note the kernel might malfunction if you disable
621 Sets the size of kernel global memory area for contiguous
622 memory allocations. For more information, see
623 include/linux/dma-contiguous.h
625 cmo_free_hint= [PPC] Format: { yes | no }
626 Specify whether pages are marked as being inactive
627 when they are freed. This is used in CMO environments
628 to determine OS memory pressure for page stealing by
632 coherent_pool=nn[KMG] [ARM,KNL]
633 Sets the size of memory pool for coherent, atomic dma
634 allocations, by default set to 256K.
636 code_bytes [X86] How many bytes of object code to print
641 com20020= [HW,NET] ARCnet - COM20020 chipset
643 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
645 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
649 ARCnet - COM90xx chipset (memory-mapped buffers)
650 Format: <io>[,<irq>[,<memstart>]]
652 condev= [HW,S390] console device
655 console= [KNL] Output console device and options.
657 tty<n> Use the virtual console device <n>.
661 Use the specified serial port. The options are of
662 the form "bbbbpnf", where "bbbb" is the baud rate,
663 "p" is parity ("n", "o", or "e"), "n" is number of
664 bits, and "f" is flow control ("r" for RTS or
665 omit it). Default is "9600n8".
667 See Documentation/serial-console.txt for more
669 Documentation/networking/netconsole.txt for an
672 uart[8250],io,<addr>[,options]
673 uart[8250],mmio,<addr>[,options]
674 Start an early, polled-mode console on the 8250/16550
675 UART at the specified I/O port or MMIO address,
676 switching to the matching ttyS device later. The
677 options are the same as for ttyS, above.
678 hvc<n> Use the hypervisor console device <n>. This is for
679 both Xen and PowerPC hypervisors.
681 If the device connected to the port is not a TTY but a braille
682 device, prepend "brl," before the device type, for instance
684 For now, only VisioBraille is supported.
686 consoleblank= [KNL] The console blank (screen saver) timeout in
687 seconds. Defaults to 10*60 = 10mins. A value of 0
688 disables the blank timer.
691 [KNL] Change the default value for
692 /proc/<pid>/coredump_filter.
693 See also Documentation/filesystems/proc.txt.
695 cpuidle.off=1 [CPU_IDLE]
696 disable the cpuidle sub-system
698 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
700 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
702 crashkernel=size[KMG][@offset[KMG]]
703 [KNL] Using kexec, Linux can switch to a 'crash kernel'
704 upon panic. This parameter reserves the physical
705 memory region [offset, offset + size] for that kernel
706 image. If '@offset' is omitted, then a suitable offset
707 is selected automatically. Check
708 Documentation/kdump/kdump.txt for further details.
710 crashkernel=range1:size1[,range2:size2,...][@offset]
711 [KNL] Same as above, but depends on the memory
712 in the running system. The syntax of range is
713 start-[end] where start and end are both
714 a memory unit (amount[KMG]). See also
715 Documentation/kdump/kdump.txt for an example.
717 crashkernel=size[KMG],high
718 [KNL, x86_64] range could be above 4G. Allow kernel
719 to allocate physical memory region from top, so could
720 be above 4G if system have more than 4G ram installed.
721 Otherwise memory region will be allocated below 4G, if
723 It will be ignored if crashkernel=X is specified.
724 crashkernel=size[KMG],low
725 [KNL, x86_64] range under 4G. When crashkernel=X,high
726 is passed, kernel could allocate physical memory region
727 above 4G, that cause second kernel crash on system
728 that require some amount of low memory, e.g. swiotlb
729 requires at least 64M+32K low memory. Kernel would
730 try to allocate 72M below 4G automatically.
731 This one let user to specify own low range under 4G
732 for second kernel instead.
733 0: to disable low allocation.
734 It will be ignored when crashkernel=X,high is not used
735 or memory reserved is below 4G.
740 cs89x0_media= [HW,NET]
741 Format: { rj45 | aui | bnc }
744 See header of drivers/s390/block/dasd_devmap.c.
746 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
747 (one device per port)
748 Format: <port#>,<type>
749 See also Documentation/input/joystick-parport.txt
751 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
752 time. See Documentation/dynamic-debug-howto.txt for
753 details. Deprecated, see dyndbg.
755 debug [KNL] Enable kernel debugging (events log level).
758 [KNL] verbose self-tests
760 Print debugging info while doing the locking API
762 We default to 0 (no extra messages), setting it to
763 1 will print _a lot_ more information - normally
764 only useful to kernel developers.
766 debug_objects [KNL] Enable object debugging
769 [KNL] Disable object debugging
771 debug_guardpage_minorder=
772 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
773 parameter allows control of the order of pages that will
774 be intentionally kept free (and hence protected) by the
775 buddy allocator. Bigger value increase the probability
776 of catching random memory corruption, but reduce the
777 amount of memory for normal system use. The maximum
778 possible value is MAX_ORDER/2. Setting this parameter
779 to 1 or 2 should be enough to identify most random
780 memory corruption problems caused by bugs in kernel or
781 driver code when a CPU writes to (or reads from) a
782 random memory location. Note that there exists a class
783 of memory corruptions problems caused by buggy H/W or
784 F/W or by drivers badly programing DMA (basically when
785 memory is written at bus level and the CPU MMU is
786 bypassed) which are not detectable by
787 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
788 tracking down these problems.
790 debugpat [X86] Enable PAT debugging
792 decnet.addr= [HW,NET]
793 Format: <area>[,<node>]
794 See also Documentation/networking/decnet.txt.
797 [same as hugepagesz=] The size of the default
798 HugeTLB page size. This is the size represented by
799 the legacy /proc/ hugepages APIs, used for SHM, and
800 default size when mounting hugetlbfs filesystems.
801 Defaults to the default architecture's huge page size
805 Set number of hash buckets for dentry cache.
808 IO parameters + enable/disable command.
810 digiepca= [HW,SERIAL]
811 See drivers/char/README.epca and
812 Documentation/serial/digiepca.txt.
815 See Documentation/networking/ipv6.txt.
817 disable_cpu_apicid= [X86,APIC,SMP]
819 The number of initial APIC ID for the
820 corresponding CPU to be disabled at boot,
821 mostly used for the kdump 2nd kernel to
822 disable BSP to wake up multiple CPUs without
823 causing system reset or hang due to sending
826 disable_ddw [PPC/PSERIES]
827 Disable Dynamic DMA Window support. Use this if
828 to workaround buggy firmware.
831 See Documentation/networking/ipv6.txt.
833 disable_mtrr_cleanup [X86]
834 The kernel tries to adjust MTRR layout from continuous
835 to discrete, to make X server driver able to add WB
836 entry later. This parameter disables that.
838 disable_mtrr_trim [X86, Intel and AMD only]
839 By default the kernel will trim any uncacheable
840 memory out of your available memory pool based on
841 MTRR settings. This parameter disables that behavior,
842 possibly causing your machine to run very slowly.
844 disable_timer_pin_1 [X86]
845 Disable PIN 1 of APIC timer
846 Can be useful to work around chipset bugs.
848 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
849 this option disables the debugging code at boot.
851 dma_debug_entries=<number>
852 This option allows to tune the number of preallocated
853 entries for DMA-API debugging code. One entry is
854 required per DMA-API allocation. Use this if the
855 DMA-API debugging code disables itself because the
856 architectural default is too low.
858 dma_debug_driver=<driver_name>
859 With this option the DMA-API debugging driver
860 filter feature can be enabled at boot time. Just
861 pass the driver to filter for as the parameter.
862 The filter can be disabled or changed to another
863 driver later using sysfs.
865 drm_kms_helper.edid_firmware=[<connector>:]<file>
866 Broken monitors, graphic adapters and KVMs may
867 send no or incorrect EDID data sets. This parameter
868 allows to specify an EDID data set in the
869 /lib/firmware directory that is used instead.
870 Generic built-in EDID data sets are used, if one of
871 edid/1024x768.bin, edid/1280x1024.bin,
872 edid/1680x1050.bin, or edid/1920x1080.bin is given
873 and no file with the same name exists. Details and
874 instructions how to build your own EDID data are
875 available in Documentation/EDID/HOWTO.txt. An EDID
876 data set will only be used for a particular connector,
877 if its name and a colon are prepended to the EDID
882 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
883 module.dyndbg[="val"]
884 Enable debug messages at boot time. See
885 Documentation/dynamic-debug-howto.txt for details.
887 earlycon= [KNL] Output early console device and options.
888 uart[8250],io,<addr>[,options]
889 uart[8250],mmio,<addr>[,options]
890 uart[8250],mmio32,<addr>[,options]
891 Start an early, polled-mode console on the 8250/16550
892 UART at the specified I/O port or MMIO address.
893 MMIO inter-register address stride is either 8-bit
894 (mmio) or 32-bit (mmio32).
895 The options are the same as for ttyS, above.
897 earlyprintk= [X86,SH,BLACKFIN,ARM]
901 earlyprintk=serial[,ttySn[,baudrate]]
902 earlyprintk=serial[,0x...[,baudrate]]
903 earlyprintk=ttySn[,baudrate]
904 earlyprintk=dbgp[debugController#]
906 earlyprintk is useful when the kernel crashes before
907 the normal console is initialized. It is not enabled by
908 default because it has some cosmetic problems.
910 Append ",keep" to not disable it when the real console
913 Only one of vga, efi, serial, or usb debug port can
916 Currently only ttyS0 and ttyS1 may be specified by
917 name. Other I/O ports may be explicitly specified
918 on some architectures (x86 and arm at least) by
919 replacing ttySn with an I/O port address, like this:
920 earlyprintk=serial,0x1008,115200
921 You can find the port for a given device in
922 /proc/tty/driver/serial:
923 2: uart:ST16650V2 port:00001008 irq:18 ...
925 Interaction with the standard serial driver is not
928 The VGA and EFI output is eventually overwritten by
931 The xen output can only be used by Xen PV guests.
933 edac_report= [HW,EDAC] Control how to report EDAC event
934 Format: {"on" | "off" | "force"}
935 on: enable EDAC to report H/W event. May be overridden
936 by other higher priority error reporting module.
937 off: disable H/W event reporting through EDAC.
938 force: enforce the use of EDAC to report H/W event.
941 ekgdboc= [X86,KGDB] Allow early kernel console debugging
944 This is designed to be used in conjunction with
945 the boot argument: earlyprintk=vga
948 Format: {"off" | "on" | "skip[mbr]"}
951 Format: { "old_map" }
952 old_map [X86-64]: switch to the old ioremap-based EFI
953 runtime services mapping. 32-bit still uses this one by
956 efi_no_storage_paranoia [EFI; X86]
957 Using this parameter you can use more than 50% of
958 your efi variable storage. Use this parameter only if
959 you are really sure that your UEFI does sane gc and
960 fulfills the spec otherwise your board may brick.
962 eisa_irq_edge= [PARISC,HW]
963 See header of drivers/parisc/eisa.c.
966 See comment before function elanfreq_setup() in
967 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
970 Format: {"cfq" | "deadline" | "noop"}
971 See Documentation/block/cfq-iosched.txt and
972 Documentation/block/deadline-iosched.txt for details.
974 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
975 Specifies physical address of start of kernel core
976 image elf header and optionally the size. Generally
977 kexec loader will pass this option to capture kernel.
978 See Documentation/kdump/kdump.txt for details.
980 enable_mtrr_cleanup [X86]
981 The kernel tries to adjust MTRR layout from continuous
982 to discrete, to make X server driver able to add WB
983 entry later. This parameter enables that.
985 enable_timer_pin_1 [X86]
986 Enable PIN 1 of APIC timer
987 Can be useful to work around chipset bugs
988 (in particular on some ATI chipsets).
989 The kernel tries to set a reasonable default.
991 enforcing [SELINUX] Set initial enforcing status.
993 See security/selinux/Kconfig help text.
994 0 -- permissive (log only, no denials).
995 1 -- enforcing (deny and log).
997 Value can be changed at runtime via /selinux/enforce.
1000 Disable Error Record Serialization Table (ERST)
1003 ether= [HW,NET] Ethernet cards parameters
1004 This option is obsoleted by the "netdev=" option, which
1005 has equivalent usage. See its documentation for details.
