source: src/linux/universal/linux-4.9/Documentation/kernel-parameters.txt @ 31859

Last change on this file since 31859 was 31859, checked in by brainslayer, 2 months ago

kernel update

File size: 160.7 KB
1                          Kernel Parameters
2                          ~~~~~~~~~~~~~~~~~
4The following is a consolidated list of the kernel parameters as
5implemented by the __setup(), core_param() and module_param() macros
6and sorted into English Dictionary order (defined as ignoring all
7punctuation and sorting digits before letters in a case insensitive
8manner), and with descriptions where known.
10The kernel parses parameters from the kernel command line up to "--";
11if it doesn't recognize a parameter and it doesn't contain a '.', the
12parameter gets passed to init: parameters with '=' go into init's
13environment, others are passed as command line arguments to init.
14Everything after "--" is passed as an argument to init.
16Module parameters can be specified in two ways: via the kernel command
17line with a module name prefix, or via modprobe, e.g.:
19        (kernel command line) usbcore.blinkenlights=1
20        (modprobe command line) modprobe usbcore blinkenlights=1
22Parameters for modules which are built into the kernel need to be
23specified on the kernel command line.  modprobe looks through the
24kernel command line (/proc/cmdline) and collects module parameters
25when it loads a module, so the kernel command line can be used for
26loadable modules too.
28Hyphens (dashes) and underscores are equivalent in parameter names, so
29        log_buf_len=1M print-fatal-signals=1
30can also be entered as
31        log-buf-len=1M print_fatal_signals=1
33Double-quotes can be used to protect spaces in values, e.g.:
34        param="spaces in here"
36cpu lists:
39Some kernel parameters take a list of CPUs as a value, e.g.  isolcpus,
40nohz_full, irqaffinity, rcu_nocbs.  The format of this list is:
42        <cpu number>,...,<cpu number>
46        <cpu number>-<cpu number>
47        (must be a positive range in ascending order)
49or a mixture
51<cpu number>,...,<cpu number>-<cpu number>
53Note that for the special case of a range one can split the range into equal
54sized groups and for each group use some amount from the beginning of that
57        <cpu number>-cpu number>:<used size>/<group size>
59For example one can add to the command line following parameter:
61        isolcpus=1,2,10-20,100-2000:2/25
63where the final item represents CPUs 100,101,125,126,150,151,...
67This document may not be entirely up to date and comprehensive. The command
68"modinfo -p ${modulename}" shows a current list of all parameters of a loadable
69module. Loadable modules, after being loaded into the running kernel, also
70reveal their parameters in /sys/module/${modulename}/parameters/. Some of these
71parameters may be changed at runtime by the command
72"echo -n ${value} > /sys/module/${modulename}/parameters/${parm}".
74The parameters listed below are only valid if certain kernel build options were
75enabled and if respective hardware is present. The text in square brackets at
76the beginning of each description states the restrictions within which a
77parameter is applicable:
79        ACPI    ACPI support is enabled.
80        AGP     AGP (Accelerated Graphics Port) is enabled.
81        ALSA    ALSA sound support is enabled.
82        APIC    APIC support is enabled.
83        APM     Advanced Power Management support is enabled.
84        ARM     ARM architecture is enabled.
85        AVR32   AVR32 architecture is enabled.
86        AX25    Appropriate AX.25 support is enabled.
87        BLACKFIN Blackfin architecture is enabled.
88        CLK     Common clock infrastructure is enabled.
89        CMA     Contiguous Memory Area support is enabled.
90        DRM     Direct Rendering Management support is enabled.
91        DYNAMIC_DEBUG Build in debug messages and enable them at runtime
92        EDD     BIOS Enhanced Disk Drive Services (EDD) is enabled
93        EFI     EFI Partitioning (GPT) is enabled
94        EIDE    EIDE/ATAPI support is enabled.
95        EVM     Extended Verification Module
96        FB      The frame buffer device is enabled.
97        FTRACE  Function tracing enabled.
98        GCOV    GCOV profiling is enabled.
99        HW      Appropriate hardware is enabled.
100        IA-64   IA-64 architecture is enabled.
101        IMA     Integrity measurement architecture is enabled.
102        IOSCHED More than one I/O scheduler is enabled.
103        IP_PNP  IP DHCP, BOOTP, or RARP is enabled.
104        IPV6    IPv6 support is enabled.
105        ISAPNP  ISA PnP code is enabled.
106        ISDN    Appropriate ISDN support is enabled.
107        JOY     Appropriate joystick support is enabled.
108        KGDB    Kernel debugger support is enabled.
109        KVM     Kernel Virtual Machine support is enabled.
110        LIBATA  Libata driver is enabled
111        LP      Printer support is enabled.
112        LOOP    Loopback device support is enabled.
113        M68k    M68k architecture is enabled.
114                        These options have more detailed description inside of
115                        Documentation/m68k/kernel-options.txt.
116        MDA     MDA console support is enabled.
117        MIPS    MIPS architecture is enabled.
118        MOUSE   Appropriate mouse support is enabled.
119        MSI     Message Signaled Interrupts (PCI).
120        MTD     MTD (Memory Technology Device) support is enabled.
121        NET     Appropriate network support is enabled.
122        NUMA    NUMA support is enabled.
123        NFS     Appropriate NFS support is enabled.
124        OSS     OSS sound support is enabled.
125        PV_OPS  A paravirtualized kernel is enabled.
126        PARIDE  The ParIDE (parallel port IDE) subsystem is enabled.
127        PARISC  The PA-RISC architecture is enabled.
128        PCI     PCI bus support is enabled.
129        PCIE    PCI Express support is enabled.
130        PCMCIA  The PCMCIA subsystem is enabled.
131        PNP     Plug & Play support is enabled.
132        PPC     PowerPC architecture is enabled.
133        PPT     Parallel port support is enabled.
134        PS2     Appropriate PS/2 support is enabled.
135        RAM     RAM disk support is enabled.
136        S390    S390 architecture is enabled.
137        SCSI    Appropriate SCSI support is enabled.
138                        A lot of drivers have their options described inside
139                        the Documentation/scsi/ sub-directory.
140        SECURITY Different security models are enabled.
141        SELINUX SELinux support is enabled.
142        APPARMOR AppArmor support is enabled.
143        SERIAL  Serial support is enabled.
144        SH      SuperH architecture is enabled.
145        SMP     The kernel is an SMP kernel.
146        SPARC   Sparc architecture is enabled.
147        SWSUSP  Software suspend (hibernation) is enabled.
148        SUSPEND System suspend states are enabled.
149        TPM     TPM drivers are enabled.
150        TS      Appropriate touchscreen support is enabled.
151        UMS     USB Mass Storage support is enabled.
152        USB     USB support is enabled.
153        USBHID  USB Human Interface Device support is enabled.
154        V4L     Video For Linux support is enabled.
155        VMMIO   Driver for memory mapped virtio devices is enabled.
156        VGA     The VGA console has been enabled.
157        VT      Virtual terminal support is enabled.
158        WDT     Watchdog support is enabled.
159        XT      IBM PC/XT MFM hard disk support is enabled.
160        X86-32  X86-32, aka i386 architecture is enabled.
161        X86-64  X86-64 architecture is enabled.
162                        More X86-64 boot options can be found in
163                        Documentation/x86/x86_64/boot-options.txt .
164        X86     Either 32-bit or 64-bit x86 (same as X86-32+X86-64)
165        X86_UV  SGI UV support is enabled.
166        XEN     Xen support is enabled
168In addition, the following text indicates that the option:
170        BUGS=   Relates to possible processor bugs on the said processor.
171        KNL     Is a kernel start-up parameter.
172        BOOT    Is a boot loader parameter.
174Parameters denoted with BOOT are actually interpreted by the boot
175loader, and have no meaning to the kernel directly.
176Do not modify the syntax of boot loader parameters without extreme
177need or coordination with <Documentation/x86/boot.txt>.
179There are also arch-specific kernel-parameters not documented here.
180See for example <Documentation/x86/x86_64/boot-options.txt>.
182Note that ALL kernel parameters listed below are CASE SENSITIVE, and that
183a trailing = on the name of any parameter states that that parameter will
184be entered as an environment variable, whereas its absence indicates that
185it will appear as a kernel argument readable via /proc/cmdline by programs
186running once the system is up.
188The number of kernel parameters is not limited, but the length of the
189complete command line (parameters including spaces etc.) is limited to
190a fixed number of characters. This limit depends on the architecture
191and is between 256 and 4096 characters. It is defined in the file
192./include/asm/setup.h as COMMAND_LINE_SIZE.
194Finally, the [KMG] suffix is commonly described after a number of kernel
195parameter values. These 'K', 'M', and 'G' letters represent the _binary_
196multipliers 'Kilo', 'Mega', and 'Giga', equalling 2^10, 2^20, and 2^30
197bytes respectively. Such letter suffixes can also be entirely omitted.
200        acpi=           [HW,ACPI,X86,ARM64]
201                        Advanced Configuration and Power Interface
202                        Format: { force | on | off | strict | noirq | rsdt |
203                                  copy_dsdt }
204                        force -- enable ACPI if default was off
205                        on -- enable ACPI but allow fallback to DT [arm64]
206                        off -- disable ACPI if default was on
207                        noirq -- do not use ACPI for IRQ routing
208                        strict -- Be less tolerant of platforms that are not
209                                strictly ACPI specification compliant.
210                        rsdt -- prefer RSDT over (default) XSDT
211                        copy_dsdt -- copy DSDT to memory
212                        For ARM64, ONLY "acpi=off", "acpi=on" or "acpi=force"
213                        are available
215                        See also Documentation/power/runtime_pm.txt, pci=noacpi
217        acpi_apic_instance=     [ACPI, IOAPIC]
218                        Format: <int>
219                        2: use 2nd APIC table, if available
220                        1,0: use 1st APIC table
221                        default: 0
223        acpi_backlight= [HW,ACPI]
224                        acpi_backlight=vendor
225                        acpi_backlight=video
226                        If set to vendor, prefer vendor specific driver
227                        (e.g. thinkpad_acpi, sony_acpi, etc.) instead
228                        of the ACPI video.ko driver.
230        acpi_force_32bit_fadt_addr
231                        force FADT to use 32 bit addresses rather than the
232                        64 bit X_* addresses. Some firmware have broken 64
233                        bit addresses for force ACPI ignore these and use
234                        the older legacy 32 bit addresses.
236        acpica_no_return_repair [HW, ACPI]
237                        Disable AML predefined validation mechanism
238                        This mechanism can repair the evaluation result to make
239                        the return objects more ACPI specification compliant.
240                        This option is useful for developers to identify the
241                        root cause of an AML interpreter issue when the issue
242                        has something to do with the repair mechanism.
244        acpi.debug_layer=       [HW,ACPI,ACPI_DEBUG]
245        acpi.debug_level=       [HW,ACPI,ACPI_DEBUG]
246                        Format: <int>
247                        CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
248                        debug output.  Bits in debug_layer correspond to a
249                        _COMPONENT in an ACPI source file, e.g.,
250                            #define _COMPONENT ACPI_PCI_COMPONENT
251                        Bits in debug_level correspond to a level in
252                        ACPI_DEBUG_PRINT statements, e.g.,
253                            ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
254                        The debug_level mask defaults to "info".  See
255                        Documentation/acpi/debug.txt for more information about
256                        debug layers and levels.
258                        Enable processor driver info messages:
259                            acpi.debug_layer=0x20000000
260                        Enable PCI/PCI interrupt routing info messages:
261                            acpi.debug_layer=0x400000
262                        Enable AML "Debug" output, i.e., stores to the Debug
263                        object while interpreting AML:
264                            acpi.debug_layer=0xffffffff acpi.debug_level=0x2
265                        Enable all messages related to ACPI hardware:
266                            acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
268                        Some values produce so much output that the system is
269                        unusable.  The "log_buf_len" parameter may be useful
270                        if you need to capture more output.
272        acpi_enforce_resources= [ACPI]
273                        { strict | lax | no }
274                        Check for resource conflicts between native drivers
275                        and ACPI OperationRegions (SystemIO and SystemMemory
276                        only). IO ports and memory declared in ACPI might be
277                        used by the ACPI subsystem in arbitrary AML code and
278                        can interfere with legacy drivers.
279                        strict (default): access to resources claimed by ACPI
280                        is denied; legacy drivers trying to access reserved
281                        resources will fail to bind to device using them.
282                        lax: access to resources claimed by ACPI is allowed;
283                        legacy drivers trying to access reserved resources
284                        will bind successfully but a warning message is logged.
285                        no: ACPI OperationRegions are not marked as reserved,
286                        no further checks are performed.
288        acpi_force_table_verification   [HW,ACPI]
289                        Enable table checksum verification during early stage.
290                        By default, this is disabled due to x86 early mapping
291                        size limitation.
293        acpi_irq_balance [HW,ACPI]
294                        ACPI will balance active IRQs
295                        default in APIC mode
297        acpi_irq_nobalance [HW,ACPI]
298                        ACPI will not move active IRQs (default)
299                        default in PIC mode
301        acpi_irq_isa=   [HW,ACPI] If irq_balance, mark listed IRQs used by ISA
302                        Format: <irq>,<irq>...
304        acpi_irq_pci=   [HW,ACPI] If irq_balance, clear listed IRQs for
305                        use by PCI
306                        Format: <irq>,<irq>...
308        acpi_mask_gpe=  [HW,ACPI]
309                        Due to the existence of _Lxx/_Exx, some GPEs triggered
310                        by unsupported hardware/firmware features can result in
311                        GPE floodings that cannot be automatically disabled by
312                        the GPE dispatcher.
313                        This facility can be used to prevent such uncontrolled
314                        GPE floodings.
315                        Format: <int>
316                        Support masking of GPEs numbered from 0x00 to 0x7f.
318        acpi_no_auto_serialize  [HW,ACPI]
319                        Disable auto-serialization of AML methods
320                        AML control methods that contain the opcodes to create
321                        named objects will be marked as "Serialized" by the
322                        auto-serialization feature.
323                        This feature is enabled by default.
324                        This option allows to turn off the feature.
326        acpi_no_memhotplug [ACPI] Disable memory hotplug.  Useful for kdump
327                           kernels.
329        acpi_no_static_ssdt     [HW,ACPI]
330                        Disable installation of static SSDTs at early boot time
331                        By default, SSDTs contained in the RSDT/XSDT will be
332                        installed automatically and they will appear under
333                        /sys/firmware/acpi/tables.
334                        This option turns off this feature.
335                        Note that specifying this option does not affect
336                        dynamic table installation which will install SSDT
337                        tables to /sys/firmware/acpi/tables/dynamic.
339        acpi_rsdp=      [ACPI,EFI,KEXEC]
340                        Pass the RSDP address to the kernel, mostly used
341                        on machines running EFI runtime service to boot the
342                        second kernel for kdump.
344        acpi_os_name=   [HW,ACPI] Tell ACPI BIOS the name of the OS
345                        Format: To spoof as Windows 98: ="Microsoft Windows"
347        acpi_rev_override [ACPI] Override the _REV object to return 5 (instead
348                        of 2 which is mandated by ACPI 6) as the supported ACPI
349                        specification revision (when using this switch, it may
350                        be necessary to carry out a cold reboot _twice_ in a
351                        row to make it take effect on the platform firmware).
353        acpi_osi=       [HW,ACPI] Modify list of supported OS interface strings
354                        acpi_osi="string1"      # add string1
355                        acpi_osi="!string2"     # remove string2
356                        acpi_osi=!*             # remove all strings
357                        acpi_osi=!              # disable all built-in OS vendor
358                                                  strings
359                        acpi_osi=!!             # enable all built-in OS vendor
360                                                  strings
361                        acpi_osi=               # disable all strings
363                        'acpi_osi=!' can be used in combination with single or
364                        multiple 'acpi_osi="string1"' to support specific OS
365                        vendor string(s).  Note that such command can only
366                        affect the default state of the OS vendor strings, thus
367                        it cannot affect the default state of the feature group
368                        strings and the current state of the OS vendor strings,
369                        specifying it multiple times through kernel command line
370                        is meaningless.  This command is useful when one do not
371                        care about the state of the feature group strings which
372                        should be controlled by the OSPM.
373                        Examples:
374                          1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
375                             to 'acpi_osi="Windows 2000" acpi_osi=!', they all
376                             can make '_OSI("Windows 2000")' TRUE.
378                        'acpi_osi=' cannot be used in combination with other
379                        'acpi_osi=' command lines, the _OSI method will not
380                        exist in the ACPI namespace.  NOTE that such command can
381                        only affect the _OSI support state, thus specifying it
382                        multiple times through kernel command line is also
383                        meaningless.
384                        Examples:
385                          1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
386                             FALSE.
388                        'acpi_osi=!*' can be used in combination with single or
389                        multiple 'acpi_osi="string1"' to support specific
390                        string(s).  Note that such command can affect the
391                        current state of both the OS vendor strings and the
392                        feature group strings, thus specifying it multiple times
393                        through kernel command line is meaningful.  But it may
394                        still not able to affect the final state of a string if
395                        there are quirks related to this string.  This command
396                        is useful when one want to control the state of the
397                        feature group strings to debug BIOS issues related to
398                        the OSPM features.
399                        Examples:
400                          1. 'acpi_osi="Module Device" acpi_osi=!*' can make
401                             '_OSI("Module Device")' FALSE.
402                          2. 'acpi_osi=!* acpi_osi="Module Device"' can make
403                             '_OSI("Module Device")' TRUE.
404                          3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
405                             equivalent to
406                             'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
407                             and
408                             'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
409                             they all will make '_OSI("Windows 2000")' TRUE.
411        acpi_pm_good    [X86]
412                        Override the pmtimer bug detection: force the kernel
413                        to assume that this machine's pmtimer latches its value
414                        and always returns good values.
416        acpi_sci=       [HW,ACPI] ACPI System Control Interrupt trigger mode
417                        Format: { level | edge | high | low }
419        acpi_skip_timer_override [HW,ACPI]
420                        Recognize and ignore IRQ0/pin2 Interrupt Override.
421                        For broken nForce2 BIOS resulting in XT-PIC timer.
423        acpi_sleep=     [HW,ACPI] Sleep options
424                        Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
425                                  old_ordering, nonvs, sci_force_enable }
426                        See Documentation/power/video.txt for information on
427                        s3_bios and s3_mode.
428                        s3_beep is for debugging; it makes the PC's speaker beep
429                        as soon as the kernel's real-mode entry point is called.
430                        s4_nohwsig prevents ACPI hardware signature from being
431                        used during resume from hibernation.
432                        old_ordering causes the ACPI 1.0 ordering of the _PTS
433                        control method, with respect to putting devices into
434                        low power states, to be enforced (the ACPI 2.0 ordering
435                        of _PTS is used by default).
436                        nonvs prevents the kernel from saving/restoring the
437                        ACPI NVS memory during suspend/hibernation and resume.
438                        sci_force_enable causes the kernel to set SCI_EN directly
439                        on resume from S1/S3 (which is against the ACPI spec,
440                        but some broken systems don't work without it).
442        acpi_use_timer_override [HW,ACPI]
443                        Use timer override. For some broken Nvidia NF5 boards
444                        that require a timer override, but don't have HPET
446        add_efi_memmap  [EFI; X86] Include EFI memory map in
447                        kernel's map of available physical RAM.
449        agp=            [AGP]
450                        { off | try_unsupported }
451                        off: disable AGP support
452                        try_unsupported: try to drive unsupported chipsets
453                                (may crash computer or cause data corruption)
455        ALSA            [HW,ALSA]
456                        See Documentation/sound/alsa/alsa-parameters.txt
458        alignment=      [KNL,ARM]
459                        Allow the default userspace alignment fault handler
460                        behaviour to be specified.  Bit 0 enables warnings,
461                        bit 1 enables fixups, and bit 2 sends a segfault.
463        align_va_addr=  [X86-64]
464                        Align virtual addresses by clearing slice [14:12] when
465                        allocating a VMA at process creation time. This option
466                        gives you up to 3% performance improvement on AMD F15h
467                        machines (where it is enabled by default) for a
468                        CPU-intensive style benchmark, and it can vary highly in
469                        a microbenchmark depending on workload and compiler.
471                        32: only for 32-bit processes
472                        64: only for 64-bit processes
473                        on: enable for both 32- and 64-bit processes
474                        off: disable for both 32- and 64-bit processes
476        alloc_snapshot  [FTRACE]
477                        Allocate the ftrace snapshot buffer on boot up when the
478                        main buffer is allocated. This is handy if debugging
479                        and you need to use tracing_snapshot() on boot up, and
480                        do not want to use tracing_snapshot_alloc() as it needs
481                        to be done where GFP_KERNEL allocations are allowed.
483        amd_iommu=      [HW,X86-64]
484                        Pass parameters to the AMD IOMMU driver in the system.
485                        Possible values are:
486                        fullflush - enable flushing of IO/TLB entries when
487                                    they are unmapped. Otherwise they are
488                                    flushed before they will be reused, which
489                                    is a lot of faster
490                        off       - do not initialize any AMD IOMMU found in
491                                    the system
492                        force_isolation - Force device isolation for all
493                                          devices. The IOMMU driver is not
494                                          allowed anymore to lift isolation
495                                          requirements as needed. This option
496                                          does not override iommu=pt
498        amd_iommu_dump= [HW,X86-64]
499                        Enable AMD IOMMU driver option to dump the ACPI table
500                        for AMD IOMMU. With this option enabled, AMD IOMMU
501                        driver will print ACPI tables for AMD IOMMU during
502                        IOMMU initialization.
504        amd_iommu_intr= [HW,X86-64]
505                        Specifies one of the following AMD IOMMU interrupt
506                        remapping modes:
507                        legacy     - Use legacy interrupt remapping mode.
508                        vapic      - Use virtual APIC mode, which allows IOMMU
509                                     to inject interrupts directly into guest.
510                                     This mode requires kvm-amd.avic=1.
511                                     (Default when IOMMU HW support is present.)
513     [HW,JOY] Amiga joystick support
514                        Map of devices attached to JOY0DAT and JOY1DAT
515                        Format: <a>,<b>
516                        See also Documentation/input/joystick.txt
518     [HW,JOY] Analog joystick and gamepad support
519                        Specifies type or capabilities of an analog joystick
520                        connected to one of 16 gameports
521                        Format: <type1>,<type2>,..<type16>
523        apc=            [HW,SPARC]
524                        Power management functions (SPARCstation-4/5 + deriv.)
525                        Format: noidle
526                        Disable APC CPU standby support. SPARCstation-Fox does
527                        not play well with APC CPU idle - disable it if you have
528                        APC and your system crashes randomly.
530        apic=           [APIC,X86-32] Advanced Programmable Interrupt Controller
531                        Change the output verbosity whilst booting
532                        Format: { quiet (default) | verbose | debug }
533                        Change the amount of debugging information output
534                        when initialising the APIC and IO-APIC components.
