4 The following is a consolidated list of the kernel parameters as
5 implemented by the __setup(), core_param() and module_param() macros
6 and sorted into English Dictionary order (defined as ignoring all
7 punctuation and sorting digits before letters in a case insensitive
8 manner), and with descriptions where known.
10 The kernel parses parameters from the kernel command line up to "--";
11 if it doesn't recognize a parameter and it doesn't contain a '.', the
12 parameter gets passed to init: parameters with '=' go into init's
13 environment, others are passed as command line arguments to init.
14 Everything after "--" is passed as an argument to init.
16 Module parameters can be specified in two ways: via the kernel command
17 line 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
22 Parameters for modules which are built into the kernel need to be
23 specified on the kernel command line. modprobe looks through the
24 kernel command line (/proc/cmdline) and collects module parameters
25 when it loads a module, so the kernel command line can be used for
28 Hyphens (dashes) and underscores are equivalent in parameter names, so
29 log_buf_len=1M print-fatal-signals=1
30 can also be entered as
31 log-buf-len=1M print_fatal_signals=1
33 Double-quotes can be used to protect spaces in values, e.g.:
34 param="spaces in here"
39 Some kernel parameters take a list of CPUs as a value, e.g. isolcpus,
40 nohz_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)
51 <cpu number>,...,<cpu number>-<cpu number>
53 Note that for the special case of a range one can split the range into equal
54 sized groups and for each group use some amount from the beginning of that
57 <cpu number>-cpu number>:<used size>/<group size>
59 For example one can add to the command line following parameter:
61 isolcpus=1,2,10-20,100-2000:2/25
63 where the final item represents CPUs 100,101,125,126,150,151,...
67 This document may not be entirely up to date and comprehensive. The command
68 "modinfo -p ${modulename}" shows a current list of all parameters of a loadable
69 module. Loadable modules, after being loaded into the running kernel, also
70 reveal their parameters in /sys/module/${modulename}/parameters/. Some of these
71 parameters may be changed at runtime by the command
72 "echo -n ${value} > /sys/module/${modulename}/parameters/${parm}".
74 The parameters listed below are only valid if certain kernel build options were
75 enabled and if respective hardware is present. The text in square brackets at
76 the beginning of each description states the restrictions within which a
77 parameter 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
168 In 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.
174 Parameters denoted with BOOT are actually interpreted by the boot
175 loader, and have no meaning to the kernel directly.
176 Do not modify the syntax of boot loader parameters without extreme
177 need or coordination with <Documentation/x86/boot.txt>.
179 There are also arch-specific kernel-parameters not documented here.
180 See for example <Documentation/x86/x86_64/boot-options.txt>.
182 Note that ALL kernel parameters listed below are CASE SENSITIVE, and that
183 a trailing = on the name of any parameter states that that parameter will
184 be entered as an environment variable, whereas its absence indicates that
185 it will appear as a kernel argument readable via /proc/cmdline by programs
186 running once the system is up.
188 The number of kernel parameters is not limited, but the length of the
189 complete command line (parameters including spaces etc.) is limited to
190 a fixed number of characters. This limit depends on the architecture
191 and is between 256 and 4096 characters. It is defined in the file
192 ./include/asm/setup.h as COMMAND_LINE_SIZE.
194 Finally, the [KMG] suffix is commonly described after a number of kernel
195 parameter values. These 'K', 'M', and 'G' letters represent the _binary_
196 multipliers 'Kilo', 'Mega', and 'Giga', equalling 2^10, 2^20, and 2^30
197 bytes 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 |
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"
215 See also Documentation/power/runtime_pm.txt, pci=noacpi
217 acpi_apic_instance= [ACPI, IOAPIC]
219 2: use 2nd APIC table, if available
220 1,0: use 1st APIC table
223 acpi_backlight= [HW,ACPI]
224 acpi_backlight=vendor
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]
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
293 acpi_irq_balance [HW,ACPI]
294 ACPI will balance active IRQs
297 acpi_irq_nobalance [HW,ACPI]
298 ACPI will not move active IRQs (default)
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
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
313 This facility can be used to prevent such uncontrolled
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
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
359 acpi_osi=!! # enable all built-in OS vendor
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.
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
385 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
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
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
406 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
408 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
409 they all will make '_OSI("Windows 2000")' TRUE.
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
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.
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)
456 See Documentation/sound/alsa/alsa-parameters.txt
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.
486 fullflush - enable flushing of IO/TLB entries when
487 they are unmapped. Otherwise they are
488 flushed before they will be reused, which
490 off - do not initialize any AMD IOMMU found in
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
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 amijoy.map= [HW,JOY] Amiga joystick support
514 Map of devices attached to JOY0DAT and JOY1DAT
516 See also Documentation/input/joystick.txt
518 analog.map= [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>
524 Power management functions (SPARCstation-4/5 + deriv.)
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
541 none: External NMI is masked for all CPUs. This is
542 useful so that a dump capture kernel won't be
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>
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
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
596 audit_backlog_limit= [KNL] Set the audit queue size limit.
597 Format: <int> (must be >=0)
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" }
605 unset - Disable the BAU.
607 baycom_epp= [HW,AX25]
610 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
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
633 bootmem_debug [KNL] Enable bootmem allocator debug messages.
636 Disable BERT OS support on buggy BIOSes.
638 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
639 bttv.radio= Most important insmod options are available as
641 bttv.pll= See Documentation/video4linux/bttv/Insmod-options
644 bulk_remove=off [PPC] This parameter disables the use of the pSeries
645 firmware feature for flushing multiple hpte entries
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
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
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
676 - foo isn't visible as an individually mountable
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.
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.
703 See Documentation/s390/CommonIO for details.
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.
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
725 Override the default clocksource and use the clocksource
726 with the name specified.
727 Some clocksource names to choose from, depending on
729 [all] jiffies (this is the base, fallback clocksource)
731 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
732 pxa_timer,timer3,32k_counter,timer0_1
734 [X86-32] pit,hpet,tsc;
735 scx200_hrt on Geode; cyclone on IBM x440
743 clocksource.arm_arch_timer.evtstrm=
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
751 clocksource.arm_arch_timer.fsl-a008585=
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
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
773 cma=nn[MG]@[start[MG][-end[MG]]]
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
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
798 com20020= [HW,NET] ARCnet - COM20020 chipset
800 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
802 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
806 ARCnet - COM90xx chipset (memory-mapped buffers)
807 Format: <io>[,<irq>[,<memstart>]]
809 condev= [HW,S390] console device
812 console= [KNL] Output console device and options.
814 tty<n> Use the virtual console device <n>.
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
826 Documentation/networking/netconsole.txt for an
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
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.
857 [KNL] Change the default value for
858 /proc/<pid>/coredump_filter.
859 See also Documentation/filesystems/proc.txt.
861 cpuidle.off=1 [CPU_IDLE]
862 disable the cpuidle sub-system
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.
870 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
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
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.
912 [KNL] Disable crypto self-tests
917 cs89x0_media= [HW,NET]
918 Format: { rj45 | aui | bnc }
921 See header of drivers/s390/block/dasd_devmap.c.
923 db9.dev[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).
935 [KNL] verbose self-tests
937 Print debugging info while doing the locking API
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
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.
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.
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
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
1000 See Documentation/networking/ipv6.txt.
1003 Disable RADIX MMU mode on POWER9
1005 disable_cpu_apicid= [X86,APIC,SMP]
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.
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
1087 module.async_probe [KNL]
1088 Enable asynchronous probe on this module.
1090 early_ioremap_debug [KNL]
1091 Enable debug messages in early_ioremap support. This
1092 is useful for tracking down temporary early mappings
1093 which are not unmapped.
1095 earlycon= [KNL] Output early console device and options.
1097 When used with no options, the early console is
1098 determined by the stdout-path property in device
1101 cdns,<addr>[,options]
1102 Start an early, polled-mode console on a Cadence
1103 (xuartps) serial port at the specified address. Only
1104 supported option is baud rate. If baud rate is not
1105 specified, the serial port must already be setup and
1108 uart[8250],io,<addr>[,options]
1109 uart[8250],mmio,<addr>[,options]
1110 uart[8250],mmio32,<addr>[,options]
1111 uart[8250],mmio32be,<addr>[,options]
1112 uart[8250],0x<addr>[,options]
1113 Start an early, polled-mode console on the 8250/16550
1114 UART at the specified I/O port or MMIO address.
1115 MMIO inter-register address stride is either 8-bit
1116 (mmio) or 32-bit (mmio32 or mmio32be).
1117 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
1118 to be equivalent to 'mmio'. 'options' are specified
1119 in the same format described for "console=ttyS<n>"; if
1120 unspecified, the h/w is not initialized.
1124 Start an early, polled-mode console on a pl011 serial
1125 port at the specified address. The pl011 serial port
1126 must already be setup and configured. Options are not
1127 yet supported. If 'mmio32' is specified, then only
1128 the driver will use only 32-bit accessors to read/write
1129 the device registers.
1132 Start an early, polled-mode console on a meson serial
1133 port at the specified address. The serial port must
1134 already be setup and configured. Options are not yet
1138 Start an early, polled-mode console on an msm serial
1139 port at the specified address. The serial port
1140 must already be setup and configured. Options are not
1143 msm_serial_dm,<addr>
1144 Start an early, polled-mode console on an msm serial
1145 dm port at the specified address. The serial port
1146 must already be setup and configured. Options are not
1149 smh Use ARM semihosting calls for early console.
1157 Use early console provided by serial driver available
1158 on Samsung SoCs, requires selecting proper type and
1159 a correct base address of the selected UART port. The
1160 serial port must already be setup and configured.
1161 Options are not yet supported.
1165 Use early console provided by Freescale LP UART driver
1166 found on Freescale Vybrid and QorIQ LS1021A processors.
1167 A valid base address must be provided, and the serial
1168 port must already be setup and configured.
1170 armada3700_uart,<addr>
1171 Start an early, polled-mode console on the
1172 Armada 3700 serial port at the specified
1173 address. The serial port must already be setup
1174 and configured. Options are not yet supported.
1176 earlyprintk= [X86,SH,BLACKFIN,ARM,M68k]
1180 earlyprintk=serial[,ttySn[,baudrate]]
1181 earlyprintk=serial[,0x...[,baudrate]]
1182 earlyprintk=ttySn[,baudrate]
1183 earlyprintk=dbgp[debugController#]
1184 earlyprintk=pciserial,bus:device.function[,baudrate]
1186 earlyprintk is useful when the kernel crashes before
1187 the normal console is initialized. It is not enabled by
1188 default because it has some cosmetic problems.
1190 Append ",keep" to not disable it when the real console
1193 Only one of vga, efi, serial, or usb debug port can
1196 Currently only ttyS0 and ttyS1 may be specified by
1197 name. Other I/O ports may be explicitly specified
1198 on some architectures (x86 and arm at least) by
1199 replacing ttySn with an I/O port address, like this:
1200 earlyprintk=serial,0x1008,115200
1201 You can find the port for a given device in
1202 /proc/tty/driver/serial:
1203 2: uart:ST16650V2 port:00001008 irq:18 ...
1205 Interaction with the standard serial driver is not
1208 The VGA and EFI output is eventually overwritten by
1211 The xen output can only be used by Xen PV guests.
1213 edac_report= [HW,EDAC] Control how to report EDAC event
1214 Format: {"on" | "off" | "force"}
1215 on: enable EDAC to report H/W event. May be overridden
1216 by other higher priority error reporting module.
1217 off: disable H/W event reporting through EDAC.
1218 force: enforce the use of EDAC to report H/W event.
1221 ekgdboc= [X86,KGDB] Allow early kernel console debugging
1224 This is designed to be used in conjunction with
1225 the boot argument: earlyprintk=vga
1228 Format: {"off" | "on" | "skip[mbr]"}
1231 Format: { "old_map", "nochunk", "noruntime", "debug" }
1232 old_map [X86-64]: switch to the old ioremap-based EFI
1233 runtime services mapping. 32-bit still uses this one by
1235 nochunk: disable reading files in "chunks" in the EFI
1236 boot stub, as chunking can cause problems with some
1237 firmware implementations.
1238 noruntime : disable EFI runtime services support
1239 debug: enable misc debug output
1241 efi_no_storage_paranoia [EFI; X86]
1242 Using this parameter you can use more than 50% of
1243 your efi variable storage. Use this parameter only if
1244 you are really sure that your UEFI does sane gc and
1245 fulfills the spec otherwise your board may brick.
1247 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1248 Add arbitrary attribute to specific memory range by
1249 updating original EFI memory map.
1250 Region of memory which aa attribute is added to is
1252 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1253 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1254 attribute is added to range 0x100000000-0x180000000 and
1255 0x10a0000000-0x1120000000.
1257 Using this parameter you can do debugging of EFI memmap
1258 related feature. For example, you can do debugging of
1259 Address Range Mirroring feature even if your box
1262 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT
1263 that is to be dynamically loaded by Linux. If there are
1264 multiple variables with the same name but with different
1265 vendor GUIDs, all of them will be loaded. See
1266 Documentation/acpi/ssdt-overlays.txt for details.
1269 eisa_irq_edge= [PARISC,HW]
1270 See header of drivers/parisc/eisa.c.
1273 See comment before function elanfreq_setup() in
1274 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1277 Format: {"cfq" | "deadline" | "noop"}
1278 See Documentation/block/cfq-iosched.txt and
1279 Documentation/block/deadline-iosched.txt for details.
1281 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1282 Specifies physical address of start of kernel core
1283 image elf header and optionally the size. Generally
1284 kexec loader will pass this option to capture kernel.
1285 See Documentation/kdump/kdump.txt for details.
1287 enable_mtrr_cleanup [X86]
1288 The kernel tries to adjust MTRR layout from continuous
1289 to discrete, to make X server driver able to add WB
1290 entry later. This parameter enables that.
1292 enable_timer_pin_1 [X86]
1293 Enable PIN 1 of APIC timer
1294 Can be useful to work around chipset bugs
1295 (in particular on some ATI chipsets).
1296 The kernel tries to set a reasonable default.
1298 enforcing [SELINUX] Set initial enforcing status.
1300 See security/selinux/Kconfig help text.
1301 0 -- permissive (log only, no denials).
1302 1 -- enforcing (deny and log).
1304 Value can be changed at runtime via /selinux/enforce.
1307 Disable Error Record Serialization Table (ERST)
1310 ether= [HW,NET] Ethernet cards parameters
1311 This option is obsoleted by the "netdev=" option, which
1312 has equivalent usage. See its documentation for details.
1316 Permit 'security.evm' to be updated regardless of
1317 current integrity status.
1321 fail_make_request=[KNL]
1322 General fault injection mechanism.
1323 Format: <interval>,<probability>,<space>,<times>
1324 See also Documentation/fault-injection/.
1327 See Documentation/blockdev/floppy.txt.
1329 force_pal_cache_flush
1330 [IA-64] Avoid check_sal_cache_flush which may hang on
1331 buggy SAL_CACHE_FLUSH implementations. Using this
1332 parameter will force ia64_sal_cache_flush to call
1333 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1336 Forcefully enable Physical Address Extension (PAE).
1337 Many Pentium M systems disable PAE but may have a
1338 functionally usable PAE implementation.
1339 Warning: use of this parameter will taint the kernel
1340 and may cause unknown problems.
1343 [FTRACE] will set and start the specified tracer
1344 as early as possible in order to facilitate early
1347 ftrace_dump_on_oops[=orig_cpu]
1348 [FTRACE] will dump the trace buffers on oops.
1349 If no parameter is passed, ftrace will dump
1350 buffers of all CPUs, but if you pass orig_cpu, it will
1351 dump only the buffer of the CPU that triggered the
1354 ftrace_filter=[function-list]
1355 [FTRACE] Limit the functions traced by the function
1356 tracer at boot up. function-list is a comma separated
1357 list of functions. This list can be changed at run
1358 time by the set_ftrace_filter file in the debugfs
1361 ftrace_notrace=[function-list]
1362 [FTRACE] Do not trace the functions specified in
1363 function-list. This list can be changed at run time
1364 by the set_ftrace_notrace file in the debugfs
1367 ftrace_graph_filter=[function-list]
1368 [FTRACE] Limit the top level callers functions traced
1369 by the function graph tracer at boot up.