1009 Permit 'security.evm' to be updated regardless of
1010 current integrity status.
1014 fail_make_request=[KNL]
1015 General fault injection mechanism.
1016 Format: <interval>,<probability>,<space>,<times>
1017 See also Documentation/fault-injection/.
1020 See Documentation/blockdev/floppy.txt.
1022 force_pal_cache_flush
1023 [IA-64] Avoid check_sal_cache_flush which may hang on
1024 buggy SAL_CACHE_FLUSH implementations. Using this
1025 parameter will force ia64_sal_cache_flush to call
1026 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1029 [FTRACE] will set and start the specified tracer
1030 as early as possible in order to facilitate early
1033 ftrace_dump_on_oops[=orig_cpu]
1034 [FTRACE] will dump the trace buffers on oops.
1035 If no parameter is passed, ftrace will dump
1036 buffers of all CPUs, but if you pass orig_cpu, it will
1037 dump only the buffer of the CPU that triggered the
1040 ftrace_filter=[function-list]
1041 [FTRACE] Limit the functions traced by the function
1042 tracer at boot up. function-list is a comma separated
1043 list of functions. This list can be changed at run
1044 time by the set_ftrace_filter file in the debugfs
1047 ftrace_notrace=[function-list]
1048 [FTRACE] Do not trace the functions specified in
1049 function-list. This list can be changed at run time
1050 by the set_ftrace_notrace file in the debugfs
1053 ftrace_graph_filter=[function-list]
1054 [FTRACE] Limit the top level callers functions traced
1055 by the function graph tracer at boot up.
1056 function-list is a comma separated list of functions
1057 that can be changed at run time by the
1058 set_graph_function file in the debugfs tracing directory.
1061 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1062 support via parallel port (up to 5 devices per port)
1063 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1064 See also Documentation/input/joystick-parport.txt
1068 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART
1072 gcov_persist= [GCOV] When non-zero (default), profiling data for
1073 kernel modules is saved and remains accessible via
1074 debugfs, even when the module is unloaded/reloaded.
1075 When zero, profiling data is discarded and associated
1076 debugfs files are removed at module unload time.
1078 gpt [EFI] Forces disk with valid GPT signature but
1079 invalid Protective MBR to be treated as GPT. If the
1080 primary GPT is corrupted, it enables the backup/alternate
1081 GPT to be used instead.
1083 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1084 the "Enable 0" bit of the configuration register.
1087 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1088 the "Enable 0" bit of the configuration register.
1091 grcan.select= [HW] Select which physical interface to use.
1094 grcan.txsize= [HW] Sets the size of the tx buffer.
1095 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1097 grcan.rxsize= [HW] Sets the size of the rx buffer.
1098 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1101 hashdist= [KNL,NUMA] Large hashes allocated during boot
1102 are distributed across NUMA nodes. Defaults on
1103 for 64-bit NUMA, off otherwise.
1104 Format: 0 | 1 (for off | on)
1106 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1108 hd= [EIDE] (E)IDE hard drive subsystem geometry
1109 Format: <cyl>,<head>,<sect>
1112 Disable Hardware Error Source Table (HEST) support;
1113 corresponding firmware-first mode error processing
1114 logic will be disabled.
1116 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1117 size of <nn>. This works even on boxes that have no
1118 highmem otherwise. This also works to reduce highmem
1119 size on bigger boxes.
1121 highres= [KNL] Enable/disable high resolution timer mode.
1122 Valid parameters: "on", "off"
1126 See Documentation/isdn/README.HiSax.
1130 hpet= [X86-32,HPET] option to control HPET usage
1131 Format: { enable (default) | disable | force |
1133 disable: disable HPET and use PIT instead
1134 force: allow force enabled of undocumented chips (ICH4,
1136 verbose: show contents of HPET registers during setup
1138 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1139 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1141 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
1142 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
1143 On x86-64 and powerpc, this option can be specified
1144 multiple times interleaved with hugepages= to reserve
1145 huge pages of different sizes. Valid pages sizes on
1146 x86-64 are 2M (when the CPU supports "pse") and 1G
1147 (when the CPU supports the "pdpe1gb" cpuinfo flag)
1148 Note that 1GB pages can only be allocated at boot time
1149 using hugepages= and not freed afterwards.
1151 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1152 terminal devices. Valid values: 0..8
1153 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1154 If specified, z/VM IUCV HVC accepts connections
1155 from listed z/VM user IDs only.
1157 hwthread_map= [METAG] Comma-separated list of Linux cpu id to
1158 hardware thread id mappings.
1159 Format: <cpu>:<hwthread>
1162 Do not unregister boot console at start. This is only
1163 useful for debugging when something happens in the window
1164 between unregistering the boot console and initializing
1167 i2c_bus= [HW] Override the default board specific I2C bus speed
1168 or register an additional I2C bus that is not
1169 registered from board initialization code.
1173 i8042.debug [HW] Toggle i8042 debug mode
1174 i8042.direct [HW] Put keyboard port into non-translated mode
1175 i8042.dumbkbd [HW] Pretend that controller can only read data from
1176 keyboard and cannot control its state
1177 (Don't attempt to blink the leds)
1178 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1179 i8042.nokbd [HW] Don't check/create keyboard port
1180 i8042.noloop [HW] Disable the AUX Loopback command while probing
1182 i8042.nomux [HW] Don't check presence of an active multiplexing
1184 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1186 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1187 i8042.reset [HW] Reset the controller during init and cleanup
1188 i8042.unlock [HW] Unlock (ignore) the keylock
1192 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
1193 indicates that the driver is running on unsupported
1195 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
1196 does not match list of supported models.
1198 [HW] Report power status in /proc/i8k
1199 (disabled by default)
1200 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
1203 i915.invert_brightness=
1204 [DRM] Invert the sense of the variable that is used to
1205 set the brightness of the panel backlight. Normally a
1206 brightness value of 0 indicates backlight switched off,
1207 and the maximum of the brightness value sets the backlight
1208 to maximum brightness. If this parameter is set to 0
1209 (default) and the machine requires it, or this parameter
1210 is set to 1, a brightness value of 0 sets the backlight
1211 to maximum brightness, and the maximum of the brightness
1212 value switches the backlight off.
1213 -1 -- never invert brightness
1214 0 -- machine default
1215 1 -- force brightness inversion
1218 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1220 ide-core.nodma= [HW] (E)IDE subsystem
1221 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1222 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1223 .cdrom .chs .ignore_cable are additional options
1224 See Documentation/ide/ide.txt.
1226 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1227 Claim all unknown PCI IDE storage controllers.
1230 Format: idle=poll, idle=halt, idle=nomwait
1231 Poll forces a polling idle loop that can slightly
1232 improve the performance of waking up a idle CPU, but
1233 will use a lot of power and make the system run hot.
1235 idle=halt: Halt is forced to be used for CPU idle.
1236 In such case C2/C3 won't be used again.
1237 idle=nomwait: Disable mwait for CPU C-states
1239 ignore_loglevel [KNL]
1240 Ignore loglevel setting - this will print /all/
1241 kernel messages to the console. Useful for debugging.
1242 We also add it as printk module parameter, so users
1243 could change it dynamically, usually by
1244 /sys/module/printk/parameters/ignore_loglevel.
1246 ihash_entries= [KNL]
1247 Set number of hash buckets for inode cache.
1249 ima_appraise= [IMA] appraise integrity measurements
1250 Format: { "off" | "enforce" | "fix" }
1253 ima_appraise_tcb [IMA]
1254 The builtin appraise policy appraises all files
1258 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1262 The list of supported hash algorithms is defined
1263 in crypto/hash_info.h.
1266 Load a policy which meets the needs of the Trusted
1267 Computing Base. This means IMA will measure all
1268 programs exec'd, files mmap'd for exec, and all files
1269 opened for read by uid=0.
1272 Select one of defined IMA measurements template formats.
1273 Formats: { "ima" | "ima-ng" }
1278 Run specified binary instead of /sbin/init as init
1281 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1282 for working out where the kernel is dying during
1285 initrd= [BOOT] Specify the location of the initial ramdisk
1287 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1290 int_pln_enable [x86] Enable power limit notification interrupt
1292 integrity_audit=[IMA]
1293 Format: { "0" | "1" }
1294 0 -- basic integrity auditing messages. (Default)
1295 1 -- additional integrity auditing messages.
1297 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
1299 Enable intel iommu driver.
1301 Disable intel iommu driver.
1302 igfx_off [Default Off]
1303 By default, gfx is mapped as normal device. If a gfx
1304 device has a dedicated DMAR unit, the DMAR unit is
1305 bypassed by not enabling DMAR with this option. In
1306 this case, gfx device will use physical address for
1309 With this option iommu will not optimize to look
1310 for io virtual address below 32-bit forcing dual
1311 address cycle on pci bus for cards supporting greater
1312 than 32-bit addressing. The default is to look
1313 for translation below 32-bit and if not available
1314 then look in the higher range.
1315 strict [Default Off]
1316 With this option on every unmap_single operation will
1317 result in a hardware IOTLB flush operation as opposed
1318 to batching them for performance.
1319 sp_off [Default Off]
1320 By default, super page will be supported if Intel IOMMU
1321 has the capability. With this option, super page will
1324 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
1325 0 disables intel_idle and fall back on acpi_idle.
1326 1 to 6 specify maximum depth of C-state.
1330 Do not enable intel_pstate as the default
1331 scaling driver for the supported processors
1333 intremap= [X86-64, Intel-IOMMU]
1334 on enable Interrupt Remapping (default)
1335 off disable Interrupt Remapping
1336 nosid disable Source ID checking
1338 BIOS x2APIC opt-out request will be ignored
1340 iomem= Disable strict checking of access to MMIO memory
1341 strict regions from userspace.
1358 io7= [HW] IO7 for Marvel based alpha systems
1359 See comment before marvel_specify_io7 in
1360 arch/alpha/kernel/core_marvel.c.
1362 io_delay= [X86] I/O delay method
1364 Standard port 0x80 based delay
1366 Alternate port 0xed based delay (needed on some systems)
1368 Simple two microseconds delay
1373 See Documentation/filesystems/nfs/nfsroot.txt.
1375 ip2= [HW] Set IO/IRQ pairs for up to 4 IntelliPort boards
1376 See comment before ip2_setup() in
1377 drivers/char/ip2/ip2base.c.
1380 When an interrupt is not handled search all handlers
1381 for it. Intended to get systems with badly broken
1385 When an interrupt is not handled search all handlers
1386 for it. Also check all handlers each timer
1387 interrupt. Intended to get systems with badly broken
1391 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
1393 isolcpus= [KNL,SMP] Isolate CPUs from the general scheduler.
1395 <cpu number>,...,<cpu number>
1397 <cpu number>-<cpu number>
1398 (must be a positive range in ascending order)
1400 <cpu number>,...,<cpu number>-<cpu number>
1402 This option can be used to specify one or more CPUs
1403 to isolate from the general SMP balancing and scheduling
1404 algorithms. You can move a process onto or off an
1405 "isolated" CPU via the CPU affinity syscalls or cpuset.
1406 <cpu number> begins at 0 and the maximum value is
1407 "number of CPUs in system - 1".
1409 This option is the preferred way to isolate CPUs. The
1410 alternative -- manually setting the CPU mask of all
1411 tasks in the system -- can cause problems and
1412 suboptimal load balancer performance.
1416 ivrs_ioapic [HW,X86_64]
1417 Provide an override to the IOAPIC-ID<->DEVICE-ID
1418 mapping provided in the IVRS ACPI table. For
1419 example, to map IOAPIC-ID decimal 10 to
1420 PCI device 00:14.0 write the parameter as:
1421 ivrs_ioapic[10]=00:14.0
1423 ivrs_hpet [HW,X86_64]
1424 Provide an override to the HPET-ID<->DEVICE-ID
1425 mapping provided in the IVRS ACPI table. For
1426 example, to map HPET-ID decimal 0 to
1427 PCI device 00:14.0 write the parameter as:
1428 ivrs_hpet[0]=00:14.0
1430 js= [HW,JOY] Analog joystick
1431 See Documentation/input/joystick.txt.
1435 kernelcore=nn[KMG] [KNL,X86,IA-64,PPC] This parameter
1436 specifies the amount of memory usable by the kernel
1437 for non-movable allocations. The requested amount is
1438 spread evenly throughout all nodes in the system. The
1439 remaining memory in each node is used for Movable
1440 pages. In the event, a node is too small to have both
1441 kernelcore and Movable pages, kernelcore pages will
1442 take priority and other nodes will have a larger number
1443 of Movable pages. The Movable zone is used for the
1444 allocation of pages that may be reclaimed or moved
1445 by the page migration subsystem. This means that
1446 HugeTLB pages may not be allocated from this zone.
1447 Note that allocations like PTEs-from-HighMem still
1448 use the HighMem zone if it exists, and the Normal
1449 zone if it does not.