536        apic_extnmi=    [APIC,X86] External NMI delivery setting
537                        Format: { bsp (default) | all | none }
538                        bsp:  External NMI is delivered only to CPU 0
539                        all:  External NMIs are broadcast to all CPUs as a
540                              backup of CPU 0
541                        none: External NMI is masked for all CPUs. This is
542                              useful so that a dump capture kernel won't be
543                              shot down by NMI
545        autoconf=       [IPV6]
546                        See Documentation/networking/ipv6.txt.
548        show_lapic=     [APIC,X86] Advanced Programmable Interrupt Controller
549                        Limit apic dumping. The parameter defines the maximal
550                        number of local apics being dumped. Also it is possible
551                        to set it to "all" by meaning -- no limit here.
552                        Format: { 1 (default) | 2 | ... | all }.
553                        The parameter valid if only apic=debug or
554                        apic=verbose is specified.
555                        Example: apic=debug show_lapic=all
557        apm=            [APM] Advanced Power Management
558                        See header of arch/x86/kernel/apm_32.c.
560        arcrimi=        [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
561                        Format: <io>,<irq>,<nodeID>
563        ataflop=        [HW,M68k]
565        atarimouse=     [HW,MOUSE] Atari Mouse
567        atkbd.extra=    [HW] Enable extra LEDs and keys on IBM RapidAccess,
568                        EzKey and similar keyboards
570        atkbd.reset=    [HW] Reset keyboard during initialization
572        atkbd.set=      [HW] Select keyboard code set
573                        Format: <int> (2 = AT (default), 3 = PS/2)
575        atkbd.scroll=   [HW] Enable scroll wheel on MS Office and similar
576                        keyboards
578        atkbd.softraw=  [HW] Choose between synthetic and real raw mode
579                        Format: <bool> (0 = real, 1 = synthetic (default))
581        atkbd.softrepeat= [HW]
582                        Use software keyboard repeat
584        audit=          [KNL] Enable the audit sub-system
585                        Format: { "0" | "1" } (0 = disabled, 1 = enabled)
586                        0 - kernel audit is disabled and can not be enabled
587                            until the next reboot
588                        unset - kernel audit is initialized but disabled and
589                            will be fully enabled by the userspace auditd.
590                        1 - kernel audit is initialized and partially enabled,
591                            storing at most audit_backlog_limit messages in
592                            RAM until it is fully enabled by the userspace
593                            auditd.
594                        Default: unset
596        audit_backlog_limit= [KNL] Set the audit queue size limit.
597                        Format: <int> (must be >=0)
598                        Default: 64
600        bau=            [X86_UV] Enable the BAU on SGI UV.  The default
601                        behavior is to disable the BAU (i.e. bau=0).
602                        Format: { "0" | "1" }
603                        0 - Disable the BAU.
604                        1 - Enable the BAU.
605                        unset - Disable the BAU.
607        baycom_epp=     [HW,AX25]
608                        Format: <io>,<mode>
610        baycom_par=     [HW,AX25] BayCom Parallel Port AX.25 Modem
611                        Format: <io>,<mode>
612                        See header of drivers/net/hamradio/baycom_par.c.
614        baycom_ser_fdx= [HW,AX25]
615                        BayCom Serial Port AX.25 Modem (Full Duplex Mode)
616                        Format: <io>,<irq>,<mode>[,<baud>]
617                        See header of drivers/net/hamradio/baycom_ser_fdx.c.
619        baycom_ser_hdx= [HW,AX25]
620                        BayCom Serial Port AX.25 Modem (Half Duplex Mode)
621                        Format: <io>,<irq>,<mode>
622                        See header of drivers/net/hamradio/baycom_ser_hdx.c.
624        blkdevparts=    Manual partition parsing of block device(s) for
625                        embedded devices based on command line input.
626                        See Documentation/block/cmdline-partition.txt
628        boot_delay=     Milliseconds to delay each printk during boot.
629                        Values larger than 10 seconds (10000) are changed to
630                        no delay (0).
631                        Format: integer
633        bootmem_debug   [KNL] Enable bootmem allocator debug messages.
635        bert_disable    [ACPI]
636                        Disable BERT OS support on buggy BIOSes.
638        bttv.card=      [HW,V4L] bttv (bt848 + bt878 based grabber cards)
639     Most important insmod options are available as
640                        kernel args too.
641        bttv.pll=       See Documentation/video4linux/bttv/Insmod-options
642        bttv.tuner=
644        bulk_remove=off [PPC]  This parameter disables the use of the pSeries
645                        firmware feature for flushing multiple hpte entries
646                        at a time.
648        c101=           [NET] Moxa C101 synchronous serial card
650        cachesize=      [BUGS=X86-32] Override level 2 CPU cache size detection.
651                        Sometimes CPU hardware bugs make them report the cache
652                        size incorrectly. The kernel will attempt work arounds
653                        to fix known problems, but for some CPUs it is not
654                        possible to determine what the correct size should be.
655                        This option provides an override for these situations.
657        ca_keys=        [KEYS] This parameter identifies a specific key(s) on
658                        the system trusted keyring to be used for certificate
659                        trust validation.
660                        format: { id:<keyid> | builtin }
662        cca=            [MIPS] Override the kernel pages' cache coherency
663                        algorithm.  Accepted values range from 0 to 7
664                        inclusive. See arch/mips/include/asm/pgtable-bits.h
665                        for platform specific values (SB1, Loongson3 and
666                        others).
668        ccw_timeout_log [S390]
669                        See Documentation/s390/CommonIO for details.
671        cgroup_disable= [KNL] Disable a particular controller
672                        Format: {name of the controller(s) to disable}
673                        The effects of cgroup_disable=foo are:
674                        - foo isn't auto-mounted if you mount all cgroups in
675                          a single hierarchy
676                        - foo isn't visible as an individually mountable
677                          subsystem
678                        {Currently only "memory" controller deal with this and
679                        cut the overhead, others just disable the usage. So
680                        only cgroup_disable=memory is actually worthy}
682        cgroup_no_v1=   [KNL] Disable one, multiple, all cgroup controllers in v1
683                        Format: { controller[,controller...] | "all" }
684                        Like cgroup_disable, but only applies to cgroup v1;
685                        the blacklisted controllers remain available in cgroup2.
687        cgroup.memory=  [KNL] Pass options to the cgroup memory controller.
688                        Format: <string>
689                        nosocket -- Disable socket memory accounting.
690                        nokmem -- Disable kernel memory accounting.
692        checkreqprot    [SELINUX] Set initial checkreqprot flag value.
693                        Format: { "0" | "1" }
694                        See security/selinux/Kconfig help text.
695                        0 -- check protection applied by kernel (includes
696                                any implied execute protection).
697                        1 -- check protection requested by application.
698                        Default value is set via a kernel config option.
699                        Value can be changed at runtime via
700                                /selinux/checkreqprot.
702        cio_ignore=     [S390]
703                        See Documentation/s390/CommonIO for details.
704        clk_ignore_unused
705                        [CLK]
706                        Prevents the clock framework from automatically gating
707                        clocks that have not been explicitly enabled by a Linux
708                        device driver but are enabled in hardware at reset or
709                        by the bootloader/firmware. Note that this does not
710                        force such clocks to be always-on nor does it reserve
711                        those clocks in any way. This parameter is useful for
712                        debug and development, but should not be needed on a
713                        platform with proper driver support.  For more
714                        information, see Documentation/clk.txt.
716        clock=          [BUGS=X86-32, HW] gettimeofday clocksource override.
717                        [Deprecated]
718                        Forces specified clocksource (if available) to be used
719                        when calculating gettimeofday(). If specified
720                        clocksource is not available, it defaults to PIT.
721                        Format: { pit | tsc | cyclone | pmtmr }
723        clocksource=    Override the default clocksource
724                        Format: <string>
725                        Override the default clocksource and use the clocksource
726                        with the name specified.
727                        Some clocksource names to choose from, depending on
728                        the platform:
729                        [all] jiffies (this is the base, fallback clocksource)
730                        [ACPI] acpi_pm
731                        [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
732                                pxa_timer,timer3,32k_counter,timer0_1
733                        [AVR32] avr32
734                        [X86-32] pit,hpet,tsc;
735                                scx200_hrt on Geode; cyclone on IBM x440
736                        [MIPS] MIPS
737                        [PARISC] cr16
738                        [S390] tod
739                        [SH] SuperH
740                        [SPARC64] tick
741                        [X86-64] hpet,tsc
743        clocksource.arm_arch_timer.evtstrm=
744                        [ARM,ARM64]
745                        Format: <bool>
746                        Enable/disable the eventstream feature of the ARM
747                        architected timer so that code using WFE-based polling
748                        loops can be debugged more effectively on production
749                        systems.
751        clocksource.arm_arch_timer.fsl-a008585=
752                        [ARM64]
753                        Format: <bool>
754                        Enable/disable the workaround of Freescale/NXP
755                        erratum A-008585.  This can be useful for KVM
756                        guests, if the guest device tree doesn't show the
757                        erratum.  If unspecified, the workaround is
758                        enabled based on the device tree.
760        clearcpuid=BITNUM [X86]
761                        Disable CPUID feature X for the kernel. See
762                        arch/x86/include/asm/cpufeatures.h for the valid bit
763                        numbers. Note the Linux specific bits are not necessarily
764                        stable over kernel options, but the vendor specific
765                        ones should be.
766                        Also note that user programs calling CPUID directly
767                        or using the feature without checking anything
768                        will still see it. This just prevents it from
769                        being used by the kernel or shown in /proc/cpuinfo.
770                        Also note the kernel might malfunction if you disable
771                        some critical bits.
773        cma=nn[MG]@[start[MG][-end[MG]]]
774                        [ARM,X86,KNL]
775                        Sets the size of kernel global memory area for
776                        contiguous memory allocations and optionally the
777                        placement constraint by the physical address range of
778                        memory allocations. A value of 0 disables CMA
779                        altogether. For more information, see
780                        include/linux/dma-contiguous.h
782        cmo_free_hint=  [PPC] Format: { yes | no }
783                        Specify whether pages are marked as being inactive
784                        when they are freed.  This is used in CMO environments
785                        to determine OS memory pressure for page stealing by
786                        a hypervisor.
787                        Default: yes
789        coherent_pool=nn[KMG]   [ARM,KNL]
790                        Sets the size of memory pool for coherent, atomic dma
791                        allocations, by default set to 256K.
793        code_bytes      [X86] How many bytes of object code to print
794                        in an oops report.
795                        Range: 0 - 8192
796                        Default: 64
798        com20020=       [HW,NET] ARCnet - COM20020 chipset
799                        Format:
800                        <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
802        com90io=        [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
803                        Format: <io>[,<irq>]
805        com90xx=        [HW,NET]
806                        ARCnet - COM90xx chipset (memory-mapped buffers)
807                        Format: <io>[,<irq>[,<memstart>]]
809        condev=         [HW,S390] console device
810        conmode=
812        console=        [KNL] Output console device and options.
814                tty<n>  Use the virtual console device <n>.
816                ttyS<n>[,options]
817                ttyUSB0[,options]
818                        Use the specified serial port.  The options are of
819                        the form "bbbbpnf", where "bbbb" is the baud rate,
820                        "p" is parity ("n", "o", or "e"), "n" is number of
821                        bits, and "f" is flow control ("r" for RTS or
822                        omit it).  Default is "9600n8".
824                        See Documentation/serial-console.txt for more
825                        information.  See
826                        Documentation/networking/netconsole.txt for an
827                        alternative.
829                uart[8250],io,<addr>[,options]
830                uart[8250],mmio,<addr>[,options]
831                uart[8250],mmio16,<addr>[,options]
832                uart[8250],mmio32,<addr>[,options]
833                uart[8250],0x<addr>[,options]
834                        Start an early, polled-mode console on the 8250/16550
835                        UART at the specified I/O port or MMIO address,
836                        switching to the matching ttyS device later.
837                        MMIO inter-register address stride is either 8-bit
838                        (mmio), 16-bit (mmio16), or 32-bit (mmio32).
839                        If none of [io|mmio|mmio16|mmio32], <addr> is assumed
840                        to be equivalent to 'mmio'. 'options' are specified in
841                        the same format described for ttyS above; if unspecified,
842                        the h/w is not re-initialized.
844                hvc<n>  Use the hypervisor console device <n>. This is for
845                        both Xen and PowerPC hypervisors.
847                If the device connected to the port is not a TTY but a braille
848                device, prepend "brl," before the device type, for instance
849                        console=brl,ttyS0
850                For now, only VisioBraille is supported.
852        consoleblank=   [KNL] The console blank (screen saver) timeout in
853                        seconds. Defaults to 10*60 = 10mins. A value of 0
854                        disables the blank timer.
856        coredump_filter=
857                        [KNL] Change the default value for
858                        /proc/<pid>/coredump_filter.
859                        See also Documentation/filesystems/proc.txt.
861   [CPU_IDLE]
862                        disable the cpuidle sub-system
864        cpu_init_udelay=N
865                        [X86] Delay for N microsec between assert and de-assert
866                        of APIC INIT to start processors.  This delay occurs
867                        on every CPU online, such as boot, and resume from suspend.
868                        Default: 10000
870        cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
871                        Format:
872                        <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
874        crashkernel=size[KMG][@offset[KMG]]
875                        [KNL] Using kexec, Linux can switch to a 'crash kernel'
876                        upon panic. This parameter reserves the physical
877                        memory region [offset, offset + size] for that kernel
878                        image. If '@offset' is omitted, then a suitable offset
879                        is selected automatically. Check
880                        Documentation/kdump/kdump.txt for further details.
882        crashkernel=range1:size1[,range2:size2,...][@offset]
883                        [KNL] Same as above, but depends on the memory
884                        in the running system. The syntax of range is
885                        start-[end] where start and end are both
886                        a memory unit (amount[KMG]). See also
887                        Documentation/kdump/kdump.txt for an example.
889        crashkernel=size[KMG],high
890                        [KNL, x86_64] range could be above 4G. Allow kernel
891                        to allocate physical memory region from top, so could
892                        be above 4G if system have more than 4G ram installed.
893                        Otherwise memory region will be allocated below 4G, if
894                        available.
895                        It will be ignored if crashkernel=X is specified.
896        crashkernel=size[KMG],low
897                        [KNL, x86_64] range under 4G. When crashkernel=X,high
898                        is passed, kernel could allocate physical memory region
899                        above 4G, that cause second kernel crash on system
900                        that require some amount of low memory, e.g. swiotlb
901                        requires at least 64M+32K low memory, also enough extra
902                        low memory is needed to make sure DMA buffers for 32-bit
903                        devices won't run out. Kernel would try to allocate at
904                        at least 256M below 4G automatically.
905                        This one let user to specify own low range under 4G
906                        for second kernel instead.
907                        0: to disable low allocation.
908                        It will be ignored when crashkernel=X,high is not used
909                        or memory reserved is below 4G.
911        cryptomgr.notests
912                        [KNL] Disable crypto self-tests
914        cs89x0_dma=     [HW,NET]
915                        Format: <dma>
917        cs89x0_media=   [HW,NET]
918                        Format: { rj45 | aui | bnc }
920        dasd=           [HW,NET]
921                        See header of drivers/s390/block/dasd_devmap.c.
923[2|3]=   [HW,JOY] Multisystem joystick support via parallel port
924                        (one device per port)
925                        Format: <port#>,<type>
926                        See also Documentation/input/joystick-parport.txt
928        ddebug_query=   [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
929                        time. See Documentation/dynamic-debug-howto.txt for
930                        details.  Deprecated, see dyndbg.
932        debug           [KNL] Enable kernel debugging (events log level).
934        debug_locks_verbose=
935                        [KNL] verbose self-tests
936                        Format=<0|1>
937                        Print debugging info while doing the locking API
938                        self-tests.
939                        We default to 0 (no extra messages), setting it to
940                        1 will print _a lot_ more information - normally
941                        only useful to kernel developers.
943        debug_objects   [KNL] Enable object debugging
945        no_debug_objects
946                        [KNL] Disable object debugging
948        debug_guardpage_minorder=
949                        [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
950                        parameter allows control of the order of pages that will
951                        be intentionally kept free (and hence protected) by the
952                        buddy allocator. Bigger value increase the probability
953                        of catching random memory corruption, but reduce the
954                        amount of memory for normal system use. The maximum
955                        possible value is MAX_ORDER/2.  Setting this parameter
956                        to 1 or 2 should be enough to identify most random
957                        memory corruption problems caused by bugs in kernel or
958                        driver code when a CPU writes to (or reads from) a
959                        random memory location. Note that there exists a class
960                        of memory corruptions problems caused by buggy H/W or
961                        F/W or by drivers badly programing DMA (basically when
962                        memory is written at bus level and the CPU MMU is
963                        bypassed) which are not detectable by
964                        CONFIG_DEBUG_PAGEALLOC, hence this option will not help
965                        tracking down these problems.
967        debug_pagealloc=
968                        [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
969                        parameter enables the feature at boot time. In
970                        default, it is disabled. We can avoid allocating huge
971                        chunk of memory for debug pagealloc if we don't enable
972                        it at boot time and the system will work mostly same
973                        with the kernel built without CONFIG_DEBUG_PAGEALLOC.
974                        on: enable the feature
976        debugpat        [X86] Enable PAT debugging
978        decnet.addr=    [HW,NET]
979                        Format: <area>[,<node>]
980                        See also Documentation/networking/decnet.txt.
982        default_hugepagesz=
983                        [same as hugepagesz=] The size of the default
984                        HugeTLB page size. This is the size represented by
985                        the legacy /proc/ hugepages APIs, used for SHM, and
986                        default size when mounting hugetlbfs filesystems.
987                        Defaults to the default architecture's huge page size
988                        if not specified.
990        dhash_entries=  [KNL]
991                        Set number of hash buckets for dentry cache.
993        disable_1tb_segments [PPC]
994                        Disables the use of 1TB hash page table segments. This
995                        causes the kernel to fall back to 256MB segments which
996                        can be useful when debugging issues that require an SLB
997                        miss to occur.
999        disable=        [IPV6]
1000                        See Documentation/networking/ipv6.txt.
1002        disable_radix   [PPC]
1003                        Disable RADIX MMU mode on POWER9
1005        disable_cpu_apicid= [X86,APIC,SMP]
1006                        Format: <int>
1007                        The number of initial APIC ID for the
1008                        corresponding CPU to be disabled at boot,
1009                        mostly used for the kdump 2nd kernel to
1010                        disable BSP to wake up multiple CPUs without
1011                        causing system reset or hang due to sending
1012                        INIT from AP to BSP.
1014        disable_ddw     [PPC/PSERIES]
1015                        Disable Dynamic DMA Window support. Use this if
1016                        to workaround buggy firmware.
1018        disable_ipv6=   [IPV6]
1019                        See Documentation/networking/ipv6.txt.
1021        disable_mtrr_cleanup [X86]
1022                        The kernel tries to adjust MTRR layout from continuous
1023                        to discrete, to make X server driver able to add WB
1024                        entry later. This parameter disables that.
1026        disable_mtrr_trim [X86, Intel and AMD only]
1027                        By default the kernel will trim any uncacheable
1028                        memory out of your available memory pool based on
1029                        MTRR settings.  This parameter disables that behavior,
1030                        possibly causing your machine to run very slowly.
1032        disable_timer_pin_1 [X86]
1033                        Disable PIN 1 of APIC timer
1034                        Can be useful to work around chipset bugs.
1036        dis_ucode_ldr   [X86] Disable the microcode loader.
1038        dma_debug=off   If the kernel is compiled with DMA_API_DEBUG support,
1039                        this option disables the debugging code at boot.
1041        dma_debug_entries=<number>
1042                        This option allows to tune the number of preallocated
1043                        entries for DMA-API debugging code. One entry is
1044                        required per DMA-API allocation. Use this if the
1045                        DMA-API debugging code disables itself because the
1046                        architectural default is too low.
1048        dma_debug_driver=<driver_name>
1049                        With this option the DMA-API debugging driver
1050                        filter feature can be enabled at boot time. Just
1051                        pass the driver to filter for as the parameter.
1052                        The filter can be disabled or changed to another
1053                        driver later using sysfs.
1055        drm_kms_helper.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
1056                        Broken monitors, graphic adapters, KVMs and EDIDless
1057                        panels may send no or incorrect EDID data sets.
1058                        This parameter allows to specify an EDID data sets
1059                        in the /lib/firmware directory that are used instead.
1060                        Generic built-in EDID data sets are used, if one of
1061                        edid/1024x768.bin, edid/1280x1024.bin,
1062                        edid/1680x1050.bin, or edid/1920x1080.bin is given
1063                        and no file with the same name exists. Details and
1064                        instructions how to build your own EDID data are
1065                        available in Documentation/EDID/HOWTO.txt. An EDID
1066                        data set will only be used for a particular connector,
1067                        if its name and a colon are prepended to the EDID
1068                        name. Each connector may use a unique EDID data
1069                        set by separating the files with a comma.  An EDID
1070                        data set with no connector name will be used for
1071                        any connectors not explicitly specified.
1073        dscc4.setup=    [NET]
1075        dyndbg[="val"]          [KNL,DYNAMIC_DEBUG]
1076        module.dyndbg[="val"]
1077                        Enable debug messages at boot time.  See
1078                        Documentation/dynamic-debug-howto.txt for details.
1080        nompx           [X86] Disables Intel Memory Protection Extensions.
1081                        See Documentation/x86/intel_mpx.txt for more
1082                        information about the feature.
1084        nopku           [X86] Disable Memory Protection Keys CPU feature found
1085                        in some Intel CPUs.
1087        eagerfpu=       [X86]
1088                        on      enable eager fpu restore
1089                        off     disable eager fpu restore
1090                        auto    selects the default scheme, which automatically
1091                                enables eagerfpu restore for xsaveopt.
1093        module.async_probe [KNL]
1094                        Enable asynchronous probe on this module.
1096        early_ioremap_debug [KNL]
1097                        Enable debug messages in early_ioremap support. This
1098                        is useful for tracking down temporary early mappings
1099                        which are not unmapped.
1101        earlycon=       [KNL] Output early console device and options.
1103                        When used with no options, the early console is
1104                        determined by the stdout-path property in device
1105                        tree's chosen node.
1107                cdns,<addr>[,options]
1108                        Start an early, polled-mode console on a Cadence
1109                        (xuartps) serial port at the specified address. Only
1110                        supported option is baud rate. If baud rate is not
1111                        specified, the serial port must already be setup and
1112                        configured.
1114                uart[8250],io,<addr>[,options]
1115                uart[8250],mmio,<addr>[,options]
1116                uart[8250],mmio32,<addr>[,options]
1117                uart[8250],mmio32be,<addr>[,options]
1118                uart[8250],0x<addr>[,options]
1119                        Start an early, polled-mode console on the 8250/16550
1120                        UART at the specified I/O port or MMIO address.
1121                        MMIO inter-register address stride is either 8-bit
1122                        (mmio) or 32-bit (mmio32 or mmio32be).
1123                        If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
1124                        to be equivalent to 'mmio'. 'options' are specified
1125                        in the same format described for "console=ttyS<n>"; if
1126                        unspecified, the h/w is not initialized.
1128                pl011,<addr>
1129                pl011,mmio32,<addr>
1130                        Start an early, polled-mode console on a pl011 serial
1131                        port at the specified address. The pl011 serial port
1132                        must already be setup and configured. Options are not
1133                        yet supported.  If 'mmio32' is specified, then only
1134                        the driver will use only 32-bit accessors to read/write
1135                        the device registers.
