1 acpi= [HW,ACPI,X86,ARM64] 2 Advanced Configuration and Power Interface 3 Format: { force | on | off | strict | noirq | rsdt | 4 copy_dsdt } 5 force -- enable ACPI if default was off 6 on -- enable ACPI but allow fallback to DT [arm64] 7 off -- disable ACPI if default was on 8 noirq -- do not use ACPI for IRQ routing 9 strict -- Be less tolerant of platforms that are not 10 strictly ACPI specification compliant. 11 rsdt -- prefer RSDT over (default) XSDT 12 copy_dsdt -- copy DSDT to memory 13 For ARM64, ONLY "acpi=off", "acpi=on" or "acpi=force" 14 are available 15 16 See also Documentation/power/runtime_pm.rst, pci=noacpi 17 18 acpi_apic_instance= [ACPI, IOAPIC] 19 Format: <int> 20 2: use 2nd APIC table, if available 21 1,0: use 1st APIC table 22 default: 0 23 24 acpi_backlight= [HW,ACPI] 25 acpi_backlight=vendor 26 acpi_backlight=video 27 If set to vendor, prefer vendor specific driver 28 (e.g. thinkpad_acpi, sony_acpi, etc.) instead 29 of the ACPI video.ko driver. 30 31 acpi_force_32bit_fadt_addr 32 force FADT to use 32 bit addresses rather than the 33 64 bit X_* addresses. Some firmware have broken 64 34 bit addresses for force ACPI ignore these and use 35 the older legacy 32 bit addresses. 36 37 acpica_no_return_repair [HW, ACPI] 38 Disable AML predefined validation mechanism 39 This mechanism can repair the evaluation result to make 40 the return objects more ACPI specification compliant. 41 This option is useful for developers to identify the 42 root cause of an AML interpreter issue when the issue 43 has something to do with the repair mechanism. 44 45 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG] 46 acpi.debug_level= [HW,ACPI,ACPI_DEBUG] 47 Format: <int> 48 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI 49 debug output. Bits in debug_layer correspond to a 50 _COMPONENT in an ACPI source file, e.g., 51 #define _COMPONENT ACPI_PCI_COMPONENT 52 Bits in debug_level correspond to a level in 53 ACPI_DEBUG_PRINT statements, e.g., 54 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ... 55 The debug_level mask defaults to "info". See 56 Documentation/firmware-guide/acpi/debug.rst for more information about 57 debug layers and levels. 58 59 Enable processor driver info messages: 60 acpi.debug_layer=0x20000000 61 Enable PCI/PCI interrupt routing info messages: 62 acpi.debug_layer=0x400000 63 Enable AML "Debug" output, i.e., stores to the Debug 64 object while interpreting AML: 65 acpi.debug_layer=0xffffffff acpi.debug_level=0x2 66 Enable all messages related to ACPI hardware: 67 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff 68 69 Some values produce so much output that the system is 70 unusable. The "log_buf_len" parameter may be useful 71 if you need to capture more output. 72 73 acpi_enforce_resources= [ACPI] 74 { strict | lax | no } 75 Check for resource conflicts between native drivers 76 and ACPI OperationRegions (SystemIO and SystemMemory 77 only). IO ports and memory declared in ACPI might be 78 used by the ACPI subsystem in arbitrary AML code and 79 can interfere with legacy drivers. 80 strict (default): access to resources claimed by ACPI 81 is denied; legacy drivers trying to access reserved 82 resources will fail to bind to device using them. 83 lax: access to resources claimed by ACPI is allowed; 84 legacy drivers trying to access reserved resources 85 will bind successfully but a warning message is logged. 86 no: ACPI OperationRegions are not marked as reserved, 87 no further checks are performed. 88 89 acpi_force_table_verification [HW,ACPI] 90 Enable table checksum verification during early stage. 91 By default, this is disabled due to x86 early mapping 92 size limitation. 93 94 acpi_irq_balance [HW,ACPI] 95 ACPI will balance active IRQs 96 default in APIC mode 97 98 acpi_irq_nobalance [HW,ACPI] 99 ACPI will not move active IRQs (default) 100 default in PIC mode 101 102 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 103 Format: <irq>,<irq>... 104 105 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 106 use by PCI 107 Format: <irq>,<irq>... 108 109 acpi_mask_gpe= [HW,ACPI] 110 Due to the existence of _Lxx/_Exx, some GPEs triggered 111 by unsupported hardware/firmware features can result in 112 GPE floodings that cannot be automatically disabled by 113 the GPE dispatcher. 114 This facility can be used to prevent such uncontrolled 115 GPE floodings. 116 Format: <int> 117 118 acpi_no_auto_serialize [HW,ACPI] 119 Disable auto-serialization of AML methods 120 AML control methods that contain the opcodes to create 121 named objects will be marked as "Serialized" by the 122 auto-serialization feature. 123 This feature is enabled by default. 124 This option allows to turn off the feature. 125 126 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 127 kernels. 128 129 acpi_no_static_ssdt [HW,ACPI] 130 Disable installation of static SSDTs at early boot time 131 By default, SSDTs contained in the RSDT/XSDT will be 132 installed automatically and they will appear under 133 /sys/firmware/acpi/tables. 134 This option turns off this feature. 135 Note that specifying this option does not affect 136 dynamic table installation which will install SSDT 137 tables to /sys/firmware/acpi/tables/dynamic. 138 139 acpi_rsdp= [ACPI,EFI,KEXEC] 140 Pass the RSDP address to the kernel, mostly used 141 on machines running EFI runtime service to boot the 142 second kernel for kdump. 143 144 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 145 Format: To spoof as Windows 98: ="Microsoft Windows" 146 147 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 148 of 2 which is mandated by ACPI 6) as the supported ACPI 149 specification revision (when using this switch, it may 150 be necessary to carry out a cold reboot _twice_ in a 151 row to make it take effect on the platform firmware). 152 153 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 154 acpi_osi="string1" # add string1 155 acpi_osi="!string2" # remove string2 156 acpi_osi=!