From Fedora Project Wiki

(Created page with ' Kernel Parameters ~~~~~~~~~~~~~~~~~ The following is a consolidated list of the kernel parameters as implemented (mostly) by ...')
 
(Blanked the page)
 
Line 1: Line 1:
                          Kernel Parameters
 
                          19:23, 20 June 2009 (UTC)19:23, 20 June 2009 (UTC)19:23, 20 June 2009 (UTC)~~
 
  
The following is a consolidated list of the kernel parameters as implemented
 
(mostly) by the __setup() macro and sorted into English Dictionary order
 
(defined as ignoring all punctuation and sorting digits before letters in a
 
case insensitive manner), and with descriptions where known.
 
 
Module parameters for loadable modules are specified only as the
 
parameter name with optional '=' and value as appropriate, such as:
 
 
modprobe usbcore blinkenlights=1
 
 
Module parameters for modules that are built into the kernel image
 
are specified on the kernel command line with the module name plus
 
'.' plus parameter name, with '=' and value if appropriate, such as:
 
 
usbcore.blinkenlights=1
 
 
This document may not be entirely up to date and comprehensive. The command
 
"modinfo -p ${modulename}" shows a current list of all parameters of a loadable
 
module. Loadable modules, after being loaded into the running kernel, also
 
reveal their parameters in /sys/module/${modulename}/parameters/. Some of these
 
parameters may be changed at runtime by the command
 
"echo -n ${value} > /sys/module/${modulename}/parameters/${parm}".
 
 
The parameters listed below are only valid if certain kernel build options were
 
enabled and if respective hardware is present. The text in square brackets at
 
the beginning of each description states the restrictions within which a
 
parameter is applicable:
 
 
ACPI ACPI support is enabled.
 
AGP AGP (Accelerated Graphics Port) is enabled.
 
ALSA ALSA sound support is enabled.
 
APIC APIC support is enabled.
 
APM Advanced Power Management support is enabled.
 
AVR32 AVR32 architecture is enabled.
 
AX25 Appropriate AX.25 support is enabled.
 
BLACKFIN Blackfin architecture is enabled.
 
DRM Direct Rendering Management support is enabled.
 
EDD BIOS Enhanced Disk Drive Services (EDD) is enabled
 
EFI EFI Partitioning (GPT) is enabled
 
EIDE EIDE/ATAPI support is enabled.
 
FB The frame buffer device is enabled.
 
HW Appropriate hardware is enabled.
 
IA-64 IA-64 architecture is enabled.
 
IOSCHED More than one I/O scheduler is enabled.
 
IP_PNP IP DHCP, BOOTP, or RARP is enabled.
 
ISAPNP ISA PnP code is enabled.
 
ISDN Appropriate ISDN support is enabled.
 
JOY Appropriate joystick support is enabled.
 
LIBATA  Libata driver is enabled
 
LP Printer support is enabled.
 
LOOP Loopback device support is enabled.
 
M68k M68k architecture is enabled.
 
These options have more detailed description inside of
 
Documentation/m68k/kernel-options.txt.
 
MCA MCA bus support is enabled.
 
MDA MDA console support is enabled.
 
MOUSE Appropriate mouse support is enabled.
 
MSI Message Signaled Interrupts (PCI).
 
MTD MTD (Memory Technology Device) support is enabled.
 
NET Appropriate network support is enabled.
 
NUMA NUMA support is enabled.
 
GENERIC_TIME The generic timeofday code is enabled.
 
NFS Appropriate NFS support is enabled.
 
OSS OSS sound support is enabled.
 
PV_OPS A paravirtualized kernel is enabled.
 
PARIDE The ParIDE (parallel port IDE) subsystem is enabled.
 
PARISC The PA-RISC architecture is enabled.
 
PCI PCI bus support is enabled.
 
PCIE PCI Express support is enabled.
 
PCMCIA The PCMCIA subsystem is enabled.
 
PNP Plug & Play support is enabled.
 
PPC PowerPC architecture is enabled.
 
PPT Parallel port support is enabled.
 
PS2 Appropriate PS/2 support is enabled.
 
RAM RAM disk support is enabled.
 
ROOTPLUG The example Root Plug LSM is enabled.
 
S390 S390 architecture is enabled.
 
SCSI Appropriate SCSI support is enabled.
 
A lot of drivers has their options described inside of
 
Documentation/scsi/.
 
SECURITY Different security models are enabled.
 
SELINUX SELinux support is enabled.
 
SERIAL Serial support is enabled.
 
SH SuperH architecture is enabled.
 
SMP The kernel is an SMP kernel.
 
SPARC Sparc architecture is enabled.
 
SWSUSP Software suspend (hibernation) is enabled.
 
SUSPEND System suspend states are enabled.
 
FTRACE Function tracing enabled.
 
TS Appropriate touchscreen support is enabled.
 
UMS USB Mass Storage support is enabled.
 
USB USB support is enabled.
 
USBHID USB Human Interface Device support is enabled.
 
V4L Video For Linux support is enabled.
 
VGA The VGA console has been enabled.
 
VT Virtual terminal support is enabled.
 
WDT Watchdog support is enabled.
 
XT IBM PC/XT MFM hard disk support is enabled.
 
X86-32 X86-32, aka i386 architecture is enabled.
 
X86-64 X86-64 architecture is enabled.
 
More X86-64 boot options can be found in
 
Documentation/x86/x86_64/boot-options.txt .
 
X86 Either 32bit or 64bit x86 (same as X86-32+X86-64)
 
 
In addition, the following text indicates that the option:
 
 
BUGS= Relates to possible processor bugs on the said processor.
 
KNL Is a kernel start-up parameter.
 
BOOT Is a boot loader parameter.
 
 
Parameters denoted with BOOT are actually interpreted by the boot
 
loader, and have no meaning to the kernel directly.
 
Do not modify the syntax of boot loader parameters without extreme
 
need or coordination with <Documentation/x86/boot.txt>.
 
 
There are also arch-specific kernel-parameters not documented here.
 
See for example <Documentation/x86/x86_64/boot-options.txt>.
 
 
Note that ALL kernel parameters listed below are CASE SENSITIVE, and that
 
a trailing = on the name of any parameter states that that parameter will
 
be entered as an environment variable, whereas its absence indicates that
 
it will appear as a kernel argument readable via /proc/cmdline by programs
 
running once the system is up.
 
 
The number of kernel parameters is not limited, but the length of the
 
complete command line (parameters including spaces etc.) is limited to
 
a fixed number of characters. This limit depends on the architecture
 
and is between 256 and 4096 characters. It is defined in the file
 
./include/asm/setup.h as COMMAND_LINE_SIZE.
 
 
 
acpi= [HW,ACPI,X86-64,i386]
 
Advanced Configuration and Power Interface
 
Format: { force | off | ht | strict | noirq | rsdt }
 
force -- enable ACPI if default was off
 
off -- disable ACPI if default was on
 
noirq -- do not use ACPI for IRQ routing
 
ht -- run only enough ACPI to enable Hyper Threading
 
strict -- Be less tolerant of platforms that are not
 
strictly ACPI specification compliant.
 
rsdt -- prefer RSDT over (default) XSDT
 
 
See also Documentation/power/pm.txt, pci=noacpi
 
 
acpi_apic_instance= [ACPI, IOAPIC]
 
Format: <int>
 
2: use 2nd APIC table, if available
 
1,0: use 1st APIC table
 
default: 0
 
 
acpi_sleep= [HW,ACPI] Sleep options
 
Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
 
  old_ordering, s4_nonvs }
 
See Documentation/power/video.txt for information on
 
s3_bios and s3_mode.
 
s3_beep is for debugging; it makes the PC's speaker beep
 
as soon as the kernel's real-mode entry point is called.
 
s4_nohwsig prevents ACPI hardware signature from being
 
used during resume from hibernation.
 
old_ordering causes the ACPI 1.0 ordering of the _PTS
 
control method, with respect to putting devices into
 
low power states, to be enforced (the ACPI 2.0 ordering
 
of _PTS is used by default).
 
s4_nonvs prevents the kernel from saving/restoring the
 
ACPI NVS memory during hibernation.
 
 
acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
 
Format: { level | edge | high | low }
 
 
acpi_irq_balance [HW,ACPI]
 
ACPI will balance active IRQs
 
default in APIC mode
 
 
acpi_irq_nobalance [HW,ACPI]
 
ACPI will not move active IRQs (default)
 
default in PIC mode
 
 
acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for
 
use by PCI
 
Format: <irq>,<irq>...
 
 
acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA
 
Format: <irq>,<irq>...
 
 
acpi_no_auto_ssdt [HW,ACPI] Disable automatic loading of SSDT
 
 
acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
 
Format: To spoof as Windows 98: ="Microsoft Windows"
 
 
acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
 
acpi_osi="string1" # add string1 -- only one string
 
acpi_osi="!string2" # remove built-in string2
 
acpi_osi= # disable all strings
 
 
acpi_serialize [HW,ACPI] force serialization of AML methods
 
 
acpi_skip_timer_override [HW,ACPI]
 
Recognize and ignore IRQ0/pin2 Interrupt Override.
 
For broken nForce2 BIOS resulting in XT-PIC timer.
 
acpi_use_timer_override [HW,ACPI]
 
Use timer override. For some broken Nvidia NF5 boards
 
that require a timer override, but don't have
 
HPET
 
 
acpi_backlight= [HW,ACPI]
 
acpi_backlight=vendor
 
acpi_backlight=video
 
If set to vendor, prefer vendor specific driver
 
(e.g. thinkpad_acpi, sony_acpi, etc.) instead
 
of the ACPI video.ko driver.
 
 
acpi_display_output= [HW,ACPI]
 
acpi_display_output=vendor
 
acpi_display_output=video
 
See above.
 
 
acpi.debug_layer= [HW,ACPI,ACPI_DEBUG]
 
acpi.debug_level= [HW,ACPI,ACPI_DEBUG]
 
Format: <int>
 
CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
 
debug output.  Bits in debug_layer correspond to a
 
_COMPONENT in an ACPI source file, e.g.,
 
    #define _COMPONENT ACPI_PCI_COMPONENT
 
Bits in debug_level correspond to a level in
 
ACPI_DEBUG_PRINT statements, e.g.,
 
    ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
 
The debug_level mask defaults to "info".  See
 
Documentation/acpi/debug.txt for more information about
 
debug layers and levels.
 
 
Enable processor driver info messages:
 
    acpi.debug_layer=0x20000000
 
Enable PCI/PCI interrupt routing info messages:
 
    acpi.debug_layer=0x400000
 
Enable AML "Debug" output, i.e., stores to the Debug
 
object while interpreting AML:
 
    acpi.debug_layer=0xffffffff acpi.debug_level=0x2
 
Enable all messages related to ACPI hardware:
 
    acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
 
 
Some values produce so much output that the system is
 
unusable.  The "log_buf_len" parameter may be useful
 
if you need to capture more output.
 
 
acpi.power_nocheck= [HW,ACPI]
 
Format: 1/0 enable/disable the check of power state.
 
On some bogus BIOS the _PSC object/_STA object of
 
power resource can't return the correct device power
 
state. In such case it is unneccessary to check its
 
power state again in power transition.
 