1137                meson,<addr>
1138                        Start an early, polled-mode console on a meson serial
1139                        port at the specified address. The serial port must
1140                        already be setup and configured. Options are not yet
1141                        supported.
1143                msm_serial,<addr>
1144                        Start an early, polled-mode console on an msm serial
1145                        port at the specified address. The serial port
1146                        must already be setup and configured. Options are not
1147                        yet supported.
1149                msm_serial_dm,<addr>
1150                        Start an early, polled-mode console on an msm serial
1151                        dm port at the specified address. The serial port
1152                        must already be setup and configured. Options are not
1153                        yet supported.
1155                smh     Use ARM semihosting calls for early console.
1157                s3c2410,<addr>
1158                s3c2412,<addr>
1159                s3c2440,<addr>
1160                s3c6400,<addr>
1161                s5pv210,<addr>
1162                exynos4210,<addr>
1163                        Use early console provided by serial driver available
1164                        on Samsung SoCs, requires selecting proper type and
1165                        a correct base address of the selected UART port. The
1166                        serial port must already be setup and configured.
1167                        Options are not yet supported.
1169                lpuart,<addr>
1170                lpuart32,<addr>
1171                        Use early console provided by Freescale LP UART driver
1172                        found on Freescale Vybrid and QorIQ LS1021A processors.
1173                        A valid base address must be provided, and the serial
1174                        port must already be setup and configured.
1176                armada3700_uart,<addr>
1177                        Start an early, polled-mode console on the
1178                        Armada 3700 serial port at the specified
1179                        address. The serial port must already be setup
1180                        and configured. Options are not yet supported.
1182        earlyprintk=    [X86,SH,BLACKFIN,ARM,M68k]
1183                        earlyprintk=vga
1184                        earlyprintk=efi
1185                        earlyprintk=xen
1186                        earlyprintk=serial[,ttySn[,baudrate]]
1187                        earlyprintk=serial[,0x...[,baudrate]]
1188                        earlyprintk=ttySn[,baudrate]
1189                        earlyprintk=dbgp[debugController#]
1190                        earlyprintk=pciserial,bus:device.function[,baudrate]
1192                        earlyprintk is useful when the kernel crashes before
1193                        the normal console is initialized. It is not enabled by
1194                        default because it has some cosmetic problems.
1196                        Append ",keep" to not disable it when the real console
1197                        takes over.
1199                        Only one of vga, efi, serial, or usb debug port can
1200                        be used at a time.
1202                        Currently only ttyS0 and ttyS1 may be specified by
1203                        name.  Other I/O ports may be explicitly specified
1204                        on some architectures (x86 and arm at least) by
1205                        replacing ttySn with an I/O port address, like this:
1206                                earlyprintk=serial,0x1008,115200
1207                        You can find the port for a given device in
1208                        /proc/tty/driver/serial:
1209                                2: uart:ST16650V2 port:00001008 irq:18 ...
1211                        Interaction with the standard serial driver is not
1212                        very good.
1214                        The VGA and EFI output is eventually overwritten by
1215                        the real console.
1217                        The xen output can only be used by Xen PV guests.
1219        edac_report=    [HW,EDAC] Control how to report EDAC event
1220                        Format: {"on" | "off" | "force"}
1221                        on: enable EDAC to report H/W event. May be overridden
1222                        by other higher priority error reporting module.
1223                        off: disable H/W event reporting through EDAC.
1224                        force: enforce the use of EDAC to report H/W event.
1225                        default: on.
1227        ekgdboc=        [X86,KGDB] Allow early kernel console debugging
1228                        ekgdboc=kbd
1230                        This is designed to be used in conjunction with
1231                        the boot argument: earlyprintk=vga
1233        edd=            [EDD]
1234                        Format: {"off" | "on" | "skip[mbr]"}
1236        efi=            [EFI]
1237                        Format: { "old_map", "nochunk", "noruntime", "debug" }
1238                        old_map [X86-64]: switch to the old ioremap-based EFI
1239                        runtime services mapping. 32-bit still uses this one by
1240                        default.
1241                        nochunk: disable reading files in "chunks" in the EFI
1242                        boot stub, as chunking can cause problems with some
1243                        firmware implementations.
1244                        noruntime : disable EFI runtime services support
1245                        debug: enable misc debug output
1247        efi_no_storage_paranoia [EFI; X86]
1248                        Using this parameter you can use more than 50% of
1249                        your efi variable storage. Use this parameter only if
1250                        you are really sure that your UEFI does sane gc and
1251                        fulfills the spec otherwise your board may brick.
1253        efi_fake_mem=   nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1254                        Add arbitrary attribute to specific memory range by
1255                        updating original EFI memory map.
1256                        Region of memory which aa attribute is added to is
1257                        from ss to ss+nn.
1258                        If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1259                        is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1260                        attribute is added to range 0x100000000-0x180000000 and
1261                        0x10a0000000-0x1120000000.
1263                        Using this parameter you can do debugging of EFI memmap
1264                        related feature. For example, you can do debugging of
1265                        Address Range Mirroring feature even if your box
1266                        doesn't support it.
1268        efivar_ssdt=    [EFI; X86] Name of an EFI variable that contains an SSDT
1269                        that is to be dynamically loaded by Linux. If there are
1270                        multiple variables with the same name but with different
1271                        vendor GUIDs, all of them will be loaded. See
1272                        Documentation/acpi/ssdt-overlays.txt for details.
1275        eisa_irq_edge=  [PARISC,HW]
1276                        See header of drivers/parisc/eisa.c.
1278        elanfreq=       [X86-32]
1279                        See comment before function elanfreq_setup() in
1280                        arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1282        elevator=       [IOSCHED]
1283                        Format: {"cfq" | "deadline" | "noop"}
1284                        See Documentation/block/cfq-iosched.txt and
1285                        Documentation/block/deadline-iosched.txt for details.
1287        elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1288                        Specifies physical address of start of kernel core
1289                        image elf header and optionally the size. Generally
1290                        kexec loader will pass this option to capture kernel.
1291                        See Documentation/kdump/kdump.txt for details.
1293        enable_mtrr_cleanup [X86]
1294                        The kernel tries to adjust MTRR layout from continuous
1295                        to discrete, to make X server driver able to add WB
1296                        entry later. This parameter enables that.
1298        enable_timer_pin_1 [X86]
1299                        Enable PIN 1 of APIC timer
1300                        Can be useful to work around chipset bugs
1301                        (in particular on some ATI chipsets).
1302                        The kernel tries to set a reasonable default.
1304        enforcing       [SELINUX] Set initial enforcing status.
1305                        Format: {"0" | "1"}
1306                        See security/selinux/Kconfig help text.
1307                        0 -- permissive (log only, no denials).
1308                        1 -- enforcing (deny and log).
1309                        Default value is 0.
1310                        Value can be changed at runtime via /selinux/enforce.
1312        erst_disable    [ACPI]
1313                        Disable Error Record Serialization Table (ERST)
1314                        support.
1316        ether=          [HW,NET] Ethernet cards parameters
1317                        This option is obsoleted by the "netdev=" option, which
1318                        has equivalent usage. See its documentation for details.
1320        evm=            [EVM]
1321                        Format: { "fix" }
1322                        Permit 'security.evm' to be updated regardless of
1323                        current integrity status.
1325        failslab=
1326        fail_page_alloc=
1327        fail_make_request=[KNL]
1328                        General fault injection mechanism.
1329                        Format: <interval>,<probability>,<space>,<times>
1330                        See also Documentation/fault-injection/.
1332        floppy=         [HW]
1333                        See Documentation/blockdev/floppy.txt.
1335        force_pal_cache_flush
1336                        [IA-64] Avoid check_sal_cache_flush which may hang on
1337                        buggy SAL_CACHE_FLUSH implementations. Using this
1338                        parameter will force ia64_sal_cache_flush to call
1339                        ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1341        forcepae [X86-32]
1342                        Forcefully enable Physical Address Extension (PAE).
1343                        Many Pentium M systems disable PAE but may have a
1344                        functionally usable PAE implementation.
1345                        Warning: use of this parameter will taint the kernel
1346                        and may cause unknown problems.
1348        ftrace=[tracer]
1349                        [FTRACE] will set and start the specified tracer
1350                        as early as possible in order to facilitate early
1351                        boot debugging.
1353        ftrace_dump_on_oops[=orig_cpu]
1354                        [FTRACE] will dump the trace buffers on oops.
1355                        If no parameter is passed, ftrace will dump
1356                        buffers of all CPUs, but if you pass orig_cpu, it will
1357                        dump only the buffer of the CPU that triggered the
1358                        oops.
1360        ftrace_filter=[function-list]
1361                        [FTRACE] Limit the functions traced by the function
1362                        tracer at boot up. function-list is a comma separated
1363                        list of functions. This list can be changed at run
1364                        time by the set_ftrace_filter file in the debugfs
1365                        tracing directory.
1367        ftrace_notrace=[function-list]
1368                        [FTRACE] Do not trace the functions specified in
1369                        function-list. This list can be changed at run time
1370                        by the set_ftrace_notrace file in the debugfs
1371                        tracing directory.
1373        ftrace_graph_filter=[function-list]
1374                        [FTRACE] Limit the top level callers functions traced
1375                        by the function graph tracer at boot up.
1376                        function-list is a comma separated list of functions
1377                        that can be changed at run time by the
1378                        set_graph_function file in the debugfs tracing directory.
1380        ftrace_graph_notrace=[function-list]
1381                        [FTRACE] Do not trace from the functions specified in
1382                        function-list.  This list is a comma separated list of
1383                        functions that can be changed at run time by the
1384                        set_graph_notrace file in the debugfs tracing directory.
1387                        [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1388                        support via parallel port (up to 5 devices per port)
1389                        Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1390                        See also Documentation/input/joystick-parport.txt
1392        gamma=          [HW,DRM]
1394        gart_fix_e820=  [X86_64] disable the fix e820 for K8 GART
1395                        Format: off | on
1396                        default: on
1398        gcov_persist=   [GCOV] When non-zero (default), profiling data for
1399                        kernel modules is saved and remains accessible via
1400                        debugfs, even when the module is unloaded/reloaded.
1401                        When zero, profiling data is discarded and associated
1402                        debugfs files are removed at module unload time.
1404        goldfish        [X86] Enable the goldfish android emulator platform.
1405                        Don't use this when you are not running on the
1406                        android emulator
1408        gpt             [EFI] Forces disk with valid GPT signature but
1409                        invalid Protective MBR to be treated as GPT. If the
1410                        primary GPT is corrupted, it enables the backup/alternate
1411                        GPT to be used instead.
1413        grcan.enable0=  [HW] Configuration of physical interface 0. Determines
1414                        the "Enable 0" bit of the configuration register.
1415                        Format: 0 | 1
1416                        Default: 0
1417        grcan.enable1=  [HW] Configuration of physical interface 1. Determines
1418                        the "Enable 0" bit of the configuration register.
1419                        Format: 0 | 1
1420                        Default: 0
1421   [HW] Select which physical interface to use.
1422                        Format: 0 | 1
1423                        Default: 0
1424        grcan.txsize=   [HW] Sets the size of the tx buffer.
1425                        Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1426                        Default: 1024
1427        grcan.rxsize=   [HW] Sets the size of the rx buffer.
1428                        Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1429                        Default: 1024
1431        gpio-mockup.gpio_mockup_ranges
1432                        [HW] Sets the ranges of gpiochip of for this device.
1433                        Format: <start1>,<end1>,<start2>,<end2>...
1435        hardlockup_all_cpu_backtrace=
1436                        [KNL] Should the hard-lockup detector generate
1437                        backtraces on all cpus.
1438                        Format: <integer>
1440        hashdist=       [KNL,NUMA] Large hashes allocated during boot
1441                        are distributed across NUMA nodes.  Defaults on
1442                        for 64-bit NUMA, off otherwise.
1443                        Format: 0 | 1 (for off | on)
1445        hcl=            [IA-64] SGI's Hardware Graph compatibility layer
1447        hd=             [EIDE] (E)IDE hard drive subsystem geometry
1448                        Format: <cyl>,<head>,<sect>
1450        hest_disable    [ACPI]
1451                        Disable Hardware Error Source Table (HEST) support;
1452                        corresponding firmware-first mode error processing
1453                        logic will be disabled.
1455        highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1456                        size of <nn>. This works even on boxes that have no
1457                        highmem otherwise. This also works to reduce highmem
1458                        size on bigger boxes.
1460        highres=        [KNL] Enable/disable high resolution timer mode.
1461                        Valid parameters: "on", "off"
1462                        Default: "on"
1464        hisax=          [HW,ISDN]
1465                        See Documentation/isdn/README.HiSax.
1467        hlt             [BUGS=ARM,SH]
1469        hpet=           [X86-32,HPET] option to control HPET usage
1470                        Format: { enable (default) | disable | force |
1471                                verbose }
1472                        disable: disable HPET and use PIT instead
1473                        force: allow force enabled of undocumented chips (ICH4,
1474                                VIA, nVidia)
1475                        verbose: show contents of HPET registers during setup
1477        hpet_mmap=      [X86, HPET_MMAP] Allow userspace to mmap HPET
1478                        registers.  Default set by CONFIG_HPET_MMAP_DEFAULT.
1480        hugepages=      [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
1481        hugepagesz=     [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
1482                        On x86-64 and powerpc, this option can be specified
1483                        multiple times interleaved with hugepages= to reserve
1484                        huge pages of different sizes. Valid pages sizes on
1485                        x86-64 are 2M (when the CPU supports "pse") and 1G
1486                        (when the CPU supports the "pdpe1gb" cpuinfo flag).
1488        hvc_iucv=       [S390] Number of z/VM IUCV hypervisor console (HVC)
1489                               terminal devices. Valid values: 0..8
1490        hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1491                               If specified, z/VM IUCV HVC accepts connections
1492                               from listed z/VM user IDs only.
1494        hwthread_map=   [METAG] Comma-separated list of Linux cpu id to
1495                                hardware thread id mappings.
1496                                Format: <cpu>:<hwthread>
1498        keep_bootcon    [KNL]
1499                        Do not unregister boot console at start. This is only
1500                        useful for debugging when something happens in the window
1501                        between unregistering the boot console and initializing
1502                        the real console.
1504        i2c_bus=        [HW] Override the default board specific I2C bus speed
1505                             or register an additional I2C bus that is not
1506                             registered from board initialization code.
1507                             Format:
1508                             <bus_id>,<clkrate>
1510        i8042.debug     [HW] Toggle i8042 debug mode
1511        i8042.unmask_kbd_data
1512                        [HW] Enable printing of interrupt data from the KBD port
1513                             (disabled by default, and as a pre-condition
1514                             requires that i8042.debug=1 be enabled)
1515    [HW] Put keyboard port into non-translated mode
1516        i8042.dumbkbd   [HW] Pretend that controller can only read data from
1517                             keyboard and cannot control its state
1518                             (Don't attempt to blink the leds)
1519        i8042.noaux     [HW] Don't check for auxiliary (== mouse) port
1520        i8042.nokbd     [HW] Don't check/create keyboard port
1521        i8042.noloop    [HW] Disable the AUX Loopback command while probing
1522                             for the AUX port
1523        i8042.nomux     [HW] Don't check presence of an active multiplexing
1524                             controller
1525        i8042.nopnp     [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1526                             controllers
1527        i8042.notimeout [HW] Ignore timeout condition signalled by controller
1528        i8042.reset     [HW] Reset the controller during init, cleanup and
1529                             suspend-to-ram transitions, only during s2r
1530                             transitions, or never reset
1531                        Format: { 1 | Y | y | 0 | N | n }
1532                        1, Y, y: always reset controller
1533                        0, N, n: don't ever reset controller
1534                        Default: only on s2r transitions on x86; most other
1535                        architectures force reset to be always executed
1536        i8042.unlock    [HW] Unlock (ignore) the keylock
1537        i8042.kbdreset  [HW] Reset device connected to KBD port
1539        i810=           [HW,DRM]
1541        i8k.ignore_dmi  [HW] Continue probing hardware even if DMI data
1542                        indicates that the driver is running on unsupported
1543                        hardware.
1544        i8k.force       [HW] Activate i8k driver even if SMM BIOS signature
1545                        does not match list of supported models.
1546        i8k.power_status
1547                        [HW] Report power status in /proc/i8k
1548                        (disabled by default)
1549        i8k.restricted  [HW] Allow controlling fans only if SYS_ADMIN
1550                        capability is set.
1552        i915.invert_brightness=
1553                        [DRM] Invert the sense of the variable that is used to
1554                        set the brightness of the panel backlight. Normally a
1555                        brightness value of 0 indicates backlight switched off,
1556                        and the maximum of the brightness value sets the backlight
1557                        to maximum brightness. If this parameter is set to 0
1558                        (default) and the machine requires it, or this parameter
1559                        is set to 1, a brightness value of 0 sets the backlight
1560                        to maximum brightness, and the maximum of the brightness
1561                        value switches the backlight off.
1562                        -1 -- never invert brightness
1563                         0 -- machine default
1564                         1 -- force brightness inversion
1566        icn=            [HW,ISDN]
1567                        Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1569        ide-core.nodma= [HW] (E)IDE subsystem
1570                        Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1571                        .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1572                        .cdrom .chs .ignore_cable are additional options
1573                        See Documentation/ide/ide.txt.
1575        ide-generic.probe-mask= [HW] (E)IDE subsystem
1576                        Format: <int>
1577                        Probe mask for legacy ISA IDE ports.  Depending on
1578                        platform up to 6 ports are supported, enabled by
1579                        setting corresponding bits in the mask to 1.  The
1580                        default value is 0x0, which has a special meaning.
1581                        On systems that have PCI, it triggers scanning the
1582                        PCI bus for the first and the second port, which
1583                        are then probed.  On systems without PCI the value
1584                        of 0x0 enables probing the two first ports as if it
1585                        was 0x3.
1587        ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1588                        Claim all unknown PCI IDE storage controllers.
1590        idle=           [X86]
1591                        Format: idle=poll, idle=halt, idle=nomwait
1592                        Poll forces a polling idle loop that can slightly
1593                        improve the performance of waking up a idle CPU, but
1594                        will use a lot of power and make the system run hot.
1595                        Not recommended.
1596                        idle=halt: Halt is forced to be used for CPU idle.
1597                        In such case C2/C3 won't be used again.
1598                        idle=nomwait: Disable mwait for CPU C-states
1600        ieee754=        [MIPS] Select IEEE Std 754 conformance mode
1601                        Format: { strict | legacy | 2008 | relaxed }
1602                        Default: strict
1604                        Choose which programs will be accepted for execution
1605                        based on the IEEE 754 NaN encoding(s) supported by
1606                        the FPU and the NaN encoding requested with the value
1607                        of an ELF file header flag individually set by each
1608                        binary.  Hardware implementations are permitted to
1609                        support either or both of the legacy and the 2008 NaN
1610                        encoding mode.
1612                        Available settings are as follows:
1613                        strict  accept binaries that request a NaN encoding
1614                                supported by the FPU
1615                        legacy  only accept legacy-NaN binaries, if supported
1616                                by the FPU
1617                        2008    only accept 2008-NaN binaries, if supported
1618                                by the FPU
1619                        relaxed accept any binaries regardless of whether
1620                                supported by the FPU
1622                        The FPU emulator is always able to support both NaN
1623                        encodings, so if no FPU hardware is present or it has
1624                        been disabled with 'nofpu', then the settings of
1625                        'legacy' and '2008' strap the emulator accordingly,
1626                        'relaxed' straps the emulator for both legacy-NaN and
1627                        2008-NaN, whereas 'strict' enables legacy-NaN only on
1628                        legacy processors and both NaN encodings on MIPS32 or
1629                        MIPS64 CPUs.
1631                        The setting for ABS.fmt/NEG.fmt instruction execution
1632                        mode generally follows that for the NaN encoding,
1633                        except where unsupported by hardware.
1635        ignore_loglevel [KNL]
1636                        Ignore loglevel setting - this will print /all/
1637                        kernel messages to the console. Useful for debugging.
1638                        We also add it as printk module parameter, so users
1639                        could change it dynamically, usually by
1640                        /sys/module/printk/parameters/ignore_loglevel.
1642        ignore_rlimit_data
1643                        Ignore RLIMIT_DATA setting for data mappings,
1644                        print warning at first misuse.  Can be changed via
1645                        /sys/module/kernel/parameters/ignore_rlimit_data.
1647        ihash_entries=  [KNL]
1648                        Set number of hash buckets for inode cache.
1650        ima_appraise=   [IMA] appraise integrity measurements
1651                        Format: { "off" | "enforce" | "fix" | "log" }
1652                        default: "enforce"
1654        ima_appraise_tcb [IMA]
1655                        The builtin appraise policy appraises all files
1656                        owned by uid=0.
1658        ima_hash=       [IMA]
1659                        Format: { md5 | sha1 | rmd160 | sha256 | sha384
1660                                   | sha512 | ... }
1661                        default: "sha1"
1663                        The list of supported hash algorithms is defined
1664                        in crypto/hash_info.h.
1666        ima_policy=     [IMA]
1667                        The builtin measurement policy to load during IMA
1668                        setup.  Specyfing "tcb" as the value, measures all
1669                        programs exec'd, files mmap'd for exec, and all files
1670                        opened with the read mode bit set by either the
1671                        effective uid (euid=0) or uid=0.
1672                        Format: "tcb"
1674        ima_tcb         [IMA] Deprecated.  Use ima_policy= instead.
1675                        Load a policy which meets the needs of the Trusted
1676                        Computing Base.  This means IMA will measure all
1677                        programs exec'd, files mmap'd for exec, and all files
1678                        opened for read by uid=0.
1680        ima_template=   [IMA]
1681                        Select one of defined IMA measurements template formats.
1682                        Formats: { "ima" | "ima-ng" | "ima-sig" }
1683                        Default: "ima-ng"
1685        ima_template_fmt=
1686                        [IMA] Define a custom template format.
1687                        Format: { "field1|...|fieldN" }
1689        ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1690                        Format: <min_file_size>
1691                        Set the minimal file size for using asynchronous hash.
1692                        If left unspecified, ahash usage is disabled.
1694                        ahash performance varies for different data sizes on
1695                        different crypto accelerators. This option can be used
1696                        to achieve the best performance for a particular HW.
1698        ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1699                        Format: <bufsize>
1700                        Set hashing buffer size. Default: 4k.
1702                        ahash performance varies for different chunk sizes on
1703                        different crypto accelerators. This option can be used
1704                        to achieve best performance for particular HW.
1706        init=           [KNL]
1707                        Format: <full_path>
1708                        Run specified binary instead of /sbin/init as init
1709                        process.
1711        initcall_debug  [KNL] Trace initcalls as they are executed.  Useful
1712                        for working out where the kernel is dying during
1713                        startup.
1715        initcall_blacklist=  [KNL] Do not execute a comma-separated list of
1716                        initcall functions.  Useful for debugging built-in
1717                        modules and initcalls.
1719        initrd=         [BOOT] Specify the location of the initial ramdisk
1721        init_pkru=      [x86] Specify the default memory protection keys rights
1722                        register contents for all processes.  0x55555554 by
1723                        default (disallow access to all but pkey 0).  Can
1724                        override in debugfs after boot.