* # remove all strings 157 acpi_osi=! # disable all built-in OS vendor 158 strings 159 acpi_osi=!! # enable all built-in OS vendor 160 strings 161 acpi_osi= # disable all strings 162 163 'acpi_osi=!' can be used in combination with single or 164 multiple 'acpi_osi="string1"' to support specific OS 165 vendor string(s). Note that such command can only 166 affect the default state of the OS vendor strings, thus 167 it cannot affect the default state of the feature group 168 strings and the current state of the OS vendor strings, 169 specifying it multiple times through kernel command line 170 is meaningless. This command is useful when one do not 171 care about the state of the feature group strings which 172 should be controlled by the OSPM. 173 Examples: 174 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 175 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 176 can make '_OSI("Windows 2000")' TRUE. 177 178 'acpi_osi=' cannot be used in combination with other 179 'acpi_osi=' command lines, the _OSI method will not 180 exist in the ACPI namespace. NOTE that such command can 181 only affect the _OSI support state, thus specifying it 182 multiple times through kernel command line is also 183 meaningless. 184 Examples: 185 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 186 FALSE. 187 188 'acpi_osi=!*' can be used in combination with single or 189 multiple 'acpi_osi="string1"' to support specific 190 string(s). Note that such command can affect the 191 current state of both the OS vendor strings and the 192 feature group strings, thus specifying it multiple times 193 through kernel command line is meaningful. But it may 194 still not able to affect the final state of a string if 195 there are quirks related to this string. This command 196 is useful when one want to control the state of the 197 feature group strings to debug BIOS issues related to 198 the OSPM features. 199 Examples: 200 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 201 '_OSI("Module Device")' FALSE. 202 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 203 '_OSI("Module Device")' TRUE. 204 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 205 equivalent to 206 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 207 and 208 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 209 they all will make '_OSI("Windows 2000")' TRUE. 210 211 acpi_pm_good [X86] 212 Override the pmtimer bug detection: force the kernel 213 to assume that this machine's pmtimer latches its value 214 and always returns good values. 215 216 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode 217 Format: { level | edge | high | low } 218 219 acpi_skip_timer_override [HW,ACPI] 220 Recognize and ignore IRQ0/pin2 Interrupt Override. 221 For broken nForce2 BIOS resulting in XT-PIC timer. 222 223 acpi_sleep= [HW,ACPI] Sleep options 224 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig, 225 old_ordering, nonvs, sci_force_enable, nobl } 226 See Documentation/power/video.rst for information on 227 s3_bios and s3_mode. 228 s3_beep is for debugging; it makes the PC's speaker beep 229 as soon as the kernel's real-mode entry point is called. 230 s4_nohwsig prevents ACPI hardware signature from being 231 used during resume from hibernation. 232 old_ordering causes the ACPI 1.0 ordering of the _PTS 233 control method, with respect to putting devices into 234 low power states, to be enforced (the ACPI 2.0 ordering 235 of _PTS is used by default). 236 nonvs prevents the kernel from saving/restoring the 237 ACPI NVS memory during suspend/hibernation and resume. 238 sci_force_enable causes the kernel to set SCI_EN directly 239 on resume from S1/S3 (which is against the ACPI spec, 240 but some broken systems don't work without it). 241 nobl causes the internal blacklist of systems known to 242 behave incorrectly in some ways with respect to system 243 suspend and resume to be ignored (use wisely). 244 245 acpi_use_timer_override [HW,ACPI] 246 Use timer override. For some broken Nvidia NF5 boards 247 that require a timer override, but don't have HPET 248 249 add_efi_memmap [EFI; X86] Include EFI memory map in 250 kernel's map of available physical RAM. 251 252 agp= [AGP] 253 { off | try_unsupported } 254 off: disable AGP support 255 try_unsupported: try to drive unsupported chipsets 256 (may crash computer or cause data corruption) 257 258 ALSA [HW,ALSA] 259 See Documentation/sound/alsa-configuration.rst 260 261 alignment= [KNL,ARM] 262 Allow the default userspace alignment fault handler 263 behaviour to be specified. Bit 0 enables warnings, 264 bit 1 enables fixups, and bit 2 sends a segfault. 265 266 align_va_addr= [X86-64] 267 Align virtual addresses by clearing slice [14:12] when 268 allocating a VMA at process creation time. This option 269 gives you up to 3% performance improvement on AMD F15h 270 machines (where it is enabled by default) for a 271 CPU-intensive style benchmark, and it can vary highly in 272 a microbenchmark depending on workload and compiler. 273 274 32: only for 32-bit processes 275 64: only for 64-bit processes 276 on: enable for both 32- and 64-bit processes 277 off: disable for both 32- and 64-bit processes 278 279 alloc_snapshot [FTRACE] 280 Allocate the ftrace snapshot buffer on boot up when the 281 main buffer is allocated. This is handy if debugging 282 and you need to use tracing_snapshot() on boot up, and 283 do not want to use tracing_snapshot_alloc() as it needs 284 to be done where GFP_KERNEL allocations are allowed. 285 286 amd_iommu= [HW,X86-64] 287 Pass parameters to the AMD IOMMU driver in the system. 288 Possible values are: 289 fullflush - enable flushing of IO/TLB entries when 290 they are unmapped. Otherwise they are 291 flushed before they will be reused, which 292 is a lot of faster 293 off - do not initialize any AMD IOMMU found in 294 the system 295 force_isolation - Force device isolation for all 296 devices. The IOMMU driver is not 297 allowed anymore to lift isolation 298 requirements as needed. This option 299 does not override iommu=pt 300 301 amd_iommu_dump= [HW,X86-64] 302 Enable AMD IOMMU driver option to dump the ACPI table 303 for AMD IOMMU. With this option enabled, AMD IOMMU 304 driver will print ACPI tables for AMD IOMMU during 305 IOMMU initialization. 