1451 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
1452 Format: <Controller#>[,poll interval]
1453 The controller # is the number of the ehci usb debug
1454 port as it is probed via PCI. The poll interval is
1455 optional and is the number seconds in between
1456 each poll cycle to the debug port in case you need
1457 the functionality for interrupting the kernel with
1458 gdb or control-c on the dbgp connection. When
1459 not using this parameter you use sysrq-g to break into
1460 the kernel debugger.
1462 kgdboc= [KGDB,HW] kgdb over consoles.
1463 Requires a tty driver that supports console polling,
1464 or a supported polling keyboard driver (non-usb).
1465 Serial only format: <serial_device>[,baud]
1466 keyboard only format: kbd
1467 keyboard and serial format: kbd,<serial_device>[,baud]
1468 Optional Kernel mode setting:
1469 kms, kbd format: kms,kbd
1470 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
1472 kgdbwait [KGDB] Stop kernel execution and enter the
1473 kernel debugger at the earliest opportunity.
1475 kmac= [MIPS] korina ethernet MAC address.
1476 Configure the RouterBoard 532 series on-chip
1477 Ethernet adapter MAC address.
1479 kmemleak= [KNL] Boot-time kmemleak enable/disable
1480 Valid arguments: on, off
1483 kmemcheck= [X86] Boot-time kmemcheck enable/disable/one-shot mode
1484 Valid arguments: 0, 1, 2
1485 kmemcheck=0 (disabled)
1486 kmemcheck=1 (enabled)
1487 kmemcheck=2 (one-shot mode)
1488 Default: 2 (one-shot mode)
1490 kstack=N [X86] Print N words from the kernel stack
1493 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
1494 Default is 0 (don't ignore, but inject #GP)
1496 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit
1500 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
1501 Default is 1 (enabled)
1503 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
1505 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
1507 kvm-intel.ept= [KVM,Intel] Disable extended page tables
1508 (virtualized MMU) support on capable Intel chips.
1509 Default is 1 (enabled)
1511 kvm-intel.emulate_invalid_guest_state=
1512 [KVM,Intel] Enable emulation of invalid guest states
1513 Default is 0 (disabled)
1515 kvm-intel.flexpriority=
1516 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
1517 Default is 1 (enabled)
1520 [KVM,Intel] Enable VMX nesting (nVMX).
1521 Default is 0 (disabled)
1523 kvm-intel.unrestricted_guest=
1524 [KVM,Intel] Disable unrestricted guest feature
1525 (virtualized real and unpaged mode) on capable
1526 Intel chips. Default is 1 (enabled)
1528 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
1529 feature (tagged TLBs) on capable Intel chips.
1530 Default is 1 (enabled)
1536 lapic [X86-32,APIC] Enable the local APIC even if BIOS
1539 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline
1540 value for LAPIC timer one-shot implementation. Default
1541 back to the programmable timer unit in the LAPIC.
1543 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
1546 libata.dma= [LIBATA] DMA control
1547 libata.dma=0 Disable all PATA and SATA DMA
1548 libata.dma=1 PATA and SATA Disk DMA only
1549 libata.dma=2 ATAPI (CDROM) DMA only
1550 libata.dma=4 Compact Flash DMA only
1551 Combinations also work, so libata.dma=3 enables DMA
1552 for disks and CDROMs, but not CFs.
1554 libata.ignore_hpa= [LIBATA] Ignore HPA limit
1555 libata.ignore_hpa=0 keep BIOS limits (default)
1556 libata.ignore_hpa=1 ignore limits, using full disk
1558 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
1562 libata.force= [LIBATA] Force configurations. The format is comma
1563 separated list of "[ID:]VAL" where ID is
1564 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
1565 matching port, link or device. Basically, it matches
1566 the ATA ID string printed on console by libata. If
1567 the whole ID part is omitted, the last PORT and DEVICE
1568 values are used. If ID hasn't been specified yet, the
1569 configuration applies to all ports, links and devices.
1571 If only DEVICE is omitted, the parameter applies to
1572 the port and all links and devices behind it. DEVICE
1573 number of 0 either selects the first device or the
1574 first fan-out link behind PMP device. It does not
1575 select the host link. DEVICE number of 15 selects the
1576 host link and device attached to it.
1578 The VAL specifies the configuration to force. As long
1579 as there's no ambiguity shortcut notation is allowed.
1580 For example, both 1.5 and 1.5G would work for 1.5Gbps.
1581 The following configurations can be forced.
1583 * Cable type: 40c, 80c, short40c, unk, ign or sata.
1584 Any ID with matching PORT is used.
1586 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
1588 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
1589 udma[/][16,25,33,44,66,100,133] notation is also
1592 * [no]ncq: Turn on or off NCQ.
1594 * nohrst, nosrst, norst: suppress hard, soft
1597 * rstonce: only attempt one reset during
1598 hot-unplug link recovery
1600 * dump_id: dump IDENTIFY data.
1602 * atapi_dmadir: Enable ATAPI DMADIR bridge support
1604 * disable: Disable this device.
1606 If there are multiple matching configurations changing
1607 the same attribute, the last one is used.
1609 memblock=debug [KNL] Enable memblock debug messages.
1611 load_ramdisk= [RAM] List of ramdisks to load from floppy
1612 See Documentation/blockdev/ramdisk.txt.
1614 lockd.nlm_grace_period=P [NFS] Assign grace period.
1617 lockd.nlm_tcpport=N [NFS] Assign TCP port.
1620 lockd.nlm_timeout=T [NFS] Assign timeout value.
1623 lockd.nlm_udpport=M [NFS] Assign UDP port.
1626 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
1629 loglevel= All Kernel Messages with a loglevel smaller than the
1630 console loglevel will be printed to the console. It can
1631 also be changed with klogd or other programs. The
1632 loglevels are defined as follows:
1634 0 (KERN_EMERG) system is unusable
1635 1 (KERN_ALERT) action must be taken immediately
1636 2 (KERN_CRIT) critical conditions
1637 3 (KERN_ERR) error conditions
1638 4 (KERN_WARNING) warning conditions
1639 5 (KERN_NOTICE) normal but significant condition
1640 6 (KERN_INFO) informational
1641 7 (KERN_DEBUG) debug-level messages
1643 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
1644 in bytes. n must be a power of two. The default
1645 size is set in the kernel config file.
1647 logo.nologo [FB] Disables display of the built-in Linux logo.
1648 This may be used to provide more screen space for
1649 kernel log messages and is useful when debugging
1650 kernel boot problems.
1652 lp=0 [LP] Specify parallel ports to use, e.g,
1653 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
1654 lp=reset first parallel port). 'lp=0' disables the
1655 lp=auto printer driver. 'lp=reset' (which can be
1656 specified in addition to the ports) causes
1657 attached printers to be reset. Using
1658 lp=port1,port2,... specifies the parallel ports
1659 to associate lp devices with, starting with
1660 lp0. A port specification may be 'none' to skip
1661 that lp device, or a parport name such as
1662 'parport0'. Specifying 'lp=auto' instead of a
1663 port specification list means that device IDs
1664 from each port should be examined, to see if
1665 an IEEE 1284-compliant printer is attached; if
1666 so, the driver will manage that printer.
1667 See also header of drivers/char/lp.c.
1670 Sets loops_per_jiffy to given constant, thus avoiding
1671 time-consuming boot-time autodetection (up to 250 ms per
1672 CPU). 0 enables autodetection (default). To determine
1673 the correct value for your kernel, boot with normal
1674 autodetection and see what value is printed. Note that
1675 on SMP systems the preset will be applied to all CPUs,
1676 which is likely to cause problems if your CPUs need
1677 significantly divergent settings. An incorrect value
1678 will cause delays in the kernel to be wrong, leading to
1679 unpredictable I/O errors and other breakage. Although
1680 unlikely, in the extreme case this might damage your
1684 Format: <io>,<irq>,<dma>
1686 machvec= [IA-64] Force the use of a particular machine-vector
1687 (machvec) in a generic kernel.
1688 Example: machvec=hpzx1_swiotlb
1690 machtype= [Loongson] Share the same kernel image file between different
1692 Example: machtype=lemote-yeeloong-2f-7inch
1694 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
1695 than or equal to this physical address is ignored.
1697 maxcpus= [SMP] Maximum number of processors that an SMP kernel
1698 should make use of. maxcpus=n : n >= 0 limits the
1699 kernel to using 'n' processors. n=0 is a special case,
1700 it is equivalent to "nosmp", which also disables
1703 max_loop= [LOOP] The number of loop block devices that get
1704 (loop.max_loop) unconditionally pre-created at init time. The default
1705 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
1706 of statically allocating a predefined number, loop
1707 devices can be requested on-demand with the
1708 /dev/loop-control interface.
1710 mce [X86-32] Machine Check Exception
1712 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.txt
1714 md= [HW] RAID subsystems devices and level
1715 See Documentation/md.txt.
1718 Format: <first>,<last>
1719 Specifies range of consoles to be captured by the MDA.
1721 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
1722 Amount of memory to be used when the kernel is not able
1723 to see the whole system memory or for test.
1724 [X86] Work as limiting max address. Use together
1725 with memmap= to avoid physical address space collisions.
1726 Without memmap= PCI devices could be placed at addresses
1727 belonging to unused RAM.
1729 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
1733 [KNL,SH] Allow user to override the default size for
1734 per-device physically contiguous DMA buffers.
1736 memmap=exactmap [KNL,X86] Enable setting of an exact
1737 E820 memory map, as specified by the user.
1738 Such memmap=exactmap lines can be constructed based on
1739 BIOS output or other requirements. See the memmap=nn@ss
1742 memmap=nn[KMG]@ss[KMG]
1743 [KNL] Force usage of a specific region of memory.
1744 Region of memory to be used is from ss to ss+nn.
1746 memmap=nn[KMG]#ss[KMG]
1747 [KNL,ACPI] Mark specific memory as ACPI data.
1748 Region of memory to be marked is from ss to ss+nn.
1750 memmap=nn[KMG]$ss[KMG]
1751 [KNL,ACPI] Mark specific memory as reserved.
1752 Region of memory to be reserved is from ss to ss+nn.
1753 Example: Exclude memory from 0x18690000-0x1869ffff
1754 memmap=64K$0x18690000
1756 memmap=0x10000$0x18690000
1758 memory_corruption_check=0/1 [X86]
1759 Some BIOSes seem to corrupt the first 64k of
1760 memory when doing things like suspend/resume.
1761 Setting this option will scan the memory
1762 looking for corruption. Enabling this will
1763 both detect corruption and prevent the kernel
1764 from using the memory being corrupted.
1765 However, its intended as a diagnostic tool; if
1766 repeatable BIOS-originated corruption always
1767 affects the same memory, you can use memmap=
1768 to prevent the kernel from using that memory.
1770 memory_corruption_check_size=size [X86]
1771 By default it checks for corruption in the low
1772 64k, making this memory unavailable for normal
1773 use. Use this parameter to scan for
1774 corruption in more or less memory.
1776 memory_corruption_check_period=seconds [X86]
1777 By default it checks for corruption every 60
1778 seconds. Use this parameter to check at some
1779 other rate. 0 disables periodic checking.
1781 memtest= [KNL,X86] Enable memtest
1783 default : 0 <disable>
1784 Specifies the number of memtest passes to be
1785 performed. Each pass selects another test
1786 pattern from a given set of patterns. Memtest
1787 fills the memory with this pattern, validates
1788 memory contents and reserves bad memory
1789 regions that are detected.