1 : disable the power state check
 
 
acpi_pm_good [X86-32,X86-64]
 
Override the pmtimer bug detection: force the kernel
 
to assume that this machine's pmtimer latches its value
 
and always returns good values.
 
 
agp= [AGP]
 
{ off | try_unsupported }
 
off: disable AGP support
 
try_unsupported: try to drive unsupported chipsets
 
(may crash computer or cause data corruption)
 
 
enable_timer_pin_1 [i386,x86-64]
 
Enable PIN 1 of APIC timer
 
Can be useful to work around chipset bugs
 
(in particular on some ATI chipsets).
 
The kernel tries to set a reasonable default.
 
 
disable_timer_pin_1 [i386,x86-64]
 
Disable PIN 1 of APIC timer
 
Can be useful to work around chipset bugs.
 
 
ad1848= [HW,OSS]
 
Format: <io>,<irq>,<dma>,<dma2>,<type>
 
 
advansys= [HW,SCSI]
 
See header of drivers/scsi/advansys.c.
 
 
advwdt= [HW,WDT] Advantech WDT
 
Format: <iostart>,<iostop>
 
 
aedsp16= [HW,OSS] Audio Excel DSP 16
 
Format: <io>,<irq>,<dma>,<mss_io>,<mpu_io>,<mpu_irq>
 
See also header of sound/oss/aedsp16.c.
 
 
aha152x= [HW,SCSI]
 
See Documentation/scsi/aha152x.txt.
 
 
aha1542= [HW,SCSI]
 
Format: <portbase>[,<buson>,<busoff>[,<dmaspeed>]]
 
 
aic7xxx= [HW,SCSI]
 
See Documentation/scsi/aic7xxx.txt.
 
 
aic79xx= [HW,SCSI]
 
See Documentation/scsi/aic79xx.txt.
 
 
amd_iommu= [HW,X86-84]
 
Pass parameters to the AMD IOMMU driver in the system.
 
Possible values are:
 
isolate - enable device isolation (each device, as far
 
          as possible, will get its own protection
 
          domain) [default]
 
share - put every device behind one IOMMU into the
 
same protection domain
 
fullflush - enable flushing of IO/TLB entries when
 
    they are unmapped. Otherwise they are
 
    flushed before they will be reused, which
 
    is a lot of faster
 
 
amd_iommu_size= [HW,X86-64]
 
Define the size of the aperture for the AMD IOMMU
 
driver. Possible values are:
 
'32M', '64M' (default), '128M', '256M', '512M', '1G'
 
 
amijoy.map= [HW,JOY] Amiga joystick support
 
Map of devices attached to JOY0DAT and JOY1DAT
 
Format: <a>,<b>
 
See also Documentation/kernel/input/joystick.txt
 
 
analog.map= [HW,JOY] Analog joystick and gamepad support
 
Specifies type or capabilities of an analog joystick
 
connected to one of 16 gameports
 
Format: <type1>,<type2>,..<type16>
 
 
apc= [HW,SPARC]
 
Power management functions (SPARCstation-4/5 + deriv.)
 
Format: noidle
 
Disable APC CPU standby support. SPARCstation-Fox does
 
not play well with APC CPU idle - disable it if you have
 
APC and your system crashes randomly.
 
 
apic= [APIC,i386] Advanced Programmable Interrupt Controller
 
Change the output verbosity whilst booting
 
Format: { quiet (default) | verbose | debug }
 
Change the amount of debugging information output
 
when initialising the APIC and IO-APIC components.
 
 
apm= [APM] Advanced Power Management
 
See header of arch/x86/kernel/apm_32.c.
 
 
arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
 
Format: <io>,<irq>,<nodeID>
 
 
ataflop= [HW,M68k]
 
 
atarimouse= [HW,MOUSE] Atari Mouse
 
 
atascsi= [HW,SCSI] Atari SCSI
 
 
atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
 
EzKey and similar keyboards
 
 
atkbd.reset= [HW] Reset keyboard during initialization
 
 
atkbd.set= [HW] Select keyboard code set
 
Format: <int> (2 = AT (default), 3 = PS/2)
 
 
atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
 
keyboards
 
 
atkbd.softraw= [HW] Choose between synthetic and real raw mode
 
Format: <bool> (0 = real, 1 = synthetic (default))
 
 
atkbd.softrepeat= [HW]
 
Use software keyboard repeat
 
 
autotest [IA64]
 
 
baycom_epp= [HW,AX25]
 
Format: <io>,<mode>
 
 
baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
 
Format: <io>,<mode>
 
See header of drivers/net/hamradio/baycom_par.c.
 
 
baycom_ser_fdx= [HW,AX25]
 
BayCom Serial Port AX.25 Modem (Full Duplex Mode)
 
Format: <io>,<irq>,<mode>[,<baud>]
 
See header of drivers/net/hamradio/baycom_ser_fdx.c.
 
 
baycom_ser_hdx= [HW,AX25]
 
BayCom Serial Port AX.25 Modem (Half Duplex Mode)
 
Format: <io>,<irq>,<mode>
 
See header of drivers/net/hamradio/baycom_ser_hdx.c.
 
 
boot_delay= Milliseconds to delay each printk during boot.
 
Values larger than 10 seconds (10000) are changed to
 
no delay (0).
 
Format: integer
 
 
bootmem_debug [KNL] Enable bootmem allocator debug messages.
 
 
bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
 
bttv.radio= Most important insmod options are available as
 
kernel args too.
 
bttv.pll= See Documentation/video4linux/bttv/Insmod-options
 
bttv.tuner= and Documentation/video4linux/bttv/CARDLIST
 
 
BusLogic= [HW,SCSI]
 
See drivers/scsi/BusLogic.c, comment before function
 
BusLogic_ParseDriverOptions().
 
 
c101= [NET] Moxa C101 synchronous serial card
 
 
cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
 
Sometimes CPU hardware bugs make them report the cache
 
size incorrectly. The kernel will attempt work arounds
 
to fix known problems, but for some CPUs it is not
 
possible to determine what the correct size should be.
 
This option provides an override for these situations.
 
 
security= [SECURITY] Choose a security module to enable at boot.
 
If this boot parameter is not specified, only the first
 
security module asking for security registration will be
 
loaded. An invalid security module name will be treated
 
as if no module has been chosen.
 
 
capability.disable=
 
[SECURITY] Disable capabilities.  This would normally
 
be used only if an alternative security model is to be
 
configured.  Potentially dangerous and should only be
 
used if you are entirely sure of the consequences.
 
 
ccw_timeout_log [S390]
 
See Documentation/s390/CommonIO for details.
 
 
cgroup_disable= [KNL] Disable a particular controller
 
Format: {name of the controller(s) to disable}
 
{Currently supported controllers - "memory"}
 
 
checkreqprot [SELINUX] Set initial checkreqprot flag value.
 
Format: { "0" | "1" }
 
See security/selinux/Kconfig help text.
 
0 -- check protection applied by kernel (includes
 
any implied execute protection).
 
1 -- check protection requested by application.
 
Default value is set via a kernel config option.
 
Value can be changed at runtime via
 
/selinux/checkreqprot.
 
 
cio_ignore= [S390]
 
See Documentation/s390/CommonIO for details.
 
 
clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
 
[Deprecated]
 
Forces specified clocksource (if available) to be used
 
when calculating gettimeofday(). If specified
 
clocksource is not available, it defaults to PIT.
 
Format: { pit | tsc | cyclone | pmtmr }
 
 
clocksource= [GENERIC_TIME] Override the default clocksource
 
Format: <string>
 
Override the default clocksource and use the clocksource
 
with the name specified.
 
Some clocksource names to choose from, depending on
 
the platform:
 
[all] jiffies (this is the base, fallback clocksource)
 
[ACPI] acpi_pm
 
[ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
 
pxa_timer,timer3,32k_counter,timer0_1
 
[AVR32] avr32
 
[X86-32] pit,hpet,tsc,vmi-timer;
 
scx200_hrt on Geode; cyclone on IBM x440
 
[MIPS] MIPS
 
[PARISC] cr16
 
[S390] tod
 
[SH] SuperH
 
[SPARC64] tick
 
[X86-64] hpet,tsc
 
 
clearcpuid=BITNUM [X86]
 
Disable CPUID feature X for the kernel. See
 
arch/x86/include/asm/cpufeature.h for the valid bit
 
numbers. Note the Linux specific bits are not necessarily
 
stable over kernel options, but the vendor specific
 
ones should be.
 
Also note that user programs calling CPUID directly
 
or using the feature without checking anything
 
will still see it. This just prevents it from
 
being used by the kernel or shown in /proc/cpuinfo.
 
Also note the kernel might malfunction if you disable
 
some critical bits.
 
 
code_bytes [IA32/X86_64] How many bytes of object code to print
 
in an oops report.
 
Range: 0 - 8192
 
Default: 64
 
 
hpet= [X86-32,HPET] option to control HPET usage
 
Format: { enable (default) | disable | force }
 
disable: disable HPET and use PIT instead
 
force: allow force enabled of undocumented chips (ICH4,
 
VIA, nVidia)
 
 
com20020= [HW,NET] ARCnet - COM20020 chipset
 
Format:
 
<io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
 
 
com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
 
Format: <io>[,<irq>]
 
 
com90xx= [HW,NET]
 
ARCnet - COM90xx chipset (memory-mapped buffers)
 
Format: <io>[,<irq>[,<memstart>]]
 
 
condev= [HW,S390] console device
 
conmode=
 
 
console= [KNL] Output console device and options.
 
 
tty<n> Use the virtual console device <n>.
 
 
ttyS<n>[,options]
 
ttyUSB0[,options]
 
Use the specified serial port.  The options are of
 
the form "bbbbpnf", where "bbbb" is the baud rate,
 
"p" is parity ("n", "o", or "e"), "n" is number of
 
bits, and "f" is flow control ("r" for RTS or
 
omit it).  Default is "9600n8".
 
 
See Documentation/serial-console.txt for more
 
information.  See
 
Documentation/networking/netconsole.txt for an
 
alternative.
 
 
uart[8250],io,<addr>[,options]
 
uart[8250],mmio,<addr>[,options]
 
Start an early, polled-mode console on the 8250/16550
 
UART at the specified I/O port or MMIO address,
 
switching to the matching ttyS device later.  The
 
options are the same as for ttyS, above.
 
 
                If the device connected to the port is not a TTY but a braille
 
                device, prepend "brl," before the device type, for instance
 
console=brl,ttyS0
 
For now, only VisioBraille is supported.
 
 
earlycon= [KNL] Output early console device and options.
 
uart[8250],io,<addr>[,options]
 
uart[8250],mmio,<addr>[,options]
 
Start an early, polled-mode console on the 8250/16550
 
UART at the specified I/O port or MMIO address.
 
The options are the same as for ttyS, above.
 
 
no_console_suspend
 
[HW] Never suspend the console
 
Disable suspending of consoles during suspend and
 
hibernate operations.  Once disabled, debugging
 
messages can reach various consoles while the rest
 
of the system is being put to sleep (ie, while
 
debugging driver suspend/resume hooks).  This may
 
not work reliably with all consoles, but is known
 
to work with serial and VGA consoles.
 
 
coredump_filter=
 
[KNL] Change the default value for
 
/proc/<pid>/coredump_filter.
 
