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 AVR32 AVR32 architecture is enabled.
44 AX25 Appropriate AX.25 support is enabled.
45 BLACKFIN Blackfin architecture is enabled.
46 DRM Direct Rendering Management support is enabled.
47 EDD BIOS Enhanced Disk Drive Services (EDD) is enabled
48 EFI EFI Partitioning (GPT) is enabled
49 EIDE EIDE/ATAPI support is enabled.
50 FB The frame buffer device is enabled.
51 GCOV GCOV profiling is enabled.
52 HW Appropriate hardware is enabled.
53 IA-64 IA-64 architecture is enabled.
54 IMA Integrity measurement architecture is enabled.
55 IOSCHED More than one I/O scheduler is enabled.
56 IP_PNP IP DHCP, BOOTP, or RARP is enabled.
57 IPV6 IPv6 support is enabled.
58 ISAPNP ISA PnP code is enabled.
59 ISDN Appropriate ISDN support is enabled.
60 JOY Appropriate joystick support is enabled.
61 KGDB Kernel debugger support is enabled.
62 KVM Kernel Virtual Machine support is enabled.
63 LIBATA Libata driver is enabled
64 LP Printer support is enabled.
65 LOOP Loopback device support is enabled.
66 M68k M68k architecture is enabled.
67 These options have more detailed description inside of
68 Documentation/m68k/kernel-options.txt.
69 MCA MCA bus support is enabled.
70 MDA MDA console support is enabled.
71 MOUSE Appropriate mouse support is enabled.
72 MSI Message Signaled Interrupts (PCI).
73 MTD MTD (Memory Technology Device) support is enabled.
74 NET Appropriate network support is enabled.
75 NUMA NUMA support is enabled.
76 GENERIC_TIME The generic timeofday code is enabled.
77 NFS Appropriate NFS support is enabled.
78 OSS OSS sound support is enabled.
79 PV_OPS A paravirtualized kernel is enabled.
80 PARIDE The ParIDE (parallel port IDE) subsystem is enabled.
81 PARISC The PA-RISC architecture is enabled.
82 PCI PCI bus support is enabled.
83 PCIE PCI Express support is enabled.
84 PCMCIA The PCMCIA subsystem is enabled.
85 PNP Plug & Play support is enabled.
86 PPC PowerPC architecture is enabled.
87 PPT Parallel port support is enabled.
88 PS2 Appropriate PS/2 support is enabled.
89 RAM RAM disk support is enabled.
90 S390 S390 architecture is enabled.
91 SCSI Appropriate SCSI support is enabled.
92 A lot of drivers has their options described inside of
94 SECURITY Different security models are enabled.
95 SELINUX SELinux support is enabled.
96 SERIAL Serial support is enabled.
97 SH SuperH architecture is enabled.
98 SMP The kernel is an SMP kernel.
99 SPARC Sparc architecture is enabled.
100 SWSUSP Software suspend (hibernation) is enabled.
101 SUSPEND System suspend states are enabled.
102 FTRACE Function tracing enabled.
103 TPM TPM drivers are enabled.
104 TS Appropriate touchscreen support is enabled.
105 UMS USB Mass Storage support is enabled.
106 USB USB support is enabled.
107 USBHID USB Human Interface Device support is enabled.
108 V4L Video For Linux support is enabled.
109 VGA The VGA console has been enabled.
110 VT Virtual terminal support is enabled.
111 WDT Watchdog support is enabled.
112 XT IBM PC/XT MFM hard disk support is enabled.
113 X86-32 X86-32, aka i386 architecture is enabled.
114 X86-64 X86-64 architecture is enabled.
115 More X86-64 boot options can be found in
116 Documentation/x86/x86_64/boot-options.txt .
117 X86 Either 32bit or 64bit x86 (same as X86-32+X86-64)
119 In addition, the following text indicates that the option:
121 BUGS= Relates to possible processor bugs on the said processor.
122 KNL Is a kernel start-up parameter.
123 BOOT Is a boot loader parameter.
125 Parameters denoted with BOOT are actually interpreted by the boot
126 loader, and have no meaning to the kernel directly.
127 Do not modify the syntax of boot loader parameters without extreme
128 need or coordination with <Documentation/x86/boot.txt>.
130 There are also arch-specific kernel-parameters not documented here.
131 See for example <Documentation/x86/x86_64/boot-options.txt>.
133 Note that ALL kernel parameters listed below are CASE SENSITIVE, and that
134 a trailing = on the name of any parameter states that that parameter will
135 be entered as an environment variable, whereas its absence indicates that
136 it will appear as a kernel argument readable via /proc/cmdline by programs
137 running once the system is up.
139 The number of kernel parameters is not limited, but the length of the
140 complete command line (parameters including spaces etc.) is limited to
141 a fixed number of characters. This limit depends on the architecture
142 and is between 256 and 4096 characters. It is defined in the file
143 ./include/asm/setup.h as COMMAND_LINE_SIZE.
147 Advanced Configuration and Power Interface
148 Format: { force | off | strict | noirq | rsdt }
149 force -- enable ACPI if default was off
150 off -- disable ACPI if default was on
151 noirq -- do not use ACPI for IRQ routing
152 strict -- Be less tolerant of platforms that are not
153 strictly ACPI specification compliant.
154 rsdt -- prefer RSDT over (default) XSDT
155 copy_dsdt -- copy DSDT to memory
157 See also Documentation/power/pm.txt, pci=noacpi
159 acpi_apic_instance= [ACPI, IOAPIC]
161 2: use 2nd APIC table, if available
162 1,0: use 1st APIC table
165 acpi_backlight= [HW,ACPI]
166 acpi_backlight=vendor
168 If set to vendor, prefer vendor specific driver
169 (e.g. thinkpad_acpi, sony_acpi, etc.) instead
170 of the ACPI video.ko driver.
172 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG]
173 acpi.debug_level= [HW,ACPI,ACPI_DEBUG]
175 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
176 debug output. Bits in debug_layer correspond to a
177 _COMPONENT in an ACPI source file, e.g.,
178 #define _COMPONENT ACPI_PCI_COMPONENT
179 Bits in debug_level correspond to a level in
180 ACPI_DEBUG_PRINT statements, e.g.,
181 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
182 The debug_level mask defaults to "info". See
183 Documentation/acpi/debug.txt for more information about
184 debug layers and levels.
186 Enable processor driver info messages:
187 acpi.debug_layer=0x20000000
188 Enable PCI/PCI interrupt routing info messages:
189 acpi.debug_layer=0x400000
190 Enable AML "Debug" output, i.e., stores to the Debug
191 object while interpreting AML:
192 acpi.debug_layer=0xffffffff acpi.debug_level=0x2
193 Enable all messages related to ACPI hardware:
194 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
196 Some values produce so much output that the system is
197 unusable. The "log_buf_len" parameter may be useful
198 if you need to capture more output.
200 acpi_display_output= [HW,ACPI]
201 acpi_display_output=vendor
202 acpi_display_output=video
205 acpi_irq_balance [HW,ACPI]
206 ACPI will balance active IRQs
209 acpi_irq_nobalance [HW,ACPI]
210 ACPI will not move active IRQs (default)
213 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA
214 Format: <irq>,<irq>...
216 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for
218 Format: <irq>,<irq>...
220 acpi_no_auto_ssdt [HW,ACPI] Disable automatic loading of SSDT
222 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
223 Format: To spoof as Windows 98: ="Microsoft Windows"
225 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
226 acpi_osi="string1" # add string1 -- only one string
227 acpi_osi="!string2" # remove built-in string2
228 acpi_osi= # disable all strings
231 Override the pmtimer bug detection: force the kernel
232 to assume that this machine's pmtimer latches its value
233 and always returns good values.
235 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
236 Format: { level | edge | high | low }
238 acpi_serialize [HW,ACPI] force serialization of AML methods
240 acpi_skip_timer_override [HW,ACPI]
241 Recognize and ignore IRQ0/pin2 Interrupt Override.
242 For broken nForce2 BIOS resulting in XT-PIC timer.
244 acpi_sleep= [HW,ACPI] Sleep options
245 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
246 old_ordering, s4_nonvs, sci_force_enable }
247 See Documentation/power/video.txt for information on
249 s3_beep is for debugging; it makes the PC's speaker beep
250 as soon as the kernel's real-mode entry point is called.
251 s4_nohwsig prevents ACPI hardware signature from being
252 used during resume from hibernation.
253 old_ordering causes the ACPI 1.0 ordering of the _PTS
254 control method, with respect to putting devices into
255 low power states, to be enforced (the ACPI 2.0 ordering
256 of _PTS is used by default).
257 nonvs prevents the kernel from saving/restoring the
258 ACPI NVS memory during suspend/hibernation and resume.
259 sci_force_enable causes the kernel to set SCI_EN directly
260 on resume from S1/S3 (which is against the ACPI spec,
261 but some broken systems don't work without it).
263 acpi_use_timer_override [HW,ACPI]
264 Use timer override. For some broken Nvidia NF5 boards
265 that require a timer override, but don't have HPET
267 acpi_enforce_resources= [ACPI]
268 { strict | lax | no }
269 Check for resource conflicts between native drivers
270 and ACPI OperationRegions (SystemIO and SystemMemory
271 only). IO ports and memory declared in ACPI might be
272 used by the ACPI subsystem in arbitrary AML code and
273 can interfere with legacy drivers.
274 strict (default): access to resources claimed by ACPI
275 is denied; legacy drivers trying to access reserved
276 resources will fail to bind to device using them.
277 lax: access to resources claimed by ACPI is allowed;
278 legacy drivers trying to access reserved resources
279 will bind successfully but a warning message is logged.
280 no: ACPI OperationRegions are not marked as reserved,
281 no further checks are performed.
284 Format: <io>,<irq>,<dma>,<dma2>,<type>
286 add_efi_memmap [EFI; X86] Include EFI memory map in
287 kernel's map of available physical RAM.
290 See header of drivers/scsi/advansys.c.
292 aedsp16= [HW,OSS] Audio Excel DSP 16
293 Format: <io>,<irq>,<dma>,<mss_io>,<mpu_io>,<mpu_irq>
294 See also header of sound/oss/aedsp16.c.
297 { off | try_unsupported }
298 off: disable AGP support
299 try_unsupported: try to drive unsupported chipsets
300 (may crash computer or cause data corruption)
303 See Documentation/scsi/aha152x.txt.
306 Format: <portbase>[,<buson>,<busoff>[,<dmaspeed>]]
309 See Documentation/scsi/aic7xxx.txt.
312 See Documentation/scsi/aic79xx.txt.
315 Allow the default userspace alignment fault handler
316 behaviour to be specified. Bit 0 enables warnings,
317 bit 1 enables fixups, and bit 2 sends a segfault.