1370 function-list is a comma separated list of functions
1371 that can be changed at run time by the
1372 set_graph_function file in the debugfs tracing directory.
1374 ftrace_graph_notrace=[function-list]
1375 [FTRACE] Do not trace from the functions specified in
1376 function-list. This list is a comma separated list of
1377 functions that can be changed at run time by the
1378 set_graph_notrace file in the debugfs tracing directory.
1381 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1382 support via parallel port (up to 5 devices per port)
1383 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1384 See also Documentation/input/joystick-parport.txt
1388 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART
1392 gcov_persist= [GCOV] When non-zero (default), profiling data for
1393 kernel modules is saved and remains accessible via
1394 debugfs, even when the module is unloaded/reloaded.
1395 When zero, profiling data is discarded and associated
1396 debugfs files are removed at module unload time.
1398 goldfish [X86] Enable the goldfish android emulator platform.
1399 Don't use this when you are not running on the
1402 gpt [EFI] Forces disk with valid GPT signature but
1403 invalid Protective MBR to be treated as GPT. If the
1404 primary GPT is corrupted, it enables the backup/alternate
1405 GPT to be used instead.
1407 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1408 the "Enable 0" bit of the configuration register.
1411 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1412 the "Enable 0" bit of the configuration register.
1415 grcan.select= [HW] Select which physical interface to use.
1418 grcan.txsize= [HW] Sets the size of the tx buffer.
1419 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1421 grcan.rxsize= [HW] Sets the size of the rx buffer.
1422 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1425 gpio-mockup.gpio_mockup_ranges
1426 [HW] Sets the ranges of gpiochip of for this device.
1427 Format: <start1>,<end1>,<start2>,<end2>...
1429 hardlockup_all_cpu_backtrace=
1430 [KNL] Should the hard-lockup detector generate
1431 backtraces on all cpus.
1434 hashdist= [KNL,NUMA] Large hashes allocated during boot
1435 are distributed across NUMA nodes. Defaults on
1436 for 64-bit NUMA, off otherwise.
1437 Format: 0 | 1 (for off | on)
1439 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1441 hd= [EIDE] (E)IDE hard drive subsystem geometry
1442 Format: <cyl>,<head>,<sect>
1445 Disable Hardware Error Source Table (HEST) support;
1446 corresponding firmware-first mode error processing
1447 logic will be disabled.
1449 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1450 size of <nn>. This works even on boxes that have no
1451 highmem otherwise. This also works to reduce highmem
1452 size on bigger boxes.
1454 highres= [KNL] Enable/disable high resolution timer mode.
1455 Valid parameters: "on", "off"
1459 See Documentation/isdn/README.HiSax.
1463 hpet= [X86-32,HPET] option to control HPET usage
1464 Format: { enable (default) | disable | force |
1466 disable: disable HPET and use PIT instead
1467 force: allow force enabled of undocumented chips (ICH4,
1469 verbose: show contents of HPET registers during setup
1471 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1472 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1474 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
1475 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
1476 On x86-64 and powerpc, this option can be specified
1477 multiple times interleaved with hugepages= to reserve
1478 huge pages of different sizes. Valid pages sizes on
1479 x86-64 are 2M (when the CPU supports "pse") and 1G
1480 (when the CPU supports the "pdpe1gb" cpuinfo flag).
1482 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1483 terminal devices. Valid values: 0..8
1484 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1485 If specified, z/VM IUCV HVC accepts connections
1486 from listed z/VM user IDs only.
1488 hwthread_map= [METAG] Comma-separated list of Linux cpu id to
1489 hardware thread id mappings.
1490 Format: <cpu>:<hwthread>
1493 Do not unregister boot console at start. This is only
1494 useful for debugging when something happens in the window
1495 between unregistering the boot console and initializing
1498 i2c_bus= [HW] Override the default board specific I2C bus speed
1499 or register an additional I2C bus that is not
1500 registered from board initialization code.
1504 i8042.debug [HW] Toggle i8042 debug mode
1505 i8042.unmask_kbd_data
1506 [HW] Enable printing of interrupt data from the KBD port
1507 (disabled by default, and as a pre-condition
1508 requires that i8042.debug=1 be enabled)
1509 i8042.direct [HW] Put keyboard port into non-translated mode
1510 i8042.dumbkbd [HW] Pretend that controller can only read data from
1511 keyboard and cannot control its state
1512 (Don't attempt to blink the leds)
1513 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1514 i8042.nokbd [HW] Don't check/create keyboard port
1515 i8042.noloop [HW] Disable the AUX Loopback command while probing
1517 i8042.nomux [HW] Don't check presence of an active multiplexing
1519 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1521 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1522 i8042.reset [HW] Reset the controller during init, cleanup and
1523 suspend-to-ram transitions, only during s2r
1524 transitions, or never reset
1525 Format: { 1 | Y | y | 0 | N | n }
1526 1, Y, y: always reset controller
1527 0, N, n: don't ever reset controller
1528 Default: only on s2r transitions on x86; most other
1529 architectures force reset to be always executed
1530 i8042.unlock [HW] Unlock (ignore) the keylock
1531 i8042.kbdreset [HW] Reset device connected to KBD port
1535 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
1536 indicates that the driver is running on unsupported
1538 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
1539 does not match list of supported models.
1541 [HW] Report power status in /proc/i8k
1542 (disabled by default)
1543 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
1546 i915.invert_brightness=
1547 [DRM] Invert the sense of the variable that is used to
1548 set the brightness of the panel backlight. Normally a
1549 brightness value of 0 indicates backlight switched off,
1550 and the maximum of the brightness value sets the backlight
1551 to maximum brightness. If this parameter is set to 0
1552 (default) and the machine requires it, or this parameter
1553 is set to 1, a brightness value of 0 sets the backlight
1554 to maximum brightness, and the maximum of the brightness
1555 value switches the backlight off.
1556 -1 -- never invert brightness
1557 0 -- machine default
1558 1 -- force brightness inversion
1561 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1563 ide-core.nodma= [HW] (E)IDE subsystem
1564 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1565 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1566 .cdrom .chs .ignore_cable are additional options
1567 See Documentation/ide/ide.txt.
1569 ide-generic.probe-mask= [HW] (E)IDE subsystem
1571 Probe mask for legacy ISA IDE ports. Depending on
1572 platform up to 6 ports are supported, enabled by
1573 setting corresponding bits in the mask to 1. The
1574 default value is 0x0, which has a special meaning.
1575 On systems that have PCI, it triggers scanning the
1576 PCI bus for the first and the second port, which
1577 are then probed. On systems without PCI the value
1578 of 0x0 enables probing the two first ports as if it
1581 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1582 Claim all unknown PCI IDE storage controllers.
1585 Format: idle=poll, idle=halt, idle=nomwait
1586 Poll forces a polling idle loop that can slightly
1587 improve the performance of waking up a idle CPU, but
1588 will use a lot of power and make the system run hot.
1590 idle=halt: Halt is forced to be used for CPU idle.
1591 In such case C2/C3 won't be used again.
1592 idle=nomwait: Disable mwait for CPU C-states
1594 ieee754= [MIPS] Select IEEE Std 754 conformance mode
1595 Format: { strict | legacy | 2008 | relaxed }
1598 Choose which programs will be accepted for execution
1599 based on the IEEE 754 NaN encoding(s) supported by
1600 the FPU and the NaN encoding requested with the value
1601 of an ELF file header flag individually set by each
1602 binary. Hardware implementations are permitted to
1603 support either or both of the legacy and the 2008 NaN
1606 Available settings are as follows:
1607 strict accept binaries that request a NaN encoding
1608 supported by the FPU
1609 legacy only accept legacy-NaN binaries, if supported
1611 2008 only accept 2008-NaN binaries, if supported
1613 relaxed accept any binaries regardless of whether
1614 supported by the FPU
1616 The FPU emulator is always able to support both NaN
1617 encodings, so if no FPU hardware is present or it has
1618 been disabled with 'nofpu', then the settings of
1619 'legacy' and '2008' strap the emulator accordingly,
1620 'relaxed' straps the emulator for both legacy-NaN and
1621 2008-NaN, whereas 'strict' enables legacy-NaN only on
1622 legacy processors and both NaN encodings on MIPS32 or
1625 The setting for ABS.fmt/NEG.fmt instruction execution
1626 mode generally follows that for the NaN encoding,
1627 except where unsupported by hardware.
1629 ignore_loglevel [KNL]
1630 Ignore loglevel setting - this will print /all/
1631 kernel messages to the console. Useful for debugging.
1632 We also add it as printk module parameter, so users
1633 could change it dynamically, usually by
1634 /sys/module/printk/parameters/ignore_loglevel.
1637 Ignore RLIMIT_DATA setting for data mappings,
1638 print warning at first misuse. Can be changed via
1639 /sys/module/kernel/parameters/ignore_rlimit_data.
1641 ihash_entries= [KNL]
1642 Set number of hash buckets for inode cache.
1644 ima_appraise= [IMA] appraise integrity measurements
1645 Format: { "off" | "enforce" | "fix" | "log" }
1648 ima_appraise_tcb [IMA]
1649 The builtin appraise policy appraises all files
1653 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1657 The list of supported hash algorithms is defined
1658 in crypto/hash_info.h.
1661 The builtin measurement policy to load during IMA
1662 setup. Specyfing "tcb" as the value, measures all
1663 programs exec'd, files mmap'd for exec, and all files
1664 opened with the read mode bit set by either the
1665 effective uid (euid=0) or uid=0.
1668 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
1669 Load a policy which meets the needs of the Trusted
1670 Computing Base. This means IMA will measure all
1671 programs exec'd, files mmap'd for exec, and all files
1672 opened for read by uid=0.
1675 Select one of defined IMA measurements template formats.
1676 Formats: { "ima" | "ima-ng" | "ima-sig" }
1680 [IMA] Define a custom template format.
1681 Format: { "field1|...|fieldN" }
1683 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1684 Format: <min_file_size>
1685 Set the minimal file size for using asynchronous hash.
1686 If left unspecified, ahash usage is disabled.
1688 ahash performance varies for different data sizes on
1689 different crypto accelerators. This option can be used
1690 to achieve the best performance for a particular HW.
1692 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1694 Set hashing buffer size. Default: 4k.
1696 ahash performance varies for different chunk sizes on
1697 different crypto accelerators. This option can be used
1698 to achieve best performance for particular HW.
1702 Run specified binary instead of /sbin/init as init
1705 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1706 for working out where the kernel is dying during
1709 initcall_blacklist= [KNL] Do not execute a comma-separated list of
1710 initcall functions. Useful for debugging built-in
1711 modules and initcalls.
1713 initrd= [BOOT] Specify the location of the initial ramdisk
1715 init_pkru= [x86] Specify the default memory protection keys rights
1716 register contents for all processes. 0x55555554 by
1717 default (disallow access to all but pkey 0). Can
1718 override in debugfs after boot.
1720 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1723 int_pln_enable [x86] Enable power limit notification interrupt
1725 integrity_audit=[IMA]
1726 Format: { "0" | "1" }
1727 0 -- basic integrity auditing messages. (Default)
1728 1 -- additional integrity auditing messages.
1730 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
1732 Enable intel iommu driver.
1734 Disable intel iommu driver.
1735 igfx_off [Default Off]
1736 By default, gfx is mapped as normal device. If a gfx
1737 device has a dedicated DMAR unit, the DMAR unit is
1738 bypassed by not enabling DMAR with this option. In
1739 this case, gfx device will use physical address for
1742 With this option iommu will not optimize to look
1743 for io virtual address below 32-bit forcing dual
1744 address cycle on pci bus for cards supporting greater
1745 than 32-bit addressing. The default is to look
1746 for translation below 32-bit and if not available
1747 then look in the higher range.
1748 strict [Default Off]
1749 With this option on every unmap_single operation will
1750 result in a hardware IOTLB flush operation as opposed
1751 to batching them for performance.
1752 sp_off [Default Off]
1753 By default, super page will be supported if Intel IOMMU
1754 has the capability. With this option, super page will
1756 ecs_off [Default Off]
1757 By default, extended context tables will be supported if
1758 the hardware advertises that it has support both for the
1759 extended tables themselves, and also PASID support. With
1760 this option set, extended tables will not be used even
1761 on hardware which claims to support them.
1763 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
1764 0 disables intel_idle and fall back on acpi_idle.
1765 1 to 9 specify maximum depth of C-state.
1769 Do not enable intel_pstate as the default
1770 scaling driver for the supported processors
1772 Enable intel_pstate on systems that prohibit it by default
1773 in favor of acpi-cpufreq. Forcing the intel_pstate driver
1774 instead of acpi-cpufreq may disable platform features, such
1775 as thermal controls and power capping, that rely on ACPI
1776 P-States information being indicated to OSPM and therefore
1777 should be used with caution. This option does not work with
1778 processors that aren't supported by the intel_pstate driver
1779 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
1781 Do not enable hardware P state control (HWP)
1784 Only load intel_pstate on systems which support
1785 hardware P state control (HWP) if available.
1787 Enforce ACPI _PPC performance limits. If the Fixed ACPI
1788 Description Table, specifies preferred power management
1789 profile as "Enterprise Server" or "Performance Server",
1790 then this feature is turned on by default.
1792 intremap= [X86-64, Intel-IOMMU]
1793 on enable Interrupt Remapping (default)
1794 off disable Interrupt Remapping
1795 nosid disable Source ID checking
1797 BIOS x2APIC opt-out request will be ignored
1798 nopost disable Interrupt Posting
1800 iomem= Disable strict checking of access to MMIO memory
1801 strict regions from userspace.
1816 nobypass [PPC/POWERNV]
1817 Disable IOMMU bypass, using IOMMU for PCI devices.
1820 io7= [HW] IO7 for Marvel based alpha systems
1821 See comment before marvel_specify_io7 in
1822 arch/alpha/kernel/core_marvel.c.
1824 io_delay= [X86] I/O delay method
1826 Standard port 0x80 based delay
1828 Alternate port 0xed based delay (needed on some systems)
1830 Simple two microseconds delay
1835 See Documentation/filesystems/nfs/nfsroot.txt.
1837 irqaffinity= [SMP] Set the default irq affinity mask
1838 The argument is a cpu list, as described above.
1841 When an interrupt is not handled search all handlers
1842 for it. Intended to get systems with badly broken
1846 When an interrupt is not handled search all handlers
1847 for it. Also check all handlers each timer
1848 interrupt. Intended to get systems with badly broken
1852 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
1854 isolcpus= [KNL,SMP] Isolate CPUs from the general scheduler.
1855 The argument is a cpu list, as described above.
1857 This option can be used to specify one or more CPUs
1858 to isolate from the general SMP balancing and scheduling
1859 algorithms. You can move a process onto or off an
1860 "isolated" CPU via the CPU affinity syscalls or cpuset.
1861 <cpu number> begins at 0 and the maximum value is
1862 "number of CPUs in system - 1".
1864 This option is the preferred way to isolate CPUs. The
1865 alternative -- manually setting the CPU mask of all
1866 tasks in the system -- can cause problems and
1867 suboptimal load balancer performance.
1871 ivrs_ioapic [HW,X86_64]
1872 Provide an override to the IOAPIC-ID<->DEVICE-ID
1873 mapping provided in the IVRS ACPI table. For
1874 example, to map IOAPIC-ID decimal 10 to
1875 PCI device 00:14.0 write the parameter as:
1876 ivrs_ioapic[10]=00:14.0
1878 ivrs_hpet [HW,X86_64]
1879 Provide an override to the HPET-ID<->DEVICE-ID
1880 mapping provided in the IVRS ACPI table. For
1881 example, to map HPET-ID decimal 0 to
1882 PCI device 00:14.0 write the parameter as:
1883 ivrs_hpet[0]=00:14.0
1885 ivrs_acpihid [HW,X86_64]
1886 Provide an override to the ACPI-HID:UID<->DEVICE-ID
1887 mapping provided in the IVRS ACPI table. For
1888 example, to map UART-HID:UID AMD0020:0 to
1889 PCI device 00:14.5 write the parameter as:
1890 ivrs_acpihid[00:14.5]=AMD0020:0
1892 js= [HW,JOY] Analog joystick
1893 See Documentation/input/joystick.txt.
1896 When CONFIG_RANDOMIZE_BASE is set, this disables
1897 kernel and module base offset ASLR (Address Space
1898 Layout Randomization).
1902 kernelcore= [KNL,X86,IA-64,PPC]
1903 Format: nn[KMGTPE] | "mirror"
1905 specifies the amount of memory usable by the kernel
1906 for non-movable allocations. The requested amount is
1907 spread evenly throughout all nodes in the system. The
1908 remaining memory in each node is used for Movable
1909 pages. In the event, a node is too small to have both
1910 kernelcore and Movable pages, kernelcore pages will
1911 take priority and other nodes will have a larger number
1912 of Movable pages. The Movable zone is used for the
1913 allocation of pages that may be reclaimed or moved
1914 by the page migration subsystem. This means that
1915 HugeTLB pages may not be allocated from this zone.
1916 Note that allocations like PTEs-from-HighMem still
1917 use the HighMem zone if it exists, and the Normal
1918 zone if it does not.