See also Documentation/filesystems/proc.txt.
 
 
cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
 
Format:
 
<first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
 
 
crashkernel=nn[KMG]@ss[KMG]
 
[KNL] Reserve a chunk of physical memory to
 
hold a kernel to switch to with kexec on panic.
 
 
crashkernel=range1:size1[,range2:size2,...][@offset]
 
[KNL] Same as above, but depends on the memory
 
in the running system. The syntax of range is
 
start-[end] where start and end are both
 
a memory unit (amount[KMG]). See also
 
Documentation/kdump/kdump.txt for a example.
 
 
cs89x0_dma= [HW,NET]
 
Format: <dma>
 
 
cs89x0_media= [HW,NET]
 
Format: { rj45 | aui | bnc }
 
 
dasd= [HW,NET]
 
See header of drivers/s390/block/dasd_devmap.c.
 
 
db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
 
(one device per port)
 
Format: <port#>,<type>
 
See also Documentation/input/joystick-parport.txt
 
 
debug [KNL] Enable kernel debugging (events log level).
 
 
debug_locks_verbose=
 
[KNL] verbose self-tests
 
Format=<0|1>
 
Print debugging info while doing the locking API
 
self-tests.
 
We default to 0 (no extra messages), setting it to
 
1 will print _a lot_ more information - normally
 
only useful to kernel developers.
 
 
debug_objects [KNL] Enable object debugging
 
 
debugpat [X86] Enable PAT debugging
 
 
decnet.addr= [HW,NET]
 
Format: <area>[,<node>]
 
See also Documentation/networking/decnet.txt.
 
 
vt.default_blu= [VT]
 
Format: <blue0>,<blue1>,<blue2>,...,<blue15>
 
Change the default blue palette of the console.
 
This is a 16-member array composed of values
 
ranging from 0-255.
 
 
vt.default_grn= [VT]
 
Format: <green0>,<green1>,<green2>,...,<green15>
 
Change the default green palette of the console.
 
This is a 16-member array composed of values
 
ranging from 0-255.
 
 
vt.default_red= [VT]
 
Format: <red0>,<red1>,<red2>,...,<red15>
 
Change the default red palette of the console.
 
This is a 16-member array composed of values
 
ranging from 0-255.
 
 
vt.default_utf8=
 
[VT]
 
Format=<0|1>
 
Set system-wide default UTF-8 mode for all tty's.
 
Default is 1, i.e. UTF-8 mode is enabled for all
 
newly opened terminals.
 
 
dhash_entries= [KNL]
 
Set number of hash buckets for dentry cache.
 
 
digi= [HW,SERIAL]
 
IO parameters + enable/disable command.
 
 
digiepca= [HW,SERIAL]
 
See drivers/char/README.epca and
 
Documentation/serial/digiepca.txt.
 
 
disable_mtrr_cleanup [X86]
 
enable_mtrr_cleanup [X86]
 
The kernel tries to adjust MTRR layout from continuous
 
to discrete, to make X server driver able to add WB
 
entry later. This parameter enables/disables that.
 
 
mtrr_chunk_size=nn[KMG] [X86]
 
used for mtrr cleanup. It is largest continous chunk
 
that could hold holes aka. UC entries.
 
 
mtrr_gran_size=nn[KMG] [X86]
 
Used for mtrr cleanup. It is granularity of mtrr block.
 
Default is 1.
 
Large value could prevent small alignment from
 
using up MTRRs.
 
 
mtrr_spare_reg_nr=n [X86]
 
Format: <integer>
 
Range: 0,7 : spare reg number
 
Default : 1
 
Used for mtrr cleanup. It is spare mtrr entries number.
 
Set to 2 or more if your graphical card needs more.
 
 
disable_mtrr_trim [X86, Intel and AMD only]
 
By default the kernel will trim any uncacheable
 
memory out of your available memory pool based on
 
MTRR settings.  This parameter disables that behavior,
 
possibly causing your machine to run very slowly.
 
 
dmasound= [HW,OSS] Sound subsystem buffers
 
 
dscc4.setup= [NET]
 
 
dtc3181e= [HW,SCSI]
 
 
earlyprintk= [X86-32,X86-64,SH,BLACKFIN]
 
earlyprintk=vga
 
earlyprintk=serial[,ttySn[,baudrate]]
 
earlyprintk=dbgp
 
 
Append ",keep" to not disable it when the real console
 
takes over.
 
 
Only vga or serial or usb debug port at a time.
 
 
Currently only ttyS0 and ttyS1 are supported.
 
 
Interaction with the standard serial driver is not
 
very good.
 
 
The VGA output is eventually overwritten by the real
 
console.
 
 
eata= [HW,SCSI]
 
 
edd= [EDD]
 
Format: {"off" | "on" | "skip[mbr]"}
 
 
eisa_irq_edge= [PARISC,HW]
 
See header of drivers/parisc/eisa.c.
 
 
elanfreq= [X86-32]
 
See comment before function elanfreq_setup() in
 
arch/x86/kernel/cpu/cpufreq/elanfreq.c.
 
 
elevator= [IOSCHED]
 
Format: {"anticipatory" | "cfq" | "deadline" | "noop"}
 
See Documentation/block/as-iosched.txt and
 
Documentation/block/deadline-iosched.txt for details.
 
 
elfcorehdr= [IA64,PPC,SH,X86-32,X86_64]
 
Specifies physical address of start of kernel core
 
image elf header. Generally kexec loader will
 
pass this option to capture kernel.
 
See Documentation/kdump/kdump.txt for details.
 
 
enforcing [SELINUX] Set initial enforcing status.
 
Format: {"0" | "1"}
 
See security/selinux/Kconfig help text.
 
0 -- permissive (log only, no denials).
 
1 -- enforcing (deny and log).
 
Default value is 0.
 
Value can be changed at runtime via /selinux/enforce.
 
 
ether= [HW,NET] Ethernet cards parameters
 
This option is obsoleted by the "netdev=" option, which
 
has equivalent usage. See its documentation for details.
 
 
eurwdt= [HW,WDT] Eurotech CPU-1220/1410 onboard watchdog.
 
Format: <io>[,<irq>]
 
 
failslab=
 
fail_page_alloc=
 
fail_make_request=[KNL]
 
General fault injection mechanism.
 
Format: <interval>,<probability>,<space>,<times>
 
See also /Documentation/fault-injection/.
 
 
fd_mcs= [HW,SCSI]
 
See header of drivers/scsi/fd_mcs.c.
 
 
fdomain= [HW,SCSI]
 
See header of drivers/scsi/fdomain.c.
 
 
floppy= [HW]
 
See Documentation/blockdev/floppy.txt.
 
 
force_pal_cache_flush
 
[IA-64] Avoid check_sal_cache_flush which may hang on
 
buggy SAL_CACHE_FLUSH implementations. Using this
 
parameter will force ia64_sal_cache_flush to call
 
ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
 
 
ftrace=[tracer]
 
[ftrace] will set and start the specified tracer
 
as early as possible in order to facilitate early
 
boot debugging.
 
 
ftrace_dump_on_oops
 
[ftrace] will dump the trace buffers on oops.
 
 
gamecon.map[2|3]=
 
[HW,JOY] Multisystem joystick and NES/SNES/PSX pad
 
support via parallel port (up to 5 devices per port)
 
Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
 
See also Documentation/input/joystick-parport.txt
 
 
gamma= [HW,DRM]
 
 
gart_fix_e820=  [X86_64] disable the fix e820 for K8 GART
 
Format: off | on
 
default: on
 
 
gdth= [HW,SCSI]
 
See header of drivers/scsi/gdth.c.
 
 
gpt [EFI] Forces disk with valid GPT signature but
 
invalid Protective MBR to be treated as GPT.
 
 
gvp11= [HW,SCSI]
 
 
hashdist= [KNL,NUMA] Large hashes allocated during boot
 
are distributed across NUMA nodes.  Defaults on
 
for IA-64, off otherwise.
 
Format: 0 | 1 (for off | on)
 
 
hcl= [IA-64] SGI's Hardware Graph compatibility layer
 
 
hd= [EIDE] (E)IDE hard drive subsystem geometry
 
Format: <cyl>,<head>,<sect>
 
 
highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
 
size of <nn>. This works even on boxes that have no
 
highmem otherwise. This also works to reduce highmem
 
size on bigger boxes.
 
 
highres= [KNL] Enable/disable high resolution timer mode.
 
Valid parameters: "on", "off"
 
Default: "on"
 
 
hisax= [HW,ISDN]
 
See Documentation/isdn/README.HiSax.
 
 
hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
 
hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
 
On x86-64 and powerpc, this option can be specified
 
multiple times interleaved with hugepages= to reserve
 
huge pages of different sizes. Valid pages sizes on
 
x86-64 are 2M (when the CPU supports "pse") and 1G
 
(when the CPU supports the "pdpe1gb" cpuinfo flag)
 
Note that 1GB pages can only be allocated at boot time
 
using hugepages= and not freed afterwards.
 
default_hugepagesz=
 
[same as hugepagesz=] The size of the default
 
HugeTLB page size. This is the size represented by
 
the legacy /proc/ hugepages APIs, used for SHM, and
 
default size when mounting hugetlbfs filesystems.
 
Defaults to the default architecture's huge page size
 
if not specified.
 
 
hlt [BUGS=ARM,SH]
 
 
hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
 
      terminal devices. Valid values: 0..8
 
 
i8042.debug [HW] Toggle i8042 debug mode
 
i8042.direct [HW] Put keyboard port into non-translated mode
 
i8042.dumbkbd [HW] Pretend that controller can only read data from
 
    keyboard and cannot control its state
 
    (Don't attempt to blink the leds)
 
i8042.noaux [HW] Don't check for auxiliary (== mouse) port
 
i8042.nokbd [HW] Don't check/create keyboard port
 
i8042.noloop [HW] Disable the AUX Loopback command while probing
 
    for the AUX port
 
i8042.nomux [HW] Don't check presence of an active multiplexing
 
    controller
 
i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
 
    controllers
 
i8042.panicblink=
 
[HW] Frequency with which keyboard LEDs should blink
 
    when kernel panics (default is 0.5 sec)
 
i8042.reset [HW] Reset the controller during init and cleanup
 
i8042.unlock [HW] Unlock (ignore) the keylock
 
 
i810= [HW,DRM]
 
 
i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
 
indicates that the driver is running on unsupported
 
hardware.
 
i8k.force [HW] Activate i8k driver even if SMM BIOS signature
 
does not match list of supported models.
 
i8k.power_status
 
[HW] Report power status in /proc/i8k
 
(disabled by default)
 
i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
 
capability is set.
 
 
ibmmcascsi= [HW,MCA,SCSI] IBM MicroChannel SCSI adapter
 
See Documentation/mca.txt.
 
 
icn= [HW,ISDN]
 
Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
 
 
ide-core.nodma= [HW] (E)IDE subsystem
 
Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
 
.vlb_clock .pci_clock .noflush .noprobe .nowerr .cdrom
 
.chs .ignore_cable are additional options
 
See Documentation/ide/ide.txt.
 
 
idebus= [HW] (E)IDE subsystem - VLB/PCI bus speed
 
See Documentation/ide/ide.txt.
 
 
idle= [X86]
 
Format: idle=poll, idle=mwait, idle=halt, idle=nomwait
 
Poll forces a polling idle loop that can slightly
 
improve the performance of waking up a idle CPU, but
 
will use a lot of power and make the system run hot.
 
Not recommended.
 
idle=mwait: On systems which support MONITOR/MWAIT but
 
the kernel chose to not use it because it doesn't save
 
as much power as a normal idle loop, use the
 
MONITOR/MWAIT idle loop anyways. Performance should be
 
the same as idle=poll.
 
idle=halt: Halt is forced to be used for CPU idle.
 