319 amd_iommu= [HW,X86-84]
320 Pass parameters to the AMD IOMMU driver in the system.
322 fullflush - enable flushing of IO/TLB entries when
323 they are unmapped. Otherwise they are
324 flushed before they will be reused, which
326 off - do not initialize any AMD IOMMU found in
329 amijoy.map= [HW,JOY] Amiga joystick support
330 Map of devices attached to JOY0DAT and JOY1DAT
332 See also Documentation/kernel/input/joystick.txt
334 analog.map= [HW,JOY] Analog joystick and gamepad support
335 Specifies type or capabilities of an analog joystick
336 connected to one of 16 gameports
337 Format: <type1>,<type2>,..<type16>
340 Power management functions (SPARCstation-4/5 + deriv.)
342 Disable APC CPU standby support. SPARCstation-Fox does
343 not play well with APC CPU idle - disable it if you have
344 APC and your system crashes randomly.
346 apic= [APIC,X86-32] Advanced Programmable Interrupt Controller
347 Change the output verbosity whilst booting
348 Format: { quiet (default) | verbose | debug }
349 Change the amount of debugging information output
350 when initialising the APIC and IO-APIC components.
353 See Documentation/networking/ipv6.txt.
355 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
356 Limit apic dumping. The parameter defines the maximal
357 number of local apics being dumped. Also it is possible
358 to set it to "all" by meaning -- no limit here.
359 Format: { 1 (default) | 2 | ... | all }.
360 The parameter valid if only apic=debug or
361 apic=verbose is specified.
362 Example: apic=debug show_lapic=all
364 apm= [APM] Advanced Power Management
365 See header of arch/x86/kernel/apm_32.c.
367 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
368 Format: <io>,<irq>,<nodeID>
372 atarimouse= [HW,MOUSE] Atari Mouse
374 atascsi= [HW,SCSI] Atari SCSI
376 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
377 EzKey and similar keyboards
379 atkbd.reset= [HW] Reset keyboard during initialization
381 atkbd.set= [HW] Select keyboard code set
382 Format: <int> (2 = AT (default), 3 = PS/2)
384 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
387 atkbd.softraw= [HW] Choose between synthetic and real raw mode
388 Format: <bool> (0 = real, 1 = synthetic (default))
390 atkbd.softrepeat= [HW]
391 Use software keyboard repeat
395 baycom_epp= [HW,AX25]
398 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
400 See header of drivers/net/hamradio/baycom_par.c.
402 baycom_ser_fdx= [HW,AX25]
403 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
404 Format: <io>,<irq>,<mode>[,<baud>]
405 See header of drivers/net/hamradio/baycom_ser_fdx.c.
407 baycom_ser_hdx= [HW,AX25]
408 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
409 Format: <io>,<irq>,<mode>
410 See header of drivers/net/hamradio/baycom_ser_hdx.c.
412 boot_delay= Milliseconds to delay each printk during boot.
413 Values larger than 10 seconds (10000) are changed to
417 bootmem_debug [KNL] Enable bootmem allocator debug messages.
419 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
420 bttv.radio= Most important insmod options are available as
422 bttv.pll= See Documentation/video4linux/bttv/Insmod-options
423 bttv.tuner= and Documentation/video4linux/bttv/CARDLIST
426 See drivers/scsi/BusLogic.c, comment before function
427 BusLogic_ParseDriverOptions().
429 c101= [NET] Moxa C101 synchronous serial card
431 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
432 Sometimes CPU hardware bugs make them report the cache
433 size incorrectly. The kernel will attempt work arounds
434 to fix known problems, but for some CPUs it is not
435 possible to determine what the correct size should be.
436 This option provides an override for these situations.
439 [SECURITY] Disable capabilities. This would normally
440 be used only if an alternative security model is to be
441 configured. Potentially dangerous and should only be
442 used if you are entirely sure of the consequences.
444 ccw_timeout_log [S390]
445 See Documentation/s390/CommonIO for details.
447 cgroup_disable= [KNL] Disable a particular controller
448 Format: {name of the controller(s) to disable}
449 {Currently supported controllers - "memory"}
451 checkreqprot [SELINUX] Set initial checkreqprot flag value.
452 Format: { "0" | "1" }
453 See security/selinux/Kconfig help text.
454 0 -- check protection applied by kernel (includes
455 any implied execute protection).
456 1 -- check protection requested by application.
457 Default value is set via a kernel config option.
458 Value can be changed at runtime via
459 /selinux/checkreqprot.
462 See Documentation/s390/CommonIO for details.
464 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
466 Forces specified clocksource (if available) to be used
467 when calculating gettimeofday(). If specified
468 clocksource is not available, it defaults to PIT.
469 Format: { pit | tsc | cyclone | pmtmr }
471 clocksource= [GENERIC_TIME] Override the default clocksource
473 Override the default clocksource and use the clocksource
474 with the name specified.
475 Some clocksource names to choose from, depending on
477 [all] jiffies (this is the base, fallback clocksource)
479 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
480 pxa_timer,timer3,32k_counter,timer0_1
482 [X86-32] pit,hpet,tsc,vmi-timer;
483 scx200_hrt on Geode; cyclone on IBM x440
491 clearcpuid=BITNUM [X86]
492 Disable CPUID feature X for the kernel. See
493 arch/x86/include/asm/cpufeature.h for the valid bit
494 numbers. Note the Linux specific bits are not necessarily
495 stable over kernel options, but the vendor specific
497 Also note that user programs calling CPUID directly
498 or using the feature without checking anything
499 will still see it. This just prevents it from
500 being used by the kernel or shown in /proc/cpuinfo.
501 Also note the kernel might malfunction if you disable
504 cmo_free_hint= [PPC] Format: { yes | no }
505 Specify whether pages are marked as being inactive
506 when they are freed. This is used in CMO environments
507 to determine OS memory pressure for page stealing by
511 code_bytes [X86] How many bytes of object code to print
516 com20020= [HW,NET] ARCnet - COM20020 chipset
518 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
520 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
524 ARCnet - COM90xx chipset (memory-mapped buffers)
525 Format: <io>[,<irq>[,<memstart>]]
527 condev= [HW,S390] console device
530 console= [KNL] Output console device and options.
532 tty<n> Use the virtual console device <n>.
536 Use the specified serial port. The options are of
537 the form "bbbbpnf", where "bbbb" is the baud rate,
538 "p" is parity ("n", "o", or "e"), "n" is number of
539 bits, and "f" is flow control ("r" for RTS or
540 omit it). Default is "9600n8".
542 See Documentation/serial-console.txt for more
544 Documentation/networking/netconsole.txt for an
547 uart[8250],io,<addr>[,options]
548 uart[8250],mmio,<addr>[,options]
549 Start an early, polled-mode console on the 8250/16550
550 UART at the specified I/O port or MMIO address,
551 switching to the matching ttyS device later. The
552 options are the same as for ttyS, above.
554 If the device connected to the port is not a TTY but a braille
555 device, prepend "brl," before the device type, for instance
557 For now, only VisioBraille is supported.
559 consoleblank= [KNL] The console blank (screen saver) timeout in
560 seconds. Defaults to 10*60 = 10mins. A value of 0
561 disables the blank timer.
564 [KNL] Change the default value for
565 /proc/<pid>/coredump_filter.
566 See also Documentation/filesystems/proc.txt.
568 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
570 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
572 crashkernel=nn[KMG]@ss[KMG]
573 [KNL] Reserve a chunk of physical memory to
574 hold a kernel to switch to with kexec on panic.
576 crashkernel=range1:size1[,range2:size2,...][@offset]
577 [KNL] Same as above, but depends on the memory
578 in the running system. The syntax of range is
579 start-[end] where start and end are both
580 a memory unit (amount[KMG]). See also
581 Documentation/kdump/kdump.txt for a example.
586 cs89x0_media= [HW,NET]
587 Format: { rj45 | aui | bnc }
590 See header of drivers/s390/block/dasd_devmap.c.
592 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
593 (one device per port)
594 Format: <port#>,<type>
595 See also Documentation/input/joystick-parport.txt
597 debug [KNL] Enable kernel debugging (events log level).
600 [KNL] verbose self-tests
602 Print debugging info while doing the locking API
604 We default to 0 (no extra messages), setting it to
605 1 will print _a lot_ more information - normally
606 only useful to kernel developers.
608 debug_objects [KNL] Enable object debugging
611 [KNL] Disable object debugging
613 debugpat [X86] Enable PAT debugging
615 decnet.addr= [HW,NET]
616 Format: <area>[,<node>]
617 See also Documentation/networking/decnet.txt.
620 [same as hugepagesz=] The size of the default
621 HugeTLB page size. This is the size represented by
622 the legacy /proc/ hugepages APIs, used for SHM, and
623 default size when mounting hugetlbfs filesystems.
624 Defaults to the default architecture's huge page size
628 Set number of hash buckets for dentry cache.
631 IO parameters + enable/disable command.
633 digiepca= [HW,SERIAL]
634 See drivers/char/README.epca and
635 Documentation/serial/digiepca.txt.
638 See Documentation/networking/ipv6.txt.
641 See Documentation/networking/ipv6.txt.
643 disable_mtrr_cleanup [X86]
644 The kernel tries to adjust MTRR layout from continuous
645 to discrete, to make X server driver able to add WB
646 entry later. This parameter disables that.
648 disable_mtrr_trim [X86, Intel and AMD only]
649 By default the kernel will trim any uncacheable
650 memory out of your available memory pool based on
651 MTRR settings. This parameter disables that behavior,
652 possibly causing your machine to run very slowly.
654 disable_timer_pin_1 [X86]
655 Disable PIN 1 of APIC timer
656 Can be useful to work around chipset bugs.
658 dmasound= [HW,OSS] Sound subsystem buffers
660 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
661 this option disables the debugging code at boot.
663 dma_debug_entries=<number>
664 This option allows to tune the number of preallocated
665 entries for DMA-API debugging code. One entry is
666 required per DMA-API allocation. Use this if the
667 DMA-API debugging code disables itself because the
668 architectural default is too low.
670 dma_debug_driver=<driver_name>
671 With this option the DMA-API debugging driver
672 filter feature can be enabled at boot time. Just
673 pass the driver to filter for as the parameter.
674 The filter can be disabled or changed to another
675 driver later using sysfs.
681 dynamic_printk Enables pr_debug()/dev_dbg() calls if
682 CONFIG_DYNAMIC_PRINTK_DEBUG has been enabled.
683 These can also be switched on/off via
684 <debugfs>/dynamic_printk/modules
686 earlycon= [KNL] Output early console device and options.
687 uart[8250],io,<addr>[,options]
688 uart[8250],mmio,<addr>[,options]
689 Start an early, polled-mode console on the 8250/16550
690 UART at the specified I/O port or MMIO address.
691 The options are the same as for ttyS, above.
693 earlyprintk= [X86,SH,BLACKFIN]
695 earlyprintk=serial[,ttySn[,baudrate]]
696 earlyprintk=ttySn[,baudrate]
697 earlyprintk=dbgp[debugController#]
699 Append ",keep" to not disable it when the real console
702 Only vga or serial or usb debug port at a time.
704 Currently only ttyS0 and ttyS1 are supported.
706 Interaction with the standard serial driver is not
709 The VGA output is eventually overwritten by the real
712 ekgdboc= [X86,KGDB] Allow early kernel console debugging
715 This is desgined to be used in conjunction with
716 the boot argument: earlyprintk=vga
721 Format: {"off" | "on" | "skip[mbr]"}
723 eisa_irq_edge= [PARISC,HW]
724 See header of drivers/parisc/eisa.c.
727 See comment before function elanfreq_setup() in
728 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
731 Format: {"anticipatory" | "cfq" | "deadline" | "noop"}
732 See Documentation/block/as-iosched.txt and
733 Documentation/block/deadline-iosched.txt for details.
735 elfcorehdr= [IA64,PPC,SH,X86]
736 Specifies physical address of start of kernel core
737 image elf header. Generally kexec loader will
738 pass this option to capture kernel.