1726        inport.irq=     [HW] Inport (ATI XL and Microsoft) busmouse driver
1727                        Format: <irq>
1729        int_pln_enable  [x86] Enable power limit notification interrupt
1731        integrity_audit=[IMA]
1732                        Format: { "0" | "1" }
1733                        0 -- basic integrity auditing messages. (Default)
1734                        1 -- additional integrity auditing messages.
1736        intel_iommu=    [DMAR] Intel IOMMU driver (DMAR) option
1737                on
1738                        Enable intel iommu driver.
1739                off
1740                        Disable intel iommu driver.
1741                igfx_off [Default Off]
1742                        By default, gfx is mapped as normal device. If a gfx
1743                        device has a dedicated DMAR unit, the DMAR unit is
1744                        bypassed by not enabling DMAR with this option. In
1745                        this case, gfx device will use physical address for
1746                        DMA.
1747                forcedac [x86_64]
1748                        With this option iommu will not optimize to look
1749                        for io virtual address below 32-bit forcing dual
1750                        address cycle on pci bus for cards supporting greater
1751                        than 32-bit addressing. The default is to look
1752                        for translation below 32-bit and if not available
1753                        then look in the higher range.
1754                strict [Default Off]
1755                        With this option on every unmap_single operation will
1756                        result in a hardware IOTLB flush operation as opposed
1757                        to batching them for performance.
1758                sp_off [Default Off]
1759                        By default, super page will be supported if Intel IOMMU
1760                        has the capability. With this option, super page will
1761                        not be supported.
1762                ecs_off [Default Off]
1763                        By default, extended context tables will be supported if
1764                        the hardware advertises that it has support both for the
1765                        extended tables themselves, and also PASID support. With
1766                        this option set, extended tables will not be used even
1767                        on hardware which claims to support them.
1769        intel_idle.max_cstate=  [KNL,HW,ACPI,X86]
1770                        0       disables intel_idle and fall back on acpi_idle.
1771                        1 to 9  specify maximum depth of C-state.
1773        intel_pstate=  [X86]
1774                       disable
1775                         Do not enable intel_pstate as the default
1776                         scaling driver for the supported processors
1777                       force
1778                         Enable intel_pstate on systems that prohibit it by default
1779                         in favor of acpi-cpufreq. Forcing the intel_pstate driver
1780                         instead of acpi-cpufreq may disable platform features, such
1781                         as thermal controls and power capping, that rely on ACPI
1782                         P-States information being indicated to OSPM and therefore
1783                         should be used with caution. This option does not work with
1784                         processors that aren't supported by the intel_pstate driver
1785                         or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
1786                       no_hwp
1787                         Do not enable hardware P state control (HWP)
1788                         if available.
1789                hwp_only
1790                        Only load intel_pstate on systems which support
1791                        hardware P state control (HWP) if available.
1792                support_acpi_ppc
1793                        Enforce ACPI _PPC performance limits. If the Fixed ACPI
1794                        Description Table, specifies preferred power management
1795                        profile as "Enterprise Server" or "Performance Server",
1796                        then this feature is turned on by default.
1798        intremap=       [X86-64, Intel-IOMMU]
1799                        on      enable Interrupt Remapping (default)
1800                        off     disable Interrupt Remapping
1801                        nosid   disable Source ID checking
1802                        no_x2apic_optout
1803                                BIOS x2APIC opt-out request will be ignored
1804                        nopost  disable Interrupt Posting
1806        iomem=          Disable strict checking of access to MMIO memory
1807                strict  regions from userspace.
1808                relaxed
1810        iommu=          [x86]
1811                off
1812                force
1813                noforce
1814                biomerge
1815                panic
1816                nopanic
1817                merge
1818                nomerge
1819                forcesac
1820                soft
1821                pt              [x86, IA-64]
1822                nobypass        [PPC/POWERNV]
1823                        Disable IOMMU bypass, using IOMMU for PCI devices.
1826        io7=            [HW] IO7 for Marvel based alpha systems
1827                        See comment before marvel_specify_io7 in
1828                        arch/alpha/kernel/core_marvel.c.
1830        io_delay=       [X86] I/O delay method
1831                0x80
1832                        Standard port 0x80 based delay
1833                0xed
1834                        Alternate port 0xed based delay (needed on some systems)
1835                udelay
1836                        Simple two microseconds delay
1837                none
1838                        No delay
1840        ip=             [IP_PNP]
1841                        See Documentation/filesystems/nfs/nfsroot.txt.
1843        irqaffinity=    [SMP] Set the default irq affinity mask
1844                        The argument is a cpu list, as described above.
1846        irqfixup        [HW]
1847                        When an interrupt is not handled search all handlers
1848                        for it. Intended to get systems with badly broken
1849                        firmware running.
1851        irqpoll         [HW]
1852                        When an interrupt is not handled search all handlers
1853                        for it. Also check all handlers each timer
1854                        interrupt. Intended to get systems with badly broken
1855                        firmware running.
1857        isapnp=         [ISAPNP]
1858                        Format: <RDP>,<reset>,<pci_scan>,<verbosity>
1860        isolcpus=       [KNL,SMP] Isolate CPUs from the general scheduler.
1861                        The argument is a cpu list, as described above.
1863                        This option can be used to specify one or more CPUs
1864                        to isolate from the general SMP balancing and scheduling
1865                        algorithms. You can move a process onto or off an
1866                        "isolated" CPU via the CPU affinity syscalls or cpuset.
1867                        <cpu number> begins at 0 and the maximum value is
1868                        "number of CPUs in system - 1".
1870                        This option is the preferred way to isolate CPUs. The
1871                        alternative -- manually setting the CPU mask of all
1872                        tasks in the system -- can cause problems and
1873                        suboptimal load balancer performance.
1875        iucv=           [HW,NET]
1877        ivrs_ioapic     [HW,X86_64]
1878                        Provide an override to the IOAPIC-ID<->DEVICE-ID
1879                        mapping provided in the IVRS ACPI table. For
1880                        example, to map IOAPIC-ID decimal 10 to
1881                        PCI device 00:14.0 write the parameter as:
1882                                ivrs_ioapic[10]=00:14.0
1884        ivrs_hpet       [HW,X86_64]
1885                        Provide an override to the HPET-ID<->DEVICE-ID
1886                        mapping provided in the IVRS ACPI table. For
1887                        example, to map HPET-ID decimal 0 to
1888                        PCI device 00:14.0 write the parameter as:
1889                                ivrs_hpet[0]=00:14.0
1891        ivrs_acpihid    [HW,X86_64]
1892                        Provide an override to the ACPI-HID:UID<->DEVICE-ID
1893                        mapping provided in the IVRS ACPI table. For
1894                        example, to map UART-HID:UID AMD0020:0 to
1895                        PCI device 00:14.5 write the parameter as:
1896                                ivrs_acpihid[00:14.5]=AMD0020:0
1898        js=             [HW,JOY] Analog joystick
1899                        See Documentation/input/joystick.txt.
1901        nokaslr         [KNL]
1902                        When CONFIG_RANDOMIZE_BASE is set, this disables
1903                        kernel and module base offset ASLR (Address Space
1904                        Layout Randomization).
1906        keepinitrd      [HW,ARM]
1908        kernelcore=     [KNL,X86,IA-64,PPC]
1909                        Format: nn[KMGTPE] | "mirror"
1910                        This parameter
1911                        specifies the amount of memory usable by the kernel
1912                        for non-movable allocations.  The requested amount is
1913                        spread evenly throughout all nodes in the system. The
1914                        remaining memory in each node is used for Movable
1915                        pages. In the event, a node is too small to have both
1916                        kernelcore and Movable pages, kernelcore pages will
1917                        take priority and other nodes will have a larger number
1918                        of Movable pages.  The Movable zone is used for the
1919                        allocation of pages that may be reclaimed or moved
1920                        by the page migration subsystem.  This means that
1921                        HugeTLB pages may not be allocated from this zone.
1922                        Note that allocations like PTEs-from-HighMem still
1923                        use the HighMem zone if it exists, and the Normal
1924                        zone if it does not.
1926                        Instead of specifying the amount of memory (nn[KMGTPE]),
1927                        you can specify "mirror" option. In case "mirror"
1928                        option is specified, mirrored (reliable) memory is used
1929                        for non-movable allocations and remaining memory is used
1930                        for Movable pages. nn[KMGTPE] and "mirror" are exclusive,
1931                        so you can NOT specify nn[KMGTPE] and "mirror" at the same
1932                        time.
1934        kgdbdbgp=       [KGDB,HW] kgdb over EHCI usb debug port.
1935                        Format: <Controller#>[,poll interval]
1936                        The controller # is the number of the ehci usb debug
1937                        port as it is probed via PCI.  The poll interval is
1938                        optional and is the number seconds in between
1939                        each poll cycle to the debug port in case you need
1940                        the functionality for interrupting the kernel with
1941                        gdb or control-c on the dbgp connection.  When
1942                        not using this parameter you use sysrq-g to break into
1943                        the kernel debugger.
1945        kgdboc=         [KGDB,HW] kgdb over consoles.
1946                        Requires a tty driver that supports console polling,
1947                        or a supported polling keyboard driver (non-usb).
1948                         Serial only format: <serial_device>[,baud]
1949                         keyboard only format: kbd
1950                         keyboard and serial format: kbd,<serial_device>[,baud]
1951                        Optional Kernel mode setting:
1952                         kms, kbd format: kms,kbd
1953                         kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
1955        kgdbwait        [KGDB] Stop kernel execution and enter the
1956                        kernel debugger at the earliest opportunity.
1958        kmac=           [MIPS] korina ethernet MAC address.
1959                        Configure the RouterBoard 532 series on-chip
1960                        Ethernet adapter MAC address.
1962        kmemleak=       [KNL] Boot-time kmemleak enable/disable
1963                        Valid arguments: on, off
1964                        Default: on
1965                        Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
1966                        the default is off.
1968        kmemcheck=      [X86] Boot-time kmemcheck enable/disable/one-shot mode
1969                        Valid arguments: 0, 1, 2
1970                        kmemcheck=0 (disabled)
1971                        kmemcheck=1 (enabled)
1972                        kmemcheck=2 (one-shot mode)
1973                        Default: 2 (one-shot mode)
1975        kstack=N        [X86] Print N words from the kernel stack
1976                        in oops dumps.
1978        kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
1979                        Default is 0 (don't ignore, but inject #GP)
1981        kvm.mmu_audit=  [KVM] This is a R/W parameter which allows audit
1982                        KVM MMU at runtime.
1983                        Default is 0 (off)
1985        kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
1986                        Default is 1 (enabled)
1988        kvm-amd.npt=    [KVM,AMD] Disable nested paging (virtualized MMU)
1989                        for all guests.
1990                        Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
1992        kvm-intel.ept=  [KVM,Intel] Disable extended page tables
1993                        (virtualized MMU) support on capable Intel chips.
1994                        Default is 1 (enabled)
1996        kvm-intel.emulate_invalid_guest_state=
1997                        [KVM,Intel] Enable emulation of invalid guest states
1998                        Default is 0 (disabled)
2000        kvm-intel.flexpriority=
2001                        [KVM,Intel] Disable FlexPriority feature (TPR shadow).
2002                        Default is 1 (enabled)
2004        kvm-intel.nested=
2005                        [KVM,Intel] Enable VMX nesting (nVMX).
2006                        Default is 0 (disabled)
2008        kvm-intel.unrestricted_guest=
2009                        [KVM,Intel] Disable unrestricted guest feature
2010                        (virtualized real and unpaged mode) on capable
2011                        Intel chips. Default is 1 (enabled)
2013        kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
2014                        feature (tagged TLBs) on capable Intel chips.
2015                        Default is 1 (enabled)
2017        l2cr=           [PPC]
2019        l3cr=           [PPC]
2021        lapic           [X86-32,APIC] Enable the local APIC even if BIOS
2022                        disabled it.
2024        lapic=          [x86,APIC] "notscdeadline" Do not use TSC deadline
2025                        value for LAPIC timer one-shot implementation. Default
2026                        back to the programmable timer unit in the LAPIC.
2028        lapic_timer_c2_ok       [X86,APIC] trust the local apic timer
2029                        in C2 power state.
2031        libata.dma=     [LIBATA] DMA control
2032                        libata.dma=0      Disable all PATA and SATA DMA
2033                        libata.dma=1      PATA and SATA Disk DMA only
2034                        libata.dma=2      ATAPI (CDROM) DMA only
2035                        libata.dma=4      Compact Flash DMA only
2036                        Combinations also work, so libata.dma=3 enables DMA
2037                        for disks and CDROMs, but not CFs.
2039        libata.ignore_hpa=      [LIBATA] Ignore HPA limit
2040                        libata.ignore_hpa=0       keep BIOS limits (default)
2041                        libata.ignore_hpa=1       ignore limits, using full disk
2043        libata.noacpi   [LIBATA] Disables use of ACPI in libata suspend/resume
2044                        when set.
2045                        Format: <int>
2047        libata.force=   [LIBATA] Force configurations.  The format is comma
2048                        separated list of "[ID:]VAL" where ID is
2049                        PORT[.DEVICE].  PORT and DEVICE are decimal numbers
2050                        matching port, link or device.  Basically, it matches
2051                        the ATA ID string printed on console by libata.  If
2052                        the whole ID part is omitted, the last PORT and DEVICE
2053                        values are used.  If ID hasn't been specified yet, the
2054                        configuration applies to all ports, links and devices.
2056                        If only DEVICE is omitted, the parameter applies to
2057                        the port and all links and devices behind it.  DEVICE
2058                        number of 0 either selects the first device or the
2059                        first fan-out link behind PMP device.  It does not
2060                        select the host link.  DEVICE number of 15 selects the
2061                        host link and device attached to it.
2063                        The VAL specifies the configuration to force.  As long
2064                        as there's no ambiguity shortcut notation is allowed.
2065                        For example, both 1.5 and 1.5G would work for 1.5Gbps.
2066                        The following configurations can be forced.
2068                        * Cable type: 40c, 80c, short40c, unk, ign or sata.
2069                          Any ID with matching PORT is used.
2071                        * SATA link speed limit: 1.5Gbps or 3.0Gbps.
2073                        * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
2074                          udma[/][16,25,33,44,66,100,133] notation is also
2075                          allowed.
2077                        * [no]ncq: Turn on or off NCQ.
2079                        * [no]ncqtrim: Turn off queued DSM TRIM.
2081                        * nohrst, nosrst, norst: suppress hard, soft
2082                          and both resets.
2084                        * rstonce: only attempt one reset during
2085                          hot-unplug link recovery
2087                        * dump_id: dump IDENTIFY data.
2089                        * atapi_dmadir: Enable ATAPI DMADIR bridge support
2091                        * disable: Disable this device.
2093                        If there are multiple matching configurations changing
2094                        the same attribute, the last one is used.
2096        memblock=debug  [KNL] Enable memblock debug messages.
2098        load_ramdisk=   [RAM] List of ramdisks to load from floppy
2099                        See Documentation/blockdev/ramdisk.txt.
2101        lockd.nlm_grace_period=P  [NFS] Assign grace period.
2102                        Format: <integer>
2104        lockd.nlm_tcpport=N     [NFS] Assign TCP port.
2105                        Format: <integer>
2107        lockd.nlm_timeout=T     [NFS] Assign timeout value.
2108                        Format: <integer>
2110        lockd.nlm_udpport=M     [NFS] Assign UDP port.
2111                        Format: <integer>
2113        locktorture.nreaders_stress= [KNL]
2114                        Set the number of locking read-acquisition kthreads.
2115                        Defaults to being automatically set based on the
2116                        number of online CPUs.
2118        locktorture.nwriters_stress= [KNL]
2119                        Set the number of locking write-acquisition kthreads.
2121        locktorture.onoff_holdoff= [KNL]
2122                        Set time (s) after boot for CPU-hotplug testing.
2124        locktorture.onoff_interval= [KNL]
2125                        Set time (s) between CPU-hotplug operations, or
2126                        zero to disable CPU-hotplug testing.
2128        locktorture.shuffle_interval= [KNL]
2129                        Set task-shuffle interval (jiffies).  Shuffling
2130                        tasks allows some CPUs to go into dyntick-idle
2131                        mode during the locktorture test.
2133        locktorture.shutdown_secs= [KNL]
2134                        Set time (s) after boot system shutdown.  This
2135                        is useful for hands-off automated testing.
2137        locktorture.stat_interval= [KNL]
2138                        Time (s) between statistics printk()s.
2140        locktorture.stutter= [KNL]
2141                        Time (s) to stutter testing, for example,
2142                        specifying five seconds causes the test to run for
2143                        five seconds, wait for five seconds, and so on.
2144                        This tests the locking primitive's ability to
2145                        transition abruptly to and from idle.
2147        locktorture.torture_runnable= [BOOT]
2148                        Start locktorture running at boot time.
2150        locktorture.torture_type= [KNL]
2151                        Specify the locking implementation to test.
2153        locktorture.verbose= [KNL]
2154                        Enable additional printk() statements.
2156        logibm.irq=     [HW,MOUSE] Logitech Bus Mouse Driver
2157                        Format: <irq>
2159        loglevel=       All Kernel Messages with a loglevel smaller than the
2160                        console loglevel will be printed to the console. It can
2161                        also be changed with klogd or other programs. The
2162                        loglevels are defined as follows:
2164                        0 (KERN_EMERG)          system is unusable
2165                        1 (KERN_ALERT)          action must be taken immediately
2166                        2 (KERN_CRIT)           critical conditions
2167                        3 (KERN_ERR)            error conditions
2168                        4 (KERN_WARNING)        warning conditions
2169                        5 (KERN_NOTICE)         normal but significant condition
2170                        6 (KERN_INFO)           informational
2171                        7 (KERN_DEBUG)          debug-level messages
2173        log_buf_len=n[KMG]      Sets the size of the printk ring buffer,
2174                        in bytes.  n must be a power of two and greater
2175                        than the minimal size. The minimal size is defined
2176                        by LOG_BUF_SHIFT kernel config parameter. There is
2177                        also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
2178                        that allows to increase the default size depending on
2179                        the number of CPUs. See init/Kconfig for more details.
2181        logo.nologo     [FB] Disables display of the built-in Linux logo.
2182                        This may be used to provide more screen space for
2183                        kernel log messages and is useful when debugging
2184                        kernel boot problems.
2186        lp=0            [LP]    Specify parallel ports to use, e.g,
2187        lp=port[,port...]       lp=none,parport0 (lp0 not configured, lp1 uses
2188        lp=reset                first parallel port). 'lp=0' disables the
2189        lp=auto                 printer driver. 'lp=reset' (which can be
2190                                specified in addition to the ports) causes
2191                                attached printers to be reset. Using
2192                                lp=port1,port2,... specifies the parallel ports
2193                                to associate lp devices with, starting with
2194                                lp0. A port specification may be 'none' to skip
2195                                that lp device, or a parport name such as
2196                                'parport0'. Specifying 'lp=auto' instead of a
2197                                port specification list means that device IDs
2198                                from each port should be examined, to see if
2199                                an IEEE 1284-compliant printer is attached; if
2200                                so, the driver will manage that printer.
2201                                See also header of drivers/char/lp.c.
2203        lpj=n           [KNL]
2204                        Sets loops_per_jiffy to given constant, thus avoiding
2205                        time-consuming boot-time autodetection (up to 250 ms per
2206                        CPU). 0 enables autodetection (default). To determine
2207                        the correct value for your kernel, boot with normal
2208                        autodetection and see what value is printed. Note that
2209                        on SMP systems the preset will be applied to all CPUs,
2210                        which is likely to cause problems if your CPUs need
2211                        significantly divergent settings. An incorrect value
2212                        will cause delays in the kernel to be wrong, leading to
2213                        unpredictable I/O errors and other breakage. Although
2214                        unlikely, in the extreme case this might damage your
2215                        hardware.
2217        ltpc=           [NET]
2218                        Format: <io>,<irq>,<dma>
2220        machvec=        [IA-64] Force the use of a particular machine-vector
2221                        (machvec) in a generic kernel.
2222                        Example: machvec=hpzx1_swiotlb
2224        machtype=       [Loongson] Share the same kernel image file between different
2225                         yeeloong laptop.
2226                        Example: machtype=lemote-yeeloong-2f-7inch
2228        max_addr=nn[KMG]        [KNL,BOOT,ia64] All physical memory greater
2229                        than or equal to this physical address is ignored.
2231        maxcpus=        [SMP] Maximum number of processors that an SMP kernel
2232                        will bring up during bootup.  maxcpus=n : n >= 0 limits
2233                        the kernel to bring up 'n' processors. Surely after
2234                        bootup you can bring up the other plugged cpu by executing
2235                        "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus
2236                        only takes effect during system bootup.
2237                        While n=0 is a special case, it is equivalent to "nosmp",
2238                        which also disables the IO APIC.
2240        max_loop=       [LOOP] The number of loop block devices that get
2241        (loop.max_loop) unconditionally pre-created at init time. The default
2242                        number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
2243                        of statically allocating a predefined number, loop
2244                        devices can be requested on-demand with the
2245                        /dev/loop-control interface.
2247        mce             [X86-32] Machine Check Exception
2249        mce=option      [X86-64] See Documentation/x86/x86_64/boot-options.txt
2251        md=             [HW] RAID subsystems devices and level
2252                        See Documentation/md.txt.
2254        mdacon=         [MDA]
2255                        Format: <first>,<last>
2256                        Specifies range of consoles to be captured by the MDA.
2258        mem=nn[KMG]     [KNL,BOOT] Force usage of a specific amount of memory
2259                        Amount of memory to be used when the kernel is not able
2260                        to see the whole system memory or for test.
2261                        [X86] Work as limiting max address. Use together
2262                        with memmap= to avoid physical address space collisions.
2263                        Without memmap= PCI devices could be placed at addresses
2264                        belonging to unused RAM.
2266        mem=nopentium   [BUGS=X86-32] Disable usage of 4MB pages for kernel
2267                        memory.
2269        memchunk=nn[KMG]
2270                        [KNL,SH] Allow user to override the default size for
2271                        per-device physically contiguous DMA buffers.
2273        memhp_default_state=online/offline
2274                        [KNL] Set the initial state for the memory hotplug
2275                        onlining policy. If not specified, the default value is
2276                        set according to the
2277                        CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config
2278                        option.
2279                        See Documentation/memory-hotplug.txt.
2281        memmap=exactmap [KNL,X86] Enable setting of an exact
2282                        E820 memory map, as specified by the user.
2283                        Such memmap=exactmap lines can be constructed based on
2284                        BIOS output or other requirements. See the memmap=nn@ss
2285                        option description.
2287        memmap=nn[KMG]@ss[KMG]
2288                        [KNL] Force usage of a specific region of memory.
2289                        Region of memory to be used is from ss to ss+nn.
2291        memmap=nn[KMG]#ss[KMG]
2292                        [KNL,ACPI] Mark specific memory as ACPI data.
2293                        Region of memory to be marked is from ss to ss+nn.
2295        memmap=nn[KMG]$ss[KMG]
2296                        [KNL,ACPI] Mark specific memory as reserved.