1920 Instead of specifying the amount of memory (nn[KMGTPE]),
1921 you can specify "mirror" option. In case "mirror"
1922 option is specified, mirrored (reliable) memory is used
1923 for non-movable allocations and remaining memory is used
1924 for Movable pages. nn[KMGTPE] and "mirror" are exclusive,
1925 so you can NOT specify nn[KMGTPE] and "mirror" at the same
1928 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
1929 Format: <Controller#>[,poll interval]
1930 The controller # is the number of the ehci usb debug
1931 port as it is probed via PCI. The poll interval is
1932 optional and is the number seconds in between
1933 each poll cycle to the debug port in case you need
1934 the functionality for interrupting the kernel with
1935 gdb or control-c on the dbgp connection. When
1936 not using this parameter you use sysrq-g to break into
1937 the kernel debugger.
1939 kgdboc= [KGDB,HW] kgdb over consoles.
1940 Requires a tty driver that supports console polling,
1941 or a supported polling keyboard driver (non-usb).
1942 Serial only format: <serial_device>[,baud]
1943 keyboard only format: kbd
1944 keyboard and serial format: kbd,<serial_device>[,baud]
1945 Optional Kernel mode setting:
1946 kms, kbd format: kms,kbd
1947 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
1949 kgdbwait [KGDB] Stop kernel execution and enter the
1950 kernel debugger at the earliest opportunity.
1952 kmac= [MIPS] korina ethernet MAC address.
1953 Configure the RouterBoard 532 series on-chip
1954 Ethernet adapter MAC address.
1956 kmemleak= [KNL] Boot-time kmemleak enable/disable
1957 Valid arguments: on, off
1959 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
1962 kmemcheck= [X86] Boot-time kmemcheck enable/disable/one-shot mode
1963 Valid arguments: 0, 1, 2
1964 kmemcheck=0 (disabled)
1965 kmemcheck=1 (enabled)
1966 kmemcheck=2 (one-shot mode)
1967 Default: 2 (one-shot mode)
1969 kstack=N [X86] Print N words from the kernel stack
1972 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
1973 Default is 0 (don't ignore, but inject #GP)
1975 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit
1979 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
1980 Default is 1 (enabled)
1982 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
1984 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
1986 kvm-intel.ept= [KVM,Intel] Disable extended page tables
1987 (virtualized MMU) support on capable Intel chips.
1988 Default is 1 (enabled)
1990 kvm-intel.emulate_invalid_guest_state=
1991 [KVM,Intel] Enable emulation of invalid guest states
1992 Default is 0 (disabled)
1994 kvm-intel.flexpriority=
1995 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
1996 Default is 1 (enabled)
1999 [KVM,Intel] Enable VMX nesting (nVMX).
2000 Default is 0 (disabled)
2002 kvm-intel.unrestricted_guest=
2003 [KVM,Intel] Disable unrestricted guest feature
2004 (virtualized real and unpaged mode) on capable
2005 Intel chips. Default is 1 (enabled)
2007 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault
2010 Valid arguments: never, cond, always
2012 always: L1D cache flush on every VMENTER.
2013 cond: Flush L1D on VMENTER only when the code between
2014 VMEXIT and VMENTER can leak host memory.
2015 never: Disables the mitigation
2017 Default is cond (do L1 cache flush in specific instances)
2019 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
2020 feature (tagged TLBs) on capable Intel chips.
2021 Default is 1 (enabled)
2023 l1tf= [X86] Control mitigation of the L1TF vulnerability on
2026 The kernel PTE inversion protection is unconditionally
2027 enabled and cannot be disabled.
2030 Provides all available mitigations for the
2031 L1TF vulnerability. Disables SMT and
2032 enables all mitigations in the
2033 hypervisors, i.e. unconditional L1D flush.
2035 SMT control and L1D flush control via the
2036 sysfs interface is still possible after
2037 boot. Hypervisors will issue a warning
2038 when the first VM is started in a
2039 potentially insecure configuration,
2040 i.e. SMT enabled or L1D flush disabled.
2043 Same as 'full', but disables SMT and L1D
2044 flush runtime control. Implies the
2045 'nosmt=force' command line option.
2046 (i.e. sysfs control of SMT is disabled.)
2049 Leaves SMT enabled and enables the default
2050 hypervisor mitigation, i.e. conditional
2053 SMT control and L1D flush control via the
2054 sysfs interface is still possible after
2055 boot. Hypervisors will issue a warning
2056 when the first VM is started in a
2057 potentially insecure configuration,
2058 i.e. SMT enabled or L1D flush disabled.
2062 Disables SMT and enables the default
2063 hypervisor mitigation.
2065 SMT control and L1D flush control via the
2066 sysfs interface is still possible after
2067 boot. Hypervisors will issue a warning
2068 when the first VM is started in a
2069 potentially insecure configuration,
2070 i.e. SMT enabled or L1D flush disabled.
2073 Same as 'flush', but hypervisors will not
2074 warn when a VM is started in a potentially
2075 insecure configuration.
2078 Disables hypervisor mitigations and doesn't
2083 For details see: Documentation/admin-guide/l1tf.rst
2089 lapic [X86-32,APIC] Enable the local APIC even if BIOS
2092 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline
2093 value for LAPIC timer one-shot implementation. Default
2094 back to the programmable timer unit in the LAPIC.
2096 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
2099 libata.dma= [LIBATA] DMA control
2100 libata.dma=0 Disable all PATA and SATA DMA
2101 libata.dma=1 PATA and SATA Disk DMA only
2102 libata.dma=2 ATAPI (CDROM) DMA only
2103 libata.dma=4 Compact Flash DMA only
2104 Combinations also work, so libata.dma=3 enables DMA
2105 for disks and CDROMs, but not CFs.
2107 libata.ignore_hpa= [LIBATA] Ignore HPA limit
2108 libata.ignore_hpa=0 keep BIOS limits (default)
2109 libata.ignore_hpa=1 ignore limits, using full disk
2111 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
2115 libata.force= [LIBATA] Force configurations. The format is comma
2116 separated list of "[ID:]VAL" where ID is
2117 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
2118 matching port, link or device. Basically, it matches
2119 the ATA ID string printed on console by libata. If
2120 the whole ID part is omitted, the last PORT and DEVICE
2121 values are used. If ID hasn't been specified yet, the
2122 configuration applies to all ports, links and devices.
2124 If only DEVICE is omitted, the parameter applies to
2125 the port and all links and devices behind it. DEVICE
2126 number of 0 either selects the first device or the
2127 first fan-out link behind PMP device. It does not
2128 select the host link. DEVICE number of 15 selects the
2129 host link and device attached to it.
2131 The VAL specifies the configuration to force. As long
2132 as there's no ambiguity shortcut notation is allowed.
2133 For example, both 1.5 and 1.5G would work for 1.5Gbps.
2134 The following configurations can be forced.
2136 * Cable type: 40c, 80c, short40c, unk, ign or sata.
2137 Any ID with matching PORT is used.
2139 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
2141 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
2142 udma[/][16,25,33,44,66,100,133] notation is also
2145 * [no]ncq: Turn on or off NCQ.
2147 * [no]ncqtrim: Turn off queued DSM TRIM.
2149 * nohrst, nosrst, norst: suppress hard, soft
2152 * rstonce: only attempt one reset during
2153 hot-unplug link recovery
2155 * dump_id: dump IDENTIFY data.
2157 * atapi_dmadir: Enable ATAPI DMADIR bridge support
2159 * disable: Disable this device.
2161 If there are multiple matching configurations changing
2162 the same attribute, the last one is used.
2164 memblock=debug [KNL] Enable memblock debug messages.
2166 load_ramdisk= [RAM] List of ramdisks to load from floppy
2167 See Documentation/blockdev/ramdisk.txt.
2169 lockd.nlm_grace_period=P [NFS] Assign grace period.
2172 lockd.nlm_tcpport=N [NFS] Assign TCP port.
2175 lockd.nlm_timeout=T [NFS] Assign timeout value.
2178 lockd.nlm_udpport=M [NFS] Assign UDP port.
2181 locktorture.nreaders_stress= [KNL]
2182 Set the number of locking read-acquisition kthreads.
2183 Defaults to being automatically set based on the
2184 number of online CPUs.
2186 locktorture.nwriters_stress= [KNL]
2187 Set the number of locking write-acquisition kthreads.
2189 locktorture.onoff_holdoff= [KNL]
2190 Set time (s) after boot for CPU-hotplug testing.
2192 locktorture.onoff_interval= [KNL]
2193 Set time (s) between CPU-hotplug operations, or
2194 zero to disable CPU-hotplug testing.
2196 locktorture.shuffle_interval= [KNL]
2197 Set task-shuffle interval (jiffies). Shuffling
2198 tasks allows some CPUs to go into dyntick-idle
2199 mode during the locktorture test.
2201 locktorture.shutdown_secs= [KNL]
2202 Set time (s) after boot system shutdown. This
2203 is useful for hands-off automated testing.
2205 locktorture.stat_interval= [KNL]
2206 Time (s) between statistics printk()s.
2208 locktorture.stutter= [KNL]
2209 Time (s) to stutter testing, for example,
2210 specifying five seconds causes the test to run for
2211 five seconds, wait for five seconds, and so on.
2212 This tests the locking primitive's ability to
2213 transition abruptly to and from idle.
2215 locktorture.torture_runnable= [BOOT]
2216 Start locktorture running at boot time.
2218 locktorture.torture_type= [KNL]
2219 Specify the locking implementation to test.
2221 locktorture.verbose= [KNL]
2222 Enable additional printk() statements.
2224 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
2227 loglevel= All Kernel Messages with a loglevel smaller than the
2228 console loglevel will be printed to the console. It can
2229 also be changed with klogd or other programs. The
2230 loglevels are defined as follows:
2232 0 (KERN_EMERG) system is unusable
2233 1 (KERN_ALERT) action must be taken immediately
2234 2 (KERN_CRIT) critical conditions
2235 3 (KERN_ERR) error conditions
2236 4 (KERN_WARNING) warning conditions
2237 5 (KERN_NOTICE) normal but significant condition
2238 6 (KERN_INFO) informational
2239 7 (KERN_DEBUG) debug-level messages
2241 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
2242 in bytes. n must be a power of two and greater
2243 than the minimal size. The minimal size is defined
2244 by LOG_BUF_SHIFT kernel config parameter. There is
2245 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
2246 that allows to increase the default size depending on
2247 the number of CPUs. See init/Kconfig for more details.
2249 logo.nologo [FB] Disables display of the built-in Linux logo.
2250 This may be used to provide more screen space for
2251 kernel log messages and is useful when debugging
2252 kernel boot problems.
2254 lp=0 [LP] Specify parallel ports to use, e.g,
2255 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
2256 lp=reset first parallel port). 'lp=0' disables the
2257 lp=auto printer driver. 'lp=reset' (which can be
2258 specified in addition to the ports) causes
2259 attached printers to be reset. Using
2260 lp=port1,port2,... specifies the parallel ports
2261 to associate lp devices with, starting with
2262 lp0. A port specification may be 'none' to skip
2263 that lp device, or a parport name such as
2264 'parport0'. Specifying 'lp=auto' instead of a
2265 port specification list means that device IDs
2266 from each port should be examined, to see if
2267 an IEEE 1284-compliant printer is attached; if
2268 so, the driver will manage that printer.
2269 See also header of drivers/char/lp.c.
2272 Sets loops_per_jiffy to given constant, thus avoiding
2273 time-consuming boot-time autodetection (up to 250 ms per
2274 CPU). 0 enables autodetection (default). To determine
2275 the correct value for your kernel, boot with normal
2276 autodetection and see what value is printed. Note that
2277 on SMP systems the preset will be applied to all CPUs,
2278 which is likely to cause problems if your CPUs need
2279 significantly divergent settings. An incorrect value
2280 will cause delays in the kernel to be wrong, leading to
2281 unpredictable I/O errors and other breakage. Although
2282 unlikely, in the extreme case this might damage your
2286 Format: <io>,<irq>,<dma>
2288 machvec= [IA-64] Force the use of a particular machine-vector
2289 (machvec) in a generic kernel.
2290 Example: machvec=hpzx1_swiotlb
2292 machtype= [Loongson] Share the same kernel image file between different
2294 Example: machtype=lemote-yeeloong-2f-7inch
2296 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
2297 than or equal to this physical address is ignored.
2299 maxcpus= [SMP] Maximum number of processors that an SMP kernel
2300 will bring up during bootup. maxcpus=n : n >= 0 limits
2301 the kernel to bring up 'n' processors. Surely after
2302 bootup you can bring up the other plugged cpu by executing
2303 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus
2304 only takes effect during system bootup.
2305 While n=0 is a special case, it is equivalent to "nosmp",
2306 which also disables the IO APIC.
2308 max_loop= [LOOP] The number of loop block devices that get
2309 (loop.max_loop) unconditionally pre-created at init time. The default
2310 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
2311 of statically allocating a predefined number, loop
2312 devices can be requested on-demand with the
2313 /dev/loop-control interface.
2315 mce [X86-32] Machine Check Exception
2317 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.txt
2319 md= [HW] RAID subsystems devices and level
2320 See Documentation/md.txt.
2323 Format: <first>,<last>
2324 Specifies range of consoles to be captured by the MDA.
2326 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
2327 Amount of memory to be used when the kernel is not able
2328 to see the whole system memory or for test.
2329 [X86] Work as limiting max address. Use together
2330 with memmap= to avoid physical address space collisions.
2331 Without memmap= PCI devices could be placed at addresses
2332 belonging to unused RAM.
2334 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
2338 [KNL,SH] Allow user to override the default size for
2339 per-device physically contiguous DMA buffers.
2341 memhp_default_state=online/offline
2342 [KNL] Set the initial state for the memory hotplug
2343 onlining policy. If not specified, the default value is
2344 set according to the
2345 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config
2347 See Documentation/memory-hotplug.txt.
2349 memmap=exactmap [KNL,X86] Enable setting of an exact
2350 E820 memory map, as specified by the user.
2351 Such memmap=exactmap lines can be constructed based on
2352 BIOS output or other requirements. See the memmap=nn@ss
2355 memmap=nn[KMG]@ss[KMG]
2356 [KNL] Force usage of a specific region of memory.
2357 Region of memory to be used is from ss to ss+nn.
2359 memmap=nn[KMG]#ss[KMG]
2360 [KNL,ACPI] Mark specific memory as ACPI data.
2361 Region of memory to be marked is from ss to ss+nn.
2363 memmap=nn[KMG]$ss[KMG]
2364 [KNL,ACPI] Mark specific memory as reserved.
2365 Region of memory to be reserved is from ss to ss+nn.
2366 Example: Exclude memory from 0x18690000-0x1869ffff
2367 memmap=64K$0x18690000
2369 memmap=0x10000$0x18690000
2371 memmap=nn[KMG]!ss[KMG]
2372 [KNL,X86] Mark specific memory as protected.
2373 Region of memory to be used, from ss to ss+nn.
2374 The memory region may be marked as e820 type 12 (0xc)
2375 and is NVDIMM or ADR memory.
2377 memory_corruption_check=0/1 [X86]
2378 Some BIOSes seem to corrupt the first 64k of
2379 memory when doing things like suspend/resume.
2380 Setting this option will scan the memory
2381 looking for corruption. Enabling this will
2382 both detect corruption and prevent the kernel
2383 from using the memory being corrupted.