739 See Documentation/kdump/kdump.txt for details.
741 enable_mtrr_cleanup [X86]
742 The kernel tries to adjust MTRR layout from continuous
743 to discrete, to make X server driver able to add WB
744 entry later. This parameter enables that.
746 enable_timer_pin_1 [X86]
747 Enable PIN 1 of APIC timer
748 Can be useful to work around chipset bugs
749 (in particular on some ATI chipsets).
750 The kernel tries to set a reasonable default.
752 enforcing [SELINUX] Set initial enforcing status.
754 See security/selinux/Kconfig help text.
755 0 -- permissive (log only, no denials).
756 1 -- enforcing (deny and log).
758 Value can be changed at runtime via /selinux/enforce.
761 Disable Error Record Serialization Table (ERST)
764 ether= [HW,NET] Ethernet cards parameters
765 This option is obsoleted by the "netdev=" option, which
766 has equivalent usage. See its documentation for details.
770 fail_make_request=[KNL]
771 General fault injection mechanism.
772 Format: <interval>,<probability>,<space>,<times>
773 See also /Documentation/fault-injection/.
776 See header of drivers/scsi/fd_mcs.c.
779 See header of drivers/scsi/fdomain.c.
782 See Documentation/blockdev/floppy.txt.
784 force_pal_cache_flush
785 [IA-64] Avoid check_sal_cache_flush which may hang on
786 buggy SAL_CACHE_FLUSH implementations. Using this
787 parameter will force ia64_sal_cache_flush to call
788 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
791 [FTRACE] will set and start the specified tracer
792 as early as possible in order to facilitate early
795 ftrace_dump_on_oops[=orig_cpu]
796 [FTRACE] will dump the trace buffers on oops.
797 If no parameter is passed, ftrace will dump
798 buffers of all CPUs, but if you pass orig_cpu, it will
799 dump only the buffer of the CPU that triggered the
802 ftrace_filter=[function-list]
803 [FTRACE] Limit the functions traced by the function
804 tracer at boot up. function-list is a comma separated
805 list of functions. This list can be changed at run
806 time by the set_ftrace_filter file in the debugfs
809 ftrace_notrace=[function-list]
810 [FTRACE] Do not trace the functions specified in
811 function-list. This list can be changed at run time
812 by the set_ftrace_notrace file in the debugfs
815 ftrace_graph_filter=[function-list]
816 [FTRACE] Limit the top level callers functions traced
817 by the function graph tracer at boot up.
818 function-list is a comma separated list of functions
819 that can be changed at run time by the
820 set_graph_function file in the debugfs tracing directory.
823 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
824 support via parallel port (up to 5 devices per port)
825 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
826 See also Documentation/input/joystick-parport.txt
830 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART
834 gcov_persist= [GCOV] When non-zero (default), profiling data for
835 kernel modules is saved and remains accessible via
836 debugfs, even when the module is unloaded/reloaded.
837 When zero, profiling data is discarded and associated
838 debugfs files are removed at module unload time.
841 See header of drivers/scsi/gdth.c.
843 gpt [EFI] Forces disk with valid GPT signature but
844 invalid Protective MBR to be treated as GPT.
848 hashdist= [KNL,NUMA] Large hashes allocated during boot
849 are distributed across NUMA nodes. Defaults on
850 for 64bit NUMA, off otherwise.
851 Format: 0 | 1 (for off | on)
853 hcl= [IA-64] SGI's Hardware Graph compatibility layer
855 hd= [EIDE] (E)IDE hard drive subsystem geometry
856 Format: <cyl>,<head>,<sect>
859 Disable Hardware Error Source Table (HEST) support;
860 corresponding firmware-first mode error processing
861 logic will be disabled.
863 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
864 size of <nn>. This works even on boxes that have no
865 highmem otherwise. This also works to reduce highmem
866 size on bigger boxes.
868 highres= [KNL] Enable/disable high resolution timer mode.
869 Valid parameters: "on", "off"
873 See Documentation/isdn/README.HiSax.
877 hpet= [X86-32,HPET] option to control HPET usage
878 Format: { enable (default) | disable | force |
880 disable: disable HPET and use PIT instead
881 force: allow force enabled of undocumented chips (ICH4,
883 verbose: show contents of HPET registers during setup
885 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
886 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
887 On x86-64 and powerpc, this option can be specified
888 multiple times interleaved with hugepages= to reserve
889 huge pages of different sizes. Valid pages sizes on
890 x86-64 are 2M (when the CPU supports "pse") and 1G
891 (when the CPU supports the "pdpe1gb" cpuinfo flag)
892 Note that 1GB pages can only be allocated at boot time
893 using hugepages= and not freed afterwards.
895 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
896 terminal devices. Valid values: 0..8
897 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
898 If specified, z/VM IUCV HVC accepts connections
899 from listed z/VM user IDs only.
901 i2c_bus= [HW] Override the default board specific I2C bus speed
902 or register an additional I2C bus that is not
903 registered from board initialization code.
907 i8042.debug [HW] Toggle i8042 debug mode
908 i8042.direct [HW] Put keyboard port into non-translated mode
909 i8042.dumbkbd [HW] Pretend that controller can only read data from
910 keyboard and cannot control its state
911 (Don't attempt to blink the leds)
912 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
913 i8042.nokbd [HW] Don't check/create keyboard port
914 i8042.noloop [HW] Disable the AUX Loopback command while probing
916 i8042.nomux [HW] Don't check presence of an active multiplexing
918 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
921 [HW] Frequency with which keyboard LEDs should blink
922 when kernel panics (default is 0.5 sec)
923 i8042.reset [HW] Reset the controller during init and cleanup
924 i8042.unlock [HW] Unlock (ignore) the keylock
928 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
929 indicates that the driver is running on unsupported
931 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
932 does not match list of supported models.
934 [HW] Report power status in /proc/i8k
935 (disabled by default)
936 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
939 ibmmcascsi= [HW,MCA,SCSI] IBM MicroChannel SCSI adapter
940 See Documentation/mca.txt.
943 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
945 ide-core.nodma= [HW] (E)IDE subsystem
946 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
947 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
948 .cdrom .chs .ignore_cable are additional options
949 See Documentation/ide/ide.txt.
951 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
952 Claim all unknown PCI IDE storage controllers.
955 Format: idle=poll, idle=mwait, idle=halt, idle=nomwait
956 Poll forces a polling idle loop that can slightly
957 improve the performance of waking up a idle CPU, but
958 will use a lot of power and make the system run hot.
960 idle=mwait: On systems which support MONITOR/MWAIT but
961 the kernel chose to not use it because it doesn't save
962 as much power as a normal idle loop, use the
963 MONITOR/MWAIT idle loop anyways. Performance should be
964 the same as idle=poll.
965 idle=halt: Halt is forced to be used for CPU idle.
966 In such case C2/C3 won't be used again.
967 idle=nomwait: Disable mwait for CPU C-states
969 ignore_loglevel [KNL]
970 Ignore loglevel setting - this will print /all/
971 kernel messages to the console. Useful for debugging.
974 Set number of hash buckets for inode cache.
977 Format: { "0" | "1" }
978 0 -- integrity auditing messages. (Default)
979 1 -- enable informational integrity auditing messages.
982 Format: { "sha1" | "md5" }
986 Load a policy which meets the needs of the Trusted
987 Computing Base. This means IMA will measure all
988 programs exec'd, files mmap'd for exec, and all files
989 opened for read by uid=0.
992 See header of drivers/scsi/in2000.c.
996 Run specified binary instead of /sbin/init as init
999 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1000 for working out where the kernel is dying during
1003 initrd= [BOOT] Specify the location of the initial ramdisk
1005 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1008 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
1010 Enable intel iommu driver.
1012 Disable intel iommu driver.
1013 igfx_off [Default Off]
1014 By default, gfx is mapped as normal device. If a gfx
1015 device has a dedicated DMAR unit, the DMAR unit is
1016 bypassed by not enabling DMAR with this option. In
1017 this case, gfx device will use physical address for
1020 With this option iommu will not optimize to look
1021 for io virtual address below 32 bit forcing dual
1022 address cycle on pci bus for cards supporting greater
1023 than 32 bit addressing. The default is to look
1024 for translation below 32 bit and if not available
1025 then look in the higher range.
1026 strict [Default Off]
1027 With this option on every unmap_single operation will
1028 result in a hardware IOTLB flush operation as opposed
1029 to batching them for performance.
1033 iomem= Disable strict checking of access to MMIO memory
1034 strict regions from userspace.
1050 io7= [HW] IO7 for Marvel based alpha systems
1051 See comment before marvel_specify_io7 in
1052 arch/alpha/kernel/core_marvel.c.
1054 io_delay= [X86] I/O delay method
1056 Standard port 0x80 based delay
1058 Alternate port 0xed based delay (needed on some systems)
1060 Simple two microseconds delay
1065 See Documentation/filesystems/nfs/nfsroot.txt.
1067 ip2= [HW] Set IO/IRQ pairs for up to 4 IntelliPort boards
1068 See comment before ip2_setup() in
1069 drivers/char/ip2/ip2base.c.
1071 ips= [HW,SCSI] Adaptec / IBM ServeRAID controller
1072 See header of drivers/scsi/ips.c.
1075 When an interrupt is not handled search all handlers
1076 for it. Intended to get systems with badly broken
1080 When an interrupt is not handled search all handlers
1081 for it. Also check all handlers each timer
1082 interrupt. Intended to get systems with badly broken
1086 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
1088 isolcpus= [KNL,SMP] Isolate CPUs from the general scheduler.
1090 <cpu number>,...,<cpu number>
1092 <cpu number>-<cpu number>
1093 (must be a positive range in ascending order)
1095 <cpu number>,...,<cpu number>-<cpu number>
1097 This option can be used to specify one or more CPUs
1098 to isolate from the general SMP balancing and scheduling
1099 algorithms. You can move a process onto or off an
1100 "isolated" CPU via the CPU affinity syscalls or cpuset.
1101 <cpu number> begins at 0 and the maximum value is
1102 "number of CPUs in system - 1".
1104 This option is the preferred way to isolate CPUs. The
1105 alternative -- manually setting the CPU mask of all
1106 tasks in the system -- can cause problems and
1107 suboptimal load balancer performance.
1111 js= [HW,JOY] Analog joystick
1112 See Documentation/input/joystick.txt.
1116 kernelcore=nn[KMG] [KNL,X86,IA-64,PPC] This parameter
1117 specifies the amount of memory usable by the kernel
1118 for non-movable allocations. The requested amount is
1119 spread evenly throughout all nodes in the system. The
1120 remaining memory in each node is used for Movable
1121 pages. In the event, a node is too small to have both
1122 kernelcore and Movable pages, kernelcore pages will
1123 take priority and other nodes will have a larger number
1124 of kernelcore pages. The Movable zone is used for the
1125 allocation of pages that may be reclaimed or moved
1126 by the page migration subsystem. This means that
1127 HugeTLB pages may not be allocated from this zone.
1128 Note that allocations like PTEs-from-HighMem still
1129 use the HighMem zone if it exists, and the Normal
1130 zone if it does not.