2297                        Region of memory to be reserved is from ss to ss+nn.
2298                        Example: Exclude memory from 0x18690000-0x1869ffff
2299                                 memmap=64K$0x18690000
2300                                 or
2301                                 memmap=0x10000$0x18690000
2303        memmap=nn[KMG]!ss[KMG]
2304                        [KNL,X86] Mark specific memory as protected.
2305                        Region of memory to be used, from ss to ss+nn.
2306                        The memory region may be marked as e820 type 12 (0xc)
2307                        and is NVDIMM or ADR memory.
2309        memory_corruption_check=0/1 [X86]
2310                        Some BIOSes seem to corrupt the first 64k of
2311                        memory when doing things like suspend/resume.
2312                        Setting this option will scan the memory
2313                        looking for corruption.  Enabling this will
2314                        both detect corruption and prevent the kernel
2315                        from using the memory being corrupted.
2316                        However, its intended as a diagnostic tool; if
2317                        repeatable BIOS-originated corruption always
2318                        affects the same memory, you can use memmap=
2319                        to prevent the kernel from using that memory.
2321        memory_corruption_check_size=size [X86]
2322                        By default it checks for corruption in the low
2323                        64k, making this memory unavailable for normal
2324                        use.  Use this parameter to scan for
2325                        corruption in more or less memory.
2327        memory_corruption_check_period=seconds [X86]
2328                        By default it checks for corruption every 60
2329                        seconds.  Use this parameter to check at some
2330                        other rate.  0 disables periodic checking.
2332        memtest=        [KNL,X86,ARM] Enable memtest
2333                        Format: <integer>
2334                        default : 0 <disable>
2335                        Specifies the number of memtest passes to be
2336                        performed. Each pass selects another test
2337                        pattern from a given set of patterns. Memtest
2338                        fills the memory with this pattern, validates
2339                        memory contents and reserves bad memory
2340                        regions that are detected.
2342        meye.*=         [HW] Set MotionEye Camera parameters
2343                        See Documentation/video4linux/meye.txt.
2345        mfgpt_irq=      [IA-32] Specify the IRQ to use for the
2346                        Multi-Function General Purpose Timers on AMD Geode
2347                        platforms.
2349        mfgptfix        [X86-32] Fix MFGPT timers on AMD Geode platforms when
2350                        the BIOS has incorrectly applied a workaround. TinyBIOS
2351                        version 0.98 is known to be affected, 0.99 fixes the
2352                        problem by letting the user disable the workaround.
2354        mga=            [HW,DRM]
2356        min_addr=nn[KMG]        [KNL,BOOT,ia64] All physical memory below this
2357                        physical address is ignored.
2359        mini2440=       [ARM,HW,KNL]
2360                        Format:[0..2][b][c][t]
2361                        Default: "0tb"
2362                        MINI2440 configuration specification:
2363                        0 - The attached screen is the 3.5" TFT
2364                        1 - The attached screen is the 7" TFT
2365                        2 - The VGA Shield is attached (1024x768)
2366                        Leaving out the screen size parameter will not load
2367                        the TFT driver, and the framebuffer will be left
2368                        unconfigured.
2369                        b - Enable backlight. The TFT backlight pin will be
2370                        linked to the kernel VESA blanking code and a GPIO
2371                        LED. This parameter is not necessary when using the
2372                        VGA shield.
2373                        c - Enable the s3c camera interface.
2374                        t - Reserved for enabling touchscreen support. The
2375                        touchscreen support is not enabled in the mainstream
2376                        kernel as of 2.6.30, a preliminary port can be found
2377                        in the "bleeding edge" mini2440 support kernel at
2380        mminit_loglevel=
2381                        [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
2382                        parameter allows control of the logging verbosity for
2383                        the additional memory initialisation checks. A value
2384                        of 0 disables mminit logging and a level of 4 will
2385                        log everything. Information is printed at KERN_DEBUG
2386                        so loglevel=8 may also need to be specified.
2388        module.sig_enforce
2389                        [KNL] When CONFIG_MODULE_SIG is set, this means that
2390                        modules without (valid) signatures will fail to load.
2391                        Note that if CONFIG_MODULE_SIG_FORCE is set, that
2392                        is always true, so this option does nothing.
2394        module_blacklist=  [KNL] Do not load a comma-separated list of
2395                        modules.  Useful for debugging problem modules.
2397        mousedev.tap_time=
2398                        [MOUSE] Maximum time between finger touching and
2399                        leaving touchpad surface for touch to be considered
2400                        a tap and be reported as a left button click (for
2401                        touchpads working in absolute mode only).
2402                        Format: <msecs>
2403        mousedev.xres=  [MOUSE] Horizontal screen resolution, used for devices
2404                        reporting absolute coordinates, such as tablets
2405        mousedev.yres=  [MOUSE] Vertical screen resolution, used for devices
2406                        reporting absolute coordinates, such as tablets
2408        movablecore=nn[KMG]     [KNL,X86,IA-64,PPC] This parameter
2409                        is similar to kernelcore except it specifies the
2410                        amount of memory used for migratable allocations.
2411                        If both kernelcore and movablecore is specified,
2412                        then kernelcore will be at *least* the specified
2413                        value but may be more. If movablecore on its own
2414                        is specified, the administrator must be careful
2415                        that the amount of memory usable for all allocations
2416                        is not too small.
2418        movable_node    [KNL,X86] Boot-time switch to enable the effects
2419                        of CONFIG_MOVABLE_NODE=y. See mm/Kconfig for details.
2421        MTD_Partition=  [MTD]
2422                        Format: <name>,<region-number>,<size>,<offset>
2424        MTD_Region=     [MTD] Format:
2425                        <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
2427        mtdparts=       [MTD]
2428                        See drivers/mtd/cmdlinepart.c.
2430        multitce=off    [PPC]  This parameter disables the use of the pSeries
2431                        firmware feature for updating multiple TCE entries
2432                        at a time.
2434        onenand.bdry=   [HW,MTD] Flex-OneNAND Boundary Configuration
2436                        Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
2438                        boundary - index of last SLC block on Flex-OneNAND.
2439                                   The remaining blocks are configured as MLC blocks.
2440                        lock     - Configure if Flex-OneNAND boundary should be locked.
2441                                   Once locked, the boundary cannot be changed.
2442                                   1 indicates lock status, 0 indicates unlock status.
2444        mtdset=         [ARM]
2445                        ARM/S3C2412 JIVE boot control
2447                        See arch/arm/mach-s3c2412/mach-jive.c
2449        mtouchusb.raw_coordinates=
2450                        [HW] Make the MicroTouch USB driver use raw coordinates
2451                        ('y', default) or cooked coordinates ('n')
2453        mtrr_chunk_size=nn[KMG] [X86]
2454                        used for mtrr cleanup. It is largest continuous chunk
2455                        that could hold holes aka. UC entries.
2457        mtrr_gran_size=nn[KMG] [X86]
2458                        Used for mtrr cleanup. It is granularity of mtrr block.
2459                        Default is 1.
2460                        Large value could prevent small alignment from
2461                        using up MTRRs.
2463        mtrr_spare_reg_nr=n [X86]
2464                        Format: <integer>
2465                        Range: 0,7 : spare reg number
2466                        Default : 1
2467                        Used for mtrr cleanup. It is spare mtrr entries number.
2468                        Set to 2 or more if your graphical card needs more.
2470        n2=             [NET] SDL Inc. RISCom/N2 synchronous serial card
2472        netdev=         [NET] Network devices parameters
2473                        Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
2474                        Note that mem_start is often overloaded to mean
2475                        something different and driver-specific.
2476                        This usage is only documented in each driver source
2477                        file if at all.
2479        nf_conntrack.acct=
2480                        [NETFILTER] Enable connection tracking flow accounting
2481                        0 to disable accounting
2482                        1 to enable accounting
2483                        Default value is 0.
2485        nfsaddrs=       [NFS] Deprecated.  Use ip= instead.
2486                        See Documentation/filesystems/nfs/nfsroot.txt.
2488        nfsroot=        [NFS] nfs root filesystem for disk-less boxes.
2489                        See Documentation/filesystems/nfs/nfsroot.txt.
2491        nfsrootdebug    [NFS] enable nfsroot debugging messages.
2492                        See Documentation/filesystems/nfs/nfsroot.txt.
2494        nfs.callback_nr_threads=
2495                        [NFSv4] set the total number of threads that the
2496                        NFS client will assign to service NFSv4 callback
2497                        requests.
2499        nfs.callback_tcpport=
2500                        [NFS] set the TCP port on which the NFSv4 callback
2501                        channel should listen.
2503        nfs.cache_getent=
2504                        [NFS] sets the pathname to the program which is used
2505                        to update the NFS client cache entries.
2507        nfs.cache_getent_timeout=
2508                        [NFS] sets the timeout after which an attempt to
2509                        update a cache entry is deemed to have failed.
2511        nfs.idmap_cache_timeout=
2512                        [NFS] set the maximum lifetime for idmapper cache
2513                        entries.
2515        nfs.enable_ino64=
2516                        [NFS] enable 64-bit inode numbers.
2517                        If zero, the NFS client will fake up a 32-bit inode
2518                        number for the readdir() and stat() syscalls instead
2519                        of returning the full 64-bit number.
2520                        The default is to return 64-bit inode numbers.
2522        nfs.max_session_cb_slots=
2523                        [NFSv4.1] Sets the maximum number of session
2524                        slots the client will assign to the callback
2525                        channel. This determines the maximum number of
2526                        callbacks the client will process in parallel for
2527                        a particular server.
2529        nfs.max_session_slots=
2530                        [NFSv4.1] Sets the maximum number of session slots
2531                        the client will attempt to negotiate with the server.
2532                        This limits the number of simultaneous RPC requests
2533                        that the client can send to the NFSv4.1 server.
2534                        Note that there is little point in setting this
2535                        value higher than the max_tcp_slot_table_limit.
2537        nfs.nfs4_disable_idmapping=
2538                        [NFSv4] When set to the default of '1', this option
2539                        ensures that both the RPC level authentication
2540                        scheme and the NFS level operations agree to use
2541                        numeric uids/gids if the mount is using the
2542                        'sec=sys' security flavour. In effect it is
2543                        disabling idmapping, which can make migration from
2544                        legacy NFSv2/v3 systems to NFSv4 easier.
2545                        Servers that do not support this mode of operation
2546                        will be autodetected by the client, and it will fall
2547                        back to using the idmapper.
2548                        To turn off this behaviour, set the value to '0'.
2549        nfs.nfs4_unique_id=
2550                        [NFS4] Specify an additional fixed unique ident-
2551                        ification string that NFSv4 clients can insert into
2552                        their nfs_client_id4 string.  This is typically a
2553                        UUID that is generated at system install time.
2555        nfs.send_implementation_id =
2556                        [NFSv4.1] Send client implementation identification
2557                        information in exchange_id requests.
2558                        If zero, no implementation identification information
2559                        will be sent.
2560                        The default is to send the implementation identification
2561                        information.
2563        nfs.recover_lost_locks =
2564                        [NFSv4] Attempt to recover locks that were lost due
2565                        to a lease timeout on the server. Please note that
2566                        doing this risks data corruption, since there are
2567                        no guarantees that the file will remain unchanged
2568                        after the locks are lost.
2569                        If you want to enable the kernel legacy behaviour of
2570                        attempting to recover these locks, then set this
2571                        parameter to '1'.
2572                        The default parameter value of '0' causes the kernel
2573                        not to attempt recovery of lost locks.
2575        nfs4.layoutstats_timer =
2576                        [NFSv4.2] Change the rate at which the kernel sends
2577                        layoutstats to the pNFS metadata server.
2579                        Setting this to value to 0 causes the kernel to use
2580                        whatever value is the default set by the layout
2581                        driver. A non-zero value sets the minimum interval
2582                        in seconds between layoutstats transmissions.
2584        nfsd.nfs4_disable_idmapping=
2585                        [NFSv4] When set to the default of '1', the NFSv4
2586                        server will return only numeric uids and gids to
2587                        clients using auth_sys, and will accept numeric uids
2588                        and gids from such clients.  This is intended to ease
2589                        migration from NFSv2/v3.
2591        objlayoutdriver.osd_login_prog=
2592                        [NFS] [OBJLAYOUT] sets the pathname to the program which
2593                        is used to automatically discover and login into new
2594                        osd-targets. Please see:
2595                        Documentation/filesystems/pnfs.txt for more explanations
2597        nmi_debug=      [KNL,AVR32,SH] Specify one or more actions to take
2598                        when a NMI is triggered.
2599                        Format: [state][,regs][,debounce][,die]
2601        nmi_watchdog=   [KNL,BUGS=X86] Debugging features for SMP kernels
2602                        Format: [panic,][nopanic,][num]
2603                        Valid num: 0 or 1
2604                        0 - turn hardlockup detector in nmi_watchdog off
2605                        1 - turn hardlockup detector in nmi_watchdog on
2606                        When panic is specified, panic when an NMI watchdog
2607                        timeout occurs (or 'nopanic' to override the opposite
2608                        default). To disable both hard and soft lockup detectors,
2609                        please see 'nowatchdog'.
2610                        This is useful when you use a panic=... timeout and
2611                        need the box quickly up again.
2613        netpoll.carrier_timeout=
2614                        [NET] Specifies amount of time (in seconds) that
2615                        netpoll should wait for a carrier. By default netpoll
2616                        waits 4 seconds.
2618        no387           [BUGS=X86-32] Tells the kernel to use the 387 maths
2619                        emulation library even if a 387 maths coprocessor
2620                        is present.
2622        no_console_suspend
2623                        [HW] Never suspend the console
2624                        Disable suspending of consoles during suspend and
2625                        hibernate operations.  Once disabled, debugging
2626                        messages can reach various consoles while the rest
2627                        of the system is being put to sleep (ie, while
2628                        debugging driver suspend/resume hooks).  This may
2629                        not work reliably with all consoles, but is known
2630                        to work with serial and VGA consoles.
2631                        To facilitate more flexible debugging, we also add
2632                        console_suspend, a printk module parameter to control
2633                        it. Users could use console_suspend (usually
2634                        /sys/module/printk/parameters/console_suspend) to
2635                        turn on/off it dynamically.
2637        noaliencache    [MM, NUMA, SLAB] Disables the allocation of alien
2638                        caches in the slab allocator.  Saves per-node memory,
2639                        but will impact performance.
2641        noalign         [KNL,ARM]
2643        noapic          [SMP,APIC] Tells the kernel to not make use of any
2644                        IOAPICs that may be present in the system.
2646        noautogroup     Disable scheduler automatic task group creation.
2648        nobats          [PPC] Do not use BATs for mapping kernel lowmem
2649                        on "Classic" PPC cores.
2651        nocache         [ARM]
2653        noclflush       [BUGS=X86] Don't use the CLFLUSH instruction
2655        nodelayacct     [KNL] Disable per-task delay accounting
2657        nodsp           [SH] Disable hardware DSP at boot time.
2659        noefi           Disable EFI runtime services support.
2661        noexec          [IA-64]
2663        noexec          [X86]
2664                        On X86-32 available only on PAE configured kernels.
2665                        noexec=on: enable non-executable mappings (default)
2666                        noexec=off: disable non-executable mappings
2668        nosmap          [X86]
2669                        Disable SMAP (Supervisor Mode Access Prevention)
2670                        even if it is supported by processor.
2672        nosmep          [X86]
2673                        Disable SMEP (Supervisor Mode Execution Prevention)
2674                        even if it is supported by processor.
2676        noexec32        [X86-64]
2677                        This affects only 32-bit executables.
2678                        noexec32=on: enable non-executable mappings (default)
2679                                read doesn't imply executable mappings
2680                        noexec32=off: disable non-executable mappings
2681                                read implies executable mappings
2683        nofpu           [MIPS,SH] Disable hardware FPU at boot time.
2685        nofxsr          [BUGS=X86-32] Disables x86 floating point extended
2686                        register save and restore. The kernel will only save
2687                        legacy floating-point registers on task switch.
2689        nohugeiomap     [KNL,x86] Disable kernel huge I/O mappings.
2691        nosmt           [KNL,S390] Disable symmetric multithreading (SMT).
2692                        Equivalent to smt=1.
2694        noxsave         [BUGS=X86] Disables x86 extended register state save
2695                        and restore using xsave. The kernel will fallback to
2696                        enabling legacy floating-point and sse state.
2698        noxsaveopt      [X86] Disables xsaveopt used in saving x86 extended
2699                        register states. The kernel will fall back to use
2700                        xsave to save the states. By using this parameter,
2701                        performance of saving the states is degraded because
2702                        xsave doesn't support modified optimization while
2703                        xsaveopt supports it on xsaveopt enabled systems.
2705        noxsaves        [X86] Disables xsaves and xrstors used in saving and
2706                        restoring x86 extended register state in compacted
2707                        form of xsave area. The kernel will fall back to use
2708                        xsaveopt and xrstor to save and restore the states
2709                        in standard form of xsave area. By using this
2710                        parameter, xsave area per process might occupy more
2711                        memory on xsaves enabled systems.
2713        nohlt           [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
2714                        wfi(ARM) instruction doesn't work correctly and not to
2715                        use it. This is also useful when using JTAG debugger.
2717        no_file_caps    Tells the kernel not to honor file capabilities.  The
2718                        only way then for a file to be executed with privilege
2719                        is to be setuid root or executed by root.
2721        nohalt          [IA-64] Tells the kernel not to use the power saving
2722                        function PAL_HALT_LIGHT when idle. This increases
2723                        power-consumption. On the positive side, it reduces
2724                        interrupt wake-up latency, which may improve performance
2725                        in certain environments such as networked servers or
2726                        real-time systems.
2728        nohibernate     [HIBERNATION] Disable hibernation and resume.
2730        nohz=           [KNL] Boottime enable/disable dynamic ticks
2731                        Valid arguments: on, off
2732                        Default: on
2734        nohz_full=      [KNL,BOOT]
2735                        The argument is a cpu list, as described above.
2736                        In kernels built with CONFIG_NO_HZ_FULL=y, set
2737                        the specified list of CPUs whose tick will be stopped
2738                        whenever possible. The boot CPU will be forced outside
2739                        the range to maintain the timekeeping.
2740                        The CPUs in this range must also be included in the
2741                        rcu_nocbs= set.
2743        noiotrap        [SH] Disables trapped I/O port accesses.
2745        noirqdebug      [X86-32] Disables the code which attempts to detect and
2746                        disable unhandled interrupt sources.
2748        no_timer_check  [X86,APIC] Disables the code which tests for
2749                        broken timer IRQ sources.
2751        noisapnp        [ISAPNP] Disables ISA PnP code.
2753        noinitrd        [RAM] Tells the kernel not to load any configured
2754                        initial RAM disk.
2756        nointremap      [X86-64, Intel-IOMMU] Do not enable interrupt
2757                        remapping.
2758                        [Deprecated - use intremap=off]
2760        nointroute      [IA-64]
2762        noinvpcid       [X86] Disable the INVPCID cpu feature.
2764        nojitter        [IA-64] Disables jitter checking for ITC timers.
2766        no-kvmclock     [X86,KVM] Disable paravirtualized KVM clock driver
2768        no-kvmapf       [X86,KVM] Disable paravirtualized asynchronous page
2769                        fault handling.
2771        no-steal-acc    [X86,KVM] Disable paravirtualized steal time accounting.
2772                        steal time is computed, but won't influence scheduler
2773                        behaviour
2775        nolapic         [X86-32,APIC] Do not enable or use the local APIC.
2777        nolapic_timer   [X86-32,APIC] Do not use the local APIC timer.
2779        noltlbs         [PPC] Do not use large page/tlb entries for kernel
2780                        lowmem mapping on PPC40x and PPC8xx
2782        nomca           [IA-64] Disable machine check abort handling
2784        nomce           [X86-32] Disable Machine Check Exception
2786        nomfgpt         [X86-32] Disable Multi-Function General Purpose
2787                        Timer usage (for AMD Geode machines).
2789        nonmi_ipi       [X86] Disable using NMI IPIs during panic/reboot to
2790                        shutdown the other cpus.  Instead use the REBOOT_VECTOR
2791                        irq.
2793        nomodule        Disable module load
2795        nopat           [X86] Disable PAT (page attribute table extension of
2796                        pagetables) support.
2798        norandmaps      Don't use address space randomization.  Equivalent to
2799                        echo 0 > /proc/sys/kernel/randomize_va_space
2801        noreplace-paravirt      [X86,IA-64,PV_OPS] Don't patch paravirt_ops
2803        noreplace-smp   [X86-32,SMP] Don't replace SMP instructions
2804                        with UP alternatives
2806        nordrand        [X86] Disable kernel use of the RDRAND and
2807                        RDSEED instructions even if they are supported
2808                        by the processor.  RDRAND and RDSEED are still
2809                        available to user space applications.
2811        noresume        [SWSUSP] Disables resume and restores original swap
2812                        space.
2814        no-scroll       [VGA] Disables scrollback.
2815                        This is required for the Braillex ib80-piezo Braille
2816                        reader made by F.H. Papenmeier (Germany).
2818        nosbagart       [IA-64]
2820        nosep           [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
2822        nosmp           [SMP] Tells an SMP kernel to act as a UP kernel,
2823                        and disable the IO APIC.  legacy for "maxcpus=0".
2825        nosoftlockup    [KNL] Disable the soft-lockup detector.
2827        nosync          [HW,M68K] Disables sync negotiation for all devices.
2829        notsc           [BUGS=X86-32] Disable Time Stamp Counter
2831        nowatchdog      [KNL] Disable both lockup detectors, i.e.
2832                        soft-lockup and NMI watchdog (hard-lockup).
2834        nowb            [ARM]
2836        nox2apic        [X86-64,APIC] Do not enable x2APIC mode.
2838        cpu0_hotplug    [X86] Turn on CPU0 hotplug feature when
2839                        CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
2840                        Some features depend on CPU0. Known dependencies are:
2841                        1. Resume from suspend/hibernate depends on CPU0.
2842                        Suspend/hibernate will fail if CPU0 is offline and you
2843                        need to online CPU0 before suspend/hibernate.
2844                        2. PIC interrupts also depend on CPU0. CPU0 can't be
2845                        removed if a PIC interrupt is detected.
2846                        It's said poweroff/reboot may depend on CPU0 on some
2847                        machines although I haven't seen such issues so far
2848                        after CPU0 is offline on a few tested machines.
2849                        If the dependencies are under your control, you can
2850                        turn on cpu0_hotplug.
2852        nptcg=          [IA-64] Override max number of concurrent global TLB
2853                        purges which is reported from either PAL_VM_SUMMARY or
2854                        SAL PALO.
2856        nr_cpus=        [SMP] Maximum number of processors that an SMP kernel
2857                        could support.  nr_cpus=n : n >= 1 limits the kernel to
2858                        support 'n' processors. It could be larger than the
2859                        number of already plugged CPU during bootup, later in
2860                        runtime you can physically add extra cpu until it reaches
2861                        n. So during boot up some boot time memory for per-cpu
2862                        variables need be pre-allocated for later physical cpu
2863                        hot plugging.
2865        nr_uarts=       [SERIAL] maximum number of UARTs to be registered.
2867        numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing.