306 307 amd_iommu_intr= [HW,X86-64] 308 Specifies one of the following AMD IOMMU interrupt 309 remapping modes: 310 legacy - Use legacy interrupt remapping mode. 311 vapic - Use virtual APIC mode, which allows IOMMU 312 to inject interrupts directly into guest. 313 This mode requires kvm-amd.avic=1. 314 (Default when IOMMU HW support is present.) 315 316 amijoy.map= [HW,JOY] Amiga joystick support 317 Map of devices attached to JOY0DAT and JOY1DAT 318 Format: <a>,<b> 319 See also Documentation/input/joydev/joystick.rst 320 321 analog.map= [HW,JOY] Analog joystick and gamepad support 322 Specifies type or capabilities of an analog joystick 323 connected to one of 16 gameports 324 Format: <type1>,<type2>,..<type16> 325 326 apc= [HW,SPARC] 327 Power management functions (SPARCstation-4/5 + deriv.) 328 Format: noidle 329 Disable APC CPU standby support. SPARCstation-Fox does 330 not play well with APC CPU idle - disable it if you have 331 APC and your system crashes randomly. 332 333 apic= [APIC,X86] Advanced Programmable Interrupt Controller 334 Change the output verbosity while booting 335 Format: { quiet (default) | verbose | debug } 336 Change the amount of debugging information output 337 when initialising the APIC and IO-APIC components. 338 For X86-32, this can also be used to specify an APIC 339 driver name. 340 Format: apic=driver_name 341 Examples: apic=bigsmp 342 343 apic_extnmi= [APIC,X86] External NMI delivery setting 344 Format: { bsp (default) | all | none } 345 bsp: External NMI is delivered only to CPU 0 346 all: External NMIs are broadcast to all CPUs as a 347 backup of CPU 0 348 none: External NMI is masked for all CPUs. This is 349 useful so that a dump capture kernel won't be 350 shot down by NMI 351 352 autoconf= [IPV6] 353 See Documentation/networking/ipv6.txt. 354 355 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 356 Limit apic dumping. The parameter defines the maximal 357 number of local apics being dumped. Also it is possible 358 to set it to "all" by meaning -- no limit here. 359 Format: { 1 (default) | 2 | ... | all }. 360 The parameter valid if only apic=debug or 361 apic=verbose is specified. 362 Example: apic=debug show_lapic=all 363 364 apm= [APM] Advanced Power Management 365 See header of arch/x86/kernel/apm_32.c. 366 367 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards 368 Format: <io>,<irq>,<nodeID> 369 370 ataflop= [HW,M68k] 371 372 atarimouse= [HW,MOUSE] Atari Mouse 373 374 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 375 EzKey and similar keyboards 376 377 atkbd.reset= [HW] Reset keyboard during initialization 378 379 atkbd.set= [HW] Select keyboard code set 380 Format: <int> (2 = AT (default), 3 = PS/2) 381 382 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 383 keyboards 384 385 atkbd.softraw= [HW] Choose between synthetic and real raw mode 386 Format: <bool> (0 = real, 1 = synthetic (default)) 387 388 atkbd.softrepeat= [HW] 389 Use software keyboard repeat 390 391 audit= [KNL] Enable the audit sub-system 392 Format: { "0" | "1" | "off" | "on" } 393 0 | off - kernel audit is disabled and can not be 394 enabled until the next reboot 395 unset - kernel audit is initialized but disabled and 396 will be fully enabled by the userspace auditd. 397 1 | on - kernel audit is initialized and partially 398 enabled, storing at most audit_backlog_limit 399 messages in RAM until it is fully enabled by the 400 userspace auditd. 401 Default: unset 402 403 audit_backlog_limit= [KNL] Set the audit queue size limit. 404 Format: <int> (must be >=0) 405 Default: 64 406 407 bau= [X86_UV] Enable the BAU on SGI UV. The default 408 behavior is to disable the BAU (i.e. bau=0). 409 Format: { "0" | "1" } 410 0 - Disable the BAU. 411 1 - Enable the BAU. 412 unset - Disable the BAU. 413 414 baycom_epp= [HW,AX25] 415 Format: <io>,<mode> 416 417 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem 418 Format: <io>,<mode> 419 See header of drivers/net/hamradio/baycom_par.c. 420 421 baycom_ser_fdx= [HW,AX25] 422 BayCom Serial Port AX.25 Modem (Full Duplex Mode) 423 Format: <io>,<irq>,<mode>[,<baud>] 424 See header of drivers/net/hamradio/baycom_ser_fdx.c. 425 426 baycom_ser_hdx= [HW,AX25] 427 BayCom Serial Port AX.25 Modem (Half Duplex Mode) 428 Format: <io>,<irq>,<mode> 429 See header of drivers/net/hamradio/baycom_ser_hdx.c. 430 431 blkdevparts= Manual partition parsing of block device(s) for 432 embedded devices based on command line input. 433 See Documentation/block/cmdline-partition.rst 434 435 boot_delay= Milliseconds to delay each printk during boot. 436 Values larger than 10 seconds (10000) are changed to 437 no delay (0). 438 Format: integer 439 440 bootmem_debug [KNL] Enable bootmem allocator debug messages. 441 442 bert_disable [ACPI] 443 Disable BERT OS support on buggy BIOSes. 444 445 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 446 bttv.radio= Most important insmod options are available as 447 kernel args too. 448 bttv.pll= See Documentation/media/v4l-drivers/bttv.rst 449 bttv.tuner= 450 451 bulk_remove=off [PPC] This parameter disables the use of the pSeries 452 firmware feature for flushing multiple hpte entries 453 at a time. 454 455 c101= [NET] Moxa C101 synchronous serial card 456 457 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 458 Sometimes CPU hardware bugs make them report the cache 459 size incorrectly. The kernel will attempt work arounds 460 to fix known problems, but for some CPUs it is not 461 possible to determine what the correct size should be. 462 This option provides an override for these situations. 463 464 carrier_timeout= 465 [NET] Specifies amount of time (in seconds) that 466 the kernel should wait for a network carrier. By default 467 it waits 120 seconds. 468 469 ca_keys= [KEYS] This parameter identifies a specific key(s) on 470 the system trusted keyring to be used for certificate 471 trust validation. 472 format: { id:<keyid> | builtin } 473 474 cca= [MIPS] Override the kernel pages' cache coherency 475 algorithm. Accepted values range from 0 to 7 476 inclusive. See arch/mips/include/asm/pgtable-bits.h 477 for platform specific values (SB1, Loongson3 and 478 others). 