1132 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
1133 Format: <Controller#>[,poll interval]
1134 The controller # is the number of the ehci usb debug
1135 port as it is probed via PCI. The poll interval is
1136 optional and is the number seconds in between
1137 each poll cycle to the debug port in case you need
1138 the functionality for interrupting the kernel with
1139 gdb or control-c on the dbgp connection. When
1140 not using this parameter you use sysrq-g to break into
1141 the kernel debugger.
1143 kgdboc= [KGDB,HW] kgdb over consoles.
1144 Requires a tty driver that supports console polling,
1145 or a supported polling keyboard driver (non-usb).
1146 Serial only format: <serial_device>[,baud]
1147 keyboard only format: kbd
1148 keyboard and serial format: kbd,<serial_device>[,baud]
1150 kgdbwait [KGDB] Stop kernel execution and enter the
1151 kernel debugger at the earliest opportunity.
1153 kmac= [MIPS] korina ethernet MAC address.
1154 Configure the RouterBoard 532 series on-chip
1155 Ethernet adapter MAC address.
1157 kmemleak= [KNL] Boot-time kmemleak enable/disable
1158 Valid arguments: on, off
1161 kstack=N [X86] Print N words from the kernel stack
1164 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
1165 Default is 0 (don't ignore, but inject #GP)
1167 kvm.oos_shadow= [KVM] Disable out-of-sync shadow paging.
1168 Default is 1 (enabled)
1170 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
1173 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
1175 Default is 1 (enabled) if in 64bit or 32bit-PAE mode
1177 kvm-intel.bypass_guest_pf=
1178 [KVM,Intel] Disables bypassing of guest page faults
1179 on Intel chips. Default is 1 (enabled)
1181 kvm-intel.ept= [KVM,Intel] Disable extended page tables
1182 (virtualized MMU) support on capable Intel chips.
1183 Default is 1 (enabled)
1185 kvm-intel.emulate_invalid_guest_state=
1186 [KVM,Intel] Enable emulation of invalid guest states
1187 Default is 0 (disabled)
1189 kvm-intel.flexpriority=
1190 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
1191 Default is 1 (enabled)
1193 kvm-intel.unrestricted_guest=
1194 [KVM,Intel] Disable unrestricted guest feature
1195 (virtualized real and unpaged mode) on capable
1196 Intel chips. Default is 1 (enabled)
1198 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
1199 feature (tagged TLBs) on capable Intel chips.
1200 Default is 1 (enabled)
1206 lapic [X86-32,APIC] Enable the local APIC even if BIOS
1209 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
1212 libata.dma= [LIBATA] DMA control
1213 libata.dma=0 Disable all PATA and SATA DMA
1214 libata.dma=1 PATA and SATA Disk DMA only
1215 libata.dma=2 ATAPI (CDROM) DMA only
1216 libata.dma=4 Compact Flash DMA only
1217 Combinations also work, so libata.dma=3 enables DMA
1218 for disks and CDROMs, but not CFs.
1220 libata.ignore_hpa= [LIBATA] Ignore HPA limit
1221 libata.ignore_hpa=0 keep BIOS limits (default)
1222 libata.ignore_hpa=1 ignore limits, using full disk
1224 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
1228 libata.force= [LIBATA] Force configurations. The format is comma
1229 separated list of "[ID:]VAL" where ID is
1230 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
1231 matching port, link or device. Basically, it matches
1232 the ATA ID string printed on console by libata. If
1233 the whole ID part is omitted, the last PORT and DEVICE
1234 values are used. If ID hasn't been specified yet, the
1235 configuration applies to all ports, links and devices.
1237 If only DEVICE is omitted, the parameter applies to
1238 the port and all links and devices behind it. DEVICE
1239 number of 0 either selects the first device or the
1240 first fan-out link behind PMP device. It does not
1241 select the host link. DEVICE number of 15 selects the
1242 host link and device attached to it.
1244 The VAL specifies the configuration to force. As long
1245 as there's no ambiguity shortcut notation is allowed.
1246 For example, both 1.5 and 1.5G would work for 1.5Gbps.
1247 The following configurations can be forced.
1249 * Cable type: 40c, 80c, short40c, unk, ign or sata.
1250 Any ID with matching PORT is used.
1252 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
1254 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
1255 udma[/][16,25,33,44,66,100,133] notation is also
1258 * [no]ncq: Turn on or off NCQ.
1260 * nohrst, nosrst, norst: suppress hard, soft
1263 * dump_id: dump IDENTIFY data.
1265 If there are multiple matching configurations changing
1266 the same attribute, the last one is used.
1268 memblock=debug [KNL] Enable memblock debug messages.
1270 load_ramdisk= [RAM] List of ramdisks to load from floppy
1271 See Documentation/blockdev/ramdisk.txt.
1273 lockd.nlm_grace_period=P [NFS] Assign grace period.
1276 lockd.nlm_tcpport=N [NFS] Assign TCP port.
1279 lockd.nlm_timeout=T [NFS] Assign timeout value.
1282 lockd.nlm_udpport=M [NFS] Assign UDP port.
1285 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
1288 loglevel= All Kernel Messages with a loglevel smaller than the
1289 console loglevel will be printed to the console. It can
1290 also be changed with klogd or other programs. The
1291 loglevels are defined as follows:
1293 0 (KERN_EMERG) system is unusable
1294 1 (KERN_ALERT) action must be taken immediately
1295 2 (KERN_CRIT) critical conditions
1296 3 (KERN_ERR) error conditions
1297 4 (KERN_WARNING) warning conditions
1298 5 (KERN_NOTICE) normal but significant condition
1299 6 (KERN_INFO) informational
1300 7 (KERN_DEBUG) debug-level messages
1302 log_buf_len=n Sets the size of the printk ring buffer, in bytes.
1303 Format: { n | nk | nM }
1304 n must be a power of two. The default size
1305 is set in the kernel config file.
1307 logo.nologo [FB] Disables display of the built-in Linux logo.
1308 This may be used to provide more screen space for
1309 kernel log messages and is useful when debugging
1310 kernel boot problems.
1312 lp=0 [LP] Specify parallel ports to use, e.g,
1313 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
1314 lp=reset first parallel port). 'lp=0' disables the
1315 lp=auto printer driver. 'lp=reset' (which can be
1316 specified in addition to the ports) causes
1317 attached printers to be reset. Using
1318 lp=port1,port2,... specifies the parallel ports
1319 to associate lp devices with, starting with
1320 lp0. A port specification may be 'none' to skip
1321 that lp device, or a parport name such as
1322 'parport0'. Specifying 'lp=auto' instead of a
1323 port specification list means that device IDs
1324 from each port should be examined, to see if
1325 an IEEE 1284-compliant printer is attached; if
1326 so, the driver will manage that printer.
1327 See also header of drivers/char/lp.c.
1330 Sets loops_per_jiffy to given constant, thus avoiding
1331 time-consuming boot-time autodetection (up to 250 ms per
1332 CPU). 0 enables autodetection (default). To determine
1333 the correct value for your kernel, boot with normal
1334 autodetection and see what value is printed. Note that
1335 on SMP systems the preset will be applied to all CPUs,
1336 which is likely to cause problems if your CPUs need
1337 significantly divergent settings. An incorrect value
1338 will cause delays in the kernel to be wrong, leading to
1339 unpredictable I/O errors and other breakage. Although
1340 unlikely, in the extreme case this might damage your
1344 Format: <io>,<irq>,<dma>
1346 mac5380= [HW,SCSI] Format:
1347 <can_queue>,<cmd_per_lun>,<sg_tablesize>,<hostid>,<use_tags>
1349 machvec= [IA64] Force the use of a particular machine-vector
1350 (machvec) in a generic kernel.
1351 Example: machvec=hpzx1_swiotlb
1353 machtype= [Loongson] Share the same kernel image file between different
1355 Example: machtype=lemote-yeeloong-2f-7inch
1357 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
1358 than or equal to this physical address is ignored.
1360 maxcpus= [SMP] Maximum number of processors that an SMP kernel
1361 should make use of. maxcpus=n : n >= 0 limits the
1362 kernel to using 'n' processors. n=0 is a special case,
1363 it is equivalent to "nosmp", which also disables
1366 max_loop= [LOOP] Maximum number of loopback devices that can
1370 max_luns= [SCSI] Maximum number of LUNs to probe.
1371 Should be between 1 and 2^32-1.
1374 [SCSI] Maximum number of LUNs received.
1375 Should be between 1 and 16384.
1379 mce [X86-32] Machine Check Exception
1381 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.txt
1383 md= [HW] RAID subsystems devices and level
1384 See Documentation/md.txt.
1387 Format: <first>,<last>
1388 Specifies range of consoles to be captured by the MDA.
1390 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
1391 Amount of memory to be used when the kernel is not able
1392 to see the whole system memory or for test.
1393 [X86-32] Use together with memmap= to avoid physical
1394 address space collisions. Without memmap= PCI devices
1395 could be placed at addresses belonging to unused RAM.
1397 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
1401 [KNL,SH] Allow user to override the default size for
1402 per-device physically contiguous DMA buffers.
1404 memmap=exactmap [KNL,X86] Enable setting of an exact
1405 E820 memory map, as specified by the user.
1406 Such memmap=exactmap lines can be constructed based on
1407 BIOS output or other requirements. See the memmap=nn@ss
1410 memmap=nn[KMG]@ss[KMG]
1411 [KNL] Force usage of a specific region of memory
1412 Region of memory to be used, from ss to ss+nn.
1414 memmap=nn[KMG]#ss[KMG]
1415 [KNL,ACPI] Mark specific memory as ACPI data.
1416 Region of memory to be used, from ss to ss+nn.
1418 memmap=nn[KMG]$ss[KMG]
1419 [KNL,ACPI] Mark specific memory as reserved.
1420 Region of memory to be used, from ss to ss+nn.
1421 Example: Exclude memory from 0x18690000-0x1869ffff
1422 memmap=64K$0x18690000
1424 memmap=0x10000$0x18690000
1426 memory_corruption_check=0/1 [X86]
1427 Some BIOSes seem to corrupt the first 64k of
1428 memory when doing things like suspend/resume.
1429 Setting this option will scan the memory
1430 looking for corruption. Enabling this will
1431 both detect corruption and prevent the kernel
1432 from using the memory being corrupted.
1433 However, its intended as a diagnostic tool; if
1434 repeatable BIOS-originated corruption always
1435 affects the same memory, you can use memmap=
1436 to prevent the kernel from using that memory.
1438 memory_corruption_check_size=size [X86]
1439 By default it checks for corruption in the low
1440 64k, making this memory unavailable for normal
1441 use. Use this parameter to scan for
1442 corruption in more or less memory.
1444 memory_corruption_check_period=seconds [X86]
1445 By default it checks for corruption every 60
1446 seconds. Use this parameter to check at some
1447 other rate. 0 disables periodic checking.
1449 memtest= [KNL,X86] Enable memtest
1451 default : 0 <disable>
1452 Specifies the number of memtest passes to be
1453 performed. Each pass selects another test
1454 pattern from a given set of patterns. Memtest
1455 fills the memory with this pattern, validates
1456 memory contents and reserves bad memory
1457 regions that are detected.