In such case C2/C3 won't be used again.
 
idle=nomwait: Disable mwait for CPU C-states
 
 
ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
 
Claim all unknown PCI IDE storage controllers.
 
 
ignore_loglevel [KNL]
 
Ignore loglevel setting - this will print /all/
 
kernel messages to the console. Useful for debugging.
 
 
ihash_entries= [KNL]
 
Set number of hash buckets for inode cache.
 
 
in2000= [HW,SCSI]
 
See header of drivers/scsi/in2000.c.
 
 
init= [KNL]
 
Format: <full_path>
 
Run specified binary instead of /sbin/init as init
 
process.
 
 
initcall_debug [KNL] Trace initcalls as they are executed.  Useful
 
for working out where the kernel is dying during
 
startup.
 
 
initrd= [BOOT] Specify the location of the initial ramdisk
 
 
inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
 
Format: <irq>
 
 
inttest= [IA64]
 
 
iomem= Disable strict checking of access to MMIO memory
 
strict regions from userspace.
 
relaxed
 
 
iommu= [x86]
 
off
 
force
 
noforce
 
biomerge
 
panic
 
nopanic
 
merge
 
nomerge
 
forcesac
 
soft
 
 
 
intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
 
on
 
Enable intel iommu driver.
 
off
 
Disable intel iommu driver.
 
igfx_off [Default Off]
 
By default, gfx is mapped as normal device. If a gfx
 
device has a dedicated DMAR unit, the DMAR unit is
 
bypassed by not enabling DMAR with this option. In
 
this case, gfx device will use physical address for
 
DMA.
 
forcedac [x86_64]
 
With this option iommu will not optimize to look
 
for io virtual address below 32 bit forcing dual
 
address cycle on pci bus for cards supporting greater
 
than 32 bit addressing. The default is to look
 
for translation below 32 bit and if not available
 
then look in the higher range.
 
strict [Default Off]
 
With this option on every unmap_single operation will
 
result in a hardware IOTLB flush operation as opposed
 
to batching them for performance.
 
 
io_delay= [X86-32,X86-64] I/O delay method
 
0x80
 
Standard port 0x80 based delay
 
0xed
 
Alternate port 0xed based delay (needed on some systems)
 
udelay
 
Simple two microseconds delay
 
none
 
No delay
 
 
io7= [HW] IO7 for Marvel based alpha systems
 
See comment before marvel_specify_io7 in
 
arch/alpha/kernel/core_marvel.c.
 
 
ip= [IP_PNP]
 
See Documentation/filesystems/nfsroot.txt.
 
 
ip2= [HW] Set IO/IRQ pairs for up to 4 IntelliPort boards
 
See comment before ip2_setup() in
 
drivers/char/ip2/ip2base.c.
 
 
ips= [HW,SCSI] Adaptec / IBM ServeRAID controller
 
See header of drivers/scsi/ips.c.
 
 
ports= [IP_VS_FTP] IPVS ftp helper module
 
Default is 21.
 
Up to 8 (IP_VS_APP_MAX_PORTS) ports
 
may be specified.
 
Format: <port>,<port>....
 
 
irqfixup [HW]
 
When an interrupt is not handled search all handlers
 
for it. Intended to get systems with badly broken
 
firmware running.
 
 
irqpoll [HW]
 
When an interrupt is not handled search all handlers
 
for it. Also check all handlers each timer
 
interrupt. Intended to get systems with badly broken
 
firmware running.
 
 
isapnp= [ISAPNP]
 
Format: <RDP>,<reset>,<pci_scan>,<verbosity>
 
 
isolcpus= [KNL,SMP] Isolate CPUs from the general scheduler.
 
Format:
 
<cpu number>,...,<cpu number>
 
or
 
<cpu number>-<cpu number>
 
(must be a positive range in ascending order)
 
or a mixture
 
<cpu number>,...,<cpu number>-<cpu number>
 
 
This option can be used to specify one or more CPUs
 
to isolate from the general SMP balancing and scheduling
 
algorithms. You can move a process onto or off an
 
"isolated" CPU via the CPU affinity syscalls or cpuset.
 
<cpu number> begins at 0 and the maximum value is
 
"number of CPUs in system - 1".
 
 
This option is the preferred way to isolate CPUs. The
 
alternative -- manually setting the CPU mask of all
 
tasks in the system -- can cause problems and
 
suboptimal load balancer performance.
 
 
iucv= [HW,NET]
 
 
js= [HW,JOY] Analog joystick
 
See Documentation/input/joystick.txt.
 
 
kernelcore=nn[KMG] [KNL,X86-32,IA-64,PPC,X86-64] This parameter
 
specifies the amount of memory usable by the kernel
 
for non-movable allocations.  The requested amount is
 
spread evenly throughout all nodes in the system. The
 
remaining memory in each node is used for Movable
 
pages. In the event, a node is too small to have both
 
kernelcore and Movable pages, kernelcore pages will
 
take priority and other nodes will have a larger number
 
of kernelcore pages.  The Movable zone is used for the
 
allocation of pages that may be reclaimed or moved
 
by the page migration subsystem.  This means that
 
HugeTLB pages may not be allocated from this zone.
 
Note that allocations like PTEs-from-HighMem still
 
use the HighMem zone if it exists, and the Normal
 
zone if it does not.
 
 
movablecore=nn[KMG] [KNL,X86-32,IA-64,PPC,X86-64] This parameter
 
is similar to kernelcore except it specifies the
 
amount of memory used for migratable allocations.
 
If both kernelcore and movablecore is specified,
 
then kernelcore will be at *least* the specified
 
value but may be more. If movablecore on its own
 
is specified, the administrator must be careful
 
that the amount of memory usable for all allocations
 
is not too small.
 
 
keepinitrd [HW,ARM]
 
 
kstack=N [X86-32,X86-64] Print N words from the kernel stack
 
in oops dumps.
 
 
kgdboc= [HW] kgdb over consoles.
 
Requires a tty driver that supports console polling.
 
(only serial suported for now)
 
Format: <serial_device>[,baud]
 
 
kmac= [MIPS] korina ethernet MAC address.
 
Configure the RouterBoard 532 series on-chip
 
Ethernet adapter MAC address.
 
 
l2cr= [PPC]
 
 
l3cr= [PPC]
 
 
lapic [X86-32,APIC] Enable the local APIC even if BIOS
 
disabled it.
 
 
lapic_timer_c2_ok [X86-32,x86-64,APIC] trust the local apic timer
 
in C2 power state.
 
 
libata.dma= [LIBATA] DMA control
 
libata.dma=0   Disable all PATA and SATA DMA
 
libata.dma=1   PATA and SATA Disk DMA only
 
libata.dma=2   ATAPI (CDROM) DMA only
 
libata.dma=4   Compact Flash DMA only
 
Combinations also work, so libata.dma=3 enables DMA
 
for disks and CDROMs, but not CFs.
 
 
libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
 
when set.
 
Format: <int>
 
 
libata.force= [LIBATA] Force configurations.  The format is comma
 
separated list of "[ID:]VAL" where ID is
 
PORT[:DEVICE].  PORT and DEVICE are decimal numbers
 
matching port, link or device.  Basically, it matches
 
the ATA ID string printed on console by libata.  If
 
the whole ID part is omitted, the last PORT and DEVICE
 
values are used.  If ID hasn't been specified yet, the
 
configuration applies to all ports, links and devices.
 
 
If only DEVICE is omitted, the parameter applies to
 
the port and all links and devices behind it.  DEVICE
 
number of 0 either selects the first device or the
 
first fan-out link behind PMP device.  It does not
 
select the host link.  DEVICE number of 15 selects the
 
host link and device attached to it.
 
 
The VAL specifies the configuration to force.  As long
 
as there's no ambiguity shortcut notation is allowed.
 
For example, both 1.5 and 1.5G would work for 1.5Gbps.
 
The following configurations can be forced.
 
 
* Cable type: 40c, 80c, short40c, unk, ign or sata.
 
  Any ID with matching PORT is used.
 
 
* SATA link speed limit: 1.5Gbps or 3.0Gbps.
 
 
* Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
 
  udma[/][16,25,33,44,66,100,133] notation is also
 
  allowed.
 
 
* [no]ncq: Turn on or off NCQ.
 
 
* nohrst, nosrst, norst: suppress hard, soft
 
                          and both resets.
 
 
If there are multiple matching configurations changing
 
the same attribute, the last one is used.
 
 
lmb=debug [KNL] Enable lmb debug messages.
 
 
load_ramdisk= [RAM] List of ramdisks to load from floppy
 
See Documentation/blockdev/ramdisk.txt.
 
 
lockd.nlm_grace_period=P  [NFS] Assign grace period.
 
Format: <integer>
 
 
lockd.nlm_tcpport=N [NFS] Assign TCP port.
 
Format: <integer>
 
 
lockd.nlm_timeout=T [NFS] Assign timeout value.
 
Format: <integer>
 
 
lockd.nlm_udpport=M [NFS] Assign UDP port.
 
Format: <integer>
 
 
logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
 
Format: <irq>
 
 
loglevel= All Kernel Messages with a loglevel smaller than the
 
console loglevel will be printed to the console. It can
 
also be changed with klogd or other programs. The
 
loglevels are defined as follows:
 
 
0 (KERN_EMERG) system is unusable
 
1 (KERN_ALERT) action must be taken immediately
 
2 (KERN_CRIT) critical conditions
 
3 (KERN_ERR) error conditions
 
4 (KERN_WARNING) warning conditions
 
5 (KERN_NOTICE) normal but significant condition
 
6 (KERN_INFO) informational
 
7 (KERN_DEBUG) debug-level messages
 
 
log_buf_len=n Sets the size of the printk ring buffer, in bytes.
 
Format: { n | nk | nM }
 
n must be a power of two.  The default size
 
is set in the kernel config file.
 
 
logo.nologo [FB] Disables display of the built-in Linux logo.
 
This may be used to provide more screen space for
 
kernel log messages and is useful when debugging
 
kernel boot problems.
 
 
lp=0 [LP] Specify parallel ports to use, e.g,
 
lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
 
lp=reset first parallel port). 'lp=0' disables the
 
lp=auto printer driver. 'lp=reset' (which can be
 
specified in addition to the ports) causes
 
attached printers to be reset. Using
 
lp=port1,port2,... specifies the parallel ports
 
to associate lp devices with, starting with
 
lp0. A port specification may be 'none' to skip
 
that lp device, or a parport name such as
 
'parport0'. Specifying 'lp=auto' instead of a
 
port specification list means that device IDs
 
from each port should be examined, to see if
 
an IEEE 1284-compliant printer is attached; if
 
so, the driver will manage that printer.
 
See also header of drivers/char/lp.c.
 
 
lpj=n [KNL]
 
Sets loops_per_jiffy to given constant, thus avoiding
 
time-consuming boot-time autodetection (up to 250 ms per
 
CPU). 0 enables autodetection (default). To determine
 
the correct value for your kernel, boot with normal
 
autodetection and see what value is printed. Note that
 
on SMP systems the preset will be applied to all CPUs,
 
which is likely to cause problems if your CPUs need
 
significantly divergent settings. An incorrect value
 
will cause delays in the kernel to be wrong, leading to
 
unpredictable I/O errors and other breakage. Although
 
unlikely, in the extreme case this might damage your
 
hardware.
 
 
ltpc= [NET]
 
Format: <io>,<irq>,<dma>
 
 
mac5380= [HW,SCSI] Format:
 
<can_queue>,<cmd_per_lun>,<sg_tablesize>,<hostid>,<use_tags>
 
 
machvec= [IA64] Force the use of a particular machine-vector
 
(machvec) in a generic kernel.
 