479 480 ccw_timeout_log [S390] 481 See Documentation/s390/common_io.rst for details. 482 483 cgroup_disable= [KNL] Disable a particular controller 484 Format: {name of the controller(s) to disable} 485 The effects of cgroup_disable=foo are: 486 - foo isn't auto-mounted if you mount all cgroups in 487 a single hierarchy 488 - foo isn't visible as an individually mountable 489 subsystem 490 {Currently only "memory" controller deal with this and 491 cut the overhead, others just disable the usage. So 492 only cgroup_disable=memory is actually worthy} 493 494 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 495 Format: { { controller | "all" | "named" } 496 [,{ controller | "all" | "named" }...] } 497 Like cgroup_disable, but only applies to cgroup v1; 498 the blacklisted controllers remain available in cgroup2. 499 "all" blacklists all controllers and "named" disables 500 named mounts. Specifying both "all" and "named" disables 501 all v1 hierarchies. 502 503 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 504 Format: <string> 505 nosocket -- Disable socket memory accounting. 506 nokmem -- Disable kernel memory accounting. 507 508 checkreqprot [SELINUX] Set initial checkreqprot flag value. 509 Format: { "0" | "1" } 510 See security/selinux/Kconfig help text. 511 0 -- check protection applied by kernel (includes 512 any implied execute protection). 513 1 -- check protection requested by application. 514 Default value is set via a kernel config option. 515 Value can be changed at runtime via 516 /selinux/checkreqprot. 517 518 cio_ignore= [S390] 519 See Documentation/s390/common_io.rst for details. 520 clk_ignore_unused 521 [CLK] 522 Prevents the clock framework from automatically gating 523 clocks that have not been explicitly enabled by a Linux 524 device driver but are enabled in hardware at reset or 525 by the bootloader/firmware. Note that this does not 526 force such clocks to be always-on nor does it reserve 527 those clocks in any way. This parameter is useful for 528 debug and development, but should not be needed on a 529 platform with proper driver support. For more 530 information, see Documentation/driver-api/clk.rst. 531 532 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 533 [Deprecated] 534 Forces specified clocksource (if available) to be used 535 when calculating gettimeofday(). If specified 536 clocksource is not available, it defaults to PIT. 537 Format: { pit | tsc | cyclone | pmtmr } 538 539 clocksource= Override the default clocksource 540 Format: <string> 541 Override the default clocksource and use the clocksource 542 with the name specified. 543 Some clocksource names to choose from, depending on 544 the platform: 545 [all] jiffies (this is the base, fallback clocksource) 546 [ACPI] acpi_pm 547 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 548 pxa_timer,timer3,32k_counter,timer0_1 549 [X86-32] pit,hpet,tsc; 550 scx200_hrt on Geode; cyclone on IBM x440 551 [MIPS] MIPS 552 [PARISC] cr16 553 [S390] tod 554 [SH] SuperH 555 [SPARC64] tick 556 [X86-64] hpet,tsc 557 558 clocksource.arm_arch_timer.evtstrm= 559 [ARM,ARM64] 560 Format: <bool> 561 Enable/disable the eventstream feature of the ARM 562 architected timer so that code using WFE-based polling 563 loops can be debugged more effectively on production 564 systems. 565 566 clearcpuid=BITNUM [X86] 567 Disable CPUID feature X for the kernel. See 568 arch/x86/include/asm/cpufeatures.h for the valid bit 569 numbers. Note the Linux specific bits are not necessarily 570 stable over kernel options, but the vendor specific 571 ones should be. 572 Also note that user programs calling CPUID directly 573 or using the feature without checking anything 574 will still see it. This just prevents it from 575 being used by the kernel or shown in /proc/cpuinfo. 576 Also note the kernel might malfunction if you disable 577 some critical bits. 578 579 cma=nn[MG]@[start[MG][-end[MG]]] 580 [ARM,X86,KNL] 581 Sets the size of kernel global memory area for 582 contiguous memory allocations and optionally the 583 placement constraint by the physical address range of 584 memory allocations. A value of 0 disables CMA 585 altogether. For more information, see 586 include/linux/dma-contiguous.h 587 588 cmo_free_hint= [PPC] Format: { yes | no } 589 Specify whether pages are marked as being inactive 590 when they are freed. This is used in CMO environments 591 to determine OS memory pressure for page stealing by 592 a hypervisor. 593 Default: yes 594 595 coherent_pool=nn[KMG] [ARM,KNL] 596 Sets the size of memory pool for coherent, atomic dma 597 allocations, by default set to 256K. 598 599 com20020= [HW,NET] ARCnet - COM20020 chipset 600 Format: 601 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 602 603 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 604 Format: <io>[,<irq>] 605 606 com90xx= [HW,NET] 607 ARCnet - COM90xx chipset (memory-mapped buffers) 608 Format: <io>[,<irq>[,<memstart>]] 609 610 condev= [HW,S390] console device 611 conmode= 612 613 console= [KNL] Output console device and options. 614 615 tty<n> Use the virtual console device <n>. 616 617 ttyS<n>[,options] 618 ttyUSB0[,options] 619 Use the specified serial port. The options are of 620 the form "bbbbpnf", where "bbbb" is the baud rate, 621 "p" is parity ("n", "o", or "e"), "n" is number of 622 bits, and "f" is flow control ("r" for RTS or 623 omit it). Default is "9600n8". 624 625 See Documentation/admin-guide/serial-console.rst for more 626 information. See 627 Documentation/networking/netconsole.txt for an 628 alternative. 629 630 uart[8250],io,<addr>[,options] 631 uart[8250],mmio,<addr>[,options] 632 uart[8250],mmio16,<addr>[,options] 633 uart[8250],mmio32,<addr>[,options] 634 uart[8250],0x<addr>[,options] 635 Start an early, polled-mode console on the 8250/16550 636 UART at the specified I/O port or MMIO address, 637 switching to the matching ttyS device later. 