1459 meye.*= [HW] Set MotionEye Camera parameters
1460 See Documentation/video4linux/meye.txt.
1462 mfgpt_irq= [IA-32] Specify the IRQ to use for the
1463 Multi-Function General Purpose Timers on AMD Geode
1466 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
1467 the BIOS has incorrectly applied a workaround. TinyBIOS
1468 version 0.98 is known to be affected, 0.99 fixes the
1469 problem by letting the user disable the workaround.
1473 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
1474 physical address is ignored.
1476 mini2440= [ARM,HW,KNL]
1477 Format:[0..2][b][c][t]
1479 MINI2440 configuration specification:
1480 0 - The attached screen is the 3.5" TFT
1481 1 - The attached screen is the 7" TFT
1482 2 - The VGA Shield is attached (1024x768)
1483 Leaving out the screen size parameter will not load
1484 the TFT driver, and the framebuffer will be left
1486 b - Enable backlight. The TFT backlight pin will be
1487 linked to the kernel VESA blanking code and a GPIO
1488 LED. This parameter is not necessary when using the
1490 c - Enable the s3c camera interface.
1491 t - Reserved for enabling touchscreen support. The
1492 touchscreen support is not enabled in the mainstream
1493 kernel as of 2.6.30, a preliminary port can be found
1494 in the "bleeding edge" mini2440 support kernel at
1495 http://repo.or.cz/w/linux-2.6/mini2440.git
1498 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
1499 parameter allows control of the logging verbosity for
1500 the additional memory initialisation checks. A value
1501 of 0 disables mminit logging and a level of 4 will
1502 log everything. Information is printed at KERN_DEBUG
1503 so loglevel=8 may also need to be specified.
1506 [MOUSE] Maximum time between finger touching and
1507 leaving touchpad surface for touch to be considered
1508 a tap and be reported as a left button click (for
1509 touchpads working in absolute mode only).
1511 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
1512 reporting absolute coordinates, such as tablets
1513 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
1514 reporting absolute coordinates, such as tablets
1516 movablecore=nn[KMG] [KNL,X86,IA-64,PPC] This parameter
1517 is similar to kernelcore except it specifies the
1518 amount of memory used for migratable allocations.
1519 If both kernelcore and movablecore is specified,
1520 then kernelcore will be at *least* the specified
1521 value but may be more. If movablecore on its own
1522 is specified, the administrator must be careful
1523 that the amount of memory usable for all allocations
1529 MTD_Partition= [MTD]
1530 Format: <name>,<region-number>,<size>,<offset>
1532 MTD_Region= [MTD] Format:
1533 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
1536 See drivers/mtd/cmdlinepart.c.
1538 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
1540 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
1542 boundary - index of last SLC block on Flex-OneNAND.
1543 The remaining blocks are configured as MLC blocks.
1544 lock - Configure if Flex-OneNAND boundary should be locked.
1545 Once locked, the boundary cannot be changed.
1546 1 indicates lock status, 0 indicates unlock status.
1549 ARM/S3C2412 JIVE boot control
1551 See arch/arm/mach-s3c2412/mach-jive.c
1553 mtouchusb.raw_coordinates=
1554 [HW] Make the MicroTouch USB driver use raw coordinates
1555 ('y', default) or cooked coordinates ('n')
1557 mtrr_chunk_size=nn[KMG] [X86]
1558 used for mtrr cleanup. It is largest continuous chunk
1559 that could hold holes aka. UC entries.
1561 mtrr_gran_size=nn[KMG] [X86]
1562 Used for mtrr cleanup. It is granularity of mtrr block.
1564 Large value could prevent small alignment from
1567 mtrr_spare_reg_nr=n [X86]
1569 Range: 0,7 : spare reg number
1571 Used for mtrr cleanup. It is spare mtrr entries number.
1572 Set to 2 or more if your graphical card needs more.
1574 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
1577 See header of drivers/scsi/NCR_D700.c.
1581 ncr53c400= [HW,SCSI]
1583 ncr53c400a= [HW,SCSI]
1585 ncr53c406a= [HW,SCSI]
1587 ncr53c8xx= [HW,SCSI]
1589 netdev= [NET] Network devices parameters
1590 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
1591 Note that mem_start is often overloaded to mean
1592 something different and driver-specific.
1593 This usage is only documented in each driver source
1597 [NETFILTER] Enable connection tracking flow accounting
1598 0 to disable accounting
1599 1 to enable accounting
1600 Default value depends on CONFIG_NF_CT_ACCT that is
1601 going to be removed in 2.6.29.
1604 See Documentation/filesystems/nfs/nfsroot.txt.
1606 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
1607 See Documentation/filesystems/nfs/nfsroot.txt.
1609 nfs.callback_tcpport=
1610 [NFS] set the TCP port on which the NFSv4 callback
1611 channel should listen.
1614 [NFS] sets the pathname to the program which is used
1615 to update the NFS client cache entries.
1617 nfs.cache_getent_timeout=
1618 [NFS] sets the timeout after which an attempt to
1619 update a cache entry is deemed to have failed.
1621 nfs.idmap_cache_timeout=
1622 [NFS] set the maximum lifetime for idmapper cache
1626 [NFS] enable 64-bit inode numbers.
1627 If zero, the NFS client will fake up a 32-bit inode
1628 number for the readdir() and stat() syscalls instead
1629 of returning the full 64-bit number.
1630 The default is to return 64-bit inode numbers.
1632 nmi_debug= [KNL,AVR32,SH] Specify one or more actions to take
1633 when a NMI is triggered.
1634 Format: [state][,regs][,debounce][,die]
1636 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
1637 Format: [panic,][num]
1639 0 - turn nmi_watchdog off
1640 1 - use the IO-APIC timer for the NMI watchdog
1641 2 - use the local APIC for the NMI watchdog using
1642 a performance counter. Note: This will use one
1643 performance counter and the local APIC's performance
1645 When panic is specified, panic when an NMI watchdog
1647 This is useful when you use a panic=... timeout and
1648 need the box quickly up again.
1649 Instead of 1 and 2 it is possible to use the following
1650 symbolic names: lapic and ioapic
1651 Example: nmi_watchdog=2 or nmi_watchdog=panic,lapic
1653 netpoll.carrier_timeout=
1654 [NET] Specifies amount of time (in seconds) that
1655 netpoll should wait for a carrier. By default netpoll
1658 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
1659 emulation library even if a 387 maths coprocessor
1663 [HW] Never suspend the console
1664 Disable suspending of consoles during suspend and
1665 hibernate operations. Once disabled, debugging
1666 messages can reach various consoles while the rest
1667 of the system is being put to sleep (ie, while
1668 debugging driver suspend/resume hooks). This may
1669 not work reliably with all consoles, but is known
1670 to work with serial and VGA consoles.
1672 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
1673 caches in the slab allocator. Saves per-node memory,
1674 but will impact performance.
1678 noapic [SMP,APIC] Tells the kernel to not make use of any
1679 IOAPICs that may be present in the system.
1681 nobats [PPC] Do not use BATs for mapping kernel lowmem
1682 on "Classic" PPC cores.
1686 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
1688 nodelayacct [KNL] Disable per-task delay accounting
1690 nodisconnect [HW,SCSI,M68K] Disables SCSI disconnects.
1692 nodsp [SH] Disable hardware DSP at boot time.
1694 noefi [X86] Disable EFI runtime services support.
1699 On X86-32 available only on PAE configured kernels.
1700 noexec=on: enable non-executable mappings (default)
1701 noexec=off: disable non-executable mappings
1704 This affects only 32-bit executables.
1705 noexec32=on: enable non-executable mappings (default)
1706 read doesn't imply executable mappings
1707 noexec32=off: disable non-executable mappings
1708 read implies executable mappings
1710 nofpu [SH] Disable hardware FPU at boot time.
1712 nofxsr [BUGS=X86-32] Disables x86 floating point extended
1713 register save and restore. The kernel will only save
1714 legacy floating-point registers on task switch.
1716 noxsave [BUGS=X86] Disables x86 extended register state save
1717 and restore using xsave. The kernel will fallback to
1718 enabling legacy floating-point and sse state.
1720 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
1721 wfi(ARM) instruction doesn't work correctly and not to
1722 use it. This is also useful when using JTAG debugger.
1724 no-hlt [BUGS=X86-32] Tells the kernel that the hlt
1725 instruction doesn't work correctly and not to
1728 no_file_caps Tells the kernel not to honor file capabilities. The
1729 only way then for a file to be executed with privilege
1730 is to be setuid root or executed by root.
1732 nohalt [IA-64] Tells the kernel not to use the power saving
1733 function PAL_HALT_LIGHT when idle. This increases
1734 power-consumption. On the positive side, it reduces
1735 interrupt wake-up latency, which may improve performance
1736 in certain environments such as networked servers or
1739 nohz= [KNL] Boottime enable/disable dynamic ticks
1740 Valid arguments: on, off
1743 noiotrap [SH] Disables trapped I/O port accesses.
1745 noirqdebug [X86-32] Disables the code which attempts to detect and
1746 disable unhandled interrupt sources.
1748 no_timer_check [X86,APIC] Disables the code which tests for
1749 broken timer IRQ sources.
1751 noisapnp [ISAPNP] Disables ISA PnP code.
1753 noinitrd [RAM] Tells the kernel not to load any configured
1756 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
1761 nojitter [IA64] Disables jitter checking for ITC timers.
1763 nolapic [X86-32,APIC] Do not enable or use the local APIC.
1765 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
1767 noltlbs [PPC] Do not use large page/tlb entries for kernel
1768 lowmem mapping on PPC40x.
1770 nomca [IA-64] Disable machine check abort handling
1772 nomce [X86-32] Machine Check Exception
1774 nomfgpt [X86-32] Disable Multi-Function General Purpose
1775 Timer usage (for AMD Geode machines).
1777 nopat [X86] Disable PAT (page attribute table extension of
1778 pagetables) support.
1780 norandmaps Don't use address space randomization. Equivalent to
1781 echo 0 > /proc/sys/kernel/randomize_va_space
1783 noreplace-paravirt [X86-32,PV_OPS] Don't patch paravirt_ops
1785 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
1786 with UP alternatives
1788 noresidual [PPC] Don't use residual data on PReP machines.
1790 noresume [SWSUSP] Disables resume and restores original swap
1793 no-scroll [VGA] Disables scrollback.
1794 This is required for the Braillex ib80-piezo Braille
1795 reader made by F.H. Papenmeier (Germany).
1799 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
1801 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
1802 and disable the IO APIC. legacy for "maxcpus=0".
1804 nosoftlockup [KNL] Disable the soft-lockup detector.
1806 noswapaccount [KNL] Disable accounting of swap in memory resource
1807 controller. (See Documentation/cgroups/memory.txt)
1809 nosync [HW,M68K] Disables sync negotiation for all devices.
1811 notsc [BUGS=X86-32] Disable Time Stamp Counter
1813 nousb [USB] Disable the USB subsystem
1817 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
1819 nptcg= [IA64] Override max number of concurrent global TLB
1820 purges which is reported from either PAL_VM_SUMMARY or
1823 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
1824 could support. nr_cpus=n : n >= 1 limits the kernel to
1825 supporting 'n' processors. Later in runtime you can not
1826 use hotplug cpu feature to put more cpu back to online.