2868                        Allowed values are enable and disable
2870        numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
2871                        one of ['zone', 'node', 'default'] can be specified
2872                        This can be set from sysctl after boot.
2873                        See Documentation/sysctl/vm.txt for details.
2875        ohci1394_dma=early      [HW] enable debugging via the ohci1394 driver.
2876                        See Documentation/debugging-via-ohci1394.txt for more
2877                        info.
2879        olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
2880                        Rather than timing out after 20 ms if an EC
2881                        command is not properly ACKed, override the length
2882                        of the timeout.  We have interrupts disabled while
2883                        waiting for the ACK, so if this is set too high
2884                        interrupts *may* be lost!
2886        omap_mux=       [OMAP] Override bootloader pin multiplexing.
2887                        Format: <mux_mode0.mode_name=value>...
2888                        For example, to override I2C bus2:
2889                        omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
2891        oprofile.timer= [HW]
2892                        Use timer interrupt instead of performance counters
2894        oprofile.cpu_type=      Force an oprofile cpu type
2895                        This might be useful if you have an older oprofile
2896                        userland or if you want common events.
2897                        Format: { arch_perfmon }
2898                        arch_perfmon: [X86] Force use of architectural
2899                                perfmon on Intel CPUs instead of the
2900                                CPU specific event set.
2901                        timer: [X86] Force use of architectural NMI
2902                                timer mode (see also oprofile.timer
2903                                for generic hr timer mode)
2905        oops=panic      Always panic on oopses. Default is to just kill the
2906                        process, but there is a small probability of
2907                        deadlocking the machine.
2908                        This will also cause panics on machine check exceptions.
2909                        Useful together with panic=30 to trigger a reboot.
2911        OSS             [HW,OSS]
2912                        See Documentation/sound/oss/oss-parameters.txt
2914        page_owner=     [KNL] Boot-time page_owner enabling option.
2915                        Storage of the information about who allocated
2916                        each page is disabled in default. With this switch,
2917                        we can turn it on.
2918                        on: enable the feature
2920        page_poison=    [KNL] Boot-time parameter changing the state of
2921                        poisoning on the buddy allocator.
2922                        off: turn off poisoning
2923                        on: turn on poisoning
2925        panic=          [KNL] Kernel behaviour on panic: delay <timeout>
2926                        timeout > 0: seconds before rebooting
2927                        timeout = 0: wait forever
2928                        timeout < 0: reboot immediately
2929                        Format: <timeout>
2931        panic_on_warn   panic() instead of WARN().  Useful to cause kdump
2932                        on a WARN().
2934        crash_kexec_post_notifiers
2935                        Run kdump after running panic-notifiers and dumping
2936                        kmsg. This only for the users who doubt kdump always
2937                        succeeds in any situation.
2938                        Note that this also increases risks of kdump failure,
2939                        because some panic notifiers can make the crashed
2940                        kernel more unstable.
2942        parkbd.port=    [HW] Parallel port number the keyboard adapter is
2943                        connected to, default is 0.
2944                        Format: <parport#>
2945        parkbd.mode=    [HW] Parallel port keyboard adapter mode of operation,
2946                        0 for XT, 1 for AT (default is AT).
2947                        Format: <mode>
2949        parport=        [HW,PPT] Specify parallel ports. 0 disables.
2950                        Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
2951                        Use 'auto' to force the driver to use any
2952                        IRQ/DMA settings detected (the default is to
2953                        ignore detected IRQ/DMA settings because of
2954                        possible conflicts). You can specify the base
2955                        address, IRQ, and DMA settings; IRQ and DMA
2956                        should be numbers, or 'auto' (for using detected
2957                        settings on that particular port), or 'nofifo'
2958                        (to avoid using a FIFO even if it is detected).
2959                        Parallel ports are assigned in the order they
2960                        are specified on the command line, starting
2961                        with parport0.
2963        parport_init_mode=      [HW,PPT]
2964                        Configure VIA parallel port to operate in
2965                        a specific mode. This is necessary on Pegasos
2966                        computer where firmware has no options for setting
2967                        up parallel port mode and sets it to spp.
2968                        Currently this function knows 686a and 8231 chips.
2969                        Format: [spp|ps2|epp|ecp|ecpepp]
2971        pause_on_oops=
2972                        Halt all CPUs after the first oops has been printed for
2973                        the specified number of seconds.  This is to be used if
2974                        your oopses keep scrolling off the screen.
2976        pcbit=          [HW,ISDN]
2978        pcd.            [PARIDE]
2979                        See header of drivers/block/paride/pcd.c.
2980                        See also Documentation/blockdev/paride.txt.
2982        pci=option[,option...]  [PCI] various PCI subsystem options:
2983                earlydump       [X86] dump PCI config space before the kernel
2984                                changes anything
2985                off             [X86] don't probe for the PCI bus
2986                bios            [X86-32] force use of PCI BIOS, don't access
2987                                the hardware directly. Use this if your machine
2988                                has a non-standard PCI host bridge.
2989                nobios          [X86-32] disallow use of PCI BIOS, only direct
2990                                hardware access methods are allowed. Use this
2991                                if you experience crashes upon bootup and you
2992                                suspect they are caused by the BIOS.
2993                conf1           [X86] Force use of PCI Configuration Access
2994                                Mechanism 1 (config address in IO port 0xCF8,
2995                                data in IO port 0xCFC, both 32-bit).
2996                conf2           [X86] Force use of PCI Configuration Access
2997                                Mechanism 2 (IO port 0xCF8 is an 8-bit port for
2998                                the function, IO port 0xCFA, also 8-bit, sets
2999                                bus number. The config space is then accessed
3000                                through ports 0xC000-0xCFFF).
3001                                See for more info
3002                                on the configuration access mechanisms.
3003                noaer           [PCIE] If the PCIEAER kernel config parameter is
3004                                enabled, this kernel boot option can be used to
3005                                disable the use of PCIE advanced error reporting.
3006                nodomains       [PCI] Disable support for multiple PCI
3007                                root domains (aka PCI segments, in ACPI-speak).
3008                nommconf        [X86] Disable use of MMCONFIG for PCI
3009                                Configuration
3010                check_enable_amd_mmconf [X86] check for and enable
3011                                properly configured MMIO access to PCI
3012                                config space on AMD family 10h CPU
3013                nomsi           [MSI] If the PCI_MSI kernel config parameter is
3014                                enabled, this kernel boot option can be used to
3015                                disable the use of MSI interrupts system-wide.
3016                noioapicquirk   [APIC] Disable all boot interrupt quirks.
3017                                Safety option to keep boot IRQs enabled. This
3018                                should never be necessary.
3019                ioapicreroute   [APIC] Enable rerouting of boot IRQs to the
3020                                primary IO-APIC for bridges that cannot disable
3021                                boot IRQs. This fixes a source of spurious IRQs
3022                                when the system masks IRQs.
3023                noioapicreroute [APIC] Disable workaround that uses the
3024                                boot IRQ equivalent of an IRQ that connects to
3025                                a chipset where boot IRQs cannot be disabled.
3026                                The opposite of ioapicreroute.
3027                biosirq         [X86-32] Use PCI BIOS calls to get the interrupt
3028                                routing table. These calls are known to be buggy
3029                                on several machines and they hang the machine
3030                                when used, but on other computers it's the only
3031                                way to get the interrupt routing table. Try
3032                                this option if the kernel is unable to allocate
3033                                IRQs or discover secondary PCI buses on your
3034                                motherboard.
3035                rom             [X86] Assign address space to expansion ROMs.
3036                                Use with caution as certain devices share
3037                                address decoders between ROMs and other
3038                                resources.
3039                norom           [X86] Do not assign address space to
3040                                expansion ROMs that do not already have
3041                                BIOS assigned address ranges.
3042                nobar           [X86] Do not assign address space to the
3043                                BARs that weren't assigned by the BIOS.
3044                irqmask=0xMMMM  [X86] Set a bit mask of IRQs allowed to be
3045                                assigned automatically to PCI devices. You can
3046                                make the kernel exclude IRQs of your ISA cards
3047                                this way.
3048                pirqaddr=0xAAAAA        [X86] Specify the physical address
3049                                of the PIRQ table (normally generated
3050                                by the BIOS) if it is outside the
3051                                F0000h-100000h range.
3052                lastbus=N       [X86] Scan all buses thru bus #N. Can be
3053                                useful if the kernel is unable to find your
3054                                secondary buses and you want to tell it
3055                                explicitly which ones they are.
3056                assign-busses   [X86] Always assign all PCI bus
3057                                numbers ourselves, overriding
3058                                whatever the firmware may have done.
3059                usepirqmask     [X86] Honor the possible IRQ mask stored
3060                                in the BIOS $PIR table. This is needed on
3061                                some systems with broken BIOSes, notably
3062                                some HP Pavilion N5400 and Omnibook XE3
3063                                notebooks. This will have no effect if ACPI
3064                                IRQ routing is enabled.
3065                noacpi          [X86] Do not use ACPI for IRQ routing
3066                                or for PCI scanning.
3067                use_crs         [X86] Use PCI host bridge window information
3068                                from ACPI.  On BIOSes from 2008 or later, this
3069                                is enabled by default.  If you need to use this,
3070                                please report a bug.
3071                nocrs           [X86] Ignore PCI host bridge windows from ACPI.
3072                                If you need to use this, please report a bug.
3073                routeirq        Do IRQ routing for all PCI devices.
3074                                This is normally done in pci_enable_device(),
3075                                so this option is a temporary workaround
3076                                for broken drivers that don't call it.
3077                skip_isa_align  [X86] do not align io start addr, so can
3078                                handle more pci cards
3079                noearly         [X86] Don't do any early type 1 scanning.
3080                                This might help on some broken boards which
3081                                machine check when some devices' config space
3082                                is read. But various workarounds are disabled
3083                                and some IOMMU drivers will not work.
3084                bfsort          Sort PCI devices into breadth-first order.
3085                                This sorting is done to get a device
3086                                order compatible with older (<= 2.4) kernels.
3087                nobfsort        Don't sort PCI devices into breadth-first order.
3088                pcie_bus_tune_off       Disable PCIe MPS (Max Payload Size)
3089                                tuning and use the BIOS-configured MPS defaults.
3090                pcie_bus_safe   Set every device's MPS to the largest value
3091                                supported by all devices below the root complex.
3092                pcie_bus_perf   Set device MPS to the largest allowable MPS
3093                                based on its parent bus. Also set MRRS (Max
3094                                Read Request Size) to the largest supported
3095                                value (no larger than the MPS that the device
3096                                or bus can support) for best performance.
3097                pcie_bus_peer2peer      Set every device's MPS to 128B, which
3098                                every device is guaranteed to support. This
3099                                configuration allows peer-to-peer DMA between
3100                                any pair of devices, possibly at the cost of
3101                                reduced performance.  This also guarantees
3102                                that hot-added devices will work.
3103                cbiosize=nn[KMG]        The fixed amount of bus space which is
3104                                reserved for the CardBus bridge's IO window.
3105                                The default value is 256 bytes.
3106                cbmemsize=nn[KMG]       The fixed amount of bus space which is
3107                                reserved for the CardBus bridge's memory
3108                                window. The default value is 64 megabytes.
3109                resource_alignment=
3110                                Format:
3111                                [<order of align>@][<domain>:]<bus>:<slot>.<func>[; ...]
3112                                [<order of align>@]pci:<vendor>:<device>\
3113                                                [:<subvendor>:<subdevice>][; ...]
3114                                Specifies alignment and device to reassign
3115                                aligned memory resources.
3116                                If <order of align> is not specified,
3117                                PAGE_SIZE is used as alignment.
3118                                PCI-PCI bridge can be specified, if resource
3119                                windows need to be expanded.
3120                                To specify the alignment for several
3121                                instances of a device, the PCI vendor,
3122                                device, subvendor, and subdevice may be
3123                                specified, e.g., 4096@pci:8086:9c22:103c:198f
3124                ecrc=           Enable/disable PCIe ECRC (transaction layer
3125                                end-to-end CRC checking).
3126                                bios: Use BIOS/firmware settings. This is the
3127                                the default.
3128                                off: Turn ECRC off
3129                                on: Turn ECRC on.
3130                hpiosize=nn[KMG]        The fixed amount of bus space which is
3131                                reserved for hotplug bridge's IO window.
3132                                Default size is 256 bytes.
3133                hpmemsize=nn[KMG]       The fixed amount of bus space which is
3134                                reserved for hotplug bridge's memory window.
3135                                Default size is 2 megabytes.
3136                hpbussize=nn    The minimum amount of additional bus numbers
3137                                reserved for buses below a hotplug bridge.
3138                                Default is 1.
3139                realloc=        Enable/disable reallocating PCI bridge resources
3140                                if allocations done by BIOS are too small to
3141                                accommodate resources required by all child
3142                                devices.
3143                                off: Turn realloc off
3144                                on: Turn realloc on
3145                realloc         same as realloc=on
3146                noari           do not use PCIe ARI.
3147                pcie_scan_all   Scan all possible PCIe devices.  Otherwise we
3148                                only look for one device below a PCIe downstream
3149                                port.
3151        pcie_aspm=      [PCIE] Forcibly enable or disable PCIe Active State Power
3152                        Management.
3153                off     Disable ASPM.
3154                force   Enable ASPM even on devices that claim not to support it.
3155                        WARNING: Forcing ASPM on may cause system lockups.
3157        pcie_hp=        [PCIE] PCI Express Hotplug driver options:
3158                nomsi   Do not use MSI for PCI Express Native Hotplug (this
3159                        makes all PCIe ports use INTx for hotplug services).
3161        pcie_ports=     [PCIE] PCIe ports handling:
3162                auto    Ask the BIOS whether or not to use native PCIe services
3163                        associated with PCIe ports (PME, hot-plug, AER).  Use
3164                        them only if that is allowed by the BIOS.
3165                native  Use native PCIe services associated with PCIe ports
3166                        unconditionally.
3167                compat  Treat PCIe ports as PCI-to-PCI bridges, disable the PCIe
3168                        ports driver.
3170        pcie_port_pm=   [PCIE] PCIe port power management handling:
3171                off     Disable power management of all PCIe ports
3172                force   Forcibly enable power management of all PCIe ports
3174        pcie_pme=       [PCIE,PM] Native PCIe PME signaling options:
3175                nomsi   Do not use MSI for native PCIe PME signaling (this makes
3176                        all PCIe root ports use INTx for all services).
3178        pcmv=           [HW,PCMCIA] BadgePAD 4
3180        pd_ignore_unused
3181                        [PM]
3182                        Keep all power-domains already enabled by bootloader on,
3183                        even if no driver has claimed them. This is useful
3184                        for debug and development, but should not be
3185                        needed on a platform with proper driver support.
3187        pd.             [PARIDE]
3188                        See Documentation/blockdev/paride.txt.
3190        pdcchassis=     [PARISC,HW] Disable/Enable PDC Chassis Status codes at
3191                        boot time.
3192                        Format: { 0 | 1 }
3193                        See arch/parisc/kernel/pdc_chassis.c
3195        percpu_alloc=   Select which percpu first chunk allocator to use.
3196                        Currently supported values are "embed" and "page".
3197                        Archs may support subset or none of the selections.
3198                        See comments in mm/percpu.c for details on each
3199                        allocator.  This parameter is primarily for debugging
3200                        and performance comparison.
3202        pf.             [PARIDE]
3203                        See Documentation/blockdev/paride.txt.
3205        pg.             [PARIDE]
3206                        See Documentation/blockdev/paride.txt.
3208        pirq=           [SMP,APIC] Manual mp-table setup
3209                        See Documentation/x86/i386/IO-APIC.txt.
3211        plip=           [PPT,NET] Parallel port network link
3212                        Format: { parport<nr> | timid | 0 }
3213                        See also Documentation/parport.txt.
3215        pmtmr=          [X86] Manual setup of pmtmr I/O Port.
3216                        Override pmtimer IOPort with a hex value.
3217                        e.g. pmtmr=0x508
3219        pnp.debug=1     [PNP]
3220                        Enable PNP debug messages (depends on the
3221                        CONFIG_PNP_DEBUG_MESSAGES option).  Change at run-time
3222                        via /sys/module/pnp/parameters/debug.  We always show
3223                        current resource usage; turning this on also shows
3224                        possible settings and some assignment information.
3226        pnpacpi=        [ACPI]
3227                        { off }
3229        pnpbios=        [ISAPNP]
3230                        { on | off | curr | res | no-curr | no-res }
3232        pnp_reserve_irq=
3233                        [ISAPNP] Exclude IRQs for the autoconfiguration
3235        pnp_reserve_dma=
3236                        [ISAPNP] Exclude DMAs for the autoconfiguration
3238        pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
3239                        Ranges are in pairs (I/O port base and size).
3241        pnp_reserve_mem=
3242                        [ISAPNP] Exclude memory regions for the
3243                        autoconfiguration.
3244                        Ranges are in pairs (memory base and size).
3246        ports=          [IP_VS_FTP] IPVS ftp helper module
3247                        Default is 21.
3248                        Up to 8 (IP_VS_APP_MAX_PORTS) ports
3249                        may be specified.
3250                        Format: <port>,<port>....
3252        ppc_strict_facility_enable
3253                        [PPC] This option catches any kernel floating point,
3254                        Altivec, VSX and SPE outside of regions specifically
3255                        allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
3256                        There is some performance impact when enabling this.
3258        print-fatal-signals=
3259                        [KNL] debug: print fatal signals
3261                        If enabled, warn about various signal handling
3262                        related application anomalies: too many signals,
3263                        too many POSIX.1 timers, fatal signals causing a
3264                        coredump - etc.
3266                        If you hit the warning due to signal overflow,
3267                        you might want to try "ulimit -i unlimited".
3269                        default: off.
3271        printk.always_kmsg_dump=
3272                        Trigger kmsg_dump for cases other than kernel oops or
3273                        panics
3274                        Format: <bool>  (1/Y/y=enable, 0/N/n=disable)
3275                        default: disabled
3277        printk.devkmsg={on,off,ratelimit}
3278                        Control writing to /dev/kmsg.
3279                        on - unlimited logging to /dev/kmsg from userspace
3280                        off - logging to /dev/kmsg disabled
3281                        ratelimit - ratelimit the logging
3282                        Default: ratelimit
3284        printk.time=    Show timing data prefixed to each printk message line
3285                        Format: <bool>  (1/Y/y=enable, 0/N/n=disable)
3287        processor.max_cstate=   [HW,ACPI]
3288                        Limit processor to maximum C-state
3289                        max_cstate=9 overrides any DMI blacklist limit.
3291        processor.nocst [HW,ACPI]
3292                        Ignore the _CST method to determine C-states,
3293                        instead using the legacy FADT method
3295        profile=        [KNL] Enable kernel profiling via /proc/profile
3296                        Format: [schedule,]<number>
3297                        Param: "schedule" - profile schedule points.
3298                        Param: <number> - step/bucket size as a power of 2 for
3299                                statistical time based profiling.
3300                        Param: "sleep" - profile D-state sleeping (millisecs).
3301                                Requires CONFIG_SCHEDSTATS
3302                        Param: "kvm" - profile VM exits.
3304        prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
3305                        before loading.
3306                        See Documentation/blockdev/ramdisk.txt.
3308        psmouse.proto=  [HW,MOUSE] Highest PS2 mouse protocol extension to
3309                        probe for; one of (bare|imps|exps|lifebook|any).
3310        psmouse.rate=   [HW,MOUSE] Set desired mouse report rate, in reports
3311                        per second.
3312        psmouse.resetafter=     [HW,MOUSE]
3313                        Try to reset the device after so many bad packets
3314                        (0 = never).
3315        psmouse.resolution=
3316                        [HW,MOUSE] Set desired mouse resolution, in dpi.
3317        psmouse.smartscroll=
3318                        [HW,MOUSE] Controls Logitech smartscroll autorepeat.
3319                        0 = disabled, 1 = enabled (default).
3321        pstore.backend= Specify the name of the pstore backend to use
3323        pt.             [PARIDE]
3324                        See Documentation/blockdev/paride.txt.
3326        pty.legacy_count=
3327                        [KNL] Number of legacy pty's. Overwrites compiled-in
3328                        default number.
3330        quiet           [KNL] Disable most log messages
3332        r128=           [HW,DRM]
3334        raid=           [HW,RAID]
3335                        See Documentation/md.txt.
3337        ramdisk_size=   [RAM] Sizes of RAM disks in kilobytes
3338                        See Documentation/blockdev/ramdisk.txt.
3340        rcu_nocbs=      [KNL]
3341                        The argument is a cpu list, as described above.
3343                        In kernels built with CONFIG_RCU_NOCB_CPU=y, set
3344                        the specified list of CPUs to be no-callback CPUs.
3345                        Invocation of these CPUs' RCU callbacks will
3346                        be offloaded to "rcuox/N" kthreads created for
3347                        that purpose, where "x" is "b" for RCU-bh, "p"
3348                        for RCU-preempt, and "s" for RCU-sched, and "N"
3349                        is the CPU number.  This reduces OS jitter on the
3350                        offloaded CPUs, which can be useful for HPC and
3351                        real-time workloads.  It can also improve energy
3352                        efficiency for asymmetric multiprocessors.
3354        rcu_nocb_poll   [KNL]
3355                        Rather than requiring that offloaded CPUs
3356                        (specified by rcu_nocbs= above) explicitly
3357                        awaken the corresponding "rcuoN" kthreads,
3358                        make these kthreads poll for callbacks.
3359                        This improves the real-time response for the
3360                        offloaded CPUs by relieving them of the need to
3361                        wake up the corresponding kthread, but degrades
3362                        energy efficiency by requiring that the kthreads
3363                        periodically wake up to do the polling.
3365        rcutree.blimit= [KNL]
3366                        Set maximum number of finished RCU callbacks to
3367                        process in one batch.
3369        rcutree.dump_tree=      [KNL]
3370                        Dump the structure of the rcu_node combining tree
3371                        out at early boot.  This is used for diagnostic
3372                        purposes, to verify correct tree setup.
3374        rcutree.gp_cleanup_delay=       [KNL]
3375                        Set the number of jiffies to delay each step of
3376                        RCU grace-period cleanup.  This only has effect
3377                        when CONFIG_RCU_TORTURE_TEST_SLOW_CLEANUP is set.
3379        rcutree.gp_init_delay=  [KNL]
3380                        Set the number of jiffies to delay each step of
3381                        RCU grace-period initialization.  This only has
3382                        effect when CONFIG_RCU_TORTURE_TEST_SLOW_INIT
3383                        is set.
3385        rcutree.gp_preinit_delay=       [KNL]
3386                        Set the number of jiffies to delay each step of
3387                        RCU grace-period pre-initialization, that is,
3388                        the propagation of recent CPU-hotplug changes up
3389                        the rcu_node combining tree.  This only has effect
3390                        when CONFIG_RCU_TORTURE_TEST_SLOW_PREINIT is set.
3392        rcutree.rcu_fanout_exact= [KNL]
3393                        Disable autobalancing of the rcu_node combining
3394                        tree.  This is used by rcutorture, and might
3395                        possibly be useful for architectures having high
3396                        cache-to-cache transfer latencies.