638 MMIO inter-register address stride is either 8-bit 639 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 640 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 641 to be equivalent to 'mmio'. 'options' are specified in 642 the same format described for ttyS above; if unspecified, 643 the h/w is not re-initialized. 644 645 hvc<n> Use the hypervisor console device <n>. This is for 646 both Xen and PowerPC hypervisors. 647 648 If the device connected to the port is not a TTY but a braille 649 device, prepend "brl," before the device type, for instance 650 console=brl,ttyS0 651 For now, only VisioBraille is supported. 652 653 console_msg_format= 654 [KNL] Change console messages format 655 default 656 By default we print messages on consoles in 657 "[time stamp] text\n" format (time stamp may not be 658 printed, depending on CONFIG_PRINTK_TIME or 659 `printk_time' param). 660 syslog 661 Switch to syslog format: "<%u>[time stamp] text\n" 662 IOW, each message will have a facility and loglevel 663 prefix. The format is similar to one used by syslog() 664 syscall, or to executing "dmesg -S --raw" or to reading 665 from /proc/kmsg. 666 667 consoleblank= [KNL] The console blank (screen saver) timeout in 668 seconds. A value of 0 disables the blank timer. 669 Defaults to 0. 670 671 coredump_filter= 672 [KNL] Change the default value for 673 /proc/<pid>/coredump_filter. 674 See also Documentation/filesystems/proc.txt. 675 676 coresight_cpu_debug.enable 677 [ARM,ARM64] 678 Format: <bool> 679 Enable/disable the CPU sampling based debugging. 680 0: default value, disable debugging 681 1: enable debugging at boot time 682 683 cpuidle.off=1 [CPU_IDLE] 684 disable the cpuidle sub-system 685 686 cpuidle.governor= 687 [CPU_IDLE] Name of the cpuidle governor to use. 688 689 cpufreq.off=1 [CPU_FREQ] 690 disable the cpufreq sub-system 691 692 cpu_init_udelay=N 693 [X86] Delay for N microsec between assert and de-assert 694 of APIC INIT to start processors. This delay occurs 695 on every CPU online, such as boot, and resume from suspend. 696 Default: 10000 697 698 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 699 Format: 700 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 701 702 crashkernel=size[KMG][@offset[KMG]] 703 [KNL] Using kexec, Linux can switch to a 'crash kernel' 704 upon panic. This parameter reserves the physical 705 memory region [offset, offset + size] for that kernel 706 image. If '@offset' is omitted, then a suitable offset 707 is selected automatically. 708 [KNL, x86_64] select a region under 4G first, and 709 fall back to reserve region above 4G when '@offset' 710 hasn't been specified. 711 See Documentation/admin-guide/kdump/kdump.rst for further details. 712 713 crashkernel=range1:size1[,range2:size2,...][@offset] 714 [KNL] Same as above, but depends on the memory 715 in the running system. The syntax of range is 716 start-[end] where start and end are both 717 a memory unit (amount[KMG]). See also 718 Documentation/admin-guide/kdump/kdump.rst for an example. 719 720 crashkernel=size[KMG],high 721 [KNL, x86_64] range could be above 4G. Allow kernel 722 to allocate physical memory region from top, so could 723 be above 4G if system have more than 4G ram installed. 724 Otherwise memory region will be allocated below 4G, if 725 available. 726 It will be ignored if crashkernel=X is specified. 727 crashkernel=size[KMG],low 728 [KNL, x86_64] range under 4G. When crashkernel=X,high 729 is passed, kernel could allocate physical memory region 730 above 4G, that cause second kernel crash on system 731 that require some amount of low memory, e.g. swiotlb 732 requires at least 64M+32K low memory, also enough extra 733 low memory is needed to make sure DMA buffers for 32-bit 734 devices won't run out. Kernel would try to allocate at 735 at least 256M below 4G automatically. 736 This one let user to specify own low range under 4G 737 for second kernel instead. 738 0: to disable low allocation. 739 It will be ignored when crashkernel=X,high is not used 740 or memory reserved is below 4G. 741 742 cryptomgr.notests 743 [KNL] Disable crypto self-tests 744 745 cs89x0_dma= [HW,NET] 746 Format: <dma> 747 748 cs89x0_media= [HW,NET] 749 Format: { rj45 | aui | bnc } 750 751 dasd= [HW,NET] 752 See header of drivers/s390/block/dasd_devmap.c. 753 754 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 755 (one device per port) 756 Format: <port#>,<type> 757 See also Documentation/input/devices/joystick-parport.rst 758 759 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot 760 time. See 761 Documentation/admin-guide/dynamic-debug-howto.rst for 762 details. Deprecated, see dyndbg. 763 764 debug [KNL] Enable kernel debugging (events log level). 765 766 debug_boot_weak_hash 767 [KNL] Enable printing [hashed] pointers early in the 768 boot sequence. If enabled, we use a weak hash instead 769 of siphash to hash pointers. Use this option if you are 770 seeing instances of '(___ptrval___)') and need to see a 771 value (hashed pointer) instead. Cryptographically 772 insecure, please do not use on production kernels. 773 774 debug_locks_verbose= 775 [KNL] verbose self-tests 776 Format=<0|1> 777 Print debugging info while doing the locking API 778 self-tests. 779 We default to 0 (no extra messages), setting it to 780 1 will print _a lot_ more information - normally 781 only useful to kernel developers. 782 783 debug_objects [KNL] Enable object debugging 784 785 no_debug_objects 786 [KNL] Disable object debugging 787 788 debug_guardpage_minorder= 789 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this 790 parameter allows control of the order of pages that will 791 be intentionally kept free (and hence protected) by the 792 buddy allocator. Bigger value increase the probability 793 of catching random memory corruption, but reduce the 794 amount of memory for normal system use. The maximum 795 possible value is MAX_ORDER/2. Setting this parameter 796 to 1 or 2 should be enough to identify most random 797 memory corruption problems caused by bugs in kernel or 798 driver code when a CPU writes to (or reads from) a 799 random memory location. Note that there exists a class 800 of memory corruptions problems caused by buggy H/W or 801 F/W or by drivers badly programing DMA (basically when 802 memory is written at bus level and the CPU MMU is 803 bypassed) which are not detectable by 804 CONFIG_DEBUG_PAGEALLOC, hence this option will not help 805 tracking down these problems. 