2384 However, its intended as a diagnostic tool; if
2385 repeatable BIOS-originated corruption always
2386 affects the same memory, you can use memmap=
2387 to prevent the kernel from using that memory.
2389 memory_corruption_check_size=size [X86]
2390 By default it checks for corruption in the low
2391 64k, making this memory unavailable for normal
2392 use. Use this parameter to scan for
2393 corruption in more or less memory.
2395 memory_corruption_check_period=seconds [X86]
2396 By default it checks for corruption every 60
2397 seconds. Use this parameter to check at some
2398 other rate. 0 disables periodic checking.
2400 memtest= [KNL,X86,ARM] Enable memtest
2402 default : 0 <disable>
2403 Specifies the number of memtest passes to be
2404 performed. Each pass selects another test
2405 pattern from a given set of patterns. Memtest
2406 fills the memory with this pattern, validates
2407 memory contents and reserves bad memory
2408 regions that are detected.
2410 meye.*= [HW] Set MotionEye Camera parameters
2411 See Documentation/video4linux/meye.txt.
2413 mfgpt_irq= [IA-32] Specify the IRQ to use for the
2414 Multi-Function General Purpose Timers on AMD Geode
2417 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
2418 the BIOS has incorrectly applied a workaround. TinyBIOS
2419 version 0.98 is known to be affected, 0.99 fixes the
2420 problem by letting the user disable the workaround.
2424 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
2425 physical address is ignored.
2427 mini2440= [ARM,HW,KNL]
2428 Format:[0..2][b][c][t]
2430 MINI2440 configuration specification:
2431 0 - The attached screen is the 3.5" TFT
2432 1 - The attached screen is the 7" TFT
2433 2 - The VGA Shield is attached (1024x768)
2434 Leaving out the screen size parameter will not load
2435 the TFT driver, and the framebuffer will be left
2437 b - Enable backlight. The TFT backlight pin will be
2438 linked to the kernel VESA blanking code and a GPIO
2439 LED. This parameter is not necessary when using the
2441 c - Enable the s3c camera interface.
2442 t - Reserved for enabling touchscreen support. The
2443 touchscreen support is not enabled in the mainstream
2444 kernel as of 2.6.30, a preliminary port can be found
2445 in the "bleeding edge" mini2440 support kernel at
2446 http://repo.or.cz/w/linux-2.6/mini2440.git
2449 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
2450 parameter allows control of the logging verbosity for
2451 the additional memory initialisation checks. A value
2452 of 0 disables mminit logging and a level of 4 will
2453 log everything. Information is printed at KERN_DEBUG
2454 so loglevel=8 may also need to be specified.
2457 [KNL] When CONFIG_MODULE_SIG is set, this means that
2458 modules without (valid) signatures will fail to load.
2459 Note that if CONFIG_MODULE_SIG_FORCE is set, that
2460 is always true, so this option does nothing.
2462 module_blacklist= [KNL] Do not load a comma-separated list of
2463 modules. Useful for debugging problem modules.
2466 [MOUSE] Maximum time between finger touching and
2467 leaving touchpad surface for touch to be considered
2468 a tap and be reported as a left button click (for
2469 touchpads working in absolute mode only).
2471 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
2472 reporting absolute coordinates, such as tablets
2473 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
2474 reporting absolute coordinates, such as tablets
2476 movablecore=nn[KMG] [KNL,X86,IA-64,PPC] This parameter
2477 is similar to kernelcore except it specifies the
2478 amount of memory used for migratable allocations.
2479 If both kernelcore and movablecore is specified,
2480 then kernelcore will be at *least* the specified
2481 value but may be more. If movablecore on its own
2482 is specified, the administrator must be careful
2483 that the amount of memory usable for all allocations
2486 movable_node [KNL,X86] Boot-time switch to enable the effects
2487 of CONFIG_MOVABLE_NODE=y. See mm/Kconfig for details.
2489 MTD_Partition= [MTD]
2490 Format: <name>,<region-number>,<size>,<offset>
2492 MTD_Region= [MTD] Format:
2493 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
2496 See drivers/mtd/cmdlinepart.c.
2498 multitce=off [PPC] This parameter disables the use of the pSeries
2499 firmware feature for updating multiple TCE entries
2502 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
2504 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
2506 boundary - index of last SLC block on Flex-OneNAND.
2507 The remaining blocks are configured as MLC blocks.
2508 lock - Configure if Flex-OneNAND boundary should be locked.
2509 Once locked, the boundary cannot be changed.
2510 1 indicates lock status, 0 indicates unlock status.
2513 ARM/S3C2412 JIVE boot control
2515 See arch/arm/mach-s3c2412/mach-jive.c
2517 mtouchusb.raw_coordinates=
2518 [HW] Make the MicroTouch USB driver use raw coordinates
2519 ('y', default) or cooked coordinates ('n')
2521 mtrr_chunk_size=nn[KMG] [X86]
2522 used for mtrr cleanup. It is largest continuous chunk
2523 that could hold holes aka. UC entries.
2525 mtrr_gran_size=nn[KMG] [X86]
2526 Used for mtrr cleanup. It is granularity of mtrr block.
2528 Large value could prevent small alignment from
2531 mtrr_spare_reg_nr=n [X86]
2533 Range: 0,7 : spare reg number
2535 Used for mtrr cleanup. It is spare mtrr entries number.
2536 Set to 2 or more if your graphical card needs more.
2538 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
2540 netdev= [NET] Network devices parameters
2541 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
2542 Note that mem_start is often overloaded to mean
2543 something different and driver-specific.
2544 This usage is only documented in each driver source
2548 [NETFILTER] Enable connection tracking flow accounting
2549 0 to disable accounting
2550 1 to enable accounting
2553 nfsaddrs= [NFS] Deprecated. Use ip= instead.
2554 See Documentation/filesystems/nfs/nfsroot.txt.
2556 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
2557 See Documentation/filesystems/nfs/nfsroot.txt.
2559 nfsrootdebug [NFS] enable nfsroot debugging messages.
2560 See Documentation/filesystems/nfs/nfsroot.txt.
2562 nfs.callback_nr_threads=
2563 [NFSv4] set the total number of threads that the
2564 NFS client will assign to service NFSv4 callback
2567 nfs.callback_tcpport=
2568 [NFS] set the TCP port on which the NFSv4 callback
2569 channel should listen.
2572 [NFS] sets the pathname to the program which is used
2573 to update the NFS client cache entries.
2575 nfs.cache_getent_timeout=
2576 [NFS] sets the timeout after which an attempt to
2577 update a cache entry is deemed to have failed.
2579 nfs.idmap_cache_timeout=
2580 [NFS] set the maximum lifetime for idmapper cache
2584 [NFS] enable 64-bit inode numbers.
2585 If zero, the NFS client will fake up a 32-bit inode
2586 number for the readdir() and stat() syscalls instead
2587 of returning the full 64-bit number.
2588 The default is to return 64-bit inode numbers.
2590 nfs.max_session_cb_slots=
2591 [NFSv4.1] Sets the maximum number of session
2592 slots the client will assign to the callback
2593 channel. This determines the maximum number of
2594 callbacks the client will process in parallel for
2595 a particular server.
2597 nfs.max_session_slots=
2598 [NFSv4.1] Sets the maximum number of session slots
2599 the client will attempt to negotiate with the server.
2600 This limits the number of simultaneous RPC requests
2601 that the client can send to the NFSv4.1 server.
2602 Note that there is little point in setting this
2603 value higher than the max_tcp_slot_table_limit.
2605 nfs.nfs4_disable_idmapping=
2606 [NFSv4] When set to the default of '1', this option
2607 ensures that both the RPC level authentication
2608 scheme and the NFS level operations agree to use
2609 numeric uids/gids if the mount is using the
2610 'sec=sys' security flavour. In effect it is
2611 disabling idmapping, which can make migration from
2612 legacy NFSv2/v3 systems to NFSv4 easier.
2613 Servers that do not support this mode of operation
2614 will be autodetected by the client, and it will fall
2615 back to using the idmapper.
2616 To turn off this behaviour, set the value to '0'.
2618 [NFS4] Specify an additional fixed unique ident-
2619 ification string that NFSv4 clients can insert into
2620 their nfs_client_id4 string. This is typically a
2621 UUID that is generated at system install time.
2623 nfs.send_implementation_id =
2624 [NFSv4.1] Send client implementation identification
2625 information in exchange_id requests.
2626 If zero, no implementation identification information
2628 The default is to send the implementation identification
2631 nfs.recover_lost_locks =
2632 [NFSv4] Attempt to recover locks that were lost due
2633 to a lease timeout on the server. Please note that
2634 doing this risks data corruption, since there are
2635 no guarantees that the file will remain unchanged
2636 after the locks are lost.
2637 If you want to enable the kernel legacy behaviour of
2638 attempting to recover these locks, then set this
2640 The default parameter value of '0' causes the kernel
2641 not to attempt recovery of lost locks.
2643 nfs4.layoutstats_timer =
2644 [NFSv4.2] Change the rate at which the kernel sends
2645 layoutstats to the pNFS metadata server.
2647 Setting this to value to 0 causes the kernel to use
2648 whatever value is the default set by the layout
2649 driver. A non-zero value sets the minimum interval
2650 in seconds between layoutstats transmissions.
2652 nfsd.nfs4_disable_idmapping=
2653 [NFSv4] When set to the default of '1', the NFSv4
2654 server will return only numeric uids and gids to
2655 clients using auth_sys, and will accept numeric uids
2656 and gids from such clients. This is intended to ease
2657 migration from NFSv2/v3.
2659 objlayoutdriver.osd_login_prog=
2660 [NFS] [OBJLAYOUT] sets the pathname to the program which
2661 is used to automatically discover and login into new
2662 osd-targets. Please see:
2663 Documentation/filesystems/pnfs.txt for more explanations
2665 nmi_debug= [KNL,AVR32,SH] Specify one or more actions to take
2666 when a NMI is triggered.
2667 Format: [state][,regs][,debounce][,die]
2669 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
2670 Format: [panic,][nopanic,][num]
2672 0 - turn hardlockup detector in nmi_watchdog off
2673 1 - turn hardlockup detector in nmi_watchdog on
2674 When panic is specified, panic when an NMI watchdog
2675 timeout occurs (or 'nopanic' to override the opposite
2676 default). To disable both hard and soft lockup detectors,
2677 please see 'nowatchdog'.
2678 This is useful when you use a panic=... timeout and
2679 need the box quickly up again.
2681 netpoll.carrier_timeout=
2682 [NET] Specifies amount of time (in seconds) that
2683 netpoll should wait for a carrier. By default netpoll
2686 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
2687 emulation library even if a 387 maths coprocessor
2691 [HW] Never suspend the console
2692 Disable suspending of consoles during suspend and
2693 hibernate operations. Once disabled, debugging
2694 messages can reach various consoles while the rest
2695 of the system is being put to sleep (ie, while
2696 debugging driver suspend/resume hooks). This may
2697 not work reliably with all consoles, but is known
2698 to work with serial and VGA consoles.
2699 To facilitate more flexible debugging, we also add
2700 console_suspend, a printk module parameter to control
2701 it. Users could use console_suspend (usually
2702 /sys/module/printk/parameters/console_suspend) to
2703 turn on/off it dynamically.
2705 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
2706 caches in the slab allocator. Saves per-node memory,
2707 but will impact performance.
2711 noaltinstr [S390] Disables alternative instructions patching
2712 (CPU alternatives feature).
2714 noapic [SMP,APIC] Tells the kernel to not make use of any
2715 IOAPICs that may be present in the system.
2717 noautogroup Disable scheduler automatic task group creation.
2719 nobats [PPC] Do not use BATs for mapping kernel lowmem
2720 on "Classic" PPC cores.
2724 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
2726 nodelayacct [KNL] Disable per-task delay accounting
2728 nodsp [SH] Disable hardware DSP at boot time.
2730 noefi Disable EFI runtime services support.
2735 On X86-32 available only on PAE configured kernels.
2736 noexec=on: enable non-executable mappings (default)
2737 noexec=off: disable non-executable mappings
2740 Disable SMAP (Supervisor Mode Access Prevention)
2741 even if it is supported by processor.
2744 Disable SMEP (Supervisor Mode Execution Prevention)
2745 even if it is supported by processor.
2748 This affects only 32-bit executables.
2749 noexec32=on: enable non-executable mappings (default)
2750 read doesn't imply executable mappings
2751 noexec32=off: disable non-executable mappings
2752 read implies executable mappings
2754 nofpu [MIPS,SH] Disable hardware FPU at boot time.
2756 nofxsr [BUGS=X86-32] Disables x86 floating point extended
2757 register save and restore. The kernel will only save
2758 legacy floating-point registers on task switch.
2760 nohugeiomap [KNL,x86] Disable kernel huge I/O mappings.
2762 nosmt [KNL,S390] Disable symmetric multithreading (SMT).
2763 Equivalent to smt=1.
2765 [KNL,x86] Disable symmetric multithreading (SMT).
2766 nosmt=force: Force disable SMT, cannot be undone
2767 via the sysfs control file.
2769 nospectre_v2 [X86] Disable all mitigations for the Spectre variant 2
2770 (indirect branch prediction) vulnerability. System may
2771 allow data leaks with this option, which is equivalent
2774 nospec_store_bypass_disable
2775 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability
2777 noxsave [BUGS=X86] Disables x86 extended register state save
2778 and restore using xsave. The kernel will fallback to
2779 enabling legacy floating-point and sse state.
2781 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
2782 register states. The kernel will fall back to use
2783 xsave to save the states. By using this parameter,
2784 performance of saving the states is degraded because
2785 xsave doesn't support modified optimization while
2786 xsaveopt supports it on xsaveopt enabled systems.
2788 noxsaves [X86] Disables xsaves and xrstors used in saving and
2789 restoring x86 extended register state in compacted
2790 form of xsave area. The kernel will fall back to use
2791 xsaveopt and xrstor to save and restore the states
2792 in standard form of xsave area. By using this
2793 parameter, xsave area per process might occupy more
2794 memory on xsaves enabled systems.
2796 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
2797 wfi(ARM) instruction doesn't work correctly and not to
2798 use it. This is also useful when using JTAG debugger.
2800 no_file_caps Tells the kernel not to honor file capabilities. The
2801 only way then for a file to be executed with privilege
2802 is to be setuid root or executed by root.
2804 nohalt [IA-64] Tells the kernel not to use the power saving
2805 function PAL_HALT_LIGHT when idle. This increases
2806 power-consumption. On the positive side, it reduces
2807 interrupt wake-up latency, which may improve performance
2808 in certain environments such as networked servers or
2811 nohibernate [HIBERNATION] Disable hibernation and resume.
2813 nohz= [KNL] Boottime enable/disable dynamic ticks
2814 Valid arguments: on, off
2817 nohz_full= [KNL,BOOT]
2818 The argument is a cpu list, as described above.
2819 In kernels built with CONFIG_NO_HZ_FULL=y, set
2820 the specified list of CPUs whose tick will be stopped
2821 whenever possible. The boot CPU will be forced outside
2822 the range to maintain the timekeeping.
2823 The CPUs in this range must also be included in the
2826 noiotrap [SH] Disables trapped I/O port accesses.
2828 noirqdebug [X86-32] Disables the code which attempts to detect and
2829 disable unhandled interrupt sources.
2831 no_timer_check [X86,APIC] Disables the code which tests for
2832 broken timer IRQ sources.
2834 noisapnp [ISAPNP] Disables ISA PnP code.
2836 noinitrd [RAM] Tells the kernel not to load any configured
2839 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
2841 [Deprecated - use intremap=off]
2845 noinvpcid [X86] Disable the INVPCID cpu feature.
2847 nojitter [IA-64] Disables jitter checking for ITC timers.
2849 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
2851 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
2854 no-steal-acc [X86,KVM] Disable paravirtualized steal time accounting.
2855 steal time is computed, but won't influence scheduler
2858 nolapic [X86-32,APIC] Do not enable or use the local APIC.
2860 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
2862 noltlbs [PPC] Do not use large page/tlb entries for kernel
2863 lowmem mapping on PPC40x and PPC8xx
2865 nomca [IA-64] Disable machine check abort handling
2867 nomce [X86-32] Disable Machine Check Exception
2869 nomfgpt [X86-32] Disable Multi-Function General Purpose
2870 Timer usage (for AMD Geode machines).
2872 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
2873 shutdown the other cpus. Instead use the REBOOT_VECTOR
2876 nomodule Disable module load
2878 nopat [X86] Disable PAT (page attribute table extension of
2879 pagetables) support.