1827 just like you compile the kernel NR_CPUS=n
1829 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
1831 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
1832 one of ['zone', 'node', 'default'] can be specified
1833 This can be set from sysctl after boot.
1834 See Documentation/sysctl/vm.txt for details.
1836 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
1837 See Documentation/debugging-via-ohci1394.txt for more
1840 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
1841 Rather than timing out after 20 ms if an EC
1842 command is not properly ACKed, override the length
1843 of the timeout. We have interrupts disabled while
1844 waiting for the ACK, so if this is set too high
1845 interrupts *may* be lost!
1847 omap_mux= [OMAP] Override bootloader pin multiplexing.
1848 Format: <mux_mode0.mode_name=value>...
1849 For example, to override I2C bus2:
1850 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
1855 oprofile.timer= [HW]
1856 Use timer interrupt instead of performance counters
1858 oprofile.cpu_type= Force an oprofile cpu type
1859 This might be useful if you have an older oprofile
1860 userland or if you want common events.
1861 Format: { arch_perfmon }
1862 arch_perfmon: [X86] Force use of architectural
1863 perfmon on Intel CPUs instead of the
1864 CPU specific event set.
1866 osst= [HW,SCSI] SCSI Tape Driver
1867 Format: <buffer_size>,<write_threshold>
1868 See also Documentation/scsi/st.txt.
1870 panic= [KNL] Kernel behaviour on panic
1873 parkbd.port= [HW] Parallel port number the keyboard adapter is
1874 connected to, default is 0.
1876 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
1877 0 for XT, 1 for AT (default is AT).
1880 parport= [HW,PPT] Specify parallel ports. 0 disables.
1881 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
1882 Use 'auto' to force the driver to use any
1883 IRQ/DMA settings detected (the default is to
1884 ignore detected IRQ/DMA settings because of
1885 possible conflicts). You can specify the base
1886 address, IRQ, and DMA settings; IRQ and DMA
1887 should be numbers, or 'auto' (for using detected
1888 settings on that particular port), or 'nofifo'
1889 (to avoid using a FIFO even if it is detected).
1890 Parallel ports are assigned in the order they
1891 are specified on the command line, starting
1894 parport_init_mode= [HW,PPT]
1895 Configure VIA parallel port to operate in
1896 a specific mode. This is necessary on Pegasos
1897 computer where firmware has no options for setting
1898 up parallel port mode and sets it to spp.
1899 Currently this function knows 686a and 8231 chips.
1900 Format: [spp|ps2|epp|ecp|ecpepp]
1902 pas2= [HW,OSS] Format:
1903 <io>,<irq>,<dma>,<dma16>,<sb_io>,<sb_irq>,<sb_dma>,<sb_dma16>
1906 See header of drivers/scsi/pas16.c.
1909 Halt all CPUs after the first oops has been printed for
1910 the specified number of seconds. This is to be used if
1911 your oopses keep scrolling off the screen.
1916 See header of drivers/block/paride/pcd.c.
1917 See also Documentation/blockdev/paride.txt.
1919 pci=option[,option...] [PCI] various PCI subsystem options:
1920 earlydump [X86] dump PCI config space before the kernel
1922 off [X86] don't probe for the PCI bus
1923 bios [X86-32] force use of PCI BIOS, don't access
1924 the hardware directly. Use this if your machine
1925 has a non-standard PCI host bridge.
1926 nobios [X86-32] disallow use of PCI BIOS, only direct
1927 hardware access methods are allowed. Use this
1928 if you experience crashes upon bootup and you
1929 suspect they are caused by the BIOS.
1930 conf1 [X86] Force use of PCI Configuration
1932 conf2 [X86] Force use of PCI Configuration
1934 noaer [PCIE] If the PCIEAER kernel config parameter is
1935 enabled, this kernel boot option can be used to
1936 disable the use of PCIE advanced error reporting.
1937 nodomains [PCI] Disable support for multiple PCI
1938 root domains (aka PCI segments, in ACPI-speak).
1939 nommconf [X86] Disable use of MMCONFIG for PCI
1941 check_enable_amd_mmconf [X86] check for and enable
1942 properly configured MMIO access to PCI
1943 config space on AMD family 10h CPU
1944 nomsi [MSI] If the PCI_MSI kernel config parameter is
1945 enabled, this kernel boot option can be used to
1946 disable the use of MSI interrupts system-wide.
1947 noioapicquirk [APIC] Disable all boot interrupt quirks.
1948 Safety option to keep boot IRQs enabled. This
1949 should never be necessary.
1950 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
1951 primary IO-APIC for bridges that cannot disable
1952 boot IRQs. This fixes a source of spurious IRQs
1953 when the system masks IRQs.
1954 noioapicreroute [APIC] Disable workaround that uses the
1955 boot IRQ equivalent of an IRQ that connects to
1956 a chipset where boot IRQs cannot be disabled.
1957 The opposite of ioapicreroute.
1958 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
1959 routing table. These calls are known to be buggy
1960 on several machines and they hang the machine
1961 when used, but on other computers it's the only
1962 way to get the interrupt routing table. Try
1963 this option if the kernel is unable to allocate
1964 IRQs or discover secondary PCI buses on your
1966 rom [X86] Assign address space to expansion ROMs.
1967 Use with caution as certain devices share
1968 address decoders between ROMs and other
1970 norom [X86] Do not assign address space to
1971 expansion ROMs that do not already have
1972 BIOS assigned address ranges.
1973 nobar [X86] Do not assign address space to the
1974 BARs that weren't assigned by the BIOS.
1975 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
1976 assigned automatically to PCI devices. You can
1977 make the kernel exclude IRQs of your ISA cards
1979 pirqaddr=0xAAAAA [X86] Specify the physical address
1980 of the PIRQ table (normally generated
1981 by the BIOS) if it is outside the
1982 F0000h-100000h range.
1983 lastbus=N [X86] Scan all buses thru bus #N. Can be
1984 useful if the kernel is unable to find your
1985 secondary buses and you want to tell it
1986 explicitly which ones they are.
1987 assign-busses [X86] Always assign all PCI bus
1988 numbers ourselves, overriding
1989 whatever the firmware may have done.
1990 usepirqmask [X86] Honor the possible IRQ mask stored
1991 in the BIOS $PIR table. This is needed on
1992 some systems with broken BIOSes, notably
1993 some HP Pavilion N5400 and Omnibook XE3
1994 notebooks. This will have no effect if ACPI
1995 IRQ routing is enabled.
1996 noacpi [X86] Do not use ACPI for IRQ routing
1997 or for PCI scanning.
1998 use_crs [X86] Use PCI host bridge window information
1999 from ACPI. On BIOSes from 2008 or later, this
2000 is enabled by default. If you need to use this,
2001 please report a bug.
2002 nocrs [X86] Ignore PCI host bridge windows from ACPI.
2003 If you need to use this, please report a bug.
2004 routeirq Do IRQ routing for all PCI devices.
2005 This is normally done in pci_enable_device(),
2006 so this option is a temporary workaround
2007 for broken drivers that don't call it.
2008 skip_isa_align [X86] do not align io start addr, so can
2009 handle more pci cards
2010 firmware [ARM] Do not re-enumerate the bus but instead
2011 just use the configuration from the
2012 bootloader. This is currently used on
2013 IXP2000 systems where the bus has to be
2014 configured a certain way for adjunct CPUs.
2015 noearly [X86] Don't do any early type 1 scanning.
2016 This might help on some broken boards which
2017 machine check when some devices' config space
2018 is read. But various workarounds are disabled
2019 and some IOMMU drivers will not work.
2020 bfsort Sort PCI devices into breadth-first order.
2021 This sorting is done to get a device
2022 order compatible with older (<= 2.4) kernels.
2023 nobfsort Don't sort PCI devices into breadth-first order.
2024 cbiosize=nn[KMG] The fixed amount of bus space which is
2025 reserved for the CardBus bridge's IO window.
2026 The default value is 256 bytes.
2027 cbmemsize=nn[KMG] The fixed amount of bus space which is
2028 reserved for the CardBus bridge's memory
2029 window. The default value is 64 megabytes.
2032 [<order of align>@][<domain>:]<bus>:<slot>.<func>[; ...]
2033 Specifies alignment and device to reassign
2034 aligned memory resources.
2035 If <order of align> is not specified,
2036 PAGE_SIZE is used as alignment.
2037 PCI-PCI bridge can be specified, if resource
2038 windows need to be expanded.
2039 ecrc= Enable/disable PCIe ECRC (transaction layer
2040 end-to-end CRC checking).
2041 bios: Use BIOS/firmware settings. This is the
2046 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
2049 force Enable ASPM even on devices that claim not to support it.
2050 WARNING: Forcing ASPM on may cause system lockups.
2052 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
2053 Format: {auto|force}[,nomsi]
2054 auto Use native PCIe PME signaling if the BIOS allows the
2055 kernel to control PCIe config registers of root ports.
2056 force Use native PCIe PME signaling even if the BIOS refuses
2057 to allow the kernel to control the relevant PCIe config
2059 nomsi Do not use MSI for native PCIe PME signaling (this makes
2060 all PCIe root ports use INTx for everything).
2062 pcmv= [HW,PCMCIA] BadgePAD 4
2065 See Documentation/blockdev/paride.txt.
2067 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
2070 See arch/parisc/kernel/pdc_chassis.c
2072 percpu_alloc= Select which percpu first chunk allocator to use.
2073 Currently supported values are "embed" and "page".
2074 Archs may support subset or none of the selections.
2075 See comments in mm/percpu.c for details on each
2076 allocator. This parameter is primarily for debugging
2077 and performance comparison.
2080 See Documentation/blockdev/paride.txt.
2083 See Documentation/blockdev/paride.txt.
2085 pirq= [SMP,APIC] Manual mp-table setup
2086 See Documentation/x86/i386/IO-APIC.txt.
2088 plip= [PPT,NET] Parallel port network link
2089 Format: { parport<nr> | timid | 0 }
2090 See also Documentation/parport.txt.
2092 pmtmr= [X86] Manual setup of pmtmr I/O Port.
2093 Override pmtimer IOPort with a hex value.
2097 Enable PNP debug messages. This depends on the
2098 CONFIG_PNP_DEBUG_MESSAGES option.
2104 { on | off | curr | res | no-curr | no-res }
2107 [ISAPNP] Exclude IRQs for the autoconfiguration
2110 [ISAPNP] Exclude DMAs for the autoconfiguration
2112 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
2113 Ranges are in pairs (I/O port base and size).
2116 [ISAPNP] Exclude memory regions for the
2118 Ranges are in pairs (memory base and size).
2120 ports= [IP_VS_FTP] IPVS ftp helper module
2122 Up to 8 (IP_VS_APP_MAX_PORTS) ports
2124 Format: <port>,<port>....
2126 print-fatal-signals=
2127 [KNL] debug: print fatal signals
2129 If enabled, warn about various signal handling
2130 related application anomalies: too many signals,
2131 too many POSIX.1 timers, fatal signals causing a
2134 If you hit the warning due to signal overflow,
2135 you might want to try "ulimit -i unlimited".
2139 printk.time= Show timing data prefixed to each printk message line
2140 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
2142 processor.max_cstate= [HW,ACPI]
2143 Limit processor to maximum C-state
2144 max_cstate=9 overrides any DMI blacklist limit.