1791 meye.*= [HW] Set MotionEye Camera parameters
1792 See Documentation/video4linux/meye.txt.
1794 mfgpt_irq= [IA-32] Specify the IRQ to use for the
1795 Multi-Function General Purpose Timers on AMD Geode
1798 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
1799 the BIOS has incorrectly applied a workaround. TinyBIOS
1800 version 0.98 is known to be affected, 0.99 fixes the
1801 problem by letting the user disable the workaround.
1805 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
1806 physical address is ignored.
1808 mini2440= [ARM,HW,KNL]
1809 Format:[0..2][b][c][t]
1811 MINI2440 configuration specification:
1812 0 - The attached screen is the 3.5" TFT
1813 1 - The attached screen is the 7" TFT
1814 2 - The VGA Shield is attached (1024x768)
1815 Leaving out the screen size parameter will not load
1816 the TFT driver, and the framebuffer will be left
1818 b - Enable backlight. The TFT backlight pin will be
1819 linked to the kernel VESA blanking code and a GPIO
1820 LED. This parameter is not necessary when using the
1822 c - Enable the s3c camera interface.
1823 t - Reserved for enabling touchscreen support. The
1824 touchscreen support is not enabled in the mainstream
1825 kernel as of 2.6.30, a preliminary port can be found
1826 in the "bleeding edge" mini2440 support kernel at
1827 http://repo.or.cz/w/linux-2.6/mini2440.git
1830 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
1831 parameter allows control of the logging verbosity for
1832 the additional memory initialisation checks. A value
1833 of 0 disables mminit logging and a level of 4 will
1834 log everything. Information is printed at KERN_DEBUG
1835 so loglevel=8 may also need to be specified.
1838 [KNL] When CONFIG_MODULE_SIG is set, this means that
1839 modules without (valid) signatures will fail to load.
1840 Note that if CONFIG_MODULE_SIG_FORCE is set, that
1841 is always true, so this option does nothing.
1844 [MOUSE] Maximum time between finger touching and
1845 leaving touchpad surface for touch to be considered
1846 a tap and be reported as a left button click (for
1847 touchpads working in absolute mode only).
1849 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
1850 reporting absolute coordinates, such as tablets
1851 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
1852 reporting absolute coordinates, such as tablets
1854 movablecore=nn[KMG] [KNL,X86,IA-64,PPC] This parameter
1855 is similar to kernelcore except it specifies the
1856 amount of memory used for migratable allocations.
1857 If both kernelcore and movablecore is specified,
1858 then kernelcore will be at *least* the specified
1859 value but may be more. If movablecore on its own
1860 is specified, the administrator must be careful
1861 that the amount of memory usable for all allocations
1864 movable_node [KNL,X86] Boot-time switch to enable the effects
1865 of CONFIG_MOVABLE_NODE=y. See mm/Kconfig for details.
1867 MTD_Partition= [MTD]
1868 Format: <name>,<region-number>,<size>,<offset>
1870 MTD_Region= [MTD] Format:
1871 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
1874 See drivers/mtd/cmdlinepart.c.
1876 multitce=off [PPC] This parameter disables the use of the pSeries
1877 firmware feature for updating multiple TCE entries
1880 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
1882 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
1884 boundary - index of last SLC block on Flex-OneNAND.
1885 The remaining blocks are configured as MLC blocks.
1886 lock - Configure if Flex-OneNAND boundary should be locked.
1887 Once locked, the boundary cannot be changed.
1888 1 indicates lock status, 0 indicates unlock status.
1891 ARM/S3C2412 JIVE boot control
1893 See arch/arm/mach-s3c2412/mach-jive.c
1895 mtouchusb.raw_coordinates=
1896 [HW] Make the MicroTouch USB driver use raw coordinates
1897 ('y', default) or cooked coordinates ('n')
1899 mtrr_chunk_size=nn[KMG] [X86]
1900 used for mtrr cleanup. It is largest continuous chunk
1901 that could hold holes aka. UC entries.
1903 mtrr_gran_size=nn[KMG] [X86]
1904 Used for mtrr cleanup. It is granularity of mtrr block.
1906 Large value could prevent small alignment from
1909 mtrr_spare_reg_nr=n [X86]
1911 Range: 0,7 : spare reg number
1913 Used for mtrr cleanup. It is spare mtrr entries number.
1914 Set to 2 or more if your graphical card needs more.
1916 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
1918 netdev= [NET] Network devices parameters
1919 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
1920 Note that mem_start is often overloaded to mean
1921 something different and driver-specific.
1922 This usage is only documented in each driver source
1926 [NETFILTER] Enable connection tracking flow accounting
1927 0 to disable accounting
1928 1 to enable accounting
1931 nfsaddrs= [NFS] Deprecated. Use ip= instead.
1932 See Documentation/filesystems/nfs/nfsroot.txt.
1934 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
1935 See Documentation/filesystems/nfs/nfsroot.txt.
1937 nfsrootdebug [NFS] enable nfsroot debugging messages.
1938 See Documentation/filesystems/nfs/nfsroot.txt.
1940 nfs.callback_tcpport=
1941 [NFS] set the TCP port on which the NFSv4 callback
1942 channel should listen.
1945 [NFS] sets the pathname to the program which is used
1946 to update the NFS client cache entries.
1948 nfs.cache_getent_timeout=
1949 [NFS] sets the timeout after which an attempt to
1950 update a cache entry is deemed to have failed.
1952 nfs.idmap_cache_timeout=
1953 [NFS] set the maximum lifetime for idmapper cache
1957 [NFS] enable 64-bit inode numbers.
1958 If zero, the NFS client will fake up a 32-bit inode
1959 number for the readdir() and stat() syscalls instead
1960 of returning the full 64-bit number.
1961 The default is to return 64-bit inode numbers.
1963 nfs.max_session_slots=
1964 [NFSv4.1] Sets the maximum number of session slots
1965 the client will attempt to negotiate with the server.
1966 This limits the number of simultaneous RPC requests
1967 that the client can send to the NFSv4.1 server.
1968 Note that there is little point in setting this
1969 value higher than the max_tcp_slot_table_limit.
1971 nfs.nfs4_disable_idmapping=
1972 [NFSv4] When set to the default of '1', this option
1973 ensures that both the RPC level authentication
1974 scheme and the NFS level operations agree to use
1975 numeric uids/gids if the mount is using the
1976 'sec=sys' security flavour. In effect it is
1977 disabling idmapping, which can make migration from
1978 legacy NFSv2/v3 systems to NFSv4 easier.
1979 Servers that do not support this mode of operation
1980 will be autodetected by the client, and it will fall
1981 back to using the idmapper.
1982 To turn off this behaviour, set the value to '0'.
1984 [NFS4] Specify an additional fixed unique ident-
1985 ification string that NFSv4 clients can insert into
1986 their nfs_client_id4 string. This is typically a
1987 UUID that is generated at system install time.
1989 nfs.send_implementation_id =
1990 [NFSv4.1] Send client implementation identification
1991 information in exchange_id requests.
1992 If zero, no implementation identification information
1994 The default is to send the implementation identification
1997 nfs.recover_lost_locks =
1998 [NFSv4] Attempt to recover locks that were lost due
1999 to a lease timeout on the server. Please note that
2000 doing this risks data corruption, since there are
2001 no guarantees that the file will remain unchanged
2002 after the locks are lost.
2003 If you want to enable the kernel legacy behaviour of
2004 attempting to recover these locks, then set this
2006 The default parameter value of '0' causes the kernel
2007 not to attempt recovery of lost locks.
2009 nfsd.nfs4_disable_idmapping=
2010 [NFSv4] When set to the default of '1', the NFSv4
2011 server will return only numeric uids and gids to
2012 clients using auth_sys, and will accept numeric uids
2013 and gids from such clients. This is intended to ease
2014 migration from NFSv2/v3.
2016 objlayoutdriver.osd_login_prog=
2017 [NFS] [OBJLAYOUT] sets the pathname to the program which
2018 is used to automatically discover and login into new
2019 osd-targets. Please see:
2020 Documentation/filesystems/pnfs.txt for more explanations
2022 nmi_debug= [KNL,AVR32,SH] Specify one or more actions to take
2023 when a NMI is triggered.
2024 Format: [state][,regs][,debounce][,die]
2026 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
2027 Format: [panic,][nopanic,][num]
2029 0 - turn nmi_watchdog off
2030 When panic is specified, panic when an NMI watchdog
2031 timeout occurs (or 'nopanic' to override the opposite
2033 This is useful when you use a panic=... timeout and
2034 need the box quickly up again.
2036 netpoll.carrier_timeout=
2037 [NET] Specifies amount of time (in seconds) that
2038 netpoll should wait for a carrier. By default netpoll
2041 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
2042 emulation library even if a 387 maths coprocessor
2046 [HW] Never suspend the console
2047 Disable suspending of consoles during suspend and
2048 hibernate operations. Once disabled, debugging
2049 messages can reach various consoles while the rest
2050 of the system is being put to sleep (ie, while
2051 debugging driver suspend/resume hooks). This may
2052 not work reliably with all consoles, but is known
2053 to work with serial and VGA consoles.
2054 To facilitate more flexible debugging, we also add
2055 console_suspend, a printk module parameter to control
2056 it. Users could use console_suspend (usually
2057 /sys/module/printk/parameters/console_suspend) to
2058 turn on/off it dynamically.
2060 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
2061 caches in the slab allocator. Saves per-node memory,
2062 but will impact performance.
2066 noapic [SMP,APIC] Tells the kernel to not make use of any
2067 IOAPICs that may be present in the system.
2070 Disable kernel base offset ASLR (Address Space
2071 Layout Randomization) if built into the kernel.
2073 noautogroup Disable scheduler automatic task group creation.
2075 nobats [PPC] Do not use BATs for mapping kernel lowmem
2076 on "Classic" PPC cores.
2080 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
2082 nodelayacct [KNL] Disable per-task delay accounting
2084 nodisconnect [HW,SCSI,M68K] Disables SCSI disconnects.
2086 nodsp [SH] Disable hardware DSP at boot time.
2088 noefi [X86] Disable EFI runtime services support.
2093 On X86-32 available only on PAE configured kernels.
2094 noexec=on: enable non-executable mappings (default)
2095 noexec=off: disable non-executable mappings
2098 Disable SMAP (Supervisor Mode Access Prevention)
2099 even if it is supported by processor.
2102 Disable SMEP (Supervisor Mode Execution Prevention)
2103 even if it is supported by processor.
2106 This affects only 32-bit executables.
2107 noexec32=on: enable non-executable mappings (default)
2108 read doesn't imply executable mappings
2109 noexec32=off: disable non-executable mappings
2110 read implies executable mappings
2112 nofpu [SH] Disable hardware FPU at boot time.
2114 nofxsr [BUGS=X86-32] Disables x86 floating point extended
2115 register save and restore. The kernel will only save
2116 legacy floating-point registers on task switch.
2118 noxsave [BUGS=X86] Disables x86 extended register state save
2119 and restore using xsave. The kernel will fallback to
2120 enabling legacy floating-point and sse state.
2123 on enable eager fpu restore
2124 off disable eager fpu restore
2125 auto selects the default scheme, which automatically
2126 enables eagerfpu restore for xsaveopt.
2128 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
2129 wfi(ARM) instruction doesn't work correctly and not to
2130 use it. This is also useful when using JTAG debugger.
2132 no_file_caps Tells the kernel not to honor file capabilities. The
2133 only way then for a file to be executed with privilege
2134 is to be setuid root or executed by root.
2136 nohalt [IA-64] Tells the kernel not to use the power saving
2137 function PAL_HALT_LIGHT when idle. This increases
2138 power-consumption. On the positive side, it reduces
2139 interrupt wake-up latency, which may improve performance
2140 in certain environments such as networked servers or
2143 nohz= [KNL] Boottime enable/disable dynamic ticks
2144 Valid arguments: on, off
2147 nohz_full= [KNL,BOOT]
2148 In kernels built with CONFIG_NO_HZ_FULL=y, set
2149 the specified list of CPUs whose tick will be stopped
2150 whenever possible. The boot CPU will be forced outside
2151 the range to maintain the timekeeping.