Example: machvec=hpzx1_swiotlb
 
 
max_loop= [LOOP] Maximum number of loopback devices that can
 
be mounted
 
Format: <1-256>
 
 
maxcpus= [SMP] Maximum number of processors that an SMP kernel
 
should make use of.  maxcpus=n : n >= 0 limits the
 
kernel to using 'n' processors.  n=0 is a special case,
 
it is equivalent to "nosmp", which also disables
 
the IO APIC.
 
 
max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater than
 
or equal to this physical address is ignored.
 
 
max_luns= [SCSI] Maximum number of LUNs to probe.
 
Should be between 1 and 2^32-1.
 
 
max_report_luns=
 
[SCSI] Maximum number of LUNs received.
 
Should be between 1 and 16384.
 
 
mcatest= [IA-64]
 
 
mce [X86-32] Machine Check Exception
 
 
mce=option [X86-64] See Documentation/x86/x86_64/boot-options.txt
 
 
md= [HW] RAID subsystems devices and level
 
See Documentation/md.txt.
 
 
mdacon= [MDA]
 
Format: <first>,<last>
 
Specifies range of consoles to be captured by the MDA.
 
 
mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
 
Amount of memory to be used when the kernel is not able
 
to see the whole system memory or for test.
 
[X86-32] Use together with memmap= to avoid physical
 
address space collisions. Without memmap= PCI devices
 
could be placed at addresses belonging to unused RAM.
 
 
mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
 
memory.
 
 
memchunk=nn[KMG]
 
[KNL,SH] Allow user to override the default size for
 
per-device physically contiguous DMA buffers.
 
 
memmap=exactmap [KNL,X86-32,X86_64] Enable setting of an exact
 
E820 memory map, as specified by the user.
 
Such memmap=exactmap lines can be constructed based on
 
BIOS output or other requirements. See the memmap=nn@ss
 
option description.
 
 
memmap=nn[KMG]@ss[KMG]
 
[KNL] Force usage of a specific region of memory
 
Region of memory to be used, from ss to ss+nn.
 
 
memmap=nn[KMG]#ss[KMG]
 
[KNL,ACPI] Mark specific memory as ACPI data.
 
Region of memory to be used, from ss to ss+nn.
 
 
memmap=nn[KMG]$ss[KMG]
 
[KNL,ACPI] Mark specific memory as reserved.
 
Region of memory to be used, from ss to ss+nn.
 
Example: Exclude memory from 0x18690000-0x1869ffff
 
        memmap=64K$0x18690000
 
        or
 
        memmap=0x10000$0x18690000
 
 
memory_corruption_check=0/1 [X86]
 
Some BIOSes seem to corrupt the first 64k of
 
memory when doing things like suspend/resume.
 
Setting this option will scan the memory
 
looking for corruption.  Enabling this will
 
both detect corruption and prevent the kernel
 
from using the memory being corrupted.
 
However, its intended as a diagnostic tool; if
 
repeatable BIOS-originated corruption always
 
affects the same memory, you can use memmap=
 
to prevent the kernel from using that memory.
 
 
memory_corruption_check_size=size [X86]
 
By default it checks for corruption in the low
 
64k, making this memory unavailable for normal
 
use.  Use this parameter to scan for
 
corruption in more or less memory.
 
 
memory_corruption_check_period=seconds [X86]
 
By default it checks for corruption every 60
 
seconds.  Use this parameter to check at some
 
other rate.  0 disables periodic checking.
 
 
memtest= [KNL,X86] Enable memtest
 
Format: <integer>
 
range: 0,4 : pattern number
 
default : 0 <disable>
 
 
meye.*= [HW] Set MotionEye Camera parameters
 
See Documentation/video4linux/meye.txt.
 
 
mfgpt_irq= [IA-32] Specify the IRQ to use for the
 
Multi-Function General Purpose Timers on AMD Geode
 
platforms.
 
 
mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
 
the BIOS has incorrectly applied a workaround. TinyBIOS
 
version 0.98 is known to be affected, 0.99 fixes the
 
problem by letting the user disable the workaround.
 
 
mga= [HW,DRM]
 
 
min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
 
physical address is ignored.
 
 
mminit_loglevel=
 
[KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
 
parameter allows control of the logging verbosity for
 
the additional memory initialisation checks. A value
 
of 0 disables mminit logging and a level of 4 will
 
log everything. Information is printed at KERN_DEBUG
 
so loglevel=8 may also need to be specified.
 
 
mousedev.tap_time=
 
[MOUSE] Maximum time between finger touching and
 
leaving touchpad surface for touch to be considered
 
a tap and be reported as a left button click (for
 
touchpads working in absolute mode only).
 
Format: <msecs>
 
mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
 
reporting absolute coordinates, such as tablets
 
mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
 
reporting absolute coordinates, such as tablets
 
 
mpu401= [HW,OSS]
 
Format: <io>,<irq>
 
 
MTD_Partition= [MTD]
 
Format: <name>,<region-number>,<size>,<offset>
 
 
MTD_Region= [MTD] Format:
 
<name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
 
 
mtdparts= [MTD]
 
See drivers/mtd/cmdlinepart.c.
 
 
mtdset= [ARM]
 
ARM/S3C2412 JIVE boot control
 
 
See arch/arm/mach-s3c2412/mach-jive.c
 
 
mtouchusb.raw_coordinates=
 
[HW] Make the MicroTouch USB driver use raw coordinates
 
('y', default) or cooked coordinates ('n')
 
 
n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
 
 
NCR_D700= [HW,SCSI]
 
See header of drivers/scsi/NCR_D700.c.
 
 
ncr5380= [HW,SCSI]
 
 
ncr53c400= [HW,SCSI]
 
 
ncr53c400a= [HW,SCSI]
 
 
ncr53c406a= [HW,SCSI]
 
 
ncr53c8xx= [HW,SCSI]
 
 
netdev= [NET] Network devices parameters
 
Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
 
Note that mem_start is often overloaded to mean
 
something different and driver-specific.
 
This usage is only documented in each driver source
 
file if at all.
 
 
nf_conntrack.acct=
 
[NETFILTER] Enable connection tracking flow accounting
 
0 to disable accounting
 
1 to enable accounting
 
Default value depends on CONFIG_NF_CT_ACCT that is
 
going to be removed in 2.6.29.
 
 
nfsaddrs= [NFS]
 
See Documentation/filesystems/nfsroot.txt.
 
 
nfsroot= [NFS] nfs root filesystem for disk-less boxes.
 
See Documentation/filesystems/nfsroot.txt.
 
 
nfs.callback_tcpport=
 
[NFS] set the TCP port on which the NFSv4 callback
 
channel should listen.
 
 
nfs.idmap_cache_timeout=
 
[NFS] set the maximum lifetime for idmapper cache
 
entries.
 
 
nfs.enable_ino64=
 
[NFS] enable 64-bit inode numbers.
 
If zero, the NFS client will fake up a 32-bit inode
 
number for the readdir() and stat() syscalls instead
 
of returning the full 64-bit number.
 
The default is to return 64-bit inode numbers.
 
 
nmi_debug= [KNL,AVR32] Specify one or more actions to take
 
when a NMI is triggered.
 
Format: [state][,regs][,debounce][,die]
 
 
nmi_watchdog= [KNL,BUGS=X86-32,X86-64] Debugging features for SMP kernels
 
Format: [panic,][num]
 
Valid num: 0,1,2
 
0 - turn nmi_watchdog off
 
1 - use the IO-APIC timer for the NMI watchdog
 
2 - use the local APIC for the NMI watchdog using
 
a performance counter. Note: This will use one performance
 
counter and the local APIC's performance vector.
 
When panic is specified panic when an NMI watchdog timeout occurs.
 
This is useful when you use a panic=... timeout and need the box
 
quickly up again.
 
Instead of 1 and 2 it is possible to use the following
 
symbolic names: lapic and ioapic
 
Example: nmi_watchdog=2 or nmi_watchdog=panic,lapic
 
 
no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
 
emulation library even if a 387 maths coprocessor
 
is present.
 
 
noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
 
caches in the slab allocator.  Saves per-node memory,
 
but will impact performance.
 
 
noalign [KNL,ARM]
 
 
noapic [SMP,APIC] Tells the kernel to not make use of any
 
IOAPICs that may be present in the system.
 
 
nobats [PPC] Do not use BATs for mapping kernel lowmem
 
on "Classic" PPC cores.
 
 
nocache [ARM]
 
 
nodelayacct [KNL] Disable per-task delay accounting
 
 
nodisconnect [HW,SCSI,M68K] Disables SCSI disconnects.
 
 
nodsp [SH] Disable hardware DSP at boot time.
 
 
noefi [X86-32,X86-64] Disable EFI runtime services support.
 
 
noexec [IA-64]
 
 
noexec [X86-32,X86-64]
 
On X86-32 available only on PAE configured kernels.
 
noexec=on: enable non-executable mappings (default)
 
noexec=off: disable non-executable mappings
 
 
noexec32 [X86-64]
 
This affects only 32-bit executables.
 
noexec32=on: enable non-executable mappings (default)
 
read doesn't imply executable mappings
 
noexec32=off: disable non-executable mappings
 
read implies executable mappings
 
 
nofpu [SH] Disable hardware FPU at boot time.
 
 
nofxsr [BUGS=X86-32] Disables x86 floating point extended
 
register save and restore. The kernel will only save
 
legacy floating-point registers on task switch.
 
 
noclflush [BUGS=X86] Don't use the CLFLUSH instruction
 
 
nohlt [BUGS=ARM,SH]
 
 
no-hlt [BUGS=X86-32] Tells the kernel that the hlt
 
instruction doesn't work correctly and not to
 
use it.
 
 
no_file_caps Tells the kernel not to honor file capabilities.  The
 
only way then for a file to be executed with privilege
 
is to be setuid root or executed by root.
 
 
nohalt [IA-64] Tells the kernel not to use the power saving
 
function PAL_HALT_LIGHT when idle. This increases
 
power-consumption. On the positive side, it reduces
 
interrupt wake-up latency, which may improve performance
 
in certain environments such as networked servers or
 
real-time systems.
 
 
nohz= [KNL] Boottime enable/disable dynamic ticks
 
Valid arguments: on, off
 
Default: on
 
 
noirqdebug [X86-32] Disables the code which attempts to detect and
 
disable unhandled interrupt sources.
 
 
no_timer_check [X86-32,X86_64,APIC] Disables the code which tests for
 
broken timer IRQ sources.
 
 
noisapnp [ISAPNP] Disables ISA PnP code.
 
 
noinitrd [RAM] Tells the kernel not to load any configured
 
initial RAM disk.
 
 
nointroute [IA-64]
 
 
nojitter [IA64] Disables jitter checking for ITC timers.
 
 
nolapic [X86-32,APIC] Do not enable or use the local APIC.
 
 
nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
 
 
nox2apic [X86-64,APIC] Do not enable x2APIC mode.
 
 
x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
 
default x2apic cluster mode on platforms
 
supporting x2apic.
 
 
noltlbs [PPC] Do not use large page/tlb entries for kernel
 
lowmem mapping on PPC40x.
 
 
nomca [IA-64] Disable machine check abort handling
 
 
nomce [X86-32] Machine Check Exception
 
 
nomfgpt [X86-32] Disable Multi-Function General Purpose
 
Timer usage (for AMD Geode machines).
 
 
noreplace-paravirt [X86-32,PV_OPS] Don't patch paravirt_ops
 
 
noreplace-smp [X86-32,SMP] Don't replace SMP instructions
 
with UP alternatives
 
 
noresidual [PPC] Don't use residual data on PReP machines.
 
 
noresume [SWSUSP] Disables resume and restores original swap
 
space.
 
 
no-scroll [VGA] Disables scrollback.
 