3398        rcutree.rcu_fanout_leaf= [KNL]
3399                        Change the number of CPUs assigned to each
3400                        leaf rcu_node structure.  Useful for very
3401                        large systems, which will choose the value 64,
3402                        and for NUMA systems with large remote-access
3403                        latencies, which will choose a value aligned
3404                        with the appropriate hardware boundaries.
3406        rcutree.jiffies_till_sched_qs= [KNL]
3407                        Set required age in jiffies for a
3408                        given grace period before RCU starts
3409                        soliciting quiescent-state help from
3410                        rcu_note_context_switch().
3412        rcutree.jiffies_till_first_fqs= [KNL]
3413                        Set delay from grace-period initialization to
3414                        first attempt to force quiescent states.
3415                        Units are jiffies, minimum value is zero,
3416                        and maximum value is HZ.
3418        rcutree.jiffies_till_next_fqs= [KNL]
3419                        Set delay between subsequent attempts to force
3420                        quiescent states.  Units are jiffies, minimum
3421                        value is one, and maximum value is HZ.
3423        rcutree.kthread_prio=    [KNL,BOOT]
3424                        Set the SCHED_FIFO priority of the RCU per-CPU
3425                        kthreads (rcuc/N). This value is also used for
3426                        the priority of the RCU boost threads (rcub/N)
3427                        and for the RCU grace-period kthreads (rcu_bh,
3428                        rcu_preempt, and rcu_sched). If RCU_BOOST is
3429                        set, valid values are 1-99 and the default is 1
3430                        (the least-favored priority).  Otherwise, when
3431                        RCU_BOOST is not set, valid values are 0-99 and
3432                        the default is zero (non-realtime operation).
3434        rcutree.rcu_nocb_leader_stride= [KNL]
3435                        Set the number of NOCB kthread groups, which
3436                        defaults to the square root of the number of
3437                        CPUs.  Larger numbers reduces the wakeup overhead
3438                        on the per-CPU grace-period kthreads, but increases
3439                        that same overhead on each group's leader.
3441        rcutree.qhimark= [KNL]
3442                        Set threshold of queued RCU callbacks beyond which
3443                        batch limiting is disabled.
3445        rcutree.qlowmark= [KNL]
3446                        Set threshold of queued RCU callbacks below which
3447                        batch limiting is re-enabled.
3449        rcutree.rcu_idle_gp_delay= [KNL]
3450                        Set wakeup interval for idle CPUs that have
3451                        RCU callbacks (RCU_FAST_NO_HZ=y).
3453        rcutree.rcu_idle_lazy_gp_delay= [KNL]
3454                        Set wakeup interval for idle CPUs that have
3455                        only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
3456                        Lazy RCU callbacks are those which RCU can
3457                        prove do nothing more than free memory.
3459        rcuperf.gp_exp= [KNL]
3460                        Measure performance of expedited synchronous
3461                        grace-period primitives.
3463        rcuperf.holdoff= [KNL]
3464                        Set test-start holdoff period.  The purpose of
3465                        this parameter is to delay the start of the
3466                        test until boot completes in order to avoid
3467                        interference.
3469        rcuperf.nreaders= [KNL]
3470                        Set number of RCU readers.  The value -1 selects
3471                        N, where N is the number of CPUs.  A value
3472                        "n" less than -1 selects N-n+1, where N is again
3473                        the number of CPUs.  For example, -2 selects N
3474                        (the number of CPUs), -3 selects N+1, and so on.
3475                        A value of "n" less than or equal to -N selects
3476                        a single reader.
3478        rcuperf.nwriters= [KNL]
3479                        Set number of RCU writers.  The values operate
3480                        the same as for rcuperf.nreaders.
3481                        N, where N is the number of CPUs
3483        rcuperf.perf_runnable= [BOOT]
3484                        Start rcuperf running at boot time.
3486        rcuperf.shutdown= [KNL]
3487                        Shut the system down after performance tests
3488                        complete.  This is useful for hands-off automated
3489                        testing.
3491        rcuperf.perf_type= [KNL]
3492                        Specify the RCU implementation to test.
3494        rcuperf.verbose= [KNL]
3495                        Enable additional printk() statements.
3497        rcutorture.cbflood_inter_holdoff= [KNL]
3498                        Set holdoff time (jiffies) between successive
3499                        callback-flood tests.
3501        rcutorture.cbflood_intra_holdoff= [KNL]
3502                        Set holdoff time (jiffies) between successive
3503                        bursts of callbacks within a given callback-flood
3504                        test.
3506        rcutorture.cbflood_n_burst= [KNL]
3507                        Set the number of bursts making up a given
3508                        callback-flood test.  Set this to zero to
3509                        disable callback-flood testing.
3511        rcutorture.cbflood_n_per_burst= [KNL]
3512                        Set the number of callbacks to be registered
3513                        in a given burst of a callback-flood test.
3515        rcutorture.fqs_duration= [KNL]
3516                        Set duration of force_quiescent_state bursts
3517                        in microseconds.
3519        rcutorture.fqs_holdoff= [KNL]
3520                        Set holdoff time within force_quiescent_state bursts
3521                        in microseconds.
3523        rcutorture.fqs_stutter= [KNL]
3524                        Set wait time between force_quiescent_state bursts
3525                        in seconds.
3527        rcutorture.gp_cond= [KNL]
3528                        Use conditional/asynchronous update-side
3529                        primitives, if available.
3531        rcutorture.gp_exp= [KNL]
3532                        Use expedited update-side primitives, if available.
3534        rcutorture.gp_normal= [KNL]
3535                        Use normal (non-expedited) asynchronous
3536                        update-side primitives, if available.
3538        rcutorture.gp_sync= [KNL]
3539                        Use normal (non-expedited) synchronous
3540                        update-side primitives, if available.  If all
3541                        of rcutorture.gp_cond=, rcutorture.gp_exp=,
3542                        rcutorture.gp_normal=, and rcutorture.gp_sync=
3543                        are zero, rcutorture acts as if is interpreted
3544                        they are all non-zero.
3546        rcutorture.n_barrier_cbs= [KNL]
3547                        Set callbacks/threads for rcu_barrier() testing.
3549        rcutorture.nfakewriters= [KNL]
3550                        Set number of concurrent RCU writers.  These just
3551                        stress RCU, they don't participate in the actual
3552                        test, hence the "fake".
3554        rcutorture.nreaders= [KNL]
3555                        Set number of RCU readers.  The value -1 selects
3556                        N-1, where N is the number of CPUs.  A value
3557                        "n" less than -1 selects N-n-2, where N is again
3558                        the number of CPUs.  For example, -2 selects N
3559                        (the number of CPUs), -3 selects N+1, and so on.
3561        rcutorture.object_debug= [KNL]
3562                        Enable debug-object double-call_rcu() testing.
3564        rcutorture.onoff_holdoff= [KNL]
3565                        Set time (s) after boot for CPU-hotplug testing.
3567        rcutorture.onoff_interval= [KNL]
3568                        Set time (s) between CPU-hotplug operations, or
3569                        zero to disable CPU-hotplug testing.
3571        rcutorture.shuffle_interval= [KNL]
3572                        Set task-shuffle interval (s).  Shuffling tasks
3573                        allows some CPUs to go into dyntick-idle mode
3574                        during the rcutorture test.
3576        rcutorture.shutdown_secs= [KNL]
3577                        Set time (s) after boot system shutdown.  This
3578                        is useful for hands-off automated testing.
3580        rcutorture.stall_cpu= [KNL]
3581                        Duration of CPU stall (s) to test RCU CPU stall
3582                        warnings, zero to disable.
3584        rcutorture.stall_cpu_holdoff= [KNL]
3585                        Time to wait (s) after boot before inducing stall.
3587        rcutorture.stat_interval= [KNL]
3588                        Time (s) between statistics printk()s.
3590        rcutorture.stutter= [KNL]
3591                        Time (s) to stutter testing, for example, specifying
3592                        five seconds causes the test to run for five seconds,
3593                        wait for five seconds, and so on.  This tests RCU's
3594                        ability to transition abruptly to and from idle.
3596        rcutorture.test_boost= [KNL]
3597                        Test RCU priority boosting?  0=no, 1=maybe, 2=yes.
3598                        "Maybe" means test if the RCU implementation
3599                        under test support RCU priority boosting.
3601        rcutorture.test_boost_duration= [KNL]
3602                        Duration (s) of each individual boost test.
3604        rcutorture.test_boost_interval= [KNL]
3605                        Interval (s) between each boost test.
3607        rcutorture.test_no_idle_hz= [KNL]
3608                        Test RCU's dyntick-idle handling.  See also the
3609                        rcutorture.shuffle_interval parameter.
3611        rcutorture.torture_runnable= [BOOT]
3612                        Start rcutorture running at boot time.
3614        rcutorture.torture_type= [KNL]
3615                        Specify the RCU implementation to test.
3617        rcutorture.verbose= [KNL]
3618                        Enable additional printk() statements.
3620        rcupdate.rcu_cpu_stall_suppress= [KNL]
3621                        Suppress RCU CPU stall warning messages.
3623        rcupdate.rcu_cpu_stall_timeout= [KNL]
3624                        Set timeout for RCU CPU stall warning messages.
3626        rcupdate.rcu_expedited= [KNL]
3627                        Use expedited grace-period primitives, for
3628                        example, synchronize_rcu_expedited() instead
3629                        of synchronize_rcu().  This reduces latency,
3630                        but can increase CPU utilization, degrade
3631                        real-time latency, and degrade energy efficiency.
3632                        No effect on CONFIG_TINY_RCU kernels.
3634        rcupdate.rcu_normal= [KNL]
3635                        Use only normal grace-period primitives,
3636                        for example, synchronize_rcu() instead of
3637                        synchronize_rcu_expedited().  This improves
3638                        real-time latency, CPU utilization, and
3639                        energy efficiency, but can expose users to
3640                        increased grace-period latency.  This parameter
3641                        overrides rcupdate.rcu_expedited.  No effect on
3642                        CONFIG_TINY_RCU kernels.
3644        rcupdate.rcu_normal_after_boot= [KNL]
3645                        Once boot has completed (that is, after
3646                        rcu_end_inkernel_boot() has been invoked), use
3647                        only normal grace-period primitives.  No effect
3648                        on CONFIG_TINY_RCU kernels.
3650        rcupdate.rcu_task_stall_timeout= [KNL]
3651                        Set timeout in jiffies for RCU task stall warning
3652                        messages.  Disable with a value less than or equal
3653                        to zero.
3655        rcupdate.rcu_self_test= [KNL]
3656                        Run the RCU early boot self tests
3658        rcupdate.rcu_self_test_bh= [KNL]
3659                        Run the RCU bh early boot self tests
3661        rcupdate.rcu_self_test_sched= [KNL]
3662                        Run the RCU sched early boot self tests
3664        rdinit=         [KNL]
3665                        Format: <full_path>
3666                        Run specified binary instead of /init from the ramdisk,
3667                        used for early userspace startup. See initrd.
3669        reboot=         [KNL]
3670                        Format (x86 or x86_64):
3671                                [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
3672                                [[,]s[mp]#### \
3673                                [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
3674                                [[,]f[orce]
3675                        Where reboot_mode is one of warm (soft) or cold (hard) or gpio,
3676                              reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
3677                              reboot_force is either force or not specified,
3678                              reboot_cpu is s[mp]#### with #### being the processor
3679                                        to be used for rebooting.
3681        relax_domain_level=
3682                        [KNL, SMP] Set scheduler's default relax_domain_level.
3683                        See Documentation/cgroup-v1/cpusets.txt.
3685        relative_sleep_states=
3686                        [SUSPEND] Use sleep state labeling where the deepest
3687                        state available other than hibernation is always "mem".
3688                        Format: { "0" | "1" }
3689                        0 -- Traditional sleep state labels.
3690                        1 -- Relative sleep state labels.
3692        reserve=        [KNL,BUGS] Force the kernel to ignore some iomem area
3694        reservetop=     [X86-32]
3695                        Format: nn[KMG]
3696                        Reserves a hole at the top of the kernel virtual
3697                        address space.
3699        reservelow=     [X86]
3700                        Format: nn[K]
3701                        Set the amount of memory to reserve for BIOS at
3702                        the bottom of the address space.
3704        reset_devices   [KNL] Force drivers to reset the underlying device
3705                        during initialization.
3707        resume=         [SWSUSP]
3708                        Specify the partition device for software suspend
3709                        Format:
3710                        {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
3712        resume_offset=  [SWSUSP]
3713                        Specify the offset from the beginning of the partition
3714                        given by "resume=" at which the swap header is located,
3715                        in <PAGE_SIZE> units (needed only for swap files).
3716                        See  Documentation/power/swsusp-and-swap-files.txt
3718        resumedelay=    [HIBERNATION] Delay (in seconds) to pause before attempting to
3719                        read the resume files
3721        resumewait      [HIBERNATION] Wait (indefinitely) for resume device to show up.
3722                        Useful for devices that are detected asynchronously
3723                        (e.g. USB and MMC devices).
3725        hibernate=      [HIBERNATION]
3726                noresume        Don't check if there's a hibernation image
3727                                present during boot.
3728                nocompress      Don't compress/decompress hibernation images.
3729                no              Disable hibernation and resume.
3730                protect_image   Turn on image protection during restoration
3731                                (that will set all pages holding image data
3732                                during restoration read-only).
3734        retain_initrd   [RAM] Keep initrd memory after extraction
3736        rfkill.default_state=
3737                0       "airplane mode".  All wifi, bluetooth, wimax, gps, fm,
3738                        etc. communication is blocked by default.
3739                1       Unblocked.
3741        rfkill.master_switch_mode=
3742                0       The "airplane mode" button does nothing.
3743                1       The "airplane mode" button toggles between everything
3744                        blocked and the previous configuration.
3745                2       The "airplane mode" button toggles between everything
3746                        blocked and everything unblocked.
3748        rhash_entries=  [KNL,NET]
3749                        Set number of hash buckets for route cache
3751        ro              [KNL] Mount root device read-only on boot
3753        rodata=         [KNL]
3754                on      Mark read-only kernel memory as read-only (default).
3755                off     Leave read-only kernel memory writable for debugging.
3757        rockchip.usb_uart
3758                        Enable the uart passthrough on the designated usb port
3759                        on Rockchip SoCs. When active, the signals of the
3760                        debug-uart get routed to the D+ and D- pins of the usb
3761                        port and the regular usb controller gets disabled.
3763        root=           [KNL] Root filesystem
3764                        See name_to_dev_t comment in init/do_mounts.c.
3766        rootdelay=      [KNL] Delay (in seconds) to pause before attempting to
3767                        mount the root filesystem
3769        rootflags=      [KNL] Set root filesystem mount option string
3771        rootfstype=     [KNL] Set root filesystem type
3773        rootwait        [KNL] Wait (indefinitely) for root device to show up.
3774                        Useful for devices that are detected asynchronously
3775                        (e.g. USB and MMC devices).
3777        rproc_mem=nn[KMG][@address]
3778                        [KNL,ARM,CMA] Remoteproc physical memory block.
3779                        Memory area to be used by remote processor image,
3780                        managed by CMA.
3782        rw              [KNL] Mount root device read-write on boot
3784        S               [KNL] Run init in single mode
3786        s390_iommu=     [HW,S390]
3787                        Set s390 IOTLB flushing mode
3788                strict
3789                        With strict flushing every unmap operation will result in
3790                        an IOTLB flush. Default is lazy flushing before reuse,
3791                        which is faster.
3793        sa1100ir        [NET]
3794                        See drivers/net/irda/sa1100_ir.c.
3796        sbni=           [NET] Granch SBNI12 leased line adapter
3798        sched_debug     [KNL] Enables verbose scheduler debug messages.
3800        schedstats=     [KNL,X86] Enable or disable scheduled statistics.
3801                        Allowed values are enable and disable. This feature
3802                        incurs a small amount of overhead in the scheduler
3803                        but is useful for debugging and performance tuning.
3805        skew_tick=      [KNL] Offset the periodic timer tick per cpu to mitigate
3806                        xtime_lock contention on larger systems, and/or RCU lock
3807                        contention on all systems with CONFIG_MAXSMP set.
3808                        Format: { "0" | "1" }
3809                        0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
3810                        1 -- enable.
3811                        Note: increases power consumption, thus should only be
3812                        enabled if running jitter sensitive (HPC/RT) workloads.
3814        security=       [SECURITY] Choose a security module to enable at boot.
3815                        If this boot parameter is not specified, only the first
3816                        security module asking for security registration will be
3817                        loaded. An invalid security module name will be treated
3818                        as if no module has been chosen.
3820        selinux=        [SELINUX] Disable or enable SELinux at boot time.
3821                        Format: { "0" | "1" }
3822                        See security/selinux/Kconfig help text.
3823                        0 -- disable.
3824                        1 -- enable.
3825                        Default value is set via kernel config option.
3826                        If enabled at boot time, /selinux/disable can be used
3827                        later to disable prior to initial policy load.
3829        apparmor=       [APPARMOR] Disable or enable AppArmor at boot time
3830                        Format: { "0" | "1" }
3831                        See security/apparmor/Kconfig help text
3832                        0 -- disable.
3833                        1 -- enable.
3834                        Default value is set via kernel config option.
3836        serialnumber    [BUGS=X86-32]
3838        shapers=        [NET]
3839                        Maximal number of shapers.
3841        show_msr=       [x86] show boot-time MSR settings
3842                        Format: { <integer> }
3843                        Show boot-time (BIOS-initialized) MSR settings.
3844                        The parameter means the number of CPUs to show,
3845                        for example 1 means boot CPU only.
3847        simeth=         [IA-64]
3848        simscsi=
3850        slram=          [HW,MTD]
3852        slab_nomerge    [MM]
3853                        Disable merging of slabs with similar size. May be
3854                        necessary if there is some reason to distinguish
3855                        allocs to different slabs. Debug options disable
3856                        merging on their own.
3857                        For more information see Documentation/vm/slub.txt.
3859        slab_max_order= [MM, SLAB]
3860                        Determines the maximum allowed order for slabs.
3861                        A high setting may cause OOMs due to memory
3862                        fragmentation.  Defaults to 1 for systems with
3863                        more than 32MB of RAM, 0 otherwise.
3865        slub_debug[=options[,slabs]]    [MM, SLUB]
3866                        Enabling slub_debug allows one to determine the
3867                        culprit if slab objects become corrupted. Enabling
3868                        slub_debug can create guard zones around objects and
3869                        may poison objects when not in use. Also tracks the
3870                        last alloc / free. For more information see
3871                        Documentation/vm/slub.txt.
3873        slub_max_order= [MM, SLUB]
3874                        Determines the maximum allowed order for slabs.
3875                        A high setting may cause OOMs due to memory
3876                        fragmentation. For more information see
3877                        Documentation/vm/slub.txt.
3879        slub_min_objects=       [MM, SLUB]
3880                        The minimum number of objects per slab. SLUB will
3881                        increase the slab order up to slub_max_order to
3882                        generate a sufficiently large slab able to contain
3883                        the number of objects indicated. The higher the number
3884                        of objects the smaller the overhead of tracking slabs
3885                        and the less frequently locks need to be acquired.
3886                        For more information see Documentation/vm/slub.txt.
3888        slub_min_order= [MM, SLUB]
3889                        Determines the minimum page order for slabs. Must be
3890                        lower than slub_max_order.
3891                        For more information see Documentation/vm/slub.txt.
3893        slub_nomerge    [MM, SLUB]
3894                        Same with slab_nomerge. This is supported for legacy.
3895                        See slab_nomerge for more information.
3897        smart2=         [HW]
3898                        Format: <io1>[,<io2>[,...,<io8>]]
3900        smsc-ircc2.nopnp        [HW] Don't use PNP to discover SMC devices
3901        smsc-ircc2.ircc_cfg=    [HW] Device configuration I/O port
3902        smsc-ircc2.ircc_sir=    [HW] SIR base I/O port
3903        smsc-ircc2.ircc_fir=    [HW] FIR base I/O port
3904        smsc-ircc2.ircc_irq=    [HW] IRQ line
3905        smsc-ircc2.ircc_dma=    [HW] DMA channel
3906        smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
3907                                0: Toshiba Satellite 1800 (GP data pin select)
3908                                1: Fast pin select (default)
3909                                2: ATC IRMode
3911        smt             [KNL,S390] Set the maximum number of threads (logical
3912                        CPUs) to use per physical CPU on systems capable of
3913                        symmetric multithreading (SMT). Will be capped to the
3914                        actual hardware limit.
3915                        Format: <integer>
3916                        Default: -1 (no limit)
3918        softlockup_panic=
3919                        [KNL] Should the soft-lockup detector generate panics.
3920                        Format: <integer>
3922        softlockup_all_cpu_backtrace=
3923                        [KNL] Should the soft-lockup detector generate
3924                        backtraces on all cpus.
3925                        Format: <integer>
3927        sonypi.*=       [HW] Sony Programmable I/O Control Device driver
3928                        See Documentation/laptops/sonypi.txt
3930        spia_io_base=   [HW,MTD]
3931        spia_fio_base=
3932        spia_pedr=
3933        spia_peddr=
3935        stacktrace      [FTRACE]
3936                        Enabled the stack tracer on boot up.
3938        stacktrace_filter=[function-list]
3939                        [FTRACE] Limit the functions that the stack tracer
3940                        will trace at boot up. function-list is a comma separated
3941                        list of functions. This list can be changed at run
3942                        time by the stack_trace_filter file in the debugfs
3943                        tracing directory. Note, this enables stack tracing
3944                        and the stacktrace above is not needed.
3946        sti=            [PARISC,HW]
3947                        Format: <num>
3948                        Set the STI (builtin display/keyboard on the HP-PARISC
3949                        machines) console (graphic card) which should be used
3950                        as the initial boot-console.
3951                        See also comment in drivers/video/console/sticore.c.
3953        sti_font=       [HW]
3954                        See comment in drivers/video/console/sticore.c.
3956        stifb=          [HW]
3957                        Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
3959        sunrpc.min_resvport=
3960        sunrpc.max_resvport=
3961                        [NFS,SUNRPC]
3962                        SunRPC servers often require that client requests
3963                        originate from a privileged port (i.e. a port in the
3964                        range 0 < portnr < 1024).
3965                        An administrator who wishes to reserve some of these
3966                        ports for other uses may adjust the range that the
3967                        kernel's sunrpc client considers to be privileged
3968                        using these two parameters to set the minimum and
3969                        maximum port values.
3971        sunrpc.svc_rpc_per_connection_limit=
3972                        [NFS,SUNRPC]
3973                        Limit the number of requests that the server will
3974                        process in parallel from a single connection.
3975                        The default value is 0 (no limit).
3977        sunrpc.pool_mode=
3978                        [NFS]
3979                        Control how the NFS server code allocates CPUs to
3980                        service thread pools.  Depending on how many NICs
3981                        you have and where their interrupts are bound, this
3982                        option will affect which CPUs will do NFS serving.