2152 The CPUs in this range must also be included in the
2155 noiotrap [SH] Disables trapped I/O port accesses.
2157 noirqdebug [X86-32] Disables the code which attempts to detect and
2158 disable unhandled interrupt sources.
2160 no_timer_check [X86,APIC] Disables the code which tests for
2161 broken timer IRQ sources.
2163 noisapnp [ISAPNP] Disables ISA PnP code.
2165 noinitrd [RAM] Tells the kernel not to load any configured
2168 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
2170 [Deprecated - use intremap=off]
2174 nojitter [IA-64] Disables jitter checking for ITC timers.
2176 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
2178 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
2181 no-steal-acc [X86,KVM] Disable paravirtualized steal time accounting.
2182 steal time is computed, but won't influence scheduler
2185 nolapic [X86-32,APIC] Do not enable or use the local APIC.
2187 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
2189 noltlbs [PPC] Do not use large page/tlb entries for kernel
2190 lowmem mapping on PPC40x.
2192 nomca [IA-64] Disable machine check abort handling
2194 nomce [X86-32] Machine Check Exception
2196 nomfgpt [X86-32] Disable Multi-Function General Purpose
2197 Timer usage (for AMD Geode machines).
2199 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
2200 shutdown the other cpus. Instead use the REBOOT_VECTOR
2203 nomodule Disable module load
2205 nopat [X86] Disable PAT (page attribute table extension of
2206 pagetables) support.
2208 norandmaps Don't use address space randomization. Equivalent to
2209 echo 0 > /proc/sys/kernel/randomize_va_space
2211 noreplace-paravirt [X86,IA-64,PV_OPS] Don't patch paravirt_ops
2213 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
2214 with UP alternatives
2216 nordrand [X86] Disable the direct use of the RDRAND
2217 instruction even if it is supported by the
2218 processor. RDRAND is still available to user
2221 noresume [SWSUSP] Disables resume and restores original swap
2224 no-scroll [VGA] Disables scrollback.
2225 This is required for the Braillex ib80-piezo Braille
2226 reader made by F.H. Papenmeier (Germany).
2230 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
2232 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
2233 and disable the IO APIC. legacy for "maxcpus=0".
2235 nosoftlockup [KNL] Disable the soft-lockup detector.
2237 nosync [HW,M68K] Disables sync negotiation for all devices.
2239 notsc [BUGS=X86-32] Disable Time Stamp Counter
2241 nousb [USB] Disable the USB subsystem
2243 nowatchdog [KNL] Disable the lockup detector (NMI watchdog).
2247 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
2249 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
2250 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
2251 Some features depend on CPU0. Known dependencies are:
2252 1. Resume from suspend/hibernate depends on CPU0.
2253 Suspend/hibernate will fail if CPU0 is offline and you
2254 need to online CPU0 before suspend/hibernate.
2255 2. PIC interrupts also depend on CPU0. CPU0 can't be
2256 removed if a PIC interrupt is detected.
2257 It's said poweroff/reboot may depend on CPU0 on some
2258 machines although I haven't seen such issues so far
2259 after CPU0 is offline on a few tested machines.
2260 If the dependencies are under your control, you can
2261 turn on cpu0_hotplug.
2263 nptcg= [IA-64] Override max number of concurrent global TLB
2264 purges which is reported from either PAL_VM_SUMMARY or
2267 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
2268 could support. nr_cpus=n : n >= 1 limits the kernel to
2269 supporting 'n' processors. Later in runtime you can not
2270 use hotplug cpu feature to put more cpu back to online.
2271 just like you compile the kernel NR_CPUS=n
2273 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
2275 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing.
2276 Allowed values are enable and disable
2278 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
2279 one of ['zone', 'node', 'default'] can be specified
2280 This can be set from sysctl after boot.
2281 See Documentation/sysctl/vm.txt for details.
2283 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
2284 See Documentation/debugging-via-ohci1394.txt for more
2287 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
2288 Rather than timing out after 20 ms if an EC
2289 command is not properly ACKed, override the length
2290 of the timeout. We have interrupts disabled while
2291 waiting for the ACK, so if this is set too high
2292 interrupts *may* be lost!
2294 omap_mux= [OMAP] Override bootloader pin multiplexing.
2295 Format: <mux_mode0.mode_name=value>...
2296 For example, to override I2C bus2:
2297 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
2299 oprofile.timer= [HW]
2300 Use timer interrupt instead of performance counters
2302 oprofile.cpu_type= Force an oprofile cpu type
2303 This might be useful if you have an older oprofile
2304 userland or if you want common events.
2305 Format: { arch_perfmon }
2306 arch_perfmon: [X86] Force use of architectural
2307 perfmon on Intel CPUs instead of the
2308 CPU specific event set.
2309 timer: [X86] Force use of architectural NMI
2310 timer mode (see also oprofile.timer
2311 for generic hr timer mode)
2312 [s390] Force legacy basic mode sampling
2313 (report cpu_type "timer")
2315 oops=panic Always panic on oopses. Default is to just kill the
2316 process, but there is a small probability of
2317 deadlocking the machine.
2318 This will also cause panics on machine check exceptions.
2319 Useful together with panic=30 to trigger a reboot.
2322 See Documentation/sound/oss/oss-parameters.txt
2324 panic= [KNL] Kernel behaviour on panic: delay <timeout>
2325 timeout > 0: seconds before rebooting
2326 timeout = 0: wait forever
2327 timeout < 0: reboot immediately
2330 parkbd.port= [HW] Parallel port number the keyboard adapter is
2331 connected to, default is 0.
2333 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
2334 0 for XT, 1 for AT (default is AT).
2337 parport= [HW,PPT] Specify parallel ports. 0 disables.
2338 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
2339 Use 'auto' to force the driver to use any
2340 IRQ/DMA settings detected (the default is to
2341 ignore detected IRQ/DMA settings because of
2342 possible conflicts). You can specify the base
2343 address, IRQ, and DMA settings; IRQ and DMA
2344 should be numbers, or 'auto' (for using detected
2345 settings on that particular port), or 'nofifo'
2346 (to avoid using a FIFO even if it is detected).
2347 Parallel ports are assigned in the order they
2348 are specified on the command line, starting
2351 parport_init_mode= [HW,PPT]
2352 Configure VIA parallel port to operate in
2353 a specific mode. This is necessary on Pegasos
2354 computer where firmware has no options for setting
2355 up parallel port mode and sets it to spp.
2356 Currently this function knows 686a and 8231 chips.
2357 Format: [spp|ps2|epp|ecp|ecpepp]
2360 Halt all CPUs after the first oops has been printed for
2361 the specified number of seconds. This is to be used if
2362 your oopses keep scrolling off the screen.
2367 See header of drivers/block/paride/pcd.c.
2368 See also Documentation/blockdev/paride.txt.
2370 pci=option[,option...] [PCI] various PCI subsystem options:
2371 earlydump [X86] dump PCI config space before the kernel
2373 off [X86] don't probe for the PCI bus
2374 bios [X86-32] force use of PCI BIOS, don't access
2375 the hardware directly. Use this if your machine
2376 has a non-standard PCI host bridge.
2377 nobios [X86-32] disallow use of PCI BIOS, only direct
2378 hardware access methods are allowed. Use this
2379 if you experience crashes upon bootup and you
2380 suspect they are caused by the BIOS.
2381 conf1 [X86] Force use of PCI Configuration
2383 conf2 [X86] Force use of PCI Configuration
2385 noaer [PCIE] If the PCIEAER kernel config parameter is
2386 enabled, this kernel boot option can be used to
2387 disable the use of PCIE advanced error reporting.
2388 nodomains [PCI] Disable support for multiple PCI
2389 root domains (aka PCI segments, in ACPI-speak).
2390 nommconf [X86] Disable use of MMCONFIG for PCI
2392 check_enable_amd_mmconf [X86] check for and enable
2393 properly configured MMIO access to PCI
2394 config space on AMD family 10h CPU
2395 nomsi [MSI] If the PCI_MSI kernel config parameter is
2396 enabled, this kernel boot option can be used to
2397 disable the use of MSI interrupts system-wide.
2398 noioapicquirk [APIC] Disable all boot interrupt quirks.
2399 Safety option to keep boot IRQs enabled. This
2400 should never be necessary.
2401 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
2402 primary IO-APIC for bridges that cannot disable
2403 boot IRQs. This fixes a source of spurious IRQs
2404 when the system masks IRQs.
2405 noioapicreroute [APIC] Disable workaround that uses the
2406 boot IRQ equivalent of an IRQ that connects to
2407 a chipset where boot IRQs cannot be disabled.
2408 The opposite of ioapicreroute.
2409 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
2410 routing table. These calls are known to be buggy
2411 on several machines and they hang the machine
2412 when used, but on other computers it's the only
2413 way to get the interrupt routing table. Try
2414 this option if the kernel is unable to allocate
2415 IRQs or discover secondary PCI buses on your
2417 rom [X86] Assign address space to expansion ROMs.
2418 Use with caution as certain devices share
2419 address decoders between ROMs and other
2421 norom [X86] Do not assign address space to
2422 expansion ROMs that do not already have
2423 BIOS assigned address ranges.
2424 nobar [X86] Do not assign address space to the
2425 BARs that weren't assigned by the BIOS.
2426 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
2427 assigned automatically to PCI devices. You can
2428 make the kernel exclude IRQs of your ISA cards
2430 pirqaddr=0xAAAAA [X86] Specify the physical address
2431 of the PIRQ table (normally generated
2432 by the BIOS) if it is outside the
2433 F0000h-100000h range.
2434 lastbus=N [X86] Scan all buses thru bus #N. Can be
2435 useful if the kernel is unable to find your
2436 secondary buses and you want to tell it
2437 explicitly which ones they are.
2438 assign-busses [X86] Always assign all PCI bus
2439 numbers ourselves, overriding
2440 whatever the firmware may have done.
2441 usepirqmask [X86] Honor the possible IRQ mask stored
2442 in the BIOS $PIR table. This is needed on
2443 some systems with broken BIOSes, notably
2444 some HP Pavilion N5400 and Omnibook XE3
2445 notebooks. This will have no effect if ACPI
2446 IRQ routing is enabled.
2447 noacpi [X86] Do not use ACPI for IRQ routing
2448 or for PCI scanning.
2449 use_crs [X86] Use PCI host bridge window information
2450 from ACPI. On BIOSes from 2008 or later, this
2451 is enabled by default. If you need to use this,
2452 please report a bug.
2453 nocrs [X86] Ignore PCI host bridge windows from ACPI.
2454 If you need to use this, please report a bug.
2455 routeirq Do IRQ routing for all PCI devices.
2456 This is normally done in pci_enable_device(),
2457 so this option is a temporary workaround
2458 for broken drivers that don't call it.
2459 skip_isa_align [X86] do not align io start addr, so can
2460 handle more pci cards
2461 firmware [ARM] Do not re-enumerate the bus but instead
2462 just use the configuration from the
2463 bootloader. This is currently used on
2464 IXP2000 systems where the bus has to be
2465 configured a certain way for adjunct CPUs.
2466 noearly [X86] Don't do any early type 1 scanning.
2467 This might help on some broken boards which
2468 machine check when some devices' config space
2469 is read. But various workarounds are disabled
2470 and some IOMMU drivers will not work.
2471 bfsort Sort PCI devices into breadth-first order.
2472 This sorting is done to get a device
2473 order compatible with older (<= 2.4) kernels.
2474 nobfsort Don't sort PCI devices into breadth-first order.
2475 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
2476 tuning and use the BIOS-configured MPS defaults.
2477 pcie_bus_safe Set every device's MPS to the largest value
2478 supported by all devices below the root complex.
2479 pcie_bus_perf Set device MPS to the largest allowable MPS
2480 based on its parent bus. Also set MRRS (Max
2481 Read Request Size) to the largest supported
2482 value (no larger than the MPS that the device
2483 or bus can support) for best performance.
2484 pcie_bus_peer2peer Set every device's MPS to 128B, which
2485 every device is guaranteed to support. This
2486 configuration allows peer-to-peer DMA between
2487 any pair of devices, possibly at the cost of
2488 reduced performance. This also guarantees
2489 that hot-added devices will work.