This is required for the Braillex ib80-piezo Braille
 
reader made by F.H. Papenmeier (Germany).
 
 
nosbagart [IA-64]
 
 
nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
 
 
nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
 
and disable the IO APIC.  legacy for "maxcpus=0".
 
 
nosoftlockup [KNL] Disable the soft-lockup detector.
 
 
noswapaccount [KNL] Disable accounting of swap in memory resource
 
controller. (See Documentation/controllers/memory.txt)
 
 
nosync [HW,M68K] Disables sync negotiation for all devices.
 
 
notsc [BUGS=X86-32] Disable Time Stamp Counter
 
 
nousb [USB] Disable the USB subsystem
 
 
nowb [ARM]
 
 
nptcg= [IA64] Override max number of concurrent global TLB
 
purges which is reported from either PAL_VM_SUMMARY or
 
SAL PALO.
 
 
numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
 
one of ['zone', 'node', 'default'] can be specified
 
This can be set from sysctl after boot.
 
See Documentation/sysctl/vm.txt for details.
 
 
nr_uarts= [SERIAL] maximum number of UARTs to be registered.
 
 
ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
 
See Documentation/debugging-via-ohci1394.txt for more
 
info.
 
 
olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
 
Rather than timing out after 20 ms if an EC
 
command is not properly ACKed, override the length
 
of the timeout.  We have interrupts disabled while
 
waiting for the ACK, so if this is set too high
 
interrupts *may* be lost!
 
 
opl3= [HW,OSS]
 
Format: <io>
 
 
oprofile.timer= [HW]
 
Use timer interrupt instead of performance counters
 
 
osst= [HW,SCSI] SCSI Tape Driver
 
Format: <buffer_size>,<write_threshold>
 
See also Documentation/scsi/st.txt.
 
 
panic= [KNL] Kernel behaviour on panic
 
Format: <timeout>
 
 
parkbd.port= [HW] Parallel port number the keyboard adapter is
 
connected to, default is 0.
 
Format: <parport#>
 
parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
 
0 for XT, 1 for AT (default is AT).
 
Format: <mode>
 
 
parport= [HW,PPT] Specify parallel ports. 0 disables.
 
Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
 
Use 'auto' to force the driver to use any
 
IRQ/DMA settings detected (the default is to
 
ignore detected IRQ/DMA settings because of
 
possible conflicts). You can specify the base
 
address, IRQ, and DMA settings; IRQ and DMA
 
should be numbers, or 'auto' (for using detected
 
settings on that particular port), or 'nofifo'
 
(to avoid using a FIFO even if it is detected).
 
Parallel ports are assigned in the order they
 
are specified on the command line, starting
 
with parport0.
 
 
parport_init_mode= [HW,PPT]
 
Configure VIA parallel port to operate in
 
a specific mode. This is necessary on Pegasos
 
computer where firmware has no options for setting
 
up parallel port mode and sets it to spp.
 
Currently this function knows 686a and 8231 chips.
 
Format: [spp|ps2|epp|ecp|ecpepp]
 
 
pas2= [HW,OSS] Format:
 
<io>,<irq>,<dma>,<dma16>,<sb_io>,<sb_irq>,<sb_dma>,<sb_dma16>
 
 
pas16= [HW,SCSI]
 
See header of drivers/scsi/pas16.c.
 
 
pause_on_oops=
 
Halt all CPUs after the first oops has been printed for
 
the specified number of seconds.  This is to be used if
 
your oopses keep scrolling off the screen.
 
 
pcbit= [HW,ISDN]
 
 
pcd. [PARIDE]
 
See header of drivers/block/paride/pcd.c.
 
See also Documentation/blockdev/paride.txt.
 
 
pci=option[,option...] [PCI] various PCI subsystem options:
 
off [X86] don't probe for the PCI bus
 
bios [X86-32] force use of PCI BIOS, don't access
 
the hardware directly. Use this if your machine
 
has a non-standard PCI host bridge.
 
nobios [X86-32] disallow use of PCI BIOS, only direct
 
hardware access methods are allowed. Use this
 
if you experience crashes upon bootup and you
 
suspect they are caused by the BIOS.
 
conf1 [X86] Force use of PCI Configuration
 
Mechanism 1.
 
conf2 [X86] Force use of PCI Configuration
 
Mechanism 2.
 
noaer [PCIE] If the PCIEAER kernel config parameter is
 
enabled, this kernel boot option can be used to
 
disable the use of PCIE advanced error reporting.
 
nodomains [PCI] Disable support for multiple PCI
 
root domains (aka PCI segments, in ACPI-speak).
 
nommconf [X86-32,X86_64] Disable use of MMCONFIG for PCI
 
Configuration
 
msi [MSI] If the PCI_MSI kernel config parameter is
 
enabled, this kernel boot option can be used to
 
enable the use of MSI interrupts system-wide.
 
nomsi [MSI] If the PCI_MSI kernel config parameter is
 
enabled, this kernel boot option can be used to
 
disable the use of MSI interrupts system-wide.
 
noioapicquirk [APIC] Disable all boot interrupt quirks.
 
Safety option to keep boot IRQs enabled. This
 
should never be necessary.
 
ioapicreroute [APIC] Enable rerouting of boot IRQs to the
 
primary IO-APIC for bridges that cannot disable
 
boot IRQs. This fixes a source of spurious IRQs
 
when the system masks IRQs.
 
noioapicreroute [APIC] Disable workaround that uses the
 
boot IRQ equivalent of an IRQ that connects to
 
a chipset where boot IRQs cannot be disabled.
 
The opposite of ioapicreroute.
 
biosirq [X86-32] Use PCI BIOS calls to get the interrupt
 
routing table. These calls are known to be buggy
 
on several machines and they hang the machine
 
when used, but on other computers it's the only
 
way to get the interrupt routing table. Try
 
this option if the kernel is unable to allocate
 
IRQs or discover secondary PCI buses on your
 
motherboard.
 
rom [X86] Assign address space to expansion ROMs.
 
Use with caution as certain devices share
 
address decoders between ROMs and other
 
resources.
 
norom [X86] Do not assign address space to
 
expansion ROMs that do not already have
 
BIOS assigned address ranges.
 
irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
 
assigned automatically to PCI devices. You can
 
make the kernel exclude IRQs of your ISA cards
 
this way.
 
pirqaddr=0xAAAAA [X86] Specify the physical address
 
of the PIRQ table (normally generated
 
by the BIOS) if it is outside the
 
F0000h-100000h range.
 
lastbus=N [X86] Scan all buses thru bus #N. Can be
 
useful if the kernel is unable to find your
 
secondary buses and you want to tell it
 
explicitly which ones they are.
 
assign-busses [X86] Always assign all PCI bus
 
numbers ourselves, overriding
 
whatever the firmware may have done.
 
usepirqmask [X86] Honor the possible IRQ mask stored
 
in the BIOS $PIR table. This is needed on
 
some systems with broken BIOSes, notably
 
some HP Pavilion N5400 and Omnibook XE3
 
notebooks. This will have no effect if ACPI
 
IRQ routing is enabled.
 
noacpi [X86] Do not use ACPI for IRQ routing
 
or for PCI scanning.
 
use_crs [X86] Use _CRS for PCI resource
 
allocation.
 
routeirq Do IRQ routing for all PCI devices.
 
This is normally done in pci_enable_device(),
 
so this option is a temporary workaround
 
for broken drivers that don't call it.
 
skip_isa_align [X86] do not align io start addr, so can
 
handle more pci cards
 
firmware [ARM] Do not re-enumerate the bus but instead
 
just use the configuration from the
 
bootloader. This is currently used on
 
IXP2000 systems where the bus has to be
 
configured a certain way for adjunct CPUs.
 
noearly [X86] Don't do any early type 1 scanning.
 
This might help on some broken boards which
 
machine check when some devices' config space
 
is read. But various workarounds are disabled
 
and some IOMMU drivers will not work.
 
bfsort Sort PCI devices into breadth-first order.
 
This sorting is done to get a device
 
order compatible with older (<= 2.4) kernels.
 
nobfsort Don't sort PCI devices into breadth-first order.
 
cbiosize=nn[KMG] The fixed amount of bus space which is
 
reserved for the CardBus bridge's IO window.
 
The default value is 256 bytes.
 
cbmemsize=nn[KMG] The fixed amount of bus space which is
 
reserved for the CardBus bridge's memory
 
window. The default value is 64 megabytes.
 
 
pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
 
Management.
 
off Disable ASPM.
 
force Enable ASPM even on devices that claim not to support it.
 
WARNING: Forcing ASPM on may cause system lockups.
 
 
pcmv= [HW,PCMCIA] BadgePAD 4
 
 
pd. [PARIDE]
 
See Documentation/blockdev/paride.txt.
 
 
pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
 
boot time.
 
Format: { 0 | 1 }
 
See arch/parisc/kernel/pdc_chassis.c
 
 
pf. [PARIDE]
 
See Documentation/blockdev/paride.txt.
 
 
pg. [PARIDE]
 
See Documentation/blockdev/paride.txt.
 
 
pirq= [SMP,APIC] Manual mp-table setup
 
See Documentation/x86/i386/IO-APIC.txt.
 
 
plip= [PPT,NET] Parallel port network link
 
Format: { parport<nr> | timid | 0 }
 
See also Documentation/parport.txt.
 
 
pmtmr= [X86] Manual setup of pmtmr I/O Port.
 
Override pmtimer IOPort with a hex value.
 
e.g. pmtmr=0x508
 
 
pnp.debug [PNP]
 
Enable PNP debug messages.  This depends on the
 
CONFIG_PNP_DEBUG_MESSAGES option.
 
 
pnpacpi= [ACPI]
 
{ off }
 
 
pnpbios= [ISAPNP]
 
{ on | off | curr | res | no-curr | no-res }
 
 
pnp_reserve_irq=
 
[ISAPNP] Exclude IRQs for the autoconfiguration
 
 
pnp_reserve_dma=
 
[ISAPNP] Exclude DMAs for the autoconfiguration
 
 
pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
 
Ranges are in pairs (I/O port base and size).
 
 
pnp_reserve_mem=
 
[ISAPNP] Exclude memory regions for the
 
autoconfiguration.
 
Ranges are in pairs (memory base and size).
 
 
dynamic_printk Enables pr_debug()/dev_dbg() calls if
 
CONFIG_DYNAMIC_PRINTK_DEBUG has been enabled.
 
These can also be switched on/off via
 
<debugfs>/dynamic_printk/modules
 
 
print-fatal-signals=
 
[KNL] debug: print fatal signals
 
print-fatal-signals=1: print segfault info to
 
the kernel console.
 
default: off.
 
 
printk.time= Show timing data prefixed to each printk message line
 
Format: <bool>  (1/Y/y=enable, 0/N/n=disable)
 
 
profile= [KNL] Enable kernel profiling via /proc/profile
 
Format: [schedule,]<number>
 
Param: "schedule" - profile schedule points.
 