2881 nopcid [X86-64] Disable the PCID cpu feature.
2883 norandmaps Don't use address space randomization. Equivalent to
2884 echo 0 > /proc/sys/kernel/randomize_va_space
2886 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
2887 with UP alternatives
2889 nordrand [X86] Disable kernel use of the RDRAND and
2890 RDSEED instructions even if they are supported
2891 by the processor. RDRAND and RDSEED are still
2892 available to user space applications.
2894 noresume [SWSUSP] Disables resume and restores original swap
2897 no-scroll [VGA] Disables scrollback.
2898 This is required for the Braillex ib80-piezo Braille
2899 reader made by F.H. Papenmeier (Germany).
2903 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
2905 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
2906 and disable the IO APIC. legacy for "maxcpus=0".
2908 nosoftlockup [KNL] Disable the soft-lockup detector.
2910 nosync [HW,M68K] Disables sync negotiation for all devices.
2912 notsc [BUGS=X86-32] Disable Time Stamp Counter
2914 nowatchdog [KNL] Disable both lockup detectors, i.e.
2915 soft-lockup and NMI watchdog (hard-lockup).
2919 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
2921 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
2922 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
2923 Some features depend on CPU0. Known dependencies are:
2924 1. Resume from suspend/hibernate depends on CPU0.
2925 Suspend/hibernate will fail if CPU0 is offline and you
2926 need to online CPU0 before suspend/hibernate.
2927 2. PIC interrupts also depend on CPU0. CPU0 can't be
2928 removed if a PIC interrupt is detected.
2929 It's said poweroff/reboot may depend on CPU0 on some
2930 machines although I haven't seen such issues so far
2931 after CPU0 is offline on a few tested machines.
2932 If the dependencies are under your control, you can
2933 turn on cpu0_hotplug.
2935 nptcg= [IA-64] Override max number of concurrent global TLB
2936 purges which is reported from either PAL_VM_SUMMARY or
2939 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
2940 could support. nr_cpus=n : n >= 1 limits the kernel to
2941 support 'n' processors. It could be larger than the
2942 number of already plugged CPU during bootup, later in
2943 runtime you can physically add extra cpu until it reaches
2944 n. So during boot up some boot time memory for per-cpu
2945 variables need be pre-allocated for later physical cpu
2948 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
2950 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing.
2951 Allowed values are enable and disable
2953 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
2954 one of ['zone', 'node', 'default'] can be specified
2955 This can be set from sysctl after boot.
2956 See Documentation/sysctl/vm.txt for details.
2958 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
2959 See Documentation/debugging-via-ohci1394.txt for more
2962 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
2963 Rather than timing out after 20 ms if an EC
2964 command is not properly ACKed, override the length
2965 of the timeout. We have interrupts disabled while
2966 waiting for the ACK, so if this is set too high
2967 interrupts *may* be lost!
2969 omap_mux= [OMAP] Override bootloader pin multiplexing.
2970 Format: <mux_mode0.mode_name=value>...
2971 For example, to override I2C bus2:
2972 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
2974 oprofile.timer= [HW]
2975 Use timer interrupt instead of performance counters
2977 oprofile.cpu_type= Force an oprofile cpu type
2978 This might be useful if you have an older oprofile
2979 userland or if you want common events.
2980 Format: { arch_perfmon }
2981 arch_perfmon: [X86] Force use of architectural
2982 perfmon on Intel CPUs instead of the
2983 CPU specific event set.
2984 timer: [X86] Force use of architectural NMI
2985 timer mode (see also oprofile.timer
2986 for generic hr timer mode)
2988 oops=panic Always panic on oopses. Default is to just kill the
2989 process, but there is a small probability of
2990 deadlocking the machine.
2991 This will also cause panics on machine check exceptions.
2992 Useful together with panic=30 to trigger a reboot.
2995 See Documentation/sound/oss/oss-parameters.txt
2997 page_owner= [KNL] Boot-time page_owner enabling option.
2998 Storage of the information about who allocated
2999 each page is disabled in default. With this switch,
3001 on: enable the feature
3003 page_poison= [KNL] Boot-time parameter changing the state of
3004 poisoning on the buddy allocator.
3005 off: turn off poisoning
3006 on: turn on poisoning
3008 panic= [KNL] Kernel behaviour on panic: delay <timeout>
3009 timeout > 0: seconds before rebooting
3010 timeout = 0: wait forever
3011 timeout < 0: reboot immediately
3014 panic_on_warn panic() instead of WARN(). Useful to cause kdump
3017 crash_kexec_post_notifiers
3018 Run kdump after running panic-notifiers and dumping
3019 kmsg. This only for the users who doubt kdump always
3020 succeeds in any situation.
3021 Note that this also increases risks of kdump failure,
3022 because some panic notifiers can make the crashed
3023 kernel more unstable.
3025 parkbd.port= [HW] Parallel port number the keyboard adapter is
3026 connected to, default is 0.
3028 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
3029 0 for XT, 1 for AT (default is AT).
3032 parport= [HW,PPT] Specify parallel ports. 0 disables.
3033 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
3034 Use 'auto' to force the driver to use any
3035 IRQ/DMA settings detected (the default is to
3036 ignore detected IRQ/DMA settings because of
3037 possible conflicts). You can specify the base
3038 address, IRQ, and DMA settings; IRQ and DMA
3039 should be numbers, or 'auto' (for using detected
3040 settings on that particular port), or 'nofifo'
3041 (to avoid using a FIFO even if it is detected).
3042 Parallel ports are assigned in the order they
3043 are specified on the command line, starting
3046 parport_init_mode= [HW,PPT]
3047 Configure VIA parallel port to operate in
3048 a specific mode. This is necessary on Pegasos
3049 computer where firmware has no options for setting
3050 up parallel port mode and sets it to spp.
3051 Currently this function knows 686a and 8231 chips.
3052 Format: [spp|ps2|epp|ecp|ecpepp]
3055 Halt all CPUs after the first oops has been printed for
3056 the specified number of seconds. This is to be used if
3057 your oopses keep scrolling off the screen.
3062 See header of drivers/block/paride/pcd.c.
3063 See also Documentation/blockdev/paride.txt.
3065 pci=option[,option...] [PCI] various PCI subsystem options:
3066 earlydump [X86] dump PCI config space before the kernel
3068 off [X86] don't probe for the PCI bus
3069 bios [X86-32] force use of PCI BIOS, don't access
3070 the hardware directly. Use this if your machine
3071 has a non-standard PCI host bridge.
3072 nobios [X86-32] disallow use of PCI BIOS, only direct
3073 hardware access methods are allowed. Use this
3074 if you experience crashes upon bootup and you
3075 suspect they are caused by the BIOS.
3076 conf1 [X86] Force use of PCI Configuration Access
3077 Mechanism 1 (config address in IO port 0xCF8,
3078 data in IO port 0xCFC, both 32-bit).
3079 conf2 [X86] Force use of PCI Configuration Access
3080 Mechanism 2 (IO port 0xCF8 is an 8-bit port for
3081 the function, IO port 0xCFA, also 8-bit, sets
3082 bus number. The config space is then accessed
3083 through ports 0xC000-0xCFFF).
3084 See http://wiki.osdev.org/PCI for more info
3085 on the configuration access mechanisms.
3086 noaer [PCIE] If the PCIEAER kernel config parameter is
3087 enabled, this kernel boot option can be used to
3088 disable the use of PCIE advanced error reporting.
3089 nodomains [PCI] Disable support for multiple PCI
3090 root domains (aka PCI segments, in ACPI-speak).
3091 nommconf [X86] Disable use of MMCONFIG for PCI
3093 check_enable_amd_mmconf [X86] check for and enable
3094 properly configured MMIO access to PCI
3095 config space on AMD family 10h CPU
3096 nomsi [MSI] If the PCI_MSI kernel config parameter is
3097 enabled, this kernel boot option can be used to
3098 disable the use of MSI interrupts system-wide.
3099 noioapicquirk [APIC] Disable all boot interrupt quirks.
3100 Safety option to keep boot IRQs enabled. This
3101 should never be necessary.
3102 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
3103 primary IO-APIC for bridges that cannot disable
3104 boot IRQs. This fixes a source of spurious IRQs
3105 when the system masks IRQs.
3106 noioapicreroute [APIC] Disable workaround that uses the
3107 boot IRQ equivalent of an IRQ that connects to
3108 a chipset where boot IRQs cannot be disabled.
3109 The opposite of ioapicreroute.
3110 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
3111 routing table. These calls are known to be buggy
3112 on several machines and they hang the machine
3113 when used, but on other computers it's the only
3114 way to get the interrupt routing table. Try
3115 this option if the kernel is unable to allocate
3116 IRQs or discover secondary PCI buses on your
3118 rom [X86] Assign address space to expansion ROMs.
3119 Use with caution as certain devices share
3120 address decoders between ROMs and other
3122 norom [X86] Do not assign address space to
3123 expansion ROMs that do not already have
3124 BIOS assigned address ranges.
3125 nobar [X86] Do not assign address space to the
3126 BARs that weren't assigned by the BIOS.
3127 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
3128 assigned automatically to PCI devices. You can
3129 make the kernel exclude IRQs of your ISA cards
3131 pirqaddr=0xAAAAA [X86] Specify the physical address
3132 of the PIRQ table (normally generated
3133 by the BIOS) if it is outside the
3134 F0000h-100000h range.
3135 lastbus=N [X86] Scan all buses thru bus #N. Can be
3136 useful if the kernel is unable to find your
3137 secondary buses and you want to tell it
3138 explicitly which ones they are.
3139 assign-busses [X86] Always assign all PCI bus
3140 numbers ourselves, overriding
3141 whatever the firmware may have done.
3142 usepirqmask [X86] Honor the possible IRQ mask stored
3143 in the BIOS $PIR table. This is needed on
3144 some systems with broken BIOSes, notably
3145 some HP Pavilion N5400 and Omnibook XE3
3146 notebooks. This will have no effect if ACPI
3147 IRQ routing is enabled.
3148 noacpi [X86] Do not use ACPI for IRQ routing
3149 or for PCI scanning.
3150 use_crs [X86] Use PCI host bridge window information
3151 from ACPI. On BIOSes from 2008 or later, this
3152 is enabled by default. If you need to use this,
3153 please report a bug.
3154 nocrs [X86] Ignore PCI host bridge windows from ACPI.
3155 If you need to use this, please report a bug.
3156 routeirq Do IRQ routing for all PCI devices.
3157 This is normally done in pci_enable_device(),
3158 so this option is a temporary workaround
3159 for broken drivers that don't call it.
3160 skip_isa_align [X86] do not align io start addr, so can
3161 handle more pci cards
3162 noearly [X86] Don't do any early type 1 scanning.
3163 This might help on some broken boards which
3164 machine check when some devices' config space
3165 is read. But various workarounds are disabled
3166 and some IOMMU drivers will not work.
3167 bfsort Sort PCI devices into breadth-first order.
3168 This sorting is done to get a device
3169 order compatible with older (<= 2.4) kernels.
3170 nobfsort Don't sort PCI devices into breadth-first order.
3171 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
3172 tuning and use the BIOS-configured MPS defaults.
3173 pcie_bus_safe Set every device's MPS to the largest value
3174 supported by all devices below the root complex.
3175 pcie_bus_perf Set device MPS to the largest allowable MPS
3176 based on its parent bus. Also set MRRS (Max
3177 Read Request Size) to the largest supported
3178 value (no larger than the MPS that the device
3179 or bus can support) for best performance.
3180 pcie_bus_peer2peer Set every device's MPS to 128B, which
3181 every device is guaranteed to support. This
3182 configuration allows peer-to-peer DMA between
3183 any pair of devices, possibly at the cost of
3184 reduced performance. This also guarantees
3185 that hot-added devices will work.
3186 cbiosize=nn[KMG] The fixed amount of bus space which is
3187 reserved for the CardBus bridge's IO window.
3188 The default value is 256 bytes.
3189 cbmemsize=nn[KMG] The fixed amount of bus space which is
3190 reserved for the CardBus bridge's memory
3191 window. The default value is 64 megabytes.
3194 [<order of align>@][<domain>:]<bus>:<slot>.<func>[; ...]
3195 [<order of align>@]pci:<vendor>:<device>\
3196 [:<subvendor>:<subdevice>][; ...]
3197 Specifies alignment and device to reassign
3198 aligned memory resources.
3199 If <order of align> is not specified,
3200 PAGE_SIZE is used as alignment.
3201 PCI-PCI bridge can be specified, if resource
3202 windows need to be expanded.
3203 To specify the alignment for several
3204 instances of a device, the PCI vendor,
3205 device, subvendor, and subdevice may be
3206 specified, e.g., 4096@pci:8086:9c22:103c:198f
3207 ecrc= Enable/disable PCIe ECRC (transaction layer
3208 end-to-end CRC checking).
3209 bios: Use BIOS/firmware settings. This is the
3213 hpiosize=nn[KMG] The fixed amount of bus space which is
3214 reserved for hotplug bridge's IO window.
3215 Default size is 256 bytes.
3216 hpmemsize=nn[KMG] The fixed amount of bus space which is
3217 reserved for hotplug bridge's memory window.
3218 Default size is 2 megabytes.
3219 hpbussize=nn The minimum amount of additional bus numbers
3220 reserved for buses below a hotplug bridge.
3222 realloc= Enable/disable reallocating PCI bridge resources
3223 if allocations done by BIOS are too small to
3224 accommodate resources required by all child
3226 off: Turn realloc off
3228 realloc same as realloc=on
3229 noari do not use PCIe ARI.
3230 pcie_scan_all Scan all possible PCIe devices. Otherwise we
3231 only look for one device below a PCIe downstream
3234 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
3237 force Enable ASPM even on devices that claim not to support it.
3238 WARNING: Forcing ASPM on may cause system lockups.
3240 pcie_hp= [PCIE] PCI Express Hotplug driver options:
3241 nomsi Do not use MSI for PCI Express Native Hotplug (this
3242 makes all PCIe ports use INTx for hotplug services).
3244 pcie_ports= [PCIE] PCIe ports handling:
3245 auto Ask the BIOS whether or not to use native PCIe services
3246 associated with PCIe ports (PME, hot-plug, AER). Use
3247 them only if that is allowed by the BIOS.
3248 native Use native PCIe services associated with PCIe ports
3250 compat Treat PCIe ports as PCI-to-PCI bridges, disable the PCIe
3253 pcie_port_pm= [PCIE] PCIe port power management handling:
3254 off Disable power management of all PCIe ports
3255 force Forcibly enable power management of all PCIe ports
3257 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
3258 nomsi Do not use MSI for native PCIe PME signaling (this makes
3259 all PCIe root ports use INTx for all services).
3261 pcmv= [HW,PCMCIA] BadgePAD 4
3265 Keep all power-domains already enabled by bootloader on,
3266 even if no driver has claimed them. This is useful
3267 for debug and development, but should not be
3268 needed on a platform with proper driver support.
3271 See Documentation/blockdev/paride.txt.
3273 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
3276 See arch/parisc/kernel/pdc_chassis.c
3278 percpu_alloc= Select which percpu first chunk allocator to use.
3279 Currently supported values are "embed" and "page".
3280 Archs may support subset or none of the selections.
3281 See comments in mm/percpu.c for details on each
3282 allocator. This parameter is primarily for debugging
3283 and performance comparison.
3286 See Documentation/blockdev/paride.txt.
3289 See Documentation/blockdev/paride.txt.
3291 pirq= [SMP,APIC] Manual mp-table setup
3292 See Documentation/x86/i386/IO-APIC.txt.
3294 plip= [PPT,NET] Parallel port network link
3295 Format: { parport<nr> | timid | 0 }
3296 See also Documentation/parport.txt.
3298 pmtmr= [X86] Manual setup of pmtmr I/O Port.
3299 Override pmtimer IOPort with a hex value.
3303 Enable PNP debug messages (depends on the
3304 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
3305 via /sys/module/pnp/parameters/debug. We always show
3306 current resource usage; turning this on also shows
3307 possible settings and some assignment information.