806 807 debug_pagealloc= 808 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 809 enables the feature at boot time. By default, it is 810 disabled and the system will work mostly the same as a 811 kernel built without CONFIG_DEBUG_PAGEALLOC. 812 Note: to get most of debug_pagealloc error reports, it's 813 useful to also enable the page_owner functionality. 814 on: enable the feature 815 816 debugpat [X86] Enable PAT debugging 817 818 decnet.addr= [HW,NET] 819 Format: <area>[,<node>] 820 See also Documentation/networking/decnet.txt. 821 822 default_hugepagesz= 823 [same as hugepagesz=] The size of the default 824 HugeTLB page size. This is the size represented by 825 the legacy /proc/ hugepages APIs, used for SHM, and 826 default size when mounting hugetlbfs filesystems. 827 Defaults to the default architecture's huge page size 828 if not specified. 829 830 deferred_probe_timeout= 831 [KNL] Debugging option to set a timeout in seconds for 832 deferred probe to give up waiting on dependencies to 833 probe. Only specific dependencies (subsystems or 834 drivers) that have opted in will be ignored. A timeout of 0 835 will timeout at the end of initcalls. This option will also 836 dump out devices still on the deferred probe list after 837 retrying. 838 839 dhash_entries= [KNL] 840 Set number of hash buckets for dentry cache. 841 842 disable_1tb_segments [PPC] 843 Disables the use of 1TB hash page table segments. This 844 causes the kernel to fall back to 256MB segments which 845 can be useful when debugging issues that require an SLB 846 miss to occur. 847 848 disable= [IPV6] 849 See Documentation/networking/ipv6.txt. 850 851 hardened_usercopy= 852 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 853 hardening is enabled for this boot. Hardened 854 usercopy checking is used to protect the kernel 855 from reading or writing beyond known memory 856 allocation boundaries as a proactive defense 857 against bounds-checking flaws in the kernel's 858 copy_to_user()/copy_from_user() interface. 859 on Perform hardened usercopy checks (default). 860 off Disable hardened usercopy checks. 861 862 disable_radix [PPC] 863 Disable RADIX MMU mode on POWER9 864 865 disable_tlbie [PPC] 866 Disable TLBIE instruction. Currently does not work 867 with KVM, with HASH MMU, or with coherent accelerators. 868 869 disable_cpu_apicid= [X86,APIC,SMP] 870 Format: <int> 871 The number of initial APIC ID for the 872 corresponding CPU to be disabled at boot, 873 mostly used for the kdump 2nd kernel to 874 disable BSP to wake up multiple CPUs without 875 causing system reset or hang due to sending 876 INIT from AP to BSP. 877 878 perf_v4_pmi= [X86,INTEL] 879 Format: <bool> 880 Disable Intel PMU counter freezing feature. 881 The feature only exists starting from 882 Arch Perfmon v4 (Skylake and newer). 883 884 disable_ddw [PPC/PSERIES] 885 Disable Dynamic DMA Window support. Use this if 886 to workaround buggy firmware. 887 888 disable_ipv6= [IPV6] 889 See Documentation/networking/ipv6.txt. 890 891 disable_mtrr_cleanup [X86] 892 The kernel tries to adjust MTRR layout from continuous 893 to discrete, to make X server driver able to add WB 894 entry later. This parameter disables that. 895 896 disable_mtrr_trim [X86, Intel and AMD only] 897 By default the kernel will trim any uncacheable 898 memory out of your available memory pool based on 899 MTRR settings. This parameter disables that behavior, 900 possibly causing your machine to run very slowly. 901 902 disable_timer_pin_1 [X86] 903 Disable PIN 1 of APIC timer 904 Can be useful to work around chipset bugs. 905 906 dis_ucode_ldr [X86] Disable the microcode loader. 907 908 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 909 this option disables the debugging code at boot. 910 911 dma_debug_entries=<number> 912 This option allows to tune the number of preallocated 913 entries for DMA-API debugging code. One entry is 914 required per DMA-API allocation. Use this if the 915 DMA-API debugging code disables itself because the 916 architectural default is too low. 917 918 dma_debug_driver=<driver_name> 919 With this option the DMA-API debugging driver 920 filter feature can be enabled at boot time. Just 921 pass the driver to filter for as the parameter. 922 The filter can be disabled or changed to another 923 driver later using sysfs. 924 925 driver_async_probe= [KNL] 926 List of driver names to be probed asynchronously. 927 Format: <driver_name1>,<driver_name2>... 928 929 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 930 Broken monitors, graphic adapters, KVMs and EDIDless 931 panels may send no or incorrect EDID data sets. 932 This parameter allows to specify an EDID data sets 933 in the /lib/firmware directory that are used instead. 934 Generic built-in EDID data sets are used, if one of 935 edid/1024x768.bin, edid/1280x1024.bin, 936 edid/1680x1050.bin, or edid/1920x1080.bin is given 937 and no file with the same name exists. Details and 938 instructions how to build your own EDID data are 939 available in Documentation/driver-api/edid.rst. An EDID 940 data set will only be used for a particular connector, 941 if its name and a colon are prepended to the EDID 942 name. Each connector may use a unique EDID data 943 set by separating the files with a comma. An EDID 944 data set with no connector name will be used for 945 any connectors not explicitly specified. 946 947 dscc4.setup= [NET] 948 949 dt_cpu_ftrs= [PPC] 950 Format: {"off" | "known"} 951 Control how the dt_cpu_ftrs device-tree binding is 952 used for CPU feature discovery and setup (if it 953 exists). 954 off: Do not use it, fall back to legacy cpu table. 955 known: Do not pass through unknown features to guests 956 or userspace, only those that the kernel is aware of. 957 958 dump_apple_properties [X86] 959 Dump name and content of EFI device properties on 960 x86 Macs. Useful for driver authors to determine 961 what data is available or for reverse-engineering. 