Param: <number> - step/bucket size as a power of 2 for
 
statistical time based profiling.
 
Param: "sleep" - profile D-state sleeping (millisecs).
 
Requires CONFIG_SCHEDSTATS
 
Param: "kvm" - profile VM exits.
 
 
processor.max_cstate= [HW,ACPI]
 
Limit processor to maximum C-state
 
max_cstate=9 overrides any DMI blacklist limit.
 
 
processor.nocst [HW,ACPI]
 
Ignore the _CST method to determine C-states,
 
instead using the legacy FADT method
 
 
prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
 
before loading.
 
See Documentation/blockdev/ramdisk.txt.
 
 
psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
 
probe for; one of (bare|imps|exps|lifebook|any).
 
psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
 
per second.
 
psmouse.resetafter= [HW,MOUSE]
 
Try to reset the device after so many bad packets
 
(0 = never).
 
psmouse.resolution=
 
[HW,MOUSE] Set desired mouse resolution, in dpi.
 
psmouse.smartscroll=
 
[HW,MOUSE] Controls Logitech smartscroll autorepeat.
 
0 = disabled, 1 = enabled (default).
 
 
pss= [HW,OSS] Personal Sound System (ECHO ESC614)
 
Format:
 
<io>,<mss_io>,<mss_irq>,<mss_dma>,<mpu_io>,<mpu_irq>
 
 
pt. [PARIDE]
 
See Documentation/blockdev/paride.txt.
 
 
pty.legacy_count=
 
[KNL] Number of legacy pty's. Overwrites compiled-in
 
default number.
 
 
quiet [KNL] Disable most log messages
 
 
r128= [HW,DRM]
 
 
raid= [HW,RAID]
 
See Documentation/md.txt.
 
 
ramdisk_blocksize= [RAM]
 
See Documentation/blockdev/ramdisk.txt.
 
 
ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
 
See Documentation/blockdev/ramdisk.txt.
 
 
rcupdate.blimit= [KNL,BOOT]
 
Set maximum number of finished RCU callbacks to process
 
in one batch.
 
 
rcupdate.qhimark= [KNL,BOOT]
 
Set threshold of queued
 
RCU callbacks over which batch limiting is disabled.
 
 
rcupdate.qlowmark= [KNL,BOOT]
 
Set threshold of queued RCU callbacks below which
 
batch limiting is re-enabled.
 
 
rdinit= [KNL]
 
Format: <full_path>
 
Run specified binary instead of /init from the ramdisk,
 
used for early userspace startup. See initrd.
 
 
reboot= [BUGS=X86-32,BUGS=ARM,BUGS=IA-64] Rebooting mode
 
Format: <reboot_mode>[,<reboot_mode2>[,...]]
 
See arch/*/kernel/reboot.c or arch/*/kernel/process.c
 
 
relax_domain_level=
 
[KNL, SMP] Set scheduler's default relax_domain_level.
 
See Documentation/cpusets.txt.
 
 
reserve= [KNL,BUGS] Force the kernel to ignore some iomem area
 
 
reservetop= [X86-32]
 
Format: nn[KMG]
 
Reserves a hole at the top of the kernel virtual
 
address space.
 
 
reset_devices [KNL] Force drivers to reset the underlying device
 
during initialization.
 
 
resume= [SWSUSP]
 
Specify the partition device for software suspend
 
 
resume_offset= [SWSUSP]
 
Specify the offset from the beginning of the partition
 
given by "resume=" at which the swap header is located,
 
in <PAGE_SIZE> units (needed only for swap files).
 
See  Documentation/power/swsusp-and-swap-files.txt
 
 
retain_initrd [RAM] Keep initrd memory after extraction
 
 
rhash_entries= [KNL,NET]
 
Set number of hash buckets for route cache
 
 
riscom8= [HW,SERIAL]
 
Format: <io_board1>[,<io_board2>[,...<io_boardN>]]
 
 
ro [KNL] Mount root device read-only on boot
 
 
root= [KNL] Root filesystem
 
 
rootdelay= [KNL] Delay (in seconds) to pause before attempting to
 
mount the root filesystem
 
 
rootflags= [KNL] Set root filesystem mount option string
 
 
rootfstype= [KNL] Set root filesystem type
 
 
rootwait [KNL] Wait (indefinitely) for root device to show up.
 
Useful for devices that are detected asynchronously
 
(e.g. USB and MMC devices).
 
 
root_plug.vendor_id=
 
[ROOTPLUG] Override the default vendor ID
 
 
root_plug.product_id=
 
[ROOTPLUG] Override the default product ID
 
 
root_plug.debug=
 
[ROOTPLUG] Enable debugging output
 
 
rw [KNL] Mount root device read-write on boot
 
 
S [KNL] Run init in single mode
 
 
sa1100ir [NET]
 
See drivers/net/irda/sa1100_ir.c.
 
 
sbni= [NET] Granch SBNI12 leased line adapter
 
 
sc1200wdt= [HW,WDT] SC1200 WDT (watchdog) driver
 
Format: <io>[,<timeout>[,<isapnp>]]
 
 
scsi_debug_*= [SCSI]
 
See drivers/scsi/scsi_debug.c.
 
 
scsi_default_dev_flags=
 
[SCSI] SCSI default device flags
 
Format: <integer>
 
 
scsi_dev_flags= [SCSI] Black/white list entry for vendor and model
 
Format: <vendor>:<model>:<flags>
 
(flags are integer value)
 
 
scsi_logging_level= [SCSI] a bit mask of logging levels
 
See drivers/scsi/scsi_logging.h for bits.  Also
 
settable via sysctl at dev.scsi.logging_level
 
(/proc/sys/dev/scsi/logging_level).
 
There is also a nice 'scsi_logging_level' script in the
 
S390-tools package, available for download at
 
http://www-128.ibm.com/developerworks/linux/linux390/s390-tools-1.5.4.html
 
 
scsi_mod.scan= [SCSI] sync (default) scans SCSI busses as they are
 
discovered.  async scans them in kernel threads,
 
allowing boot to proceed.  none ignores them, expecting
 
user space to do the scan.
 
 
selinux [SELINUX] Disable or enable SELinux at boot time.
 
Format: { "0" | "1" }
 
See security/selinux/Kconfig help text.
 
0 -- disable.
 
1 -- enable.
 
Default value is set via kernel config option.
 
If enabled at boot time, /selinux/disable can be used
 
later to disable prior to initial policy load.
 
 
selinux_compat_net =
 
[SELINUX] Set initial selinux_compat_net flag value.
 
                        Format: { "0" | "1" }
 
                        0 -- use new secmark-based packet controls
 
                        1 -- use legacy packet controls
 
                        Default value is 0 (preferred).
 
                        Value can be changed at runtime via
 
                        /selinux/compat_net.
 
 
serialnumber [BUGS=X86-32]
 
 
shapers= [NET]
 
Maximal number of shapers.
 
 
show_msr= [x86] show boot-time MSR settings
 
Format: { <integer> }
 
Show boot-time (BIOS-initialized) MSR settings.
 
The parameter means the number of CPUs to show,
 
for example 1 means boot CPU only.
 
 
sim710= [SCSI,HW]
 
See header of drivers/scsi/sim710.c.
 
 
simeth= [IA-64]
 
simscsi=
 
 
slram= [HW,MTD]
 
 
slub_debug[=options[,slabs]] [MM, SLUB]
 
Enabling slub_debug allows one to determine the
 
culprit if slab objects become corrupted. Enabling
 
slub_debug can create guard zones around objects and
 
may poison objects when not in use. Also tracks the
 
last alloc / free. For more information see
 
Documentation/vm/slub.txt.
 
 
slub_max_order= [MM, SLUB]
 
Determines the maximum allowed order for slabs.
 
A high setting may cause OOMs due to memory
 
fragmentation. For more information see
 
Documentation/vm/slub.txt.
 
 
slub_min_objects= [MM, SLUB]
 
The minimum number of objects per slab. SLUB will
 
increase the slab order up to slub_max_order to
 
generate a sufficiently large slab able to contain
 
the number of objects indicated. The higher the number
 
of objects the smaller the overhead of tracking slabs
 
and the less frequently locks need to be acquired.
 
For more information see Documentation/vm/slub.txt.
 
 
slub_min_order= [MM, SLUB]
 
Determines the mininum page order for slabs. Must be
 
lower than slub_max_order.
 
For more information see Documentation/vm/slub.txt.
 
 
slub_nomerge [MM, SLUB]
 
Disable merging of slabs with similar size. May be
 
necessary if there is some reason to distinguish
 
allocs to different slabs. Debug options disable
 
merging on their own.
 
For more information see Documentation/vm/slub.txt.
 
 
smart2= [HW]
 
Format: <io1>[,<io2>[,...,<io8>]]
 
 
smp-alt-once [X86-32,SMP] On a hotplug CPU system, only
 
attempt to substitute SMP alternatives once at boot.
 
 
smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
 
smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
 
smsc-ircc2.ircc_sir= [HW] SIR base I/O port
 
smsc-ircc2.ircc_fir= [HW] FIR base I/O port
 
smsc-ircc2.ircc_irq= [HW] IRQ line
 
smsc-ircc2.ircc_dma= [HW] DMA channel
 
smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
 
0: Toshiba Satellite 1800 (GP data pin select)
 
1: Fast pin select (default)
 
2: ATC IRMode
 
 
snd-ad1816a= [HW,ALSA]
 
 
snd-ad1848= [HW,ALSA]
 
 
snd-ali5451= [HW,ALSA]
 
 
snd-als100= [HW,ALSA]
 
 
snd-als4000= [HW,ALSA]
 
 
snd-azt2320= [HW,ALSA]
 
 
snd-cmi8330= [HW,ALSA]
 
 
snd-cmipci= [HW,ALSA]
 
 
snd-cs4231= [HW,ALSA]
 
 
snd-cs4232= [HW,ALSA]
 
 
snd-cs4236= [HW,ALSA]
 
 
snd-cs4281= [HW,ALSA]
 
 
snd-cs46xx= [HW,ALSA]
 
 
snd-dt019x= [HW,ALSA]
 
 
snd-dummy= [HW,ALSA]
 
 
snd-emu10k1= [HW,ALSA]
 
 
snd-ens1370= [HW,ALSA]
 
 
snd-ens1371= [HW,ALSA]
 
 
snd-es968= [HW,ALSA]
 
 
snd-es1688= [HW,ALSA]
 
 
snd-es18xx= [HW,ALSA]
 
 
snd-es1938= [HW,ALSA]
 
 
snd-es1968= [HW,ALSA]
 
 
snd-fm801= [HW,ALSA]
 
 
snd-gusclassic= [HW,ALSA]
 
 
snd-gusextreme= [HW,ALSA]
 
 
snd-gusmax= [HW,ALSA]
 
 
snd-hdsp= [HW,ALSA]
 
 
snd-ice1712= [HW,ALSA]
 
 
snd-intel8x0= [HW,ALSA]
 
 
snd-interwave= [HW,ALSA]
 
 
snd-interwave-stb=
 
[HW,ALSA]
 
 
snd-korg1212= [HW,ALSA]
 
 
snd-maestro3= [HW,ALSA]
 
 
snd-mpu401= [HW,ALSA]
 
 
snd-mtpav= [HW,ALSA]
 
 
snd-nm256= [HW,ALSA]
 
 
snd-opl3sa2= [HW,ALSA]
 
 
snd-opti92x-ad1848=
 
[HW,ALSA]
 
 
snd-opti92x-cs4231=
 
[HW,ALSA]
 
 
snd-opti93x= [HW,ALSA]
 
 
snd-pmac= [HW,ALSA]
 
 
snd-rme32= [HW,ALSA]
 
 
snd-rme96= [HW,ALSA]
 
 
snd-rme9652= [HW,ALSA]
 
 
snd-sb8= [HW,ALSA]
 
 
snd-sb16= [HW,ALSA]
 
 
snd-sbawe= [HW,ALSA]
 
 
snd-serial= [HW,ALSA]
 
 
snd-sgalaxy= [HW,ALSA]
 
 
snd-sonicvibes= [HW,ALSA]
 
 
snd-sun-amd7930=
 
[HW,ALSA]
 
 
snd-sun-cs4231= [HW,ALSA]
 
 
snd-trident= [HW,ALSA]
 
 
snd-usb-audio= [HW,ALSA,USB]
 
 
snd-via82xx= [HW,ALSA]
 
 
snd-virmidi= [HW,ALSA]
 
 
snd-wavefront= [HW,ALSA]
 
 
snd-ymfpci= [HW,ALSA]
 
 
softlockup_panic=
 
[KNL] Should the soft-lockup detector generate panics.
 