2146 processor.nocst [HW,ACPI]
2147 Ignore the _CST method to determine C-states,
2148 instead using the legacy FADT method
2150 profile= [KNL] Enable kernel profiling via /proc/profile
2151 Format: [schedule,]<number>
2152 Param: "schedule" - profile schedule points.
2153 Param: <number> - step/bucket size as a power of 2 for
2154 statistical time based profiling.
2155 Param: "sleep" - profile D-state sleeping (millisecs).
2156 Requires CONFIG_SCHEDSTATS
2157 Param: "kvm" - profile VM exits.
2159 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
2161 See Documentation/blockdev/ramdisk.txt.
2163 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
2164 probe for; one of (bare|imps|exps|lifebook|any).
2165 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
2167 psmouse.resetafter= [HW,MOUSE]
2168 Try to reset the device after so many bad packets
2171 [HW,MOUSE] Set desired mouse resolution, in dpi.
2172 psmouse.smartscroll=
2173 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
2174 0 = disabled, 1 = enabled (default).
2176 pss= [HW,OSS] Personal Sound System (ECHO ESC614)
2178 <io>,<mss_io>,<mss_irq>,<mss_dma>,<mpu_io>,<mpu_irq>
2181 See Documentation/blockdev/paride.txt.
2184 [KNL] Number of legacy pty's. Overwrites compiled-in
2187 quiet [KNL] Disable most log messages
2192 See Documentation/md.txt.
2194 ramdisk_blocksize= [RAM]
2195 See Documentation/blockdev/ramdisk.txt.
2197 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
2198 See Documentation/blockdev/ramdisk.txt.
2200 rcupdate.blimit= [KNL,BOOT]
2201 Set maximum number of finished RCU callbacks to process
2204 rcupdate.qhimark= [KNL,BOOT]
2205 Set threshold of queued
2206 RCU callbacks over which batch limiting is disabled.
2208 rcupdate.qlowmark= [KNL,BOOT]
2209 Set threshold of queued RCU callbacks below which
2210 batch limiting is re-enabled.
2214 Run specified binary instead of /init from the ramdisk,
2215 used for early userspace startup. See initrd.
2217 reboot= [BUGS=X86-32,BUGS=ARM,BUGS=IA-64] Rebooting mode
2218 Format: <reboot_mode>[,<reboot_mode2>[,...]]
2219 See arch/*/kernel/reboot.c or arch/*/kernel/process.c
2222 [KNL, SMP] Set scheduler's default relax_domain_level.
2223 See Documentation/cgroups/cpusets.txt.
2225 reserve= [KNL,BUGS] Force the kernel to ignore some iomem area
2227 reservetop= [X86-32]
2229 Reserves a hole at the top of the kernel virtual
2232 reset_devices [KNL] Force drivers to reset the underlying device
2233 during initialization.
2236 Specify the partition device for software suspend
2238 resume_offset= [SWSUSP]
2239 Specify the offset from the beginning of the partition
2240 given by "resume=" at which the swap header is located,
2241 in <PAGE_SIZE> units (needed only for swap files).
2242 See Documentation/power/swsusp-and-swap-files.txt
2244 retain_initrd [RAM] Keep initrd memory after extraction
2246 rhash_entries= [KNL,NET]
2247 Set number of hash buckets for route cache
2249 riscom8= [HW,SERIAL]
2250 Format: <io_board1>[,<io_board2>[,...<io_boardN>]]
2252 ro [KNL] Mount root device read-only on boot
2254 root= [KNL] Root filesystem
2256 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
2257 mount the root filesystem
2259 rootflags= [KNL] Set root filesystem mount option string
2261 rootfstype= [KNL] Set root filesystem type
2263 rootwait [KNL] Wait (indefinitely) for root device to show up.
2264 Useful for devices that are detected asynchronously
2265 (e.g. USB and MMC devices).
2267 rw [KNL] Mount root device read-write on boot
2269 S [KNL] Run init in single mode
2272 See drivers/net/irda/sa1100_ir.c.
2274 sbni= [NET] Granch SBNI12 leased line adapter
2276 sched_debug [KNL] Enables verbose scheduler debug messages.
2278 scsi_debug_*= [SCSI]
2279 See drivers/scsi/scsi_debug.c.
2281 scsi_default_dev_flags=
2282 [SCSI] SCSI default device flags
2285 scsi_dev_flags= [SCSI] Black/white list entry for vendor and model
2286 Format: <vendor>:<model>:<flags>
2287 (flags are integer value)
2289 scsi_logging_level= [SCSI] a bit mask of logging levels
2290 See drivers/scsi/scsi_logging.h for bits. Also
2291 settable via sysctl at dev.scsi.logging_level
2292 (/proc/sys/dev/scsi/logging_level).
2293 There is also a nice 'scsi_logging_level' script in the
2294 S390-tools package, available for download at
2295 http://www-128.ibm.com/developerworks/linux/linux390/s390-tools-1.5.4.html
2297 scsi_mod.scan= [SCSI] sync (default) scans SCSI busses as they are
2298 discovered. async scans them in kernel threads,
2299 allowing boot to proceed. none ignores them, expecting
2300 user space to do the scan.
2302 security= [SECURITY] Choose a security module to enable at boot.
2303 If this boot parameter is not specified, only the first
2304 security module asking for security registration will be
2305 loaded. An invalid security module name will be treated
2306 as if no module has been chosen.
2308 selinux= [SELINUX] Disable or enable SELinux at boot time.
2309 Format: { "0" | "1" }
2310 See security/selinux/Kconfig help text.
2313 Default value is set via kernel config option.
2314 If enabled at boot time, /selinux/disable can be used
2315 later to disable prior to initial policy load.
2317 serialnumber [BUGS=X86-32]
2320 Maximal number of shapers.
2322 show_msr= [x86] show boot-time MSR settings
2323 Format: { <integer> }
2324 Show boot-time (BIOS-initialized) MSR settings.
2325 The parameter means the number of CPUs to show,
2326 for example 1 means boot CPU only.
2329 See header of drivers/scsi/sim710.c.
2336 slub_debug[=options[,slabs]] [MM, SLUB]
2337 Enabling slub_debug allows one to determine the
2338 culprit if slab objects become corrupted. Enabling
2339 slub_debug can create guard zones around objects and
2340 may poison objects when not in use. Also tracks the
2341 last alloc / free. For more information see
2342 Documentation/vm/slub.txt.
2344 slub_max_order= [MM, SLUB]
2345 Determines the maximum allowed order for slabs.
2346 A high setting may cause OOMs due to memory
2347 fragmentation. For more information see
2348 Documentation/vm/slub.txt.
2350 slub_min_objects= [MM, SLUB]
2351 The minimum number of objects per slab. SLUB will
2352 increase the slab order up to slub_max_order to
2353 generate a sufficiently large slab able to contain
2354 the number of objects indicated. The higher the number
2355 of objects the smaller the overhead of tracking slabs
2356 and the less frequently locks need to be acquired.
2357 For more information see Documentation/vm/slub.txt.
2359 slub_min_order= [MM, SLUB]
2360 Determines the mininum page order for slabs. Must be
2361 lower than slub_max_order.
2362 For more information see Documentation/vm/slub.txt.
2364 slub_nomerge [MM, SLUB]
2365 Disable merging of slabs with similar size. May be
2366 necessary if there is some reason to distinguish
2367 allocs to different slabs. Debug options disable
2368 merging on their own.
2369 For more information see Documentation/vm/slub.txt.
2372 Format: <io1>[,<io2>[,...,<io8>]]
2374 smp-alt-once [X86-32,SMP] On a hotplug CPU system, only
2375 attempt to substitute SMP alternatives once at boot.
2377 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
2378 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
2379 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
2380 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
2381 smsc-ircc2.ircc_irq= [HW] IRQ line
2382 smsc-ircc2.ircc_dma= [HW] DMA channel
2383 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
2384 0: Toshiba Satellite 1800 (GP data pin select)
2385 1: Fast pin select (default)
2388 snd-ad1816a= [HW,ALSA]
2390 snd-ad1848= [HW,ALSA]
2392 snd-ali5451= [HW,ALSA]
2394 snd-als100= [HW,ALSA]
2396 snd-als4000= [HW,ALSA]
2398 snd-azt2320= [HW,ALSA]
2400 snd-cmi8330= [HW,ALSA]
2402 snd-cmipci= [HW,ALSA]
2404 snd-cs4231= [HW,ALSA]
2406 snd-cs4232= [HW,ALSA]
2408 snd-cs4236= [HW,ALSA]
2410 snd-cs4281= [HW,ALSA]
2412 snd-cs46xx= [HW,ALSA]
2414 snd-dt019x= [HW,ALSA]
2416 snd-dummy= [HW,ALSA]
2418 snd-emu10k1= [HW,ALSA]
2420 snd-ens1370= [HW,ALSA]
2422 snd-ens1371= [HW,ALSA]
2424 snd-es968= [HW,ALSA]
2426 snd-es1688= [HW,ALSA]
2428 snd-es18xx= [HW,ALSA]
2430 snd-es1938= [HW,ALSA]
2432 snd-es1968= [HW,ALSA]
2434 snd-fm801= [HW,ALSA]
2436 snd-gusclassic= [HW,ALSA]
2438 snd-gusextreme= [HW,ALSA]
2440 snd-gusmax= [HW,ALSA]
2444 snd-ice1712= [HW,ALSA]
2446 snd-intel8x0= [HW,ALSA]
2448 snd-interwave= [HW,ALSA]
2453 snd-korg1212= [HW,ALSA]
2455 snd-maestro3= [HW,ALSA]
2457 snd-mpu401= [HW,ALSA]
2459 snd-mtpav= [HW,ALSA]
2461 snd-nm256= [HW,ALSA]
2463 snd-opl3sa2= [HW,ALSA]
2471 snd-opti93x= [HW,ALSA]
2475 snd-rme32= [HW,ALSA]
2477 snd-rme96= [HW,ALSA]
2479 snd-rme9652= [HW,ALSA]
2485 snd-sbawe= [HW,ALSA]
2487 snd-serial= [HW,ALSA]
2489 snd-sgalaxy= [HW,ALSA]
2491 snd-sonicvibes= [HW,ALSA]
2496 snd-sun-cs4231= [HW,ALSA]
2498 snd-trident= [HW,ALSA]
2500 snd-usb-audio= [HW,ALSA,USB]
2502 snd-via82xx= [HW,ALSA]
2504 snd-virmidi= [HW,ALSA]
2506 snd-wavefront= [HW,ALSA]
2508 snd-ymfpci= [HW,ALSA]
2511 [KNL] Should the soft-lockup detector generate panics.
2513 sonypi.*= [HW] Sony Programmable I/O Control Device driver
2514 See Documentation/sonypi.txt
2516 specialix= [HW,SERIAL] Specialix multi-serial port adapter
2517 See Documentation/serial/specialix.txt.
2519 spia_io_base= [HW,MTD]
2525 Format: <io>,<irq>,<dma>,<mpu_io>,<mpu_irq>
2527 st= [HW,SCSI] SCSI tape parameters (buffers, etc.)