962 963 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 964 module.dyndbg[="val"] 965 Enable debug messages at boot time. See 966 Documentation/admin-guide/dynamic-debug-howto.rst 967 for details. 968 969 nompx [X86] Disables Intel Memory Protection Extensions. 970 See Documentation/x86/intel_mpx.rst for more 971 information about the feature. 972 973 nopku [X86] Disable Memory Protection Keys CPU feature found 974 in some Intel CPUs. 975 976 module.async_probe [KNL] 977 Enable asynchronous probe on this module. 978 979 early_ioremap_debug [KNL] 980 Enable debug messages in early_ioremap support. This 981 is useful for tracking down temporary early mappings 982 which are not unmapped. 983 984 earlycon= [KNL] Output early console device and options. 985 986 [ARM64] The early console is determined by the 987 stdout-path property in device tree's chosen node, 988 or determined by the ACPI SPCR table. 989 990 [X86] When used with no options the early console is 991 determined by the ACPI SPCR table. 992 993 cdns,<addr>[,options] 994 Start an early, polled-mode console on a Cadence 995 (xuartps) serial port at the specified address. Only 996 supported option is baud rate. If baud rate is not 997 specified, the serial port must already be setup and 998 configured. 999 1000 uart[8250],io,<addr>[,options] 1001 uart[8250],mmio,<addr>[,options] 1002 uart[8250],mmio32,<addr>[,options] 1003 uart[8250],mmio32be,<addr>[,options] 1004 uart[8250],0x<addr>[,options] 1005 Start an early, polled-mode console on the 8250/16550 1006 UART at the specified I/O port or MMIO address. 1007 MMIO inter-register address stride is either 8-bit 1008 (mmio) or 32-bit (mmio32 or mmio32be). 1009 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1010 to be equivalent to 'mmio'. 'options' are specified 1011 in the same format described for "console=ttyS<n>"; if 1012 unspecified, the h/w is not initialized. 1013 1014 pl011,<addr> 1015 pl011,mmio32,<addr> 1016 Start an early, polled-mode console on a pl011 serial 1017 port at the specified address. The pl011 serial port 1018 must already be setup and configured. Options are not 1019 yet supported. If 'mmio32' is specified, then only 1020 the driver will use only 32-bit accessors to read/write 1021 the device registers. 1022 1023 meson,<addr> 1024 Start an early, polled-mode console on a meson serial 1025 port at the specified address. The serial port must 1026 already be setup and configured. Options are not yet 1027 supported. 1028 1029 msm_serial,<addr> 1030 Start an early, polled-mode console on an msm serial 1031 port at the specified address. The serial port 1032 must already be setup and configured. Options are not 1033 yet supported. 1034 1035 msm_serial_dm,<addr> 1036 Start an early, polled-mode console on an msm serial 1037 dm port at the specified address. The serial port 1038 must already be setup and configured. Options are not 1039 yet supported. 1040 1041 owl,<addr> 1042 Start an early, polled-mode console on a serial port 1043 of an Actions Semi SoC, such as S500 or S900, at the 1044 specified address. The serial port must already be 1045 setup and configured. Options are not yet supported. 1046 1047 rda,<addr> 1048 Start an early, polled-mode console on a serial port 1049 of an RDA Micro SoC, such as RDA8810PL, at the 1050 specified address. The serial port must already be 1051 setup and configured. Options are not yet supported. 1052 1053 sbi 1054 Use RISC-V SBI (Supervisor Binary Interface) for early 1055 console. 1056 1057 smh Use ARM semihosting calls for early console. 1058 1059 s3c2410,<addr> 1060 s3c2412,<addr> 1061 s3c2440,<addr> 1062 s3c6400,<addr> 1063 s5pv210,<addr> 1064 exynos4210,<addr> 1065 Use early console provided by serial driver available 1066 on Samsung SoCs, requires selecting proper type and 1067 a correct base address of the selected UART port. The 1068 serial port must already be setup and configured. 1069 Options are not yet supported. 1070 1071 lantiq,<addr> 1072 Start an early, polled-mode console on a lantiq serial 1073 (lqasc) port at the specified address. The serial port 1074 must already be setup and configured. Options are not 1075 yet supported. 1076 1077 lpuart,<addr> 1078 lpuart32,<addr> 1079 Use early console provided by Freescale LP UART driver 1080 found on Freescale Vybrid and QorIQ LS1021A processors. 1081 A valid base address must be provided, and the serial 1082 port must already be setup and configured. 1083 1084 ar3700_uart,<addr> 1085 Start an early, polled-mode console on the 1086 Armada 3700 serial port at the specified 1087 address. The serial port must already be setup 1088 and configured. Options are not yet supported. 1089 1090 qcom_geni,<addr> 1091 Start an early, polled-mode console on a Qualcomm 1092 Generic Interface (GENI) based serial port at the 1093 specified address. The serial port must already be 1094 setup and configured. Options are not yet supported. 1095 1096 efifb,[options] 1097 Start an early, unaccelerated console on the EFI 1098 memory mapped framebuffer (if available). On cache 1099 coherent non-x86 systems that use system memory for 1100 the framebuffer, pass the 'ram' option so that it is 1101 mapped with the correct attributes. 1102 1103 linflex,<addr> 1104 Use early console provided by Freescale LinFlex UART 1105 serial driver for NXP S32V234 SoCs. A valid base 1106 address must be provided, and the serial port must 1107 already be setup and configured. 1108 1109 earlyprintk= [X86,SH,ARM,M68k,S390] 1110 earlyprintk=vga 1111 earlyprintk=sclp 1112 earlyprintk=xen 1113 earlyprintk=serial[,ttySn[,baudrate]] 1114 earlyprintk=serial[,0x...[,baudrate]] 1115 earlyprintk=ttySn[,baudrate] 1116 earlyprintk=dbgp[debugController#] 1117 earlyprintk=pciserial[,force],bus:device.function[,baudrate] 1118 earlyprintk=xdbc[xhciController#] 1119 1120 earlyprintk is useful when the kernel crashes before 1121 the normal console is initialized. It is not enabled by 1122 default because it has some cosmetic problems. 