3313 { on | off | curr | res | no-curr | no-res }
3316 [ISAPNP] Exclude IRQs for the autoconfiguration
3319 [ISAPNP] Exclude DMAs for the autoconfiguration
3321 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
3322 Ranges are in pairs (I/O port base and size).
3325 [ISAPNP] Exclude memory regions for the
3327 Ranges are in pairs (memory base and size).
3329 ports= [IP_VS_FTP] IPVS ftp helper module
3331 Up to 8 (IP_VS_APP_MAX_PORTS) ports
3333 Format: <port>,<port>....
3335 ppc_strict_facility_enable
3336 [PPC] This option catches any kernel floating point,
3337 Altivec, VSX and SPE outside of regions specifically
3338 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
3339 There is some performance impact when enabling this.
3341 print-fatal-signals=
3342 [KNL] debug: print fatal signals
3344 If enabled, warn about various signal handling
3345 related application anomalies: too many signals,
3346 too many POSIX.1 timers, fatal signals causing a
3349 If you hit the warning due to signal overflow,
3350 you might want to try "ulimit -i unlimited".
3354 printk.always_kmsg_dump=
3355 Trigger kmsg_dump for cases other than kernel oops or
3357 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3360 printk.devkmsg={on,off,ratelimit}
3361 Control writing to /dev/kmsg.
3362 on - unlimited logging to /dev/kmsg from userspace
3363 off - logging to /dev/kmsg disabled
3364 ratelimit - ratelimit the logging
3367 printk.time= Show timing data prefixed to each printk message line
3368 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3370 processor.max_cstate= [HW,ACPI]
3371 Limit processor to maximum C-state
3372 max_cstate=9 overrides any DMI blacklist limit.
3374 processor.nocst [HW,ACPI]
3375 Ignore the _CST method to determine C-states,
3376 instead using the legacy FADT method
3378 profile= [KNL] Enable kernel profiling via /proc/profile
3379 Format: [schedule,]<number>
3380 Param: "schedule" - profile schedule points.
3381 Param: <number> - step/bucket size as a power of 2 for
3382 statistical time based profiling.
3383 Param: "sleep" - profile D-state sleeping (millisecs).
3384 Requires CONFIG_SCHEDSTATS
3385 Param: "kvm" - profile VM exits.
3387 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
3389 See Documentation/blockdev/ramdisk.txt.
3391 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
3392 probe for; one of (bare|imps|exps|lifebook|any).
3393 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
3395 psmouse.resetafter= [HW,MOUSE]
3396 Try to reset the device after so many bad packets
3399 [HW,MOUSE] Set desired mouse resolution, in dpi.
3400 psmouse.smartscroll=
3401 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
3402 0 = disabled, 1 = enabled (default).
3404 pstore.backend= Specify the name of the pstore backend to use
3407 See Documentation/blockdev/paride.txt.
3409 pti= [X86_64] Control Page Table Isolation of user and
3410 kernel address spaces. Disabling this feature
3411 removes hardening, but improves performance of
3412 system calls and interrupts.
3414 on - unconditionally enable
3415 off - unconditionally disable
3416 auto - kernel detects whether your CPU model is
3417 vulnerable to issues that PTI mitigates
3419 Not specifying this option is equivalent to pti=auto.
3422 Equivalent to pti=off
3425 [KNL] Number of legacy pty's. Overwrites compiled-in
3428 quiet [KNL] Disable most log messages
3433 See Documentation/md.txt.
3435 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
3436 See Documentation/blockdev/ramdisk.txt.
3439 The argument is a cpu list, as described above.
3441 In kernels built with CONFIG_RCU_NOCB_CPU=y, set
3442 the specified list of CPUs to be no-callback CPUs.
3443 Invocation of these CPUs' RCU callbacks will
3444 be offloaded to "rcuox/N" kthreads created for
3445 that purpose, where "x" is "b" for RCU-bh, "p"
3446 for RCU-preempt, and "s" for RCU-sched, and "N"
3447 is the CPU number. This reduces OS jitter on the
3448 offloaded CPUs, which can be useful for HPC and
3449 real-time workloads. It can also improve energy
3450 efficiency for asymmetric multiprocessors.
3453 Rather than requiring that offloaded CPUs
3454 (specified by rcu_nocbs= above) explicitly
3455 awaken the corresponding "rcuoN" kthreads,
3456 make these kthreads poll for callbacks.
3457 This improves the real-time response for the
3458 offloaded CPUs by relieving them of the need to
3459 wake up the corresponding kthread, but degrades
3460 energy efficiency by requiring that the kthreads
3461 periodically wake up to do the polling.
3463 rcutree.blimit= [KNL]
3464 Set maximum number of finished RCU callbacks to
3465 process in one batch.
3467 rcutree.dump_tree= [KNL]
3468 Dump the structure of the rcu_node combining tree
3469 out at early boot. This is used for diagnostic
3470 purposes, to verify correct tree setup.
3472 rcutree.gp_cleanup_delay= [KNL]
3473 Set the number of jiffies to delay each step of
3474 RCU grace-period cleanup. This only has effect
3475 when CONFIG_RCU_TORTURE_TEST_SLOW_CLEANUP is set.
3477 rcutree.gp_init_delay= [KNL]
3478 Set the number of jiffies to delay each step of
3479 RCU grace-period initialization. This only has
3480 effect when CONFIG_RCU_TORTURE_TEST_SLOW_INIT
3483 rcutree.gp_preinit_delay= [KNL]
3484 Set the number of jiffies to delay each step of
3485 RCU grace-period pre-initialization, that is,
3486 the propagation of recent CPU-hotplug changes up
3487 the rcu_node combining tree. This only has effect
3488 when CONFIG_RCU_TORTURE_TEST_SLOW_PREINIT is set.
3490 rcutree.rcu_fanout_exact= [KNL]
3491 Disable autobalancing of the rcu_node combining
3492 tree. This is used by rcutorture, and might
3493 possibly be useful for architectures having high
3494 cache-to-cache transfer latencies.
3496 rcutree.rcu_fanout_leaf= [KNL]
3497 Change the number of CPUs assigned to each
3498 leaf rcu_node structure. Useful for very
3499 large systems, which will choose the value 64,
3500 and for NUMA systems with large remote-access
3501 latencies, which will choose a value aligned
3502 with the appropriate hardware boundaries.
3504 rcutree.jiffies_till_sched_qs= [KNL]
3505 Set required age in jiffies for a
3506 given grace period before RCU starts
3507 soliciting quiescent-state help from
3508 rcu_note_context_switch().
3510 rcutree.jiffies_till_first_fqs= [KNL]
3511 Set delay from grace-period initialization to
3512 first attempt to force quiescent states.
3513 Units are jiffies, minimum value is zero,
3514 and maximum value is HZ.
3516 rcutree.jiffies_till_next_fqs= [KNL]
3517 Set delay between subsequent attempts to force
3518 quiescent states. Units are jiffies, minimum
3519 value is one, and maximum value is HZ.
3521 rcutree.kthread_prio= [KNL,BOOT]
3522 Set the SCHED_FIFO priority of the RCU per-CPU
3523 kthreads (rcuc/N). This value is also used for
3524 the priority of the RCU boost threads (rcub/N)
3525 and for the RCU grace-period kthreads (rcu_bh,
3526 rcu_preempt, and rcu_sched). If RCU_BOOST is
3527 set, valid values are 1-99 and the default is 1
3528 (the least-favored priority). Otherwise, when
3529 RCU_BOOST is not set, valid values are 0-99 and
3530 the default is zero (non-realtime operation).
3532 rcutree.rcu_nocb_leader_stride= [KNL]
3533 Set the number of NOCB kthread groups, which
3534 defaults to the square root of the number of
3535 CPUs. Larger numbers reduces the wakeup overhead
3536 on the per-CPU grace-period kthreads, but increases
3537 that same overhead on each group's leader.
3539 rcutree.qhimark= [KNL]
3540 Set threshold of queued RCU callbacks beyond which
3541 batch limiting is disabled.
3543 rcutree.qlowmark= [KNL]
3544 Set threshold of queued RCU callbacks below which
3545 batch limiting is re-enabled.
3547 rcutree.rcu_idle_gp_delay= [KNL]
3548 Set wakeup interval for idle CPUs that have
3549 RCU callbacks (RCU_FAST_NO_HZ=y).
3551 rcutree.rcu_idle_lazy_gp_delay= [KNL]
3552 Set wakeup interval for idle CPUs that have
3553 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
3554 Lazy RCU callbacks are those which RCU can
3555 prove do nothing more than free memory.
3557 rcuperf.gp_exp= [KNL]
3558 Measure performance of expedited synchronous
3559 grace-period primitives.
3561 rcuperf.holdoff= [KNL]
3562 Set test-start holdoff period. The purpose of
3563 this parameter is to delay the start of the
3564 test until boot completes in order to avoid
3567 rcuperf.nreaders= [KNL]
3568 Set number of RCU readers. The value -1 selects
3569 N, where N is the number of CPUs. A value
3570 "n" less than -1 selects N-n+1, where N is again
3571 the number of CPUs. For example, -2 selects N
3572 (the number of CPUs), -3 selects N+1, and so on.
3573 A value of "n" less than or equal to -N selects
3576 rcuperf.nwriters= [KNL]
3577 Set number of RCU writers. The values operate
3578 the same as for rcuperf.nreaders.
3579 N, where N is the number of CPUs
3581 rcuperf.perf_runnable= [BOOT]
3582 Start rcuperf running at boot time.
3584 rcuperf.shutdown= [KNL]
3585 Shut the system down after performance tests
3586 complete. This is useful for hands-off automated
3589 rcuperf.perf_type= [KNL]
3590 Specify the RCU implementation to test.
3592 rcuperf.verbose= [KNL]
3593 Enable additional printk() statements.
3595 rcutorture.cbflood_inter_holdoff= [KNL]
3596 Set holdoff time (jiffies) between successive
3597 callback-flood tests.
3599 rcutorture.cbflood_intra_holdoff= [KNL]
3600 Set holdoff time (jiffies) between successive
3601 bursts of callbacks within a given callback-flood
3604 rcutorture.cbflood_n_burst= [KNL]
3605 Set the number of bursts making up a given
3606 callback-flood test. Set this to zero to
3607 disable callback-flood testing.
3609 rcutorture.cbflood_n_per_burst= [KNL]
3610 Set the number of callbacks to be registered
3611 in a given burst of a callback-flood test.
3613 rcutorture.fqs_duration= [KNL]
3614 Set duration of force_quiescent_state bursts
3617 rcutorture.fqs_holdoff= [KNL]
3618 Set holdoff time within force_quiescent_state bursts
3621 rcutorture.fqs_stutter= [KNL]
3622 Set wait time between force_quiescent_state bursts
3625 rcutorture.gp_cond= [KNL]
3626 Use conditional/asynchronous update-side
3627 primitives, if available.
3629 rcutorture.gp_exp= [KNL]
3630 Use expedited update-side primitives, if available.
3632 rcutorture.gp_normal= [KNL]
3633 Use normal (non-expedited) asynchronous
3634 update-side primitives, if available.
3636 rcutorture.gp_sync= [KNL]
3637 Use normal (non-expedited) synchronous
3638 update-side primitives, if available. If all
3639 of rcutorture.gp_cond=, rcutorture.gp_exp=,
3640 rcutorture.gp_normal=, and rcutorture.gp_sync=
3641 are zero, rcutorture acts as if is interpreted
3642 they are all non-zero.
3644 rcutorture.n_barrier_cbs= [KNL]
3645 Set callbacks/threads for rcu_barrier() testing.
3647 rcutorture.nfakewriters= [KNL]
3648 Set number of concurrent RCU writers. These just
3649 stress RCU, they don't participate in the actual
3650 test, hence the "fake".
3652 rcutorture.nreaders= [KNL]
3653 Set number of RCU readers. The value -1 selects
3654 N-1, where N is the number of CPUs. A value
3655 "n" less than -1 selects N-n-2, where N is again
3656 the number of CPUs. For example, -2 selects N
3657 (the number of CPUs), -3 selects N+1, and so on.
3659 rcutorture.object_debug= [KNL]
3660 Enable debug-object double-call_rcu() testing.
3662 rcutorture.onoff_holdoff= [KNL]
3663 Set time (s) after boot for CPU-hotplug testing.
3665 rcutorture.onoff_interval= [KNL]
3666 Set time (s) between CPU-hotplug operations, or
3667 zero to disable CPU-hotplug testing.
3669 rcutorture.shuffle_interval= [KNL]
3670 Set task-shuffle interval (s). Shuffling tasks
3671 allows some CPUs to go into dyntick-idle mode
3672 during the rcutorture test.
3674 rcutorture.shutdown_secs= [KNL]
3675 Set time (s) after boot system shutdown. This
3676 is useful for hands-off automated testing.
3678 rcutorture.stall_cpu= [KNL]
3679 Duration of CPU stall (s) to test RCU CPU stall
3680 warnings, zero to disable.
3682 rcutorture.stall_cpu_holdoff= [KNL]
3683 Time to wait (s) after boot before inducing stall.
3685 rcutorture.stat_interval= [KNL]
3686 Time (s) between statistics printk()s.
3688 rcutorture.stutter= [KNL]
3689 Time (s) to stutter testing, for example, specifying
3690 five seconds causes the test to run for five seconds,
3691 wait for five seconds, and so on. This tests RCU's
3692 ability to transition abruptly to and from idle.
3694 rcutorture.test_boost= [KNL]
3695 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
3696 "Maybe" means test if the RCU implementation
3697 under test support RCU priority boosting.
3699 rcutorture.test_boost_duration= [KNL]
3700 Duration (s) of each individual boost test.
3702 rcutorture.test_boost_interval= [KNL]
3703 Interval (s) between each boost test.
3705 rcutorture.test_no_idle_hz= [KNL]
3706 Test RCU's dyntick-idle handling. See also the
3707 rcutorture.shuffle_interval parameter.
3709 rcutorture.torture_runnable= [BOOT]
3710 Start rcutorture running at boot time.
3712 rcutorture.torture_type= [KNL]
3713 Specify the RCU implementation to test.
3715 rcutorture.verbose= [KNL]
3716 Enable additional printk() statements.
3718 rcupdate.rcu_cpu_stall_suppress= [KNL]
3719 Suppress RCU CPU stall warning messages.
3721 rcupdate.rcu_cpu_stall_timeout= [KNL]
3722 Set timeout for RCU CPU stall warning messages.
3724 rcupdate.rcu_expedited= [KNL]
3725 Use expedited grace-period primitives, for
3726 example, synchronize_rcu_expedited() instead
3727 of synchronize_rcu(). This reduces latency,
3728 but can increase CPU utilization, degrade
3729 real-time latency, and degrade energy efficiency.
3730 No effect on CONFIG_TINY_RCU kernels.
3732 rcupdate.rcu_normal= [KNL]
3733 Use only normal grace-period primitives,
3734 for example, synchronize_rcu() instead of
3735 synchronize_rcu_expedited(). This improves
3736 real-time latency, CPU utilization, and
3737 energy efficiency, but can expose users to
3738 increased grace-period latency. This parameter
3739 overrides rcupdate.rcu_expedited. No effect on
3740 CONFIG_TINY_RCU kernels.
3742 rcupdate.rcu_normal_after_boot= [KNL]
3743 Once boot has completed (that is, after
3744 rcu_end_inkernel_boot() has been invoked), use
3745 only normal grace-period primitives. No effect
3746 on CONFIG_TINY_RCU kernels.
3748 rcupdate.rcu_task_stall_timeout= [KNL]
3749 Set timeout in jiffies for RCU task stall warning
3750 messages. Disable with a value less than or equal
3753 rcupdate.rcu_self_test= [KNL]
3754 Run the RCU early boot self tests
3756 rcupdate.rcu_self_test_bh= [KNL]
3757 Run the RCU bh early boot self tests
3759 rcupdate.rcu_self_test_sched= [KNL]
3760 Run the RCU sched early boot self tests
3764 Run specified binary instead of /init from the ramdisk,
3765 used for early userspace startup. See initrd.