2490 cbiosize=nn[KMG] The fixed amount of bus space which is
2491 reserved for the CardBus bridge's IO window.
2492 The default value is 256 bytes.
2493 cbmemsize=nn[KMG] The fixed amount of bus space which is
2494 reserved for the CardBus bridge's memory
2495 window. The default value is 64 megabytes.
2498 [<order of align>@][<domain>:]<bus>:<slot>.<func>[; ...]
2499 Specifies alignment and device to reassign
2500 aligned memory resources.
2501 If <order of align> is not specified,
2502 PAGE_SIZE is used as alignment.
2503 PCI-PCI bridge can be specified, if resource
2504 windows need to be expanded.
2505 ecrc= Enable/disable PCIe ECRC (transaction layer
2506 end-to-end CRC checking).
2507 bios: Use BIOS/firmware settings. This is the
2511 hpiosize=nn[KMG] The fixed amount of bus space which is
2512 reserved for hotplug bridge's IO window.
2513 Default size is 256 bytes.
2514 hpmemsize=nn[KMG] The fixed amount of bus space which is
2515 reserved for hotplug bridge's memory window.
2516 Default size is 2 megabytes.
2517 realloc= Enable/disable reallocating PCI bridge resources
2518 if allocations done by BIOS are too small to
2519 accommodate resources required by all child
2521 off: Turn realloc off
2523 realloc same as realloc=on
2524 noari do not use PCIe ARI.
2525 pcie_scan_all Scan all possible PCIe devices. Otherwise we
2526 only look for one device below a PCIe downstream
2529 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
2532 force Enable ASPM even on devices that claim not to support it.
2533 WARNING: Forcing ASPM on may cause system lockups.
2535 pcie_hp= [PCIE] PCI Express Hotplug driver options:
2536 nomsi Do not use MSI for PCI Express Native Hotplug (this
2537 makes all PCIe ports use INTx for hotplug services).
2539 pcie_ports= [PCIE] PCIe ports handling:
2540 auto Ask the BIOS whether or not to use native PCIe services
2541 associated with PCIe ports (PME, hot-plug, AER). Use
2542 them only if that is allowed by the BIOS.
2543 native Use native PCIe services associated with PCIe ports
2545 compat Treat PCIe ports as PCI-to-PCI bridges, disable the PCIe
2548 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
2549 nomsi Do not use MSI for native PCIe PME signaling (this makes
2550 all PCIe root ports use INTx for all services).
2552 pcmv= [HW,PCMCIA] BadgePAD 4
2555 See Documentation/blockdev/paride.txt.
2557 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
2560 See arch/parisc/kernel/pdc_chassis.c
2562 percpu_alloc= Select which percpu first chunk allocator to use.
2563 Currently supported values are "embed" and "page".
2564 Archs may support subset or none of the selections.
2565 See comments in mm/percpu.c for details on each
2566 allocator. This parameter is primarily for debugging
2567 and performance comparison.
2570 See Documentation/blockdev/paride.txt.
2573 See Documentation/blockdev/paride.txt.
2575 pirq= [SMP,APIC] Manual mp-table setup
2576 See Documentation/x86/i386/IO-APIC.txt.
2578 plip= [PPT,NET] Parallel port network link
2579 Format: { parport<nr> | timid | 0 }
2580 See also Documentation/parport.txt.
2582 pmtmr= [X86] Manual setup of pmtmr I/O Port.
2583 Override pmtimer IOPort with a hex value.
2587 Enable PNP debug messages (depends on the
2588 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
2589 via /sys/module/pnp/parameters/debug. We always show
2590 current resource usage; turning this on also shows
2591 possible settings and some assignment information.
2597 { on | off | curr | res | no-curr | no-res }
2600 [ISAPNP] Exclude IRQs for the autoconfiguration
2603 [ISAPNP] Exclude DMAs for the autoconfiguration
2605 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
2606 Ranges are in pairs (I/O port base and size).
2609 [ISAPNP] Exclude memory regions for the
2611 Ranges are in pairs (memory base and size).
2613 ports= [IP_VS_FTP] IPVS ftp helper module
2615 Up to 8 (IP_VS_APP_MAX_PORTS) ports
2617 Format: <port>,<port>....
2619 print-fatal-signals=
2620 [KNL] debug: print fatal signals
2622 If enabled, warn about various signal handling
2623 related application anomalies: too many signals,
2624 too many POSIX.1 timers, fatal signals causing a
2627 If you hit the warning due to signal overflow,
2628 you might want to try "ulimit -i unlimited".
2632 printk.always_kmsg_dump=
2633 Trigger kmsg_dump for cases other than kernel oops or
2635 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
2638 printk.time= Show timing data prefixed to each printk message line
2639 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
2641 processor.max_cstate= [HW,ACPI]
2642 Limit processor to maximum C-state
2643 max_cstate=9 overrides any DMI blacklist limit.
2645 processor.nocst [HW,ACPI]
2646 Ignore the _CST method to determine C-states,
2647 instead using the legacy FADT method
2649 profile= [KNL] Enable kernel profiling via /proc/profile
2650 Format: [schedule,]<number>
2651 Param: "schedule" - profile schedule points.
2652 Param: <number> - step/bucket size as a power of 2 for
2653 statistical time based profiling.
2654 Param: "sleep" - profile D-state sleeping (millisecs).
2655 Requires CONFIG_SCHEDSTATS
2656 Param: "kvm" - profile VM exits.
2658 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
2660 See Documentation/blockdev/ramdisk.txt.
2662 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
2663 probe for; one of (bare|imps|exps|lifebook|any).
2664 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
2666 psmouse.resetafter= [HW,MOUSE]
2667 Try to reset the device after so many bad packets
2670 [HW,MOUSE] Set desired mouse resolution, in dpi.
2671 psmouse.smartscroll=
2672 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
2673 0 = disabled, 1 = enabled (default).
2675 pstore.backend= Specify the name of the pstore backend to use
2678 See Documentation/blockdev/paride.txt.
2681 [KNL] Number of legacy pty's. Overwrites compiled-in
2684 quiet [KNL] Disable most log messages
2689 See Documentation/md.txt.
2691 ramdisk_blocksize= [RAM]
2692 See Documentation/blockdev/ramdisk.txt.
2694 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
2695 See Documentation/blockdev/ramdisk.txt.
2698 In kernels built with CONFIG_RCU_NOCB_CPU=y, set
2699 the specified list of CPUs to be no-callback CPUs.
2700 Invocation of these CPUs' RCU callbacks will
2701 be offloaded to "rcuox/N" kthreads created for
2702 that purpose, where "x" is "b" for RCU-bh, "p"
2703 for RCU-preempt, and "s" for RCU-sched, and "N"
2704 is the CPU number. This reduces OS jitter on the
2705 offloaded CPUs, which can be useful for HPC and
2706 real-time workloads. It can also improve energy
2707 efficiency for asymmetric multiprocessors.
2710 Rather than requiring that offloaded CPUs
2711 (specified by rcu_nocbs= above) explicitly
2712 awaken the corresponding "rcuoN" kthreads,
2713 make these kthreads poll for callbacks.
2714 This improves the real-time response for the
2715 offloaded CPUs by relieving them of the need to
2716 wake up the corresponding kthread, but degrades
2717 energy efficiency by requiring that the kthreads
2718 periodically wake up to do the polling.
2720 rcutree.blimit= [KNL]
2721 Set maximum number of finished RCU callbacks to
2722 process in one batch.
2724 rcutree.rcu_fanout_leaf= [KNL]
2725 Increase the number of CPUs assigned to each
2726 leaf rcu_node structure. Useful for very large
2729 rcutree.jiffies_till_first_fqs= [KNL]
2730 Set delay from grace-period initialization to
2731 first attempt to force quiescent states.
2732 Units are jiffies, minimum value is zero,
2733 and maximum value is HZ.
2735 rcutree.jiffies_till_next_fqs= [KNL]
2736 Set delay between subsequent attempts to force
2737 quiescent states. Units are jiffies, minimum
2738 value is one, and maximum value is HZ.
2740 rcutree.qhimark= [KNL]
2741 Set threshold of queued RCU callbacks beyond which
2742 batch limiting is disabled.
2744 rcutree.qlowmark= [KNL]
2745 Set threshold of queued RCU callbacks below which
2746 batch limiting is re-enabled.
2748 rcutree.rcu_idle_gp_delay= [KNL]
2749 Set wakeup interval for idle CPUs that have
2750 RCU callbacks (RCU_FAST_NO_HZ=y).
2752 rcutree.rcu_idle_lazy_gp_delay= [KNL]
2753 Set wakeup interval for idle CPUs that have
2754 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
2755 Lazy RCU callbacks are those which RCU can
2756 prove do nothing more than free memory.
2758 rcutorture.fqs_duration= [KNL]
2759 Set duration of force_quiescent_state bursts.
2761 rcutorture.fqs_holdoff= [KNL]
2762 Set holdoff time within force_quiescent_state bursts.
2764 rcutorture.fqs_stutter= [KNL]
2765 Set wait time between force_quiescent_state bursts.
2767 rcutorture.gp_exp= [KNL]
2768 Use expedited update-side primitives.
2770 rcutorture.gp_normal= [KNL]
2771 Use normal (non-expedited) update-side primitives.
2772 If both gp_exp and gp_normal are set, do both.
2773 If neither gp_exp nor gp_normal are set, still
2776 rcutorture.n_barrier_cbs= [KNL]
2777 Set callbacks/threads for rcu_barrier() testing.
2779 rcutorture.nfakewriters= [KNL]
2780 Set number of concurrent RCU writers. These just
2781 stress RCU, they don't participate in the actual
2782 test, hence the "fake".
2784 rcutorture.nreaders= [KNL]
2785 Set number of RCU readers.
2787 rcutorture.object_debug= [KNL]
2788 Enable debug-object double-call_rcu() testing.
2790 rcutorture.onoff_holdoff= [KNL]
2791 Set time (s) after boot for CPU-hotplug testing.
2793 rcutorture.onoff_interval= [KNL]
2794 Set time (s) between CPU-hotplug operations, or
2795 zero to disable CPU-hotplug testing.
2797 rcutorture.rcutorture_runnable= [BOOT]
2798 Start rcutorture running at boot time.
2800 rcutorture.shuffle_interval= [KNL]
2801 Set task-shuffle interval (s). Shuffling tasks
2802 allows some CPUs to go into dyntick-idle mode
2803 during the rcutorture test.
2805 rcutorture.shutdown_secs= [KNL]
2806 Set time (s) after boot system shutdown. This
2807 is useful for hands-off automated testing.
2809 rcutorture.stall_cpu= [KNL]
2810 Duration of CPU stall (s) to test RCU CPU stall
2811 warnings, zero to disable.
2813 rcutorture.stall_cpu_holdoff= [KNL]
2814 Time to wait (s) after boot before inducing stall.
2816 rcutorture.stat_interval= [KNL]
2817 Time (s) between statistics printk()s.
2819 rcutorture.stutter= [KNL]
2820 Time (s) to stutter testing, for example, specifying
2821 five seconds causes the test to run for five seconds,
2822 wait for five seconds, and so on. This tests RCU's
2823 ability to transition abruptly to and from idle.
2825 rcutorture.test_boost= [KNL]
2826 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
2827 "Maybe" means test if the RCU implementation
2828 under test support RCU priority boosting.
2830 rcutorture.test_boost_duration= [KNL]
2831 Duration (s) of each individual boost test.
2833 rcutorture.test_boost_interval= [KNL]
2834 Interval (s) between each boost test.
2836 rcutorture.test_no_idle_hz= [KNL]
2837 Test RCU's dyntick-idle handling. See also the
2838 rcutorture.shuffle_interval parameter.
2840 rcutorture.torture_type= [KNL]
2841 Specify the RCU implementation to test.
2843 rcutorture.verbose= [KNL]
2844 Enable additional printk() statements.
2846 rcupdate.rcu_expedited= [KNL]
2847 Use expedited grace-period primitives, for
2848 example, synchronize_rcu_expedited() instead
2849 of synchronize_rcu(). This reduces latency,
2850 but can increase CPU utilization, degrade
2851 real-time latency, and degrade energy efficiency.