3983                        Note: this parameter cannot be changed while the
3984                        NFS server is running.
3986                        auto        the server chooses an appropriate mode
3987                                    automatically using heuristics
3988                        global      a single global pool contains all CPUs
3989                        percpu      one pool for each CPU
3990                        pernode     one pool for each NUMA node (equivalent
3991                                    to global on non-NUMA machines)
3993        sunrpc.tcp_slot_table_entries=
3994        sunrpc.udp_slot_table_entries=
3995                        [NFS,SUNRPC]
3996                        Sets the upper limit on the number of simultaneous
3997                        RPC calls that can be sent from the client to a
3998                        server. Increasing these values may allow you to
3999                        improve throughput, but will also increase the
4000                        amount of memory reserved for use by the client.
4002        suspend.pm_test_delay=
4003                        [SUSPEND]
4004                        Sets the number of seconds to remain in a suspend test
4005                        mode before resuming the system (see
4006                        /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
4007                        is set. Default value is 5.
4009        swapaccount=[0|1]
4010                        [KNL] Enable accounting of swap in memory resource
4011                        controller if no parameter or 1 is given or disable
4012                        it if 0 is given (See Documentation/cgroup-v1/memory.txt)
4014        swiotlb=        [ARM,IA-64,PPC,MIPS,X86]
4015                        Format: { <int> | force | noforce }
4016                        <int> -- Number of I/O TLB slabs
4017                        force -- force using of bounce buffers even if they
4018                                 wouldn't be automatically used by the kernel
4019                        noforce -- Never use bounce buffers (for debugging)
4021        switches=       [HW,M68k]
4023        sysfs.deprecated=0|1 [KNL]
4024                        Enable/disable old style sysfs layout for old udev
4025                        on older distributions. When this option is enabled
4026                        very new udev will not work anymore. When this option
4027                        is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
4028                        in older udev will not work anymore.
4029                        Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
4030                        the kernel configuration.
4032        sysrq_always_enabled
4033                        [KNL]
4034                        Ignore sysrq setting - this boot parameter will
4035                        neutralize any effect of /proc/sys/kernel/sysrq.
4036                        Useful for debugging.
4038        tcpmhash_entries= [KNL,NET]
4039                        Set the number of tcp_metrics_hash slots.
4040                        Default value is 8192 or 16384 depending on total
4041                        ram pages. This is used to specify the TCP metrics
4042                        cache size. See Documentation/networking/ip-sysctl.txt
4043                        "tcp_no_metrics_save" section for more details.
4045        tdfx=           [HW,DRM]
4047        test_suspend=   [SUSPEND][,N]
4048                        Specify "mem" (for Suspend-to-RAM) or "standby" (for
4049                        standby suspend) or "freeze" (for suspend type freeze)
4050                        as the system sleep state during system startup with
4051                        the optional capability to repeat N number of times.
4052                        The system is woken from this state using a
4053                        wakeup-capable RTC alarm.
4055        thash_entries=  [KNL,NET]
4056                        Set number of hash buckets for TCP connection
4058        thermal.act=    [HW,ACPI]
4059                        -1: disable all active trip points in all thermal zones
4060                        <degrees C>: override all lowest active trip points
4062        thermal.crt=    [HW,ACPI]
4063                        -1: disable all critical trip points in all thermal zones
4064                        <degrees C>: override all critical trip points
4066        thermal.nocrt=  [HW,ACPI]
4067                        Set to disable actions on ACPI thermal zone
4068                        critical and hot trip points.
4070    [HW,ACPI]
4071                        1: disable ACPI thermal control
4073        thermal.psv=    [HW,ACPI]
4074                        -1: disable all passive trip points
4075                        <degrees C>: override all passive trip points to this
4076                        value
4078        thermal.tzp=    [HW,ACPI]
4079                        Specify global default ACPI thermal zone polling rate
4080                        <deci-seconds>: poll all this frequency
4081                        0: no polling (default)
4083        threadirqs      [KNL]
4084                        Force threading of all interrupt handlers except those
4085                        marked explicitly IRQF_NO_THREAD.
4087        tmem            [KNL,XEN]
4088                        Enable the Transcendent memory driver if built-in.
4090        tmem.cleancache=0|1 [KNL, XEN]
4091                        Default is on (1). Disable the usage of the cleancache
4092                        API to send anonymous pages to the hypervisor.
4094        tmem.frontswap=0|1 [KNL, XEN]
4095                        Default is on (1). Disable the usage of the frontswap
4096                        API to send swap pages to the hypervisor. If disabled
4097                        the selfballooning and selfshrinking are force disabled.
4099        tmem.selfballooning=0|1 [KNL, XEN]
4100                        Default is on (1). Disable the driving of swap pages
4101                        to the hypervisor.
4103        tmem.selfshrinking=0|1 [KNL, XEN]
4104                        Default is on (1). Partial swapoff that immediately
4105                        transfers pages from Xen hypervisor back to the
4106                        kernel based on different criteria.
4108        topology=       [S390]
4109                        Format: {off | on}
4110                        Specify if the kernel should make use of the cpu
4111                        topology information if the hardware supports this.
4112                        The scheduler will make use of this information and
4113                        e.g. base its process migration decisions on it.
4114                        Default is on.
4116        topology_updates= [KNL, PPC, NUMA]
4117                        Format: {off}
4118                        Specify if the kernel should ignore (off)
4119                        topology updates sent by the hypervisor to this
4120                        LPAR.
4122        tp720=          [HW,PS2]
4124        tpm_suspend_pcr=[HW,TPM]
4125                        Format: integer pcr id
4126                        Specify that at suspend time, the tpm driver
4127                        should extend the specified pcr with zeros,
4128                        as a workaround for some chips which fail to
4129                        flush the last written pcr on TPM_SaveState.
4130                        This will guarantee that all the other pcrs
4131                        are saved.
4133        trace_buf_size=nn[KMG]
4134                        [FTRACE] will set tracing buffer size on each cpu.
4136        trace_event=[event-list]
4137                        [FTRACE] Set and start specified trace events in order
4138                        to facilitate early boot debugging. The event-list is a
4139                        comma separated list of trace events to enable. See
4140                        also Documentation/trace/events.txt
4142        trace_options=[option-list]
4143                        [FTRACE] Enable or disable tracer options at boot.
4144                        The option-list is a comma delimited list of options
4145                        that can be enabled or disabled just as if you were
4146                        to echo the option name into
4148                            /sys/kernel/debug/tracing/trace_options
4150                        For example, to enable stacktrace option (to dump the
4151                        stack trace of each event), add to the command line:
4153                              trace_options=stacktrace
4155                        See also Documentation/trace/ftrace.txt "trace options"
4156                        section.
4158        tp_printk[FTRACE]
4159                        Have the tracepoints sent to printk as well as the
4160                        tracing ring buffer. This is useful for early boot up
4161                        where the system hangs or reboots and does not give the
4162                        option for reading the tracing buffer or performing a
4163                        ftrace_dump_on_oops.
4165                        To turn off having tracepoints sent to printk,
4166                         echo 0 > /proc/sys/kernel/tracepoint_printk
4167                        Note, echoing 1 into this file without the
4168                        tracepoint_printk kernel cmdline option has no effect.
4170                        ** CAUTION **
4172                        Having tracepoints sent to printk() and activating high
4173                        frequency tracepoints such as irq or sched, can cause
4174                        the system to live lock.
4176        traceoff_on_warning
4177                        [FTRACE] enable this option to disable tracing when a
4178                        warning is hit. This turns off "tracing_on". Tracing can
4179                        be enabled again by echoing '1' into the "tracing_on"
4180                        file located in /sys/kernel/debug/tracing/
4182                        This option is useful, as it disables the trace before
4183                        the WARNING dump is called, which prevents the trace to
4184                        be filled with content caused by the warning output.
4186                        This option can also be set at run time via the sysctl
4187                        option:  kernel/traceoff_on_warning
4189        transparent_hugepage=
4190                        [KNL]
4191                        Format: [always|madvise|never]
4192                        Can be used to control the default behavior of the system
4193                        with respect to transparent hugepages.
4194                        See Documentation/vm/transhuge.txt for more details.
4196        tsc=            Disable clocksource stability checks for TSC.
4197                        Format: <string>
4198                        [x86] reliable: mark tsc clocksource as reliable, this
4199                        disables clocksource verification at runtime, as well
4200                        as the stability checks done at bootup. Used to enable
4201                        high-resolution timer mode on older hardware, and in
4202                        virtualized environment.
4203                        [x86] noirqtime: Do not use TSC to do irq accounting.
4204                        Used to run time disable IRQ_TIME_ACCOUNTING on any
4205                        platforms where RDTSC is slow and this accounting
4206                        can add overhead.
4208[2|3]=    [HW,JOY]
4209                        TurboGraFX parallel port interface
4210                        Format:
4211                        <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
4212                        See also Documentation/input/joystick-parport.txt
4214        udbg-immortal   [PPC] When debugging early kernel crashes that
4215                        happen after console_init() and before a proper
4216                        console driver takes over, this boot options might
4217                        help "seeing" what's going on.
4219        uhash_entries=  [KNL,NET]
4220                        Set number of hash buckets for UDP/UDP-Lite connections
4222        uhci-hcd.ignore_oc=
4223                        [USB] Ignore overcurrent events (default N).
4224                        Some badly-designed motherboards generate lots of
4225                        bogus events, for ports that aren't wired to
4226                        anything.  Set this parameter to avoid log spamming.
4227                        Note that genuine overcurrent events won't be
4228                        reported either.
4230        unknown_nmi_panic
4231                        [X86] Cause panic on unknown NMI.
4233        usbcore.authorized_default=
4234                        [USB] Default USB device authorization:
4235                        (default -1 = authorized except for wireless USB,
4236                        0 = not authorized, 1 = authorized)
4238        usbcore.autosuspend=
4239                        [USB] The autosuspend time delay (in seconds) used
4240                        for newly-detected USB devices (default 2).  This
4241                        is the time required before an idle device will be
4242                        autosuspended.  Devices for which the delay is set
4243                        to a negative value won't be autosuspended at all.
4245        usbcore.usbfs_snoop=
4246                        [USB] Set to log all usbfs traffic (default 0 = off).
4248        usbcore.usbfs_snoop_max=
4249                        [USB] Maximum number of bytes to snoop in each URB
4250                        (default = 65536).
4252        usbcore.blinkenlights=
4253                        [USB] Set to cycle leds on hubs (default 0 = off).
4255        usbcore.old_scheme_first=
4256                        [USB] Start with the old device initialization
4257                        scheme (default 0 = off).
4259        usbcore.usbfs_memory_mb=
4260                        [USB] Memory limit (in MB) for buffers allocated by
4261                        usbfs (default = 16, 0 = max = 2047).
4263        usbcore.use_both_schemes=
4264                        [USB] Try the other device initialization scheme
4265                        if the first one fails (default 1 = enabled).
4267        usbcore.initial_descriptor_timeout=
4268                        [USB] Specifies timeout for the initial 64-byte
4269                        USB_REQ_GET_DESCRIPTOR request in milliseconds
4270                        (default 5000 = 5.0 seconds).
4272        usbcore.nousb   [USB] Disable the USB subsystem
4274        usbhid.mousepoll=
4275                        [USBHID] The interval which mice are to be polled at.
4277        usb-storage.delay_use=
4278                        [UMS] The delay in seconds before a new device is
4279                        scanned for Logical Units (default 1).
4281        usb-storage.quirks=
4282                        [UMS] A list of quirks entries to supplement or
4283                        override the built-in unusual_devs list.  List
4284                        entries are separated by commas.  Each entry has
4285                        the form VID:PID:Flags where VID and PID are Vendor
4286                        and Product ID values (4-digit hex numbers) and
4287                        Flags is a set of characters, each corresponding
4288                        to a common usb-storage quirk flag as follows:
4289                                a = SANE_SENSE (collect more than 18 bytes
4290                                        of sense data);
4291                                b = BAD_SENSE (don't collect more than 18
4292                                        bytes of sense data);
4293                                c = FIX_CAPACITY (decrease the reported
4294                                        device capacity by one sector);
4295                                d = NO_READ_DISC_INFO (don't use
4296                                        READ_DISC_INFO command);
4297                                e = NO_READ_CAPACITY_16 (don't use
4298                                        READ_CAPACITY_16 command);
4299                                f = NO_REPORT_OPCODES (don't use report opcodes
4300                                        command, uas only);
4301                                g = MAX_SECTORS_240 (don't transfer more than
4302                                        240 sectors at a time, uas only);
4303                                h = CAPACITY_HEURISTICS (decrease the
4304                                        reported device capacity by one
4305                                        sector if the number is odd);
4306                                i = IGNORE_DEVICE (don't bind to this
4307                                        device);
4308                                j = NO_REPORT_LUNS (don't use report luns
4309                                        command, uas only);
4310                                l = NOT_LOCKABLE (don't try to lock and
4311                                        unlock ejectable media);
4312                                m = MAX_SECTORS_64 (don't transfer more
4313                                        than 64 sectors = 32 KB at a time);
4314                                n = INITIAL_READ10 (force a retry of the
4315                                        initial READ(10) command);
4316                                o = CAPACITY_OK (accept the capacity
4317                                        reported by the device);
4318                                p = WRITE_CACHE (the device cache is ON
4319                                        by default);
4320                                r = IGNORE_RESIDUE (the device reports
4321                                        bogus residue values);
4322                                s = SINGLE_LUN (the device has only one
4323                                        Logical Unit);
4324                                t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
4325                                        commands, uas only);
4326                                u = IGNORE_UAS (don't bind to the uas driver);
4327                                w = NO_WP_DETECT (don't test whether the
4328                                        medium is write-protected).
4329                                y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE
4330                                        even if the device claims no cache)
4331                        Example: quirks=0419:aaf5:rl,0421:0433:rc
4333        user_debug=     [KNL,ARM]
4334                        Format: <int>
4335                        See arch/arm/Kconfig.debug help text.
4336                                 1 - undefined instruction events
4337                                 2 - system calls
4338                                 4 - invalid data aborts
4339                                 8 - SIGSEGV faults
4340                                16 - SIGBUS faults
4341                        Example: user_debug=31
4343        userpte=
4344                        [X86] Flags controlling user PTE allocations.
4346                                nohigh = do not allocate PTE pages in
4347                                        HIGHMEM regardless of setting
4348                                        of CONFIG_HIGHPTE.
4350        vdso=           [X86,SH]
4351                        On X86_32, this is an alias for vdso32=.  Otherwise:
4353                        vdso=1: enable VDSO (the default)
4354                        vdso=0: disable VDSO mapping
4356        vdso32=         [X86] Control the 32-bit vDSO
4357                        vdso32=1: enable 32-bit VDSO
4358                        vdso32=0 or vdso32=2: disable 32-bit VDSO
4360                        See the help text for CONFIG_COMPAT_VDSO for more
4361                        details.  If CONFIG_COMPAT_VDSO is set, the default is
4362                        vdso32=0; otherwise, the default is vdso32=1.
4364                        For compatibility with older kernels, vdso32=2 is an
4365                        alias for vdso32=0.
4367                        Try vdso32=0 if you encounter an error that says:
4368                        dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
4370        vector=         [IA-64,SMP]
4371                        vector=percpu: enable percpu vector domain
4373        video=          [FB] Frame buffer configuration
4374                        See Documentation/fb/modedb.txt.
4376        video.brightness_switch_enabled= [0,1]
4377                        If set to 1, on receiving an ACPI notify event
4378                        generated by hotkey, video driver will adjust brightness
4379                        level and then send out the event to user space through
4380                        the allocated input device; If set to 0, video driver
4381                        will only send out the event without touching backlight
4382                        brightness level.
4383                        default: 1
4385        virtio_mmio.device=
4386                        [VMMIO] Memory mapped virtio (platform) device.
4388                                <size>@<baseaddr>:<irq>[:<id>]
4389                        where:
4390                                <size>     := size (can use standard suffixes
4391                                                like K, M and G)
4392                                <baseaddr> := physical base address
4393                                <irq>      := interrupt number (as passed to
4394                                                request_irq())
4395                                <id>       := (optional) platform device id
4396                        example:
4397                                virtio_mmio.device=1K@0x100b0000:48:7
4399                        Can be used multiple times for multiple devices.
4401        vga=            [BOOT,X86-32] Select a particular video mode
4402                        See Documentation/x86/boot.txt and
4403                        Documentation/svga.txt.
4404                        Use vga=ask for menu.
4405                        This is actually a boot loader parameter; the value is
4406                        passed to the kernel using a special protocol.
4408        vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
4409                        size of <nn>. This can be used to increase the
4410                        minimum size (128MB on x86). It can also be used to
4411                        decrease the size and leave more room for directly
4412                        mapped kernel RAM.
4414        vmhalt=         [KNL,S390] Perform z/VM CP command after system halt.
4415                        Format: <command>
4417        vmpanic=        [KNL,S390] Perform z/VM CP command after kernel panic.
4418                        Format: <command>
4420        vmpoff=         [KNL,S390] Perform z/VM CP command after power off.
4421                        Format: <command>
4423        vsyscall=       [X86-64]
4424                        Controls the behavior of vsyscalls (i.e. calls to
4425                        fixed addresses of 0xffffffffff600x00 from legacy
4426                        code).  Most statically-linked binaries and older
4427                        versions of glibc use these calls.  Because these
4428                        functions are at fixed addresses, they make nice
4429                        targets for exploits that can control RIP.
4431                        emulate     [default] Vsyscalls turn into traps and are
4432                                    emulated reasonably safely.
4434                        native      Vsyscalls are native syscall instructions.
4435                                    This is a little bit faster than trapping
4436                                    and makes a few dynamic recompilers work
4437                                    better than they would in emulation mode.
4438                                    It also makes exploits much easier to write.
4440                        none        Vsyscalls don't work at all.  This makes
4441                                    them quite hard to use for exploits but
4442                                    might break your system.
4444        vt.color=       [VT] Default text color.
4445                        Format: 0xYX, X = foreground, Y = background.
4446                        Default: 0x07 = light gray on black.
4448        vt.cur_default= [VT] Default cursor shape.
4449                        Format: 0xCCBBAA, where AA, BB, and CC are the same as
4450                        the parameters of the <Esc>[?A;B;Cc escape sequence;
4451                        see VGA-softcursor.txt. Default: 2 = underline.
4453        vt.default_blu= [VT]
4454                        Format: <blue0>,<blue1>,<blue2>,...,<blue15>
4455                        Change the default blue palette of the console.
4456                        This is a 16-member array composed of values
4457                        ranging from 0-255.
4459        vt.default_grn= [VT]
4460                        Format: <green0>,<green1>,<green2>,...,<green15>
4461                        Change the default green palette of the console.
4462                        This is a 16-member array composed of values
4463                        ranging from 0-255.
4465        vt.default_red= [VT]
4466                        Format: <red0>,<red1>,<red2>,...,<red15>
4467                        Change the default red palette of the console.
4468                        This is a 16-member array composed of values
4469                        ranging from 0-255.
4471        vt.default_utf8=
4472                        [VT]
4473                        Format=<0|1>
4474                        Set system-wide default UTF-8 mode for all tty's.
4475                        Default is 1, i.e. UTF-8 mode is enabled for all
4476                        newly opened terminals.
4478        vt.global_cursor_default=
4479                        [VT]
4480                        Format=<-1|0|1>
4481                        Set system-wide default for whether a cursor
4482                        is shown on new VTs. Default is -1,
4483                        i.e. cursors will be created by default unless
4484                        overridden by individual drivers. 0 will hide
4485                        cursors, 1 will display them.
4487        vt.italic=      [VT] Default color for italic text; 0-15.
4488                        Default: 2 = green.
4490        vt.underline=   [VT] Default color for underlined text; 0-15.
4491                        Default: 3 = cyan.
4493        watchdog timers [HW,WDT] For information on watchdog timers,
4494                        see Documentation/watchdog/watchdog-parameters.txt
4495                        or other driver-specific files in the
4496                        Documentation/watchdog/ directory.
4498        workqueue.watchdog_thresh=
4499                        If CONFIG_WQ_WATCHDOG is configured, workqueue can
4500                        warn stall conditions and dump internal state to
4501                        help debugging.  0 disables workqueue stall
4502                        detection; otherwise, it's the stall threshold
4503                        duration in seconds.  The default value is 30 and
4504                        it can be updated at runtime by writing to the
4505                        corresponding sysfs file.
4507        workqueue.disable_numa
4508                        By default, all work items queued to unbound
4509                        workqueues are affine to the NUMA nodes they're
4510                        issued on, which results in better behavior in
4511                        general.  If NUMA affinity needs to be disabled for
4512                        whatever reason, this option can be used.  Note
4513                        that this also can be controlled per-workqueue for
4514                        workqueues visible under /sys/bus/workqueue/.
4516        workqueue.power_efficient
4517                        Per-cpu workqueues are generally preferred because
4518                        they show better performance thanks to cache
4519                        locality; unfortunately, per-cpu workqueues tend to
4520                        be more power hungry than unbound workqueues.
4522                        Enabling this makes the per-cpu workqueues which
4523                        were observed to contribute significantly to power
4524                        consumption unbound, leading to measurably lower
4525                        power usage at the cost of small performance
4526                        overhead.
4528                        The default value of this parameter is determined by
4529                        the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
4531        workqueue.debug_force_rr_cpu
4532                        Workqueue used to implicitly guarantee that work
4533                        items queued without explicit CPU specified are put
4534                        on the local CPU.  This guarantee is no longer true
4535                        and while local CPU is still preferred work items
4536                        may be put on foreign CPUs.  This debug option
4537                        forces round-robin CPU selection to flush out
4538                        usages which depend on the now broken guarantee.
4539                        When enabled, memory and cache locality will be
4540                        impacted.
4542        x2apic_phys     [X86-64,APIC] Use x2apic physical mode instead of
4543                        default x2apic cluster mode on platforms
4544                        supporting x2apic.
4546        x86_intel_mid_timer= [X86-32,APBT]
4547                        Choose timer option for x86 Intel MID platform.
4548                        Two valid options are apbt timer only and lapic timer
4549                        plus one apbt timer for broadcast timer.
4550                        x86_intel_mid_timer=apbt_only | lapic_and_apbt
4552        xen_512gb_limit         [KNL,X86-64,XEN]
4553                        Restricts the kernel running paravirtualized under Xen
4554                        to use only up to 512 GB of RAM. The reason to do so is
4555                        crash analysis tools and Xen tools for doing domain
4556                        save/restore/migration must be enabled to handle larger
4557                        domains.
4559        xen_emul_unplug=                [HW,X86,XEN]
4560                        Unplug Xen emulated devices
4561                        Format: [unplug0,][unplug1]
4562                        ide-disks -- unplug primary master IDE devices
4563                        aux-ide-disks -- unplug non-primary-master IDE devices
4564                        nics -- unplug network devices
4565                        all -- unplug all emulated devices (NICs and IDE disks)
4566                        unnecessary -- unplugging emulated devices is
4567                                unnecessary even if the host did not respond to
4568                                the unplug protocol
4569                        never -- do not unplug even if version check succeeds
4571        xen_nopvspin    [X86,XEN]
4572                        Disables the ticketlock slowpath using Xen PV
4573                        optimizations.
4575        xen_nopv        [X86]
4576                        Disables the PV optimizations forcing the HVM guest to
4577                        run as generic HVM guest with no PV drivers.
4579        xirc2ps_cs=     [NET,PCMCIA]
4580                        Format:
4581                        <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
4587        Add more DRM drivers.
Note: See TracBrowser for help on using the repository browser.