3768 Format (x86 or x86_64):
3769 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
3771 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
3773 Where reboot_mode is one of warm (soft) or cold (hard) or gpio,
3774 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
3775 reboot_force is either force or not specified,
3776 reboot_cpu is s[mp]#### with #### being the processor
3777 to be used for rebooting.
3780 [KNL, SMP] Set scheduler's default relax_domain_level.
3781 See Documentation/cgroup-v1/cpusets.txt.
3783 relative_sleep_states=
3784 [SUSPEND] Use sleep state labeling where the deepest
3785 state available other than hibernation is always "mem".
3786 Format: { "0" | "1" }
3787 0 -- Traditional sleep state labels.
3788 1 -- Relative sleep state labels.
3790 reserve= [KNL,BUGS] Force the kernel to ignore some iomem area
3792 reservetop= [X86-32]
3794 Reserves a hole at the top of the kernel virtual
3799 Set the amount of memory to reserve for BIOS at
3800 the bottom of the address space.
3802 reset_devices [KNL] Force drivers to reset the underlying device
3803 during initialization.
3806 Specify the partition device for software suspend
3808 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
3810 resume_offset= [SWSUSP]
3811 Specify the offset from the beginning of the partition
3812 given by "resume=" at which the swap header is located,
3813 in <PAGE_SIZE> units (needed only for swap files).
3814 See Documentation/power/swsusp-and-swap-files.txt
3816 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
3817 read the resume files
3819 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
3820 Useful for devices that are detected asynchronously
3821 (e.g. USB and MMC devices).
3823 hibernate= [HIBERNATION]
3824 noresume Don't check if there's a hibernation image
3825 present during boot.
3826 nocompress Don't compress/decompress hibernation images.
3827 no Disable hibernation and resume.
3828 protect_image Turn on image protection during restoration
3829 (that will set all pages holding image data
3830 during restoration read-only).
3832 retain_initrd [RAM] Keep initrd memory after extraction
3834 rfkill.default_state=
3835 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
3836 etc. communication is blocked by default.
3839 rfkill.master_switch_mode=
3840 0 The "airplane mode" button does nothing.
3841 1 The "airplane mode" button toggles between everything
3842 blocked and the previous configuration.
3843 2 The "airplane mode" button toggles between everything
3844 blocked and everything unblocked.
3846 rhash_entries= [KNL,NET]
3847 Set number of hash buckets for route cache
3849 ro [KNL] Mount root device read-only on boot
3852 on Mark read-only kernel memory as read-only (default).
3853 off Leave read-only kernel memory writable for debugging.
3856 Enable the uart passthrough on the designated usb port
3857 on Rockchip SoCs. When active, the signals of the
3858 debug-uart get routed to the D+ and D- pins of the usb
3859 port and the regular usb controller gets disabled.
3861 root= [KNL] Root filesystem
3862 See name_to_dev_t comment in init/do_mounts.c.
3864 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
3865 mount the root filesystem
3867 rootflags= [KNL] Set root filesystem mount option string
3869 rootfstype= [KNL] Set root filesystem type
3871 rootwait [KNL] Wait (indefinitely) for root device to show up.
3872 Useful for devices that are detected asynchronously
3873 (e.g. USB and MMC devices).
3875 rproc_mem=nn[KMG][@address]
3876 [KNL,ARM,CMA] Remoteproc physical memory block.
3877 Memory area to be used by remote processor image,
3880 rw [KNL] Mount root device read-write on boot
3882 S [KNL] Run init in single mode
3884 s390_iommu= [HW,S390]
3885 Set s390 IOTLB flushing mode
3887 With strict flushing every unmap operation will result in
3888 an IOTLB flush. Default is lazy flushing before reuse,
3892 See drivers/net/irda/sa1100_ir.c.
3894 sbni= [NET] Granch SBNI12 leased line adapter
3896 sched_debug [KNL] Enables verbose scheduler debug messages.
3898 schedstats= [KNL,X86] Enable or disable scheduled statistics.
3899 Allowed values are enable and disable. This feature
3900 incurs a small amount of overhead in the scheduler
3901 but is useful for debugging and performance tuning.
3903 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
3904 xtime_lock contention on larger systems, and/or RCU lock
3905 contention on all systems with CONFIG_MAXSMP set.
3906 Format: { "0" | "1" }
3907 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
3909 Note: increases power consumption, thus should only be
3910 enabled if running jitter sensitive (HPC/RT) workloads.
3912 security= [SECURITY] Choose a security module to enable at boot.
3913 If this boot parameter is not specified, only the first
3914 security module asking for security registration will be
3915 loaded. An invalid security module name will be treated
3916 as if no module has been chosen.
3918 selinux= [SELINUX] Disable or enable SELinux at boot time.
3919 Format: { "0" | "1" }
3920 See security/selinux/Kconfig help text.
3923 Default value is set via kernel config option.
3924 If enabled at boot time, /selinux/disable can be used
3925 later to disable prior to initial policy load.
3927 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
3928 Format: { "0" | "1" }
3929 See security/apparmor/Kconfig help text
3932 Default value is set via kernel config option.
3934 serialnumber [BUGS=X86-32]
3937 Maximal number of shapers.
3939 show_msr= [x86] show boot-time MSR settings
3940 Format: { <integer> }
3941 Show boot-time (BIOS-initialized) MSR settings.
3942 The parameter means the number of CPUs to show,
3943 for example 1 means boot CPU only.
3951 Disable merging of slabs with similar size. May be
3952 necessary if there is some reason to distinguish
3953 allocs to different slabs. Debug options disable
3954 merging on their own.
3955 For more information see Documentation/vm/slub.txt.
3957 slab_max_order= [MM, SLAB]
3958 Determines the maximum allowed order for slabs.
3959 A high setting may cause OOMs due to memory
3960 fragmentation. Defaults to 1 for systems with
3961 more than 32MB of RAM, 0 otherwise.
3963 slub_debug[=options[,slabs]] [MM, SLUB]
3964 Enabling slub_debug allows one to determine the
3965 culprit if slab objects become corrupted. Enabling
3966 slub_debug can create guard zones around objects and
3967 may poison objects when not in use. Also tracks the
3968 last alloc / free. For more information see
3969 Documentation/vm/slub.txt.
3971 slub_max_order= [MM, SLUB]
3972 Determines the maximum allowed order for slabs.
3973 A high setting may cause OOMs due to memory
3974 fragmentation. For more information see
3975 Documentation/vm/slub.txt.
3977 slub_min_objects= [MM, SLUB]
3978 The minimum number of objects per slab. SLUB will
3979 increase the slab order up to slub_max_order to
3980 generate a sufficiently large slab able to contain
3981 the number of objects indicated. The higher the number
3982 of objects the smaller the overhead of tracking slabs
3983 and the less frequently locks need to be acquired.
3984 For more information see Documentation/vm/slub.txt.
3986 slub_min_order= [MM, SLUB]
3987 Determines the minimum page order for slabs. Must be
3988 lower than slub_max_order.
3989 For more information see Documentation/vm/slub.txt.
3991 slub_nomerge [MM, SLUB]
3992 Same with slab_nomerge. This is supported for legacy.
3993 See slab_nomerge for more information.
3996 Format: <io1>[,<io2>[,...,<io8>]]
3998 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
3999 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
4000 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
4001 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
4002 smsc-ircc2.ircc_irq= [HW] IRQ line
4003 smsc-ircc2.ircc_dma= [HW] DMA channel
4004 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
4005 0: Toshiba Satellite 1800 (GP data pin select)
4006 1: Fast pin select (default)
4009 smt [KNL,S390] Set the maximum number of threads (logical
4010 CPUs) to use per physical CPU on systems capable of
4011 symmetric multithreading (SMT). Will be capped to the
4012 actual hardware limit.
4014 Default: -1 (no limit)
4017 [KNL] Should the soft-lockup detector generate panics.
4020 softlockup_all_cpu_backtrace=
4021 [KNL] Should the soft-lockup detector generate
4022 backtraces on all cpus.
4025 sonypi.*= [HW] Sony Programmable I/O Control Device driver
4026 See Documentation/laptops/sonypi.txt
4028 spectre_v2= [X86] Control mitigation of Spectre variant 2
4029 (indirect branch speculation) vulnerability.
4031 on - unconditionally enable
4032 off - unconditionally disable
4033 auto - kernel detects whether your CPU model is
4036 Selecting 'on' will, and 'auto' may, choose a
4037 mitigation method at run time according to the
4038 CPU, the available microcode, the setting of the
4039 CONFIG_RETPOLINE configuration option, and the
4040 compiler with which the kernel was built.
4042 Specific mitigations can also be selected manually:
4044 retpoline - replace indirect branches
4045 retpoline,generic - google's original retpoline
4046 retpoline,amd - AMD-specific minimal thunk
4048 Not specifying this option is equivalent to
4051 spec_store_bypass_disable=
4052 [HW] Control Speculative Store Bypass (SSB) Disable mitigation
4053 (Speculative Store Bypass vulnerability)
4055 Certain CPUs are vulnerable to an exploit against a
4056 a common industry wide performance optimization known
4057 as "Speculative Store Bypass" in which recent stores
4058 to the same memory location may not be observed by
4059 later loads during speculative execution. The idea
4060 is that such stores are unlikely and that they can
4061 be detected prior to instruction retirement at the
4062 end of a particular speculation execution window.
4064 In vulnerable processors, the speculatively forwarded
4065 store can be used in a cache side channel attack, for
4066 example to read memory to which the attacker does not
4067 directly have access (e.g. inside sandboxed code).
4069 This parameter controls whether the Speculative Store
4070 Bypass optimization is used.
4072 on - Unconditionally disable Speculative Store Bypass
4073 off - Unconditionally enable Speculative Store Bypass
4074 auto - Kernel detects whether the CPU model contains an
4075 implementation of Speculative Store Bypass and
4076 picks the most appropriate mitigation. If the
4077 CPU is not vulnerable, "off" is selected. If the
4078 CPU is vulnerable the default mitigation is
4079 architecture and Kconfig dependent. See below.
4080 prctl - Control Speculative Store Bypass per thread
4081 via prctl. Speculative Store Bypass is enabled
4082 for a process by default. The state of the control
4083 is inherited on fork.
4084 seccomp - Same as "prctl" above, but all seccomp threads
4085 will disable SSB unless they explicitly opt out.
4087 Not specifying this option is equivalent to
4088 spec_store_bypass_disable=auto.
4090 Default mitigations:
4091 X86: If CONFIG_SECCOMP=y "seccomp", otherwise "prctl"
4093 spia_io_base= [HW,MTD]
4099 Speculative Store Bypass Disable control
4101 On CPUs that are vulnerable to the Speculative
4102 Store Bypass vulnerability and offer a
4103 firmware based mitigation, this parameter
4104 indicates how the mitigation should be used:
4106 force-on: Unconditionally enable mitigation for
4107 for both kernel and userspace
4108 force-off: Unconditionally disable mitigation for
4109 for both kernel and userspace
4110 kernel: Always enable mitigation in the
4111 kernel, and offer a prctl interface
4112 to allow userspace to register its
4113 interest in being mitigated too.
4115 stack_guard_gap= [MM]
4116 override the default stack gap protection. The value
4117 is in page units and it defines how many pages prior
4118 to (for stacks growing down) resp. after (for stacks
4119 growing up) the main stack are reserved for no other
4120 mapping. Default value is 256 pages.
4123 Enabled the stack tracer on boot up.
4125 stacktrace_filter=[function-list]
4126 [FTRACE] Limit the functions that the stack tracer
4127 will trace at boot up. function-list is a comma separated
4128 list of functions. This list can be changed at run
4129 time by the stack_trace_filter file in the debugfs
4130 tracing directory. Note, this enables stack tracing
4131 and the stacktrace above is not needed.
4135 Set the STI (builtin display/keyboard on the HP-PARISC
4136 machines) console (graphic card) which should be used
4137 as the initial boot-console.
4138 See also comment in drivers/video/console/sticore.c.
4141 See comment in drivers/video/console/sticore.c.
4144 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
4146 sunrpc.min_resvport=
4147 sunrpc.max_resvport=
4149 SunRPC servers often require that client requests
4150 originate from a privileged port (i.e. a port in the
4151 range 0 < portnr < 1024).
4152 An administrator who wishes to reserve some of these
4153 ports for other uses may adjust the range that the
4154 kernel's sunrpc client considers to be privileged
4155 using these two parameters to set the minimum and
4156 maximum port values.
4158 sunrpc.svc_rpc_per_connection_limit=
4160 Limit the number of requests that the server will
4161 process in parallel from a single connection.
4162 The default value is 0 (no limit).
4166 Control how the NFS server code allocates CPUs to
4167 service thread pools. Depending on how many NICs
4168 you have and where their interrupts are bound, this
4169 option will affect which CPUs will do NFS serving.
4170 Note: this parameter cannot be changed while the
4171 NFS server is running.
4173 auto the server chooses an appropriate mode
4174 automatically using heuristics
4175 global a single global pool contains all CPUs
4176 percpu one pool for each CPU
4177 pernode one pool for each NUMA node (equivalent
4178 to global on non-NUMA machines)
4180 sunrpc.tcp_slot_table_entries=
4181 sunrpc.udp_slot_table_entries=
4183 Sets the upper limit on the number of simultaneous
4184 RPC calls that can be sent from the client to a
4185 server. Increasing these values may allow you to
4186 improve throughput, but will also increase the
4187 amount of memory reserved for use by the client.
4189 suspend.pm_test_delay=
4191 Sets the number of seconds to remain in a suspend test
4192 mode before resuming the system (see
4193 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
4194 is set. Default value is 5.
4197 [KNL] Enable accounting of swap in memory resource
4198 controller if no parameter or 1 is given or disable
4199 it if 0 is given (See Documentation/cgroup-v1/memory.txt)
4201 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
4202 Format: { <int> | force | noforce }
4203 <int> -- Number of I/O TLB slabs
4204 force -- force using of bounce buffers even if they
4205 wouldn't be automatically used by the kernel
4206 noforce -- Never use bounce buffers (for debugging)
4210 sysfs.deprecated=0|1 [KNL]
4211 Enable/disable old style sysfs layout for old udev
4212 on older distributions. When this option is enabled
4213 very new udev will not work anymore. When this option
4214 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
4215 in older udev will not work anymore.
4216 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
4217 the kernel configuration.
4219 sysrq_always_enabled
4221 Ignore sysrq setting - this boot parameter will
4222 neutralize any effect of /proc/sys/kernel/sysrq.
4223 Useful for debugging.
4225 tcpmhash_entries= [KNL,NET]
4226 Set the number of tcp_metrics_hash slots.
4227 Default value is 8192 or 16384 depending on total
4228 ram pages. This is used to specify the TCP metrics
4229 cache size. See Documentation/networking/ip-sysctl.txt
4230 "tcp_no_metrics_save" section for more details.
4234 test_suspend= [SUSPEND][,N]
4235 Specify "mem" (for Suspend-to-RAM) or "standby" (for
4236 standby suspend) or "freeze" (for suspend type freeze)
4237 as the system sleep state during system startup with
4238 the optional capability to repeat N number of times.
4239 The system is woken from this state using a
4240 wakeup-capable RTC alarm.
4242 thash_entries= [KNL,NET]
4243 Set number of hash buckets for TCP connection
4245 thermal.act= [HW,ACPI]
4246 -1: disable all active trip points in all thermal zones
4247 <degrees C>: override all lowest active trip points
4249 thermal.crt= [HW,ACPI]
4250 -1: disable all critical trip points in all thermal zones
4251 <degrees C>: override all critical trip points
4253 thermal.nocrt= [HW,ACPI]
4254 Set to disable actions on ACPI thermal zone
4255 critical and hot trip points.
4257 thermal.off= [HW,ACPI]
4258 1: disable ACPI thermal control
4260 thermal.psv= [HW,ACPI]
4261 -1: disable all passive trip points
4262 <degrees C>: override all passive trip points to this
4265 thermal.tzp= [HW,ACPI]
4266 Specify global default ACPI thermal zone polling rate
4267 <deci-seconds>: poll all this frequency
4268 0: no polling (default)
4271 Force threading of all interrupt handlers except those
4272 marked explicitly IRQF_NO_THREAD.