1123 1124 Append ",keep" to not disable it when the real console 1125 takes over. 1126 1127 Only one of vga, efi, serial, or usb debug port can 1128 be used at a time. 1129 1130 Currently only ttyS0 and ttyS1 may be specified by 1131 name. Other I/O ports may be explicitly specified 1132 on some architectures (x86 and arm at least) by 1133 replacing ttySn with an I/O port address, like this: 1134 earlyprintk=serial,0x1008,115200 1135 You can find the port for a given device in 1136 /proc/tty/driver/serial: 1137 2: uart:ST16650V2 port:00001008 irq:18 ... 1138 1139 Interaction with the standard serial driver is not 1140 very good. 1141 1142 The VGA and EFI output is eventually overwritten by 1143 the real console. 1144 1145 The xen output can only be used by Xen PV guests. 1146 1147 The sclp output can only be used on s390. 1148 1149 The optional "force" to "pciserial" enables use of a 1150 PCI device even when its classcode is not of the 1151 UART class. 1152 1153 edac_report= [HW,EDAC] Control how to report EDAC event 1154 Format: {"on" | "off" | "force"} 1155 on: enable EDAC to report H/W event. May be overridden 1156 by other higher priority error reporting module. 1157 off: disable H/W event reporting through EDAC. 1158 force: enforce the use of EDAC to report H/W event. 1159 default: on. 1160 1161 ekgdboc= [X86,KGDB] Allow early kernel console debugging 1162 ekgdboc=kbd 1163 1164 This is designed to be used in conjunction with 1165 the boot argument: earlyprintk=vga 1166 1167 edd= [EDD] 1168 Format: {"off" | "on" | "skip[mbr]"} 1169 1170 efi= [EFI] 1171 Format: { "old_map", "nochunk", "noruntime", "debug", 1172 "nosoftreserve" } 1173 old_map [X86-64]: switch to the old ioremap-based EFI 1174 runtime services mapping. 32-bit still uses this one by 1175 default. 1176 nochunk: disable reading files in "chunks" in the EFI 1177 boot stub, as chunking can cause problems with some 1178 firmware implementations. 1179 noruntime : disable EFI runtime services support 1180 debug: enable misc debug output 1181 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1182 attribute may cause the kernel to reserve the 1183 memory range for a memory mapping driver to 1184 claim. Specify efi=nosoftreserve to disable this 1185 reservation and treat the memory by its base type 1186 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1187 1188 efi_no_storage_paranoia [EFI; X86] 1189 Using this parameter you can use more than 50% of 1190 your efi variable storage. Use this parameter only if 1191 you are really sure that your UEFI does sane gc and 1192 fulfills the spec otherwise your board may brick. 1193 1194 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86] 1195 Add arbitrary attribute to specific memory range by 1196 updating original EFI memory map. 1197 Region of memory which aa attribute is added to is 1198 from ss to ss+nn. 1199 1200 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000 1201 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000) 1202 attribute is added to range 0x100000000-0x180000000 and 1203 0x10a0000000-0x1120000000. 1204 1205 If efi_fake_mem=8G@9G:0x40000 is specified, the 1206 EFI_MEMORY_SP(0x40000) attribute is added to 1207 range 0x240000000-0x43fffffff. 1208 1209 Using this parameter you can do debugging of EFI memmap 1210 related features. For example, you can do debugging of 1211 Address Range Mirroring feature even if your box 1212 doesn't support it, or mark specific memory as 1213 "soft reserved". 1214 1215 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1216 that is to be dynamically loaded by Linux. If there are 1217 multiple variables with the same name but with different 1218 vendor GUIDs, all of them will be loaded. See 1219 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1220 1221 1222 eisa_irq_edge= [PARISC,HW] 1223 See header of drivers/parisc/eisa.c. 1224 1225 elanfreq= [X86-32] 1226 See comment before function elanfreq_setup() in 1227 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1228 1229 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390] 1230 Specifies physical address of start of kernel core 1231 image elf header and optionally the size. Generally 1232 kexec loader will pass this option to capture kernel. 1233 See Documentation/admin-guide/kdump/kdump.rst for details. 1234 1235 enable_mtrr_cleanup [X86] 1236 The kernel tries to adjust MTRR layout from continuous 1237 to discrete, to make X server driver able to add WB 1238 entry later. This parameter enables that. 1239 1240 enable_timer_pin_1 [X86] 1241 Enable PIN 1 of APIC timer 1242 Can be useful to work around chipset bugs 1243 (in particular on some ATI chipsets). 1244 The kernel tries to set a reasonable default. 1245 1246 enforcing [SELINUX] Set initial enforcing status. 1247 Format: {"0" | "1"} 1248 See security/selinux/Kconfig help text. 1249 0 -- permissive (log only, no denials). 1250 1 -- enforcing (deny and log). 1251 Default value is 0. 1252 Value can be changed at runtime via /selinux/enforce. 1253 1254 erst_disable [ACPI] 1255 Disable Error Record Serialization Table (ERST) 1256 support. 1257 1258 ether= [HW,NET] Ethernet cards parameters 1259 This option is obsoleted by the "netdev=" option, which 1260 has equivalent usage. See its documentation for details. 1261 1262 evm= [EVM] 1263 Format: { "fix" } 1264 Permit 'security.evm' to be updated regardless of 1265 current integrity status. 1266 1267 failslab= 1268 fail_page_alloc= 1269 fail_make_request=[KNL] 1270 General fault injection mechanism. 1271 Format: <interval>,<probability>,<space>,<times> 1272 See also Documentation/fault-injection/. 1273 1274 floppy= [HW] 1275 See Documentation/admin-guide/blockdev/floppy.rst. 1276 1277 force_pal_cache_flush 1278 [IA-64] Avoid check_sal_cache_flush which may hang on 1279 buggy SAL_CACHE_FLUSH implementations. Using this 1280 parameter will force ia64_sal_cache_flush to call 1281 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH. 1282 1283 forcepae [X86-32] 1284 Forcefully enable Physical Address Extension (PAE). 1285 Many Pentium M systems disable PAE