 
sonypi.*= [HW] Sony Programmable I/O Control Device driver
 
See Documentation/sonypi.txt
 
 
specialix= [HW,SERIAL] Specialix multi-serial port adapter
 
See Documentation/serial/specialix.txt.
 
 
spia_io_base= [HW,MTD]
 
spia_fio_base=
 
spia_pedr=
 
spia_peddr=
 
 
sscape= [HW,OSS]
 
Format: <io>,<irq>,<dma>,<mpu_io>,<mpu_irq>
 
 
st= [HW,SCSI] SCSI tape parameters (buffers, etc.)
 
See Documentation/scsi/st.txt.
 
 
stacktrace [FTRACE]
 
Enabled the stack tracer on boot up.
 
 
sti= [PARISC,HW]
 
Format: <num>
 
Set the STI (builtin display/keyboard on the HP-PARISC
 
machines) console (graphic card) which should be used
 
as the initial boot-console.
 
See also comment in drivers/video/console/sticore.c.
 
 
sti_font= [HW]
 
See comment in drivers/video/console/sticore.c.
 
 
stifb= [HW]
 
Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
 
 
sunrpc.pool_mode=
 
[NFS]
 
Control how the NFS server code allocates CPUs to
 
service thread pools.  Depending on how many NICs
 
you have and where their interrupts are bound, this
 
option will affect which CPUs will do NFS serving.
 
Note: this parameter cannot be changed while the
 
NFS server is running.
 
 
auto     the server chooses an appropriate mode
 
    automatically using heuristics
 
global     a single global pool contains all CPUs
 
percpu     one pool for each CPU
 
pernode     one pool for each NUMA node (equivalent
 
    to global on non-NUMA machines)
 
 
swiotlb= [IA-64] Number of I/O TLB slabs
 
 
switches= [HW,M68k]
 
 
sym53c416= [HW,SCSI]
 
See header of drivers/scsi/sym53c416.c.
 
 
sysrq_always_enabled
 
[KNL]
 
Ignore sysrq setting - this boot parameter will
 
neutralize any effect of /proc/sys/kernel/sysrq.
 
Useful for debugging.
 
 
t128= [HW,SCSI]
 
See header of drivers/scsi/t128.c.
 
 
tdfx= [HW,DRM]
 
 
test_suspend= [SUSPEND]
 
Specify "mem" (for Suspend-to-RAM) or "standby" (for
 
standby suspend) as the system sleep state to briefly
 
enter during system startup.  The system is woken from
 
this state using a wakeup-capable RTC alarm.
 
 
thash_entries= [KNL,NET]
 
Set number of hash buckets for TCP connection
 
 
thermal.act= [HW,ACPI]
 
-1: disable all active trip points in all thermal zones
 
<degrees C>: override all lowest active trip points
 
 
thermal.crt= [HW,ACPI]
 
-1: disable all critical trip points in all thermal zones
 
<degrees C>: override all critical trip points
 
 
thermal.nocrt= [HW,ACPI]
 
Set to disable actions on ACPI thermal zone
 
critical and hot trip points.
 
 
thermal.off= [HW,ACPI]
 
1: disable ACPI thermal control
 
 
thermal.psv= [HW,ACPI]
 
-1: disable all passive trip points
 
<degrees C>: override all passive trip points to this
 
value
 
 
thermal.tzp= [HW,ACPI]
 
Specify global default ACPI thermal zone polling rate
 
<deci-seconds>: poll all this frequency
 
0: no polling (default)
 
 
tmscsim= [HW,SCSI]
 
See comment before function dc390_setup() in
 
drivers/scsi/tmscsim.c.
 
 
topology= [S390]
 
Format: {off | on}
 
Specify if the kernel should make use of the cpu
 
topology informations if the hardware supports these.
 
The scheduler will make use of these informations and
 
e.g. base its process migration decisions on it.
 
Default is off.
 
 
tp720= [HW,PS2]
 
 
trix= [HW,OSS] MediaTrix AudioTrix Pro
 
Format:
 
<io>,<irq>,<dma>,<dma2>,<sb_io>,<sb_irq>,<sb_dma>,<mpu_io>,<mpu_irq>
 
 
tsc= Disable clocksource-must-verify flag for TSC.
 
Format: <string>
 
[x86] reliable: mark tsc clocksource as reliable, this
 
disables clocksource verification at runtime.
 
Used to enable high-resolution timer mode on older
 
hardware, and in virtualized environment.
 
 
turbografx.map[2|3]= [HW,JOY]
 
TurboGraFX parallel port interface
 
Format:
 
<port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
 
See also Documentation/input/joystick-parport.txt
 
 
u14-34f= [HW,SCSI] UltraStor 14F/34F SCSI host adapter
 
See header of drivers/scsi/u14-34f.c.
 
 
uart401= [HW,OSS]
 
Format: <io>,<irq>
 
 
uart6850= [HW,OSS]
 
Format: <io>,<irq>
 
 
uhci-hcd.ignore_oc=
 
[USB] Ignore overcurrent events (default N).
 
Some badly-designed motherboards generate lots of
 
bogus events, for ports that aren't wired to
 
anything.  Set this parameter to avoid log spamming.
 
Note that genuine overcurrent events won't be
 
reported either.
 
 
unknown_nmi_panic
 
[X86-32,X86-64]
 
Set unknown_nmi_panic=1 early on boot.
 
 
usbcore.autosuspend=
 
[USB] The autosuspend time delay (in seconds) used
 
for newly-detected USB devices (default 2).  This
 
is the time required before an idle device will be
 
autosuspended.  Devices for which the delay is set
 
to a negative value won't be autosuspended at all.
 
 
usbcore.usbfs_snoop=
 
[USB] Set to log all usbfs traffic (default 0 = off).
 
 
usbcore.blinkenlights=
 
[USB] Set to cycle leds on hubs (default 0 = off).
 
 
usbcore.old_scheme_first=
 
[USB] Start with the old device initialization
 
scheme (default 0 = off).
 
 
usbcore.use_both_schemes=
 
[USB] Try the other device initialization scheme
 
if the first one fails (default 1 = enabled).
 
 
usbcore.initial_descriptor_timeout=
 
[USB] Specifies timeout for the initial 64-byte
 
                        USB_REQ_GET_DESCRIPTOR request in milliseconds
 
(default 5000 = 5.0 seconds).
 
 
usbhid.mousepoll=
 
[USBHID] The interval which mice are to be polled at.
 
 
usb-storage.delay_use=
 
[UMS] The delay in seconds before a new device is
 
scanned for Logical Units (default 5).
 
 
usb-storage.quirks=
 
[UMS] A list of quirks entries to supplement or
 
override the built-in unusual_devs list.  List
 
entries are separated by commas.  Each entry has
 
the form VID:PID:Flags where VID and PID are Vendor
 
and Product ID values (4-digit hex numbers) and
 
Flags is a set of characters, each corresponding
 
to a common usb-storage quirk flag as follows:
 
a = SANE_SENSE (collect more than 18 bytes
 
of sense data);
 
c = FIX_CAPACITY (decrease the reported
 
device capacity by one sector);
 
h = CAPACITY_HEURISTICS (decrease the
 
reported device capacity by one
 
sector if the number is odd);
 
i = IGNORE_DEVICE (don't bind to this
 
device);
 
l = NOT_LOCKABLE (don't try to lock and
 
unlock ejectable media);
 
m = MAX_SECTORS_64 (don't transfer more
 
than 64 sectors = 32 KB at a time);
 
o = CAPACITY_OK (accept the capacity
 
reported by the device);
 
r = IGNORE_RESIDUE (the device reports
 
bogus residue values);
 
s = SINGLE_LUN (the device has only one
 
Logical Unit);
 
w = NO_WP_DETECT (don't test whether the
 
medium is write-protected).
 
Example: quirks=0419:aaf5:rl,0421:0433:rc
 
 
add_efi_memmap [EFI; x86-32,X86-64] Include EFI memory map in
 
kernel's map of available physical RAM.
 
 
vdso= [X86-32,SH,x86-64]
 
vdso=2: enable compat VDSO (default with COMPAT_VDSO)
 
vdso=1: enable VDSO (default)
 
vdso=0: disable VDSO mapping
 
 
vdso32= [X86-32,X86-64]
 
vdso32=2: enable compat VDSO (default with COMPAT_VDSO)
 
vdso32=1: enable 32-bit VDSO (default)
 
vdso32=0: disable 32-bit VDSO mapping
 
 
vector= [IA-64,SMP]
 
vector=percpu: enable percpu vector domain
 
 
video= [FB] Frame buffer configuration
 
See Documentation/fb/modedb.txt.
 
 
vga= [BOOT,X86-32] Select a particular video mode
 
See Documentation/x86/boot.txt and
 
Documentation/svga.txt.
 
Use vga=ask for menu.
 
This is actually a boot loader parameter; the value is
 
passed to the kernel using a special protocol.
 
 
vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
 
size of <nn>. This can be used to increase the
 
minimum size (128MB on x86). It can also be used to
 
decrease the size and leave more room for directly
 
mapped kernel RAM.
 
 
vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
 
Format: <command>
 
 
vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
 
Format: <command>
 
 
vmpoff= [KNL,S390] Perform z/VM CP command after power off.
 
Format: <command>
 
 
waveartist= [HW,OSS]
 
Format: <io>,<irq>,<dma>,<dma2>
 
 
wd33c93= [HW,SCSI]
 
See header of drivers/scsi/wd33c93.c.
 
 
wd7000= [HW,SCSI]
 
See header of drivers/scsi/wd7000.c.
 
 
wdt= [WDT] Watchdog
 
See Documentation/watchdog/wdt.txt.
 
 
xd= [HW,XT] Original XT pre-IDE (RLL encoded) disks.
 
xd_geo= See header of drivers/block/xd.c.
 
 
xirc2ps_cs= [NET,PCMCIA]
 
Format:
 
<irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
 
 
norandmaps Don't use address space randomization.  Equivalent to
 
echo 0 > /proc/sys/kernel/randomize_va_space
 
 
______________________________________________________________________
 
 
TODO:
 
 
Add documentation for ALSA options.
 
Add more DRM drivers.
 

Latest revision as of 19:27, 20 June 2009