4275 Enable the Transcendent memory driver if built-in.
4277 tmem.cleancache=0|1 [KNL, XEN]
4278 Default is on (1). Disable the usage of the cleancache
4279 API to send anonymous pages to the hypervisor.
4281 tmem.frontswap=0|1 [KNL, XEN]
4282 Default is on (1). Disable the usage of the frontswap
4283 API to send swap pages to the hypervisor. If disabled
4284 the selfballooning and selfshrinking are force disabled.
4286 tmem.selfballooning=0|1 [KNL, XEN]
4287 Default is on (1). Disable the driving of swap pages
4290 tmem.selfshrinking=0|1 [KNL, XEN]
4291 Default is on (1). Partial swapoff that immediately
4292 transfers pages from Xen hypervisor back to the
4293 kernel based on different criteria.
4297 Specify if the kernel should make use of the cpu
4298 topology information if the hardware supports this.
4299 The scheduler will make use of this information and
4300 e.g. base its process migration decisions on it.
4303 topology_updates= [KNL, PPC, NUMA]
4305 Specify if the kernel should ignore (off)
4306 topology updates sent by the hypervisor to this
4311 tpm_suspend_pcr=[HW,TPM]
4312 Format: integer pcr id
4313 Specify that at suspend time, the tpm driver
4314 should extend the specified pcr with zeros,
4315 as a workaround for some chips which fail to
4316 flush the last written pcr on TPM_SaveState.
4317 This will guarantee that all the other pcrs
4320 trace_buf_size=nn[KMG]
4321 [FTRACE] will set tracing buffer size on each cpu.
4323 trace_event=[event-list]
4324 [FTRACE] Set and start specified trace events in order
4325 to facilitate early boot debugging. The event-list is a
4326 comma separated list of trace events to enable. See
4327 also Documentation/trace/events.txt
4329 trace_options=[option-list]
4330 [FTRACE] Enable or disable tracer options at boot.
4331 The option-list is a comma delimited list of options
4332 that can be enabled or disabled just as if you were
4333 to echo the option name into
4335 /sys/kernel/debug/tracing/trace_options
4337 For example, to enable stacktrace option (to dump the
4338 stack trace of each event), add to the command line:
4340 trace_options=stacktrace
4342 See also Documentation/trace/ftrace.txt "trace options"
4346 Have the tracepoints sent to printk as well as the
4347 tracing ring buffer. This is useful for early boot up
4348 where the system hangs or reboots and does not give the
4349 option for reading the tracing buffer or performing a
4350 ftrace_dump_on_oops.
4352 To turn off having tracepoints sent to printk,
4353 echo 0 > /proc/sys/kernel/tracepoint_printk
4354 Note, echoing 1 into this file without the
4355 tracepoint_printk kernel cmdline option has no effect.
4359 Having tracepoints sent to printk() and activating high
4360 frequency tracepoints such as irq or sched, can cause
4361 the system to live lock.
4364 [FTRACE] enable this option to disable tracing when a
4365 warning is hit. This turns off "tracing_on". Tracing can
4366 be enabled again by echoing '1' into the "tracing_on"
4367 file located in /sys/kernel/debug/tracing/
4369 This option is useful, as it disables the trace before
4370 the WARNING dump is called, which prevents the trace to
4371 be filled with content caused by the warning output.
4373 This option can also be set at run time via the sysctl
4374 option: kernel/traceoff_on_warning
4376 transparent_hugepage=
4378 Format: [always|madvise|never]
4379 Can be used to control the default behavior of the system
4380 with respect to transparent hugepages.
4381 See Documentation/vm/transhuge.txt for more details.
4383 tsc= Disable clocksource stability checks for TSC.
4385 [x86] reliable: mark tsc clocksource as reliable, this
4386 disables clocksource verification at runtime, as well
4387 as the stability checks done at bootup. Used to enable
4388 high-resolution timer mode on older hardware, and in
4389 virtualized environment.
4390 [x86] noirqtime: Do not use TSC to do irq accounting.
4391 Used to run time disable IRQ_TIME_ACCOUNTING on any
4392 platforms where RDTSC is slow and this accounting
4395 turbografx.map[2|3]= [HW,JOY]
4396 TurboGraFX parallel port interface
4398 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
4399 See also Documentation/input/joystick-parport.txt
4401 udbg-immortal [PPC] When debugging early kernel crashes that
4402 happen after console_init() and before a proper
4403 console driver takes over, this boot options might
4404 help "seeing" what's going on.
4406 uhash_entries= [KNL,NET]
4407 Set number of hash buckets for UDP/UDP-Lite connections
4410 [USB] Ignore overcurrent events (default N).
4411 Some badly-designed motherboards generate lots of
4412 bogus events, for ports that aren't wired to
4413 anything. Set this parameter to avoid log spamming.
4414 Note that genuine overcurrent events won't be
4418 [X86] Cause panic on unknown NMI.
4420 usbcore.authorized_default=
4421 [USB] Default USB device authorization:
4422 (default -1 = authorized except for wireless USB,
4423 0 = not authorized, 1 = authorized)
4425 usbcore.autosuspend=
4426 [USB] The autosuspend time delay (in seconds) used
4427 for newly-detected USB devices (default 2). This
4428 is the time required before an idle device will be
4429 autosuspended. Devices for which the delay is set
4430 to a negative value won't be autosuspended at all.
4432 usbcore.usbfs_snoop=
4433 [USB] Set to log all usbfs traffic (default 0 = off).
4435 usbcore.usbfs_snoop_max=
4436 [USB] Maximum number of bytes to snoop in each URB
4439 usbcore.blinkenlights=
4440 [USB] Set to cycle leds on hubs (default 0 = off).
4442 usbcore.old_scheme_first=
4443 [USB] Start with the old device initialization
4444 scheme (default 0 = off).
4446 usbcore.usbfs_memory_mb=
4447 [USB] Memory limit (in MB) for buffers allocated by
4448 usbfs (default = 16, 0 = max = 2047).
4450 usbcore.use_both_schemes=
4451 [USB] Try the other device initialization scheme
4452 if the first one fails (default 1 = enabled).
4454 usbcore.initial_descriptor_timeout=
4455 [USB] Specifies timeout for the initial 64-byte
4456 USB_REQ_GET_DESCRIPTOR request in milliseconds
4457 (default 5000 = 5.0 seconds).
4459 usbcore.nousb [USB] Disable the USB subsystem
4462 [USBHID] The interval which mice are to be polled at.
4464 usb-storage.delay_use=
4465 [UMS] The delay in seconds before a new device is
4466 scanned for Logical Units (default 1).
4469 [UMS] A list of quirks entries to supplement or
4470 override the built-in unusual_devs list. List
4471 entries are separated by commas. Each entry has
4472 the form VID:PID:Flags where VID and PID are Vendor
4473 and Product ID values (4-digit hex numbers) and
4474 Flags is a set of characters, each corresponding
4475 to a common usb-storage quirk flag as follows:
4476 a = SANE_SENSE (collect more than 18 bytes
4478 b = BAD_SENSE (don't collect more than 18
4479 bytes of sense data);
4480 c = FIX_CAPACITY (decrease the reported
4481 device capacity by one sector);
4482 d = NO_READ_DISC_INFO (don't use
4483 READ_DISC_INFO command);
4484 e = NO_READ_CAPACITY_16 (don't use
4485 READ_CAPACITY_16 command);
4486 f = NO_REPORT_OPCODES (don't use report opcodes
4488 g = MAX_SECTORS_240 (don't transfer more than
4489 240 sectors at a time, uas only);
4490 h = CAPACITY_HEURISTICS (decrease the
4491 reported device capacity by one
4492 sector if the number is odd);
4493 i = IGNORE_DEVICE (don't bind to this
4495 j = NO_REPORT_LUNS (don't use report luns
4497 l = NOT_LOCKABLE (don't try to lock and
4498 unlock ejectable media);
4499 m = MAX_SECTORS_64 (don't transfer more
4500 than 64 sectors = 32 KB at a time);
4501 n = INITIAL_READ10 (force a retry of the
4502 initial READ(10) command);
4503 o = CAPACITY_OK (accept the capacity
4504 reported by the device);
4505 p = WRITE_CACHE (the device cache is ON
4507 r = IGNORE_RESIDUE (the device reports
4508 bogus residue values);
4509 s = SINGLE_LUN (the device has only one
4511 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
4512 commands, uas only);
4513 u = IGNORE_UAS (don't bind to the uas driver);
4514 w = NO_WP_DETECT (don't test whether the
4515 medium is write-protected).
4516 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE
4517 even if the device claims no cache)
4518 Example: quirks=0419:aaf5:rl,0421:0433:rc
4520 user_debug= [KNL,ARM]
4522 See arch/arm/Kconfig.debug help text.
4523 1 - undefined instruction events
4525 4 - invalid data aborts
4528 Example: user_debug=31
4531 [X86] Flags controlling user PTE allocations.
4533 nohigh = do not allocate PTE pages in
4534 HIGHMEM regardless of setting
4538 On X86_32, this is an alias for vdso32=. Otherwise:
4540 vdso=1: enable VDSO (the default)
4541 vdso=0: disable VDSO mapping
4543 vdso32= [X86] Control the 32-bit vDSO
4544 vdso32=1: enable 32-bit VDSO
4545 vdso32=0 or vdso32=2: disable 32-bit VDSO
4547 See the help text for CONFIG_COMPAT_VDSO for more
4548 details. If CONFIG_COMPAT_VDSO is set, the default is
4549 vdso32=0; otherwise, the default is vdso32=1.
4551 For compatibility with older kernels, vdso32=2 is an
4554 Try vdso32=0 if you encounter an error that says:
4555 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
4558 vector=percpu: enable percpu vector domain
4560 video= [FB] Frame buffer configuration
4561 See Documentation/fb/modedb.txt.
4563 video.brightness_switch_enabled= [0,1]
4564 If set to 1, on receiving an ACPI notify event
4565 generated by hotkey, video driver will adjust brightness
4566 level and then send out the event to user space through
4567 the allocated input device; If set to 0, video driver
4568 will only send out the event without touching backlight
4573 [VMMIO] Memory mapped virtio (platform) device.
4575 <size>@<baseaddr>:<irq>[:<id>]
4577 <size> := size (can use standard suffixes
4579 <baseaddr> := physical base address
4580 <irq> := interrupt number (as passed to
4582 <id> := (optional) platform device id
4584 virtio_mmio.device=1K@0x100b0000:48:7
4586 Can be used multiple times for multiple devices.
4588 vga= [BOOT,X86-32] Select a particular video mode
4589 See Documentation/x86/boot.txt and
4590 Documentation/svga.txt.
4591 Use vga=ask for menu.
4592 This is actually a boot loader parameter; the value is
4593 passed to the kernel using a special protocol.
4595 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
4596 size of <nn>. This can be used to increase the
4597 minimum size (128MB on x86). It can also be used to
4598 decrease the size and leave more room for directly
4601 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
4604 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
4607 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
4611 Controls the behavior of vsyscalls (i.e. calls to
4612 fixed addresses of 0xffffffffff600x00 from legacy
4613 code). Most statically-linked binaries and older
4614 versions of glibc use these calls. Because these
4615 functions are at fixed addresses, they make nice
4616 targets for exploits that can control RIP.
4618 emulate [default] Vsyscalls turn into traps and are
4619 emulated reasonably safely.
4621 native Vsyscalls are native syscall instructions.
4622 This is a little bit faster than trapping
4623 and makes a few dynamic recompilers work
4624 better than they would in emulation mode.
4625 It also makes exploits much easier to write.
4627 none Vsyscalls don't work at all. This makes
4628 them quite hard to use for exploits but
4629 might break your system.
4631 vt.color= [VT] Default text color.
4632 Format: 0xYX, X = foreground, Y = background.
4633 Default: 0x07 = light gray on black.
4635 vt.cur_default= [VT] Default cursor shape.
4636 Format: 0xCCBBAA, where AA, BB, and CC are the same as
4637 the parameters of the <Esc>[?A;B;Cc escape sequence;
4638 see VGA-softcursor.txt. Default: 2 = underline.
4640 vt.default_blu= [VT]
4641 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
4642 Change the default blue palette of the console.
4643 This is a 16-member array composed of values
4646 vt.default_grn= [VT]
4647 Format: <green0>,<green1>,<green2>,...,<green15>
4648 Change the default green palette of the console.
4649 This is a 16-member array composed of values
4652 vt.default_red= [VT]
4653 Format: <red0>,<red1>,<red2>,...,<red15>
4654 Change the default red palette of the console.
4655 This is a 16-member array composed of values
4661 Set system-wide default UTF-8 mode for all tty's.
4662 Default is 1, i.e. UTF-8 mode is enabled for all
4663 newly opened terminals.
4665 vt.global_cursor_default=
4668 Set system-wide default for whether a cursor
4669 is shown on new VTs. Default is -1,
4670 i.e. cursors will be created by default unless
4671 overridden by individual drivers. 0 will hide
4672 cursors, 1 will display them.
4674 vt.italic= [VT] Default color for italic text; 0-15.
4677 vt.underline= [VT] Default color for underlined text; 0-15.
4680 watchdog timers [HW,WDT] For information on watchdog timers,
4681 see Documentation/watchdog/watchdog-parameters.txt
4682 or other driver-specific files in the
4683 Documentation/watchdog/ directory.
4685 workqueue.watchdog_thresh=
4686 If CONFIG_WQ_WATCHDOG is configured, workqueue can
4687 warn stall conditions and dump internal state to
4688 help debugging. 0 disables workqueue stall
4689 detection; otherwise, it's the stall threshold
4690 duration in seconds. The default value is 30 and
4691 it can be updated at runtime by writing to the
4692 corresponding sysfs file.
4694 workqueue.disable_numa
4695 By default, all work items queued to unbound
4696 workqueues are affine to the NUMA nodes they're
4697 issued on, which results in better behavior in
4698 general. If NUMA affinity needs to be disabled for
4699 whatever reason, this option can be used. Note
4700 that this also can be controlled per-workqueue for
4701 workqueues visible under /sys/bus/workqueue/.
4703 workqueue.power_efficient
4704 Per-cpu workqueues are generally preferred because
4705 they show better performance thanks to cache
4706 locality; unfortunately, per-cpu workqueues tend to
4707 be more power hungry than unbound workqueues.
4709 Enabling this makes the per-cpu workqueues which
4710 were observed to contribute significantly to power
4711 consumption unbound, leading to measurably lower
4712 power usage at the cost of small performance
4715 The default value of this parameter is determined by
4716 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
4718 workqueue.debug_force_rr_cpu
4719 Workqueue used to implicitly guarantee that work
4720 items queued without explicit CPU specified are put
4721 on the local CPU. This guarantee is no longer true
4722 and while local CPU is still preferred work items
4723 may be put on foreign CPUs. This debug option
4724 forces round-robin CPU selection to flush out
4725 usages which depend on the now broken guarantee.
4726 When enabled, memory and cache locality will be
4729 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
4730 default x2apic cluster mode on platforms
4733 x86_intel_mid_timer= [X86-32,APBT]
4734 Choose timer option for x86 Intel MID platform.
4735 Two valid options are apbt timer only and lapic timer
4736 plus one apbt timer for broadcast timer.
4737 x86_intel_mid_timer=apbt_only | lapic_and_apbt
4739 xen_512gb_limit [KNL,X86-64,XEN]
4740 Restricts the kernel running paravirtualized under Xen
4741 to use only up to 512 GB of RAM. The reason to do so is
4742 crash analysis tools and Xen tools for doing domain
4743 save/restore/migration must be enabled to handle larger
4746 xen_emul_unplug= [HW,X86,XEN]
4747 Unplug Xen emulated devices
4748 Format: [unplug0,][unplug1]
4749 ide-disks -- unplug primary master IDE devices
4750 aux-ide-disks -- unplug non-primary-master IDE devices
4751 nics -- unplug network devices
4752 all -- unplug all emulated devices (NICs and IDE disks)
4753 unnecessary -- unplugging emulated devices is
4754 unnecessary even if the host did not respond to
4756 never -- do not unplug even if version check succeeds
4758 xen_nopvspin [X86,XEN]
4759 Disables the ticketlock slowpath using Xen PV
4763 Disables the PV optimizations forcing the HVM guest to
4764 run as generic HVM guest with no PV drivers.
4766 xirc2ps_cs= [NET,PCMCIA]
4768 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
4770 ______________________________________________________________________
4774 Add more DRM drivers.