2853 rcupdate.rcu_cpu_stall_suppress= [KNL]
2854 Suppress RCU CPU stall warning messages.
2856 rcupdate.rcu_cpu_stall_timeout= [KNL]
2857 Set timeout for RCU CPU stall warning messages.
2861 Run specified binary instead of /init from the ramdisk,
2862 used for early userspace startup. See initrd.
2865 Format (x86 or x86_64):
2866 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
2868 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
2870 Where reboot_mode is one of warm (soft) or cold (hard) or gpio,
2871 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
2872 reboot_force is either force or not specified,
2873 reboot_cpu is s[mp]#### with #### being the processor
2874 to be used for rebooting.
2877 [KNL, SMP] Set scheduler's default relax_domain_level.
2878 See Documentation/cgroups/cpusets.txt.
2880 reserve= [KNL,BUGS] Force the kernel to ignore some iomem area
2882 reservetop= [X86-32]
2884 Reserves a hole at the top of the kernel virtual
2889 Set the amount of memory to reserve for BIOS at
2890 the bottom of the address space.
2892 reset_devices [KNL] Force drivers to reset the underlying device
2893 during initialization.
2896 Specify the partition device for software suspend
2898 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
2900 resume_offset= [SWSUSP]
2901 Specify the offset from the beginning of the partition
2902 given by "resume=" at which the swap header is located,
2903 in <PAGE_SIZE> units (needed only for swap files).
2904 See Documentation/power/swsusp-and-swap-files.txt
2906 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
2907 read the resume files
2909 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
2910 Useful for devices that are detected asynchronously
2911 (e.g. USB and MMC devices).
2913 hibernate= [HIBERNATION]
2914 noresume Don't check if there's a hibernation image
2915 present during boot.
2916 nocompress Don't compress/decompress hibernation images.
2918 retain_initrd [RAM] Keep initrd memory after extraction
2920 rhash_entries= [KNL,NET]
2921 Set number of hash buckets for route cache
2923 riscom8= [HW,SERIAL]
2924 Format: <io_board1>[,<io_board2>[,...<io_boardN>]]
2926 ro [KNL] Mount root device read-only on boot
2928 root= [KNL] Root filesystem
2929 See name_to_dev_t comment in init/do_mounts.c.
2931 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
2932 mount the root filesystem
2934 rootflags= [KNL] Set root filesystem mount option string
2936 rootfstype= [KNL] Set root filesystem type
2938 rootwait [KNL] Wait (indefinitely) for root device to show up.
2939 Useful for devices that are detected asynchronously
2940 (e.g. USB and MMC devices).
2942 rproc_mem=nn[KMG][@address]
2943 [KNL,ARM,CMA] Remoteproc physical memory block.
2944 Memory area to be used by remote processor image,
2947 rw [KNL] Mount root device read-write on boot
2949 S [KNL] Run init in single mode
2952 See drivers/net/irda/sa1100_ir.c.
2954 sbni= [NET] Granch SBNI12 leased line adapter
2956 sched_debug [KNL] Enables verbose scheduler debug messages.
2958 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
2959 xtime_lock contention on larger systems, and/or RCU lock
2960 contention on all systems with CONFIG_MAXSMP set.
2961 Format: { "0" | "1" }
2962 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
2964 Note: increases power consumption, thus should only be
2965 enabled if running jitter sensitive (HPC/RT) workloads.
2967 security= [SECURITY] Choose a security module to enable at boot.
2968 If this boot parameter is not specified, only the first
2969 security module asking for security registration will be
2970 loaded. An invalid security module name will be treated
2971 as if no module has been chosen.
2973 selinux= [SELINUX] Disable or enable SELinux at boot time.
2974 Format: { "0" | "1" }
2975 See security/selinux/Kconfig help text.
2978 Default value is set via kernel config option.
2979 If enabled at boot time, /selinux/disable can be used
2980 later to disable prior to initial policy load.
2982 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
2983 Format: { "0" | "1" }
2984 See security/apparmor/Kconfig help text
2987 Default value is set via kernel config option.
2989 serialnumber [BUGS=X86-32]
2992 Maximal number of shapers.
2994 show_msr= [x86] show boot-time MSR settings
2995 Format: { <integer> }
2996 Show boot-time (BIOS-initialized) MSR settings.
2997 The parameter means the number of CPUs to show,
2998 for example 1 means boot CPU only.
3005 slab_max_order= [MM, SLAB]
3006 Determines the maximum allowed order for slabs.
3007 A high setting may cause OOMs due to memory
3008 fragmentation. Defaults to 1 for systems with
3009 more than 32MB of RAM, 0 otherwise.
3011 slub_debug[=options[,slabs]] [MM, SLUB]
3012 Enabling slub_debug allows one to determine the
3013 culprit if slab objects become corrupted. Enabling
3014 slub_debug can create guard zones around objects and
3015 may poison objects when not in use. Also tracks the
3016 last alloc / free. For more information see
3017 Documentation/vm/slub.txt.
3019 slub_max_order= [MM, SLUB]
3020 Determines the maximum allowed order for slabs.
3021 A high setting may cause OOMs due to memory
3022 fragmentation. For more information see
3023 Documentation/vm/slub.txt.
3025 slub_min_objects= [MM, SLUB]
3026 The minimum number of objects per slab. SLUB will
3027 increase the slab order up to slub_max_order to
3028 generate a sufficiently large slab able to contain
3029 the number of objects indicated. The higher the number
3030 of objects the smaller the overhead of tracking slabs
3031 and the less frequently locks need to be acquired.
3032 For more information see Documentation/vm/slub.txt.
3034 slub_min_order= [MM, SLUB]
3035 Determines the minimum page order for slabs. Must be
3036 lower than slub_max_order.
3037 For more information see Documentation/vm/slub.txt.
3039 slub_nomerge [MM, SLUB]
3040 Disable merging of slabs with similar size. May be
3041 necessary if there is some reason to distinguish
3042 allocs to different slabs. Debug options disable
3043 merging on their own.
3044 For more information see Documentation/vm/slub.txt.
3047 Format: <io1>[,<io2>[,...,<io8>]]
3049 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
3050 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
3051 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
3052 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
3053 smsc-ircc2.ircc_irq= [HW] IRQ line
3054 smsc-ircc2.ircc_dma= [HW] DMA channel
3055 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
3056 0: Toshiba Satellite 1800 (GP data pin select)
3057 1: Fast pin select (default)
3061 [KNL] Should the soft-lockup detector generate panics.
3064 sonypi.*= [HW] Sony Programmable I/O Control Device driver
3065 See Documentation/laptops/sonypi.txt
3067 specialix= [HW,SERIAL] Specialix multi-serial port adapter
3068 See Documentation/serial/specialix.txt.
3070 spia_io_base= [HW,MTD]
3076 Enabled the stack tracer on boot up.
3078 stacktrace_filter=[function-list]
3079 [FTRACE] Limit the functions that the stack tracer
3080 will trace at boot up. function-list is a comma separated
3081 list of functions. This list can be changed at run
3082 time by the stack_trace_filter file in the debugfs
3083 tracing directory. Note, this enables stack tracing
3084 and the stacktrace above is not needed.
3088 Set the STI (builtin display/keyboard on the HP-PARISC
3089 machines) console (graphic card) which should be used
3090 as the initial boot-console.
3091 See also comment in drivers/video/console/sticore.c.
3094 See comment in drivers/video/console/sticore.c.
3097 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
3099 sunrpc.min_resvport=
3100 sunrpc.max_resvport=
3102 SunRPC servers often require that client requests
3103 originate from a privileged port (i.e. a port in the
3104 range 0 < portnr < 1024).
3105 An administrator who wishes to reserve some of these
3106 ports for other uses may adjust the range that the
3107 kernel's sunrpc client considers to be privileged
3108 using these two parameters to set the minimum and
3109 maximum port values.
3113 Control how the NFS server code allocates CPUs to
3114 service thread pools. Depending on how many NICs
3115 you have and where their interrupts are bound, this
3116 option will affect which CPUs will do NFS serving.
3117 Note: this parameter cannot be changed while the
3118 NFS server is running.
3120 auto the server chooses an appropriate mode
3121 automatically using heuristics
3122 global a single global pool contains all CPUs
3123 percpu one pool for each CPU
3124 pernode one pool for each NUMA node (equivalent
3125 to global on non-NUMA machines)
3127 sunrpc.tcp_slot_table_entries=
3128 sunrpc.udp_slot_table_entries=
3130 Sets the upper limit on the number of simultaneous
3131 RPC calls that can be sent from the client to a
3132 server. Increasing these values may allow you to
3133 improve throughput, but will also increase the
3134 amount of memory reserved for use by the client.
3137 [KNL] Enable accounting of swap in memory resource
3138 controller if no parameter or 1 is given or disable
3139 it if 0 is given (See Documentation/cgroups/memory.txt)
3141 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
3142 Format: { <int> | force }
3143 <int> -- Number of I/O TLB slabs
3144 force -- force using of bounce buffers even if they
3145 wouldn't be automatically used by the kernel
3149 sysfs.deprecated=0|1 [KNL]
3150 Enable/disable old style sysfs layout for old udev
3151 on older distributions. When this option is enabled
3152 very new udev will not work anymore. When this option
3153 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
3154 in older udev will not work anymore.
3155 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
3156 the kernel configuration.
3158 sysrq_always_enabled
3160 Ignore sysrq setting - this boot parameter will
3161 neutralize any effect of /proc/sys/kernel/sysrq.
3162 Useful for debugging.
3166 test_suspend= [SUSPEND]
3167 Specify "mem" (for Suspend-to-RAM) or "standby" (for
3168 standby suspend) as the system sleep state to briefly
3169 enter during system startup. The system is woken from
3170 this state using a wakeup-capable RTC alarm.
3172 thash_entries= [KNL,NET]
3173 Set number of hash buckets for TCP connection
3175 thermal.act= [HW,ACPI]
3176 -1: disable all active trip points in all thermal zones
3177 <degrees C>: override all lowest active trip points
3179 thermal.crt= [HW,ACPI]
3180 -1: disable all critical trip points in all thermal zones
3181 <degrees C>: override all critical trip points
3183 thermal.nocrt= [HW,ACPI]
3184 Set to disable actions on ACPI thermal zone
3185 critical and hot trip points.
3187 thermal.off= [HW,ACPI]
3188 1: disable ACPI thermal control
3190 thermal.psv= [HW,ACPI]
3191 -1: disable all passive trip points
3192 <degrees C>: override all passive trip points to this
3195 thermal.tzp= [HW,ACPI]
3196 Specify global default ACPI thermal zone polling rate
3197 <deci-seconds>: poll all this frequency
3198 0: no polling (default)
3201 Force threading of all interrupt handlers except those
3202 marked explicitly IRQF_NO_THREAD.
3205 Enable the Transcendent memory driver if built-in.
3207 tmem.cleancache=0|1 [KNL, XEN]
3208 Default is on (1). Disable the usage of the cleancache
3209 API to send anonymous pages to the hypervisor.
3211 tmem.frontswap=0|1 [KNL, XEN]
3212 Default is on (1). Disable the usage of the frontswap
3213 API to send swap pages to the hypervisor. If disabled
3214 the selfballooning and selfshrinking are force disabled.
3216 tmem.selfballooning=0|1 [KNL, XEN]
3217 Default is on (1). Disable the driving of swap pages
3220 tmem.selfshrinking=0|1 [KNL, XEN]
3221 Default is on (1). Partial swapoff that immediately
3222 transfers pages from Xen hypervisor back to the
3223 kernel based on different criteria.
3227 Specify if the kernel should make use of the cpu
3228 topology information if the hardware supports this.
3229 The scheduler will make use of this information and
3230 e.g. base its process migration decisions on it.
3235 tpm_suspend_pcr=[HW,TPM]
3236 Format: integer pcr id
3237 Specify that at suspend time, the tpm driver
3238 should extend the specified pcr with zeros,
3239 as a workaround for some chips which fail to
3240 flush the last written pcr on TPM_SaveState.
3241 This will guarantee that all the other pcrs
3244 trace_buf_size=nn[KMG]
3245 [FTRACE] will set tracing buffer size.