2528 See Documentation/scsi/st.txt.
2531 Enabled the stack tracer on boot up.
2535 Set the STI (builtin display/keyboard on the HP-PARISC
2536 machines) console (graphic card) which should be used
2537 as the initial boot-console.
2538 See also comment in drivers/video/console/sticore.c.
2541 See comment in drivers/video/console/sticore.c.
2544 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
2546 sunrpc.min_resvport=
2547 sunrpc.max_resvport=
2549 SunRPC servers often require that client requests
2550 originate from a privileged port (i.e. a port in the
2551 range 0 < portnr < 1024).
2552 An administrator who wishes to reserve some of these
2553 ports for other uses may adjust the range that the
2554 kernel's sunrpc client considers to be privileged
2555 using these two parameters to set the minimum and
2556 maximum port values.
2560 Control how the NFS server code allocates CPUs to
2561 service thread pools. Depending on how many NICs
2562 you have and where their interrupts are bound, this
2563 option will affect which CPUs will do NFS serving.
2564 Note: this parameter cannot be changed while the
2565 NFS server is running.
2567 auto the server chooses an appropriate mode
2568 automatically using heuristics
2569 global a single global pool contains all CPUs
2570 percpu one pool for each CPU
2571 pernode one pool for each NUMA node (equivalent
2572 to global on non-NUMA machines)
2574 sunrpc.tcp_slot_table_entries=
2575 sunrpc.udp_slot_table_entries=
2577 Sets the upper limit on the number of simultaneous
2578 RPC calls that can be sent from the client to a
2579 server. Increasing these values may allow you to
2580 improve throughput, but will also increase the
2581 amount of memory reserved for use by the client.
2583 swiotlb= [IA-64] Number of I/O TLB slabs
2587 sym53c416= [HW,SCSI]
2588 See header of drivers/scsi/sym53c416.c.
2590 sysrq_always_enabled
2592 Ignore sysrq setting - this boot parameter will
2593 neutralize any effect of /proc/sys/kernel/sysrq.
2594 Useful for debugging.
2597 See header of drivers/scsi/t128.c.
2601 test_suspend= [SUSPEND]
2602 Specify "mem" (for Suspend-to-RAM) or "standby" (for
2603 standby suspend) as the system sleep state to briefly
2604 enter during system startup. The system is woken from
2605 this state using a wakeup-capable RTC alarm.
2607 thash_entries= [KNL,NET]
2608 Set number of hash buckets for TCP connection
2610 thermal.act= [HW,ACPI]
2611 -1: disable all active trip points in all thermal zones
2612 <degrees C>: override all lowest active trip points
2614 thermal.crt= [HW,ACPI]
2615 -1: disable all critical trip points in all thermal zones
2616 <degrees C>: override all critical trip points
2618 thermal.nocrt= [HW,ACPI]
2619 Set to disable actions on ACPI thermal zone
2620 critical and hot trip points.
2622 thermal.off= [HW,ACPI]
2623 1: disable ACPI thermal control
2625 thermal.psv= [HW,ACPI]
2626 -1: disable all passive trip points
2627 <degrees C>: override all passive trip points to this
2630 thermal.tzp= [HW,ACPI]
2631 Specify global default ACPI thermal zone polling rate
2632 <deci-seconds>: poll all this frequency
2633 0: no polling (default)
2636 See comment before function dc390_setup() in
2637 drivers/scsi/tmscsim.c.
2641 Specify if the kernel should make use of the cpu
2642 topology informations if the hardware supports these.
2643 The scheduler will make use of these informations and
2644 e.g. base its process migration decisions on it.
2649 tpm_suspend_pcr=[HW,TPM]
2650 Format: integer pcr id
2651 Specify that at suspend time, the tpm driver
2652 should extend the specified pcr with zeros,
2653 as a workaround for some chips which fail to
2654 flush the last written pcr on TPM_SaveState.
2655 This will guarantee that all the other pcrs
2658 trace_buf_size=nn[KMG]
2659 [FTRACE] will set tracing buffer size.
2661 trace_event=[event-list]
2662 [FTRACE] Set and start specified trace events in order
2663 to facilitate early boot debugging.
2664 See also Documentation/trace/events.txt
2666 trix= [HW,OSS] MediaTrix AudioTrix Pro
2668 <io>,<irq>,<dma>,<dma2>,<sb_io>,<sb_irq>,<sb_dma>,<mpu_io>,<mpu_irq>
2670 tsc= Disable clocksource-must-verify flag for TSC.
2672 [x86] reliable: mark tsc clocksource as reliable, this
2673 disables clocksource verification at runtime.
2674 Used to enable high-resolution timer mode on older
2675 hardware, and in virtualized environment.
2677 turbografx.map[2|3]= [HW,JOY]
2678 TurboGraFX parallel port interface
2680 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
2681 See also Documentation/input/joystick-parport.txt
2683 u14-34f= [HW,SCSI] UltraStor 14F/34F SCSI host adapter
2684 See header of drivers/scsi/u14-34f.c.
2692 uhash_entries= [KNL,NET]
2693 Set number of hash buckets for UDP/UDP-Lite connections
2696 [USB] Ignore overcurrent events (default N).
2697 Some badly-designed motherboards generate lots of
2698 bogus events, for ports that aren't wired to
2699 anything. Set this parameter to avoid log spamming.
2700 Note that genuine overcurrent events won't be
2705 Set unknown_nmi_panic=1 early on boot.
2707 usbcore.autosuspend=
2708 [USB] The autosuspend time delay (in seconds) used
2709 for newly-detected USB devices (default 2). This
2710 is the time required before an idle device will be
2711 autosuspended. Devices for which the delay is set
2712 to a negative value won't be autosuspended at all.
2714 usbcore.usbfs_snoop=
2715 [USB] Set to log all usbfs traffic (default 0 = off).
2717 usbcore.blinkenlights=
2718 [USB] Set to cycle leds on hubs (default 0 = off).
2720 usbcore.old_scheme_first=
2721 [USB] Start with the old device initialization
2722 scheme (default 0 = off).
2724 usbcore.use_both_schemes=
2725 [USB] Try the other device initialization scheme
2726 if the first one fails (default 1 = enabled).
2728 usbcore.initial_descriptor_timeout=
2729 [USB] Specifies timeout for the initial 64-byte
2730 USB_REQ_GET_DESCRIPTOR request in milliseconds
2731 (default 5000 = 5.0 seconds).
2734 [USBHID] The interval which mice are to be polled at.
2736 usb-storage.delay_use=
2737 [UMS] The delay in seconds before a new device is
2738 scanned for Logical Units (default 5).
2741 [UMS] A list of quirks entries to supplement or
2742 override the built-in unusual_devs list. List
2743 entries are separated by commas. Each entry has
2744 the form VID:PID:Flags where VID and PID are Vendor
2745 and Product ID values (4-digit hex numbers) and
2746 Flags is a set of characters, each corresponding
2747 to a common usb-storage quirk flag as follows:
2748 a = SANE_SENSE (collect more than 18 bytes
2750 b = BAD_SENSE (don't collect more than 18
2751 bytes of sense data);
2752 c = FIX_CAPACITY (decrease the reported
2753 device capacity by one sector);
2754 h = CAPACITY_HEURISTICS (decrease the
2755 reported device capacity by one
2756 sector if the number is odd);
2757 i = IGNORE_DEVICE (don't bind to this
2759 l = NOT_LOCKABLE (don't try to lock and
2760 unlock ejectable media);
2761 m = MAX_SECTORS_64 (don't transfer more
2762 than 64 sectors = 32 KB at a time);
2763 o = CAPACITY_OK (accept the capacity
2764 reported by the device);
2765 r = IGNORE_RESIDUE (the device reports
2766 bogus residue values);
2767 s = SINGLE_LUN (the device has only one
2769 w = NO_WP_DETECT (don't test whether the
2770 medium is write-protected).
2771 Example: quirks=0419:aaf5:rl,0421:0433:rc
2774 [X86] Flags controlling user PTE allocations.
2776 nohigh = do not allocate PTE pages in
2777 HIGHMEM regardless of setting
2781 vdso=2: enable compat VDSO (default with COMPAT_VDSO)
2782 vdso=1: enable VDSO (default)
2783 vdso=0: disable VDSO mapping
2786 vdso32=2: enable compat VDSO (default with COMPAT_VDSO)
2787 vdso32=1: enable 32-bit VDSO (default)
2788 vdso32=0: disable 32-bit VDSO mapping
2791 vector=percpu: enable percpu vector domain
2793 video= [FB] Frame buffer configuration
2794 See Documentation/fb/modedb.txt.
2796 vga= [BOOT,X86-32] Select a particular video mode
2797 See Documentation/x86/boot.txt and
2798 Documentation/svga.txt.
2799 Use vga=ask for menu.
2800 This is actually a boot loader parameter; the value is
2801 passed to the kernel using a special protocol.
2803 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
2804 size of <nn>. This can be used to increase the
2805 minimum size (128MB on x86). It can also be used to
2806 decrease the size and leave more room for directly
2809 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
2812 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
2815 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
2818 vt.cur_default= [VT] Default cursor shape.
2819 Format: 0xCCBBAA, where AA, BB, and CC are the same as
2820 the parameters of the <Esc>[?A;B;Cc escape sequence;
2821 see VGA-softcursor.txt. Default: 2 = underline.
2823 vt.default_blu= [VT]
2824 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
2825 Change the default blue palette of the console.
2826 This is a 16-member array composed of values
2829 vt.default_grn= [VT]
2830 Format: <green0>,<green1>,<green2>,...,<green15>
2831 Change the default green palette of the console.
2832 This is a 16-member array composed of values
2835 vt.default_red= [VT]
2836 Format: <red0>,<red1>,<red2>,...,<red15>
2837 Change the default red palette of the console.
2838 This is a 16-member array composed of values
2844 Set system-wide default UTF-8 mode for all tty's.
2845 Default is 1, i.e. UTF-8 mode is enabled for all
2846 newly opened terminals.
2848 vt.global_cursor_default=
2851 Set system-wide default for whether a cursor
2852 is shown on new VTs. Default is -1,
2853 i.e. cursors will be created by default unless
2854 overridden by individual drivers. 0 will hide
2855 cursors, 1 will display them.
2857 waveartist= [HW,OSS]
2858 Format: <io>,<irq>,<dma>,<dma2>
2861 See header of drivers/scsi/wd33c93.c.
2864 See header of drivers/scsi/wd7000.c.
2866 watchdog timers [HW,WDT] For information on watchdog timers,
2867 see Documentation/watchdog/watchdog-parameters.txt
2868 or other driver-specific files in the
2869 Documentation/watchdog/ directory.
2871 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
2872 default x2apic cluster mode on platforms
2875 x86_mrst_timer= [X86-32,APBT]
2876 Choose timer option for x86 Moorestown MID platform.
2877 Two valid options are apbt timer only and lapic timer
2878 plus one apbt timer for broadcast timer.
2879 x86_mrst_timer=apbt_only | lapic_and_apbt
2881 xd= [HW,XT] Original XT pre-IDE (RLL encoded) disks.
2882 xd_geo= See header of drivers/block/xd.c.
2884 xirc2ps_cs= [NET,PCMCIA]
2886 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
2888 ______________________________________________________________________
2892 Add documentation for ALSA options.
2893 Add more DRM drivers.