3247 trace_event=[event-list]
3248 [FTRACE] Set and start specified trace events in order
3249 to facilitate early boot debugging.
3250 See also Documentation/trace/events.txt
3252 trace_options=[option-list]
3253 [FTRACE] Enable or disable tracer options at boot.
3254 The option-list is a comma delimited list of options
3255 that can be enabled or disabled just as if you were
3256 to echo the option name into
3258 /sys/kernel/debug/tracing/trace_options
3260 For example, to enable stacktrace option (to dump the
3261 stack trace of each event), add to the command line:
3263 trace_options=stacktrace
3265 See also Documentation/trace/ftrace.txt "trace options"
3269 [FTRACE] enable this option to disable tracing when a
3270 warning is hit. This turns off "tracing_on". Tracing can
3271 be enabled again by echoing '1' into the "tracing_on"
3272 file located in /sys/kernel/debug/tracing/
3274 This option is useful, as it disables the trace before
3275 the WARNING dump is called, which prevents the trace to
3276 be filled with content caused by the warning output.
3278 This option can also be set at run time via the sysctl
3279 option: kernel/traceoff_on_warning
3281 transparent_hugepage=
3283 Format: [always|madvise|never]
3284 Can be used to control the default behavior of the system
3285 with respect to transparent hugepages.
3286 See Documentation/vm/transhuge.txt for more details.
3288 tsc= Disable clocksource stability checks for TSC.
3290 [x86] reliable: mark tsc clocksource as reliable, this
3291 disables clocksource verification at runtime, as well
3292 as the stability checks done at bootup. Used to enable
3293 high-resolution timer mode on older hardware, and in
3294 virtualized environment.
3295 [x86] noirqtime: Do not use TSC to do irq accounting.
3296 Used to run time disable IRQ_TIME_ACCOUNTING on any
3297 platforms where RDTSC is slow and this accounting
3300 turbografx.map[2|3]= [HW,JOY]
3301 TurboGraFX parallel port interface
3303 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
3304 See also Documentation/input/joystick-parport.txt
3306 udbg-immortal [PPC] When debugging early kernel crashes that
3307 happen after console_init() and before a proper
3308 console driver takes over, this boot options might
3309 help "seeing" what's going on.
3311 uhash_entries= [KNL,NET]
3312 Set number of hash buckets for UDP/UDP-Lite connections
3315 [USB] Ignore overcurrent events (default N).
3316 Some badly-designed motherboards generate lots of
3317 bogus events, for ports that aren't wired to
3318 anything. Set this parameter to avoid log spamming.
3319 Note that genuine overcurrent events won't be
3323 [X86] Cause panic on unknown NMI.
3325 usbcore.authorized_default=
3326 [USB] Default USB device authorization:
3327 (default -1 = authorized except for wireless USB,
3328 0 = not authorized, 1 = authorized)
3330 usbcore.autosuspend=
3331 [USB] The autosuspend time delay (in seconds) used
3332 for newly-detected USB devices (default 2). This
3333 is the time required before an idle device will be
3334 autosuspended. Devices for which the delay is set
3335 to a negative value won't be autosuspended at all.
3337 usbcore.usbfs_snoop=
3338 [USB] Set to log all usbfs traffic (default 0 = off).
3340 usbcore.blinkenlights=
3341 [USB] Set to cycle leds on hubs (default 0 = off).
3343 usbcore.old_scheme_first=
3344 [USB] Start with the old device initialization
3345 scheme (default 0 = off).
3347 usbcore.usbfs_memory_mb=
3348 [USB] Memory limit (in MB) for buffers allocated by
3349 usbfs (default = 16, 0 = max = 2047).
3351 usbcore.use_both_schemes=
3352 [USB] Try the other device initialization scheme
3353 if the first one fails (default 1 = enabled).
3355 usbcore.initial_descriptor_timeout=
3356 [USB] Specifies timeout for the initial 64-byte
3357 USB_REQ_GET_DESCRIPTOR request in milliseconds
3358 (default 5000 = 5.0 seconds).
3361 [USBHID] The interval which mice are to be polled at.
3363 usb-storage.delay_use=
3364 [UMS] The delay in seconds before a new device is
3365 scanned for Logical Units (default 5).
3368 [UMS] A list of quirks entries to supplement or
3369 override the built-in unusual_devs list. List
3370 entries are separated by commas. Each entry has
3371 the form VID:PID:Flags where VID and PID are Vendor
3372 and Product ID values (4-digit hex numbers) and
3373 Flags is a set of characters, each corresponding
3374 to a common usb-storage quirk flag as follows:
3375 a = SANE_SENSE (collect more than 18 bytes
3377 b = BAD_SENSE (don't collect more than 18
3378 bytes of sense data);
3379 c = FIX_CAPACITY (decrease the reported
3380 device capacity by one sector);
3381 d = NO_READ_DISC_INFO (don't use
3382 READ_DISC_INFO command);
3383 e = NO_READ_CAPACITY_16 (don't use
3384 READ_CAPACITY_16 command);
3385 h = CAPACITY_HEURISTICS (decrease the
3386 reported device capacity by one
3387 sector if the number is odd);
3388 i = IGNORE_DEVICE (don't bind to this
3390 l = NOT_LOCKABLE (don't try to lock and
3391 unlock ejectable media);
3392 m = MAX_SECTORS_64 (don't transfer more
3393 than 64 sectors = 32 KB at a time);
3394 n = INITIAL_READ10 (force a retry of the
3395 initial READ(10) command);
3396 o = CAPACITY_OK (accept the capacity
3397 reported by the device);
3398 p = WRITE_CACHE (the device cache is ON
3400 r = IGNORE_RESIDUE (the device reports
3401 bogus residue values);
3402 s = SINGLE_LUN (the device has only one
3404 w = NO_WP_DETECT (don't test whether the
3405 medium is write-protected).
3406 Example: quirks=0419:aaf5:rl,0421:0433:rc
3408 user_debug= [KNL,ARM]
3410 See arch/arm/Kconfig.debug help text.
3411 1 - undefined instruction events
3413 4 - invalid data aborts
3416 Example: user_debug=31
3419 [X86] Flags controlling user PTE allocations.
3421 nohigh = do not allocate PTE pages in
3422 HIGHMEM regardless of setting
3426 vdso=2: enable compat VDSO (default with COMPAT_VDSO)
3427 vdso=1: enable VDSO (default)
3428 vdso=0: disable VDSO mapping
3431 vdso32=2: enable compat VDSO (default with COMPAT_VDSO)
3432 vdso32=1: enable 32-bit VDSO (default)
3433 vdso32=0: disable 32-bit VDSO mapping
3436 vector=percpu: enable percpu vector domain
3438 video= [FB] Frame buffer configuration
3439 See Documentation/fb/modedb.txt.
3441 video.brightness_switch_enabled= [0,1]
3442 If set to 1, on receiving an ACPI notify event
3443 generated by hotkey, video driver will adjust brightness
3444 level and then send out the event to user space through
3445 the allocated input device; If set to 0, video driver
3446 will only send out the event without touching backlight
3451 [VMMIO] Memory mapped virtio (platform) device.
3453 <size>@<baseaddr>:<irq>[:<id>]
3455 <size> := size (can use standard suffixes
3457 <baseaddr> := physical base address
3458 <irq> := interrupt number (as passed to
3460 <id> := (optional) platform device id
3462 virtio_mmio.device=1K@0x100b0000:48:7
3464 Can be used multiple times for multiple devices.
3466 vga= [BOOT,X86-32] Select a particular video mode
3467 See Documentation/x86/boot.txt and
3468 Documentation/svga.txt.
3469 Use vga=ask for menu.
3470 This is actually a boot loader parameter; the value is
3471 passed to the kernel using a special protocol.
3473 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
3474 size of <nn>. This can be used to increase the
3475 minimum size (128MB on x86). It can also be used to
3476 decrease the size and leave more room for directly
3479 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
3482 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
3485 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
3489 Controls the behavior of vsyscalls (i.e. calls to
3490 fixed addresses of 0xffffffffff600x00 from legacy
3491 code). Most statically-linked binaries and older
3492 versions of glibc use these calls. Because these
3493 functions are at fixed addresses, they make nice
3494 targets for exploits that can control RIP.
3496 emulate [default] Vsyscalls turn into traps and are
3497 emulated reasonably safely.
3499 native Vsyscalls are native syscall instructions.
3500 This is a little bit faster than trapping
3501 and makes a few dynamic recompilers work
3502 better than they would in emulation mode.
3503 It also makes exploits much easier to write.
3505 none Vsyscalls don't work at all. This makes
3506 them quite hard to use for exploits but
3507 might break your system.
3509 vt.color= [VT] Default text color.
3510 Format: 0xYX, X = foreground, Y = background.
3511 Default: 0x07 = light gray on black.
3513 vt.cur_default= [VT] Default cursor shape.
3514 Format: 0xCCBBAA, where AA, BB, and CC are the same as
3515 the parameters of the <Esc>[?A;B;Cc escape sequence;
3516 see VGA-softcursor.txt. Default: 2 = underline.
3518 vt.default_blu= [VT]
3519 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
3520 Change the default blue palette of the console.
3521 This is a 16-member array composed of values
3524 vt.default_grn= [VT]
3525 Format: <green0>,<green1>,<green2>,...,<green15>
3526 Change the default green palette of the console.
3527 This is a 16-member array composed of values
3530 vt.default_red= [VT]
3531 Format: <red0>,<red1>,<red2>,...,<red15>
3532 Change the default red palette of the console.
3533 This is a 16-member array composed of values
3539 Set system-wide default UTF-8 mode for all tty's.
3540 Default is 1, i.e. UTF-8 mode is enabled for all
3541 newly opened terminals.
3543 vt.global_cursor_default=
3546 Set system-wide default for whether a cursor
3547 is shown on new VTs. Default is -1,
3548 i.e. cursors will be created by default unless
3549 overridden by individual drivers. 0 will hide
3550 cursors, 1 will display them.
3552 vt.italic= [VT] Default color for italic text; 0-15.
3555 vt.underline= [VT] Default color for underlined text; 0-15.
3558 watchdog timers [HW,WDT] For information on watchdog timers,
3559 see Documentation/watchdog/watchdog-parameters.txt
3560 or other driver-specific files in the
3561 Documentation/watchdog/ directory.
3563 workqueue.disable_numa
3564 By default, all work items queued to unbound
3565 workqueues are affine to the NUMA nodes they're
3566 issued on, which results in better behavior in
3567 general. If NUMA affinity needs to be disabled for
3568 whatever reason, this option can be used. Note
3569 that this also can be controlled per-workqueue for
3570 workqueues visible under /sys/bus/workqueue/.
3572 workqueue.power_efficient
3573 Per-cpu workqueues are generally preferred because
3574 they show better performance thanks to cache
3575 locality; unfortunately, per-cpu workqueues tend to
3576 be more power hungry than unbound workqueues.
3578 Enabling this makes the per-cpu workqueues which
3579 were observed to contribute significantly to power
3580 consumption unbound, leading to measurably lower
3581 power usage at the cost of small performance
3584 The default value of this parameter is determined by
3585 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
3587 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
3588 default x2apic cluster mode on platforms
3591 x86_intel_mid_timer= [X86-32,APBT]
3592 Choose timer option for x86 Intel MID platform.
3593 Two valid options are apbt timer only and lapic timer
3594 plus one apbt timer for broadcast timer.
3595 x86_intel_mid_timer=apbt_only | lapic_and_apbt
3597 xen_emul_unplug= [HW,X86,XEN]
3598 Unplug Xen emulated devices
3599 Format: [unplug0,][unplug1]
3600 ide-disks -- unplug primary master IDE devices
3601 aux-ide-disks -- unplug non-primary-master IDE devices
3602 nics -- unplug network devices
3603 all -- unplug all emulated devices (NICs and IDE disks)
3604 unnecessary -- unplugging emulated devices is
3605 unnecessary even if the host did not respond to
3607 never -- do not unplug even if version check succeeds
3609 xen_nopvspin [X86,XEN]
3610 Disables the ticketlock slowpath using Xen PV
3613 xirc2ps_cs= [NET,PCMCIA]
3615 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
3617 ______________________________________________________________________
3621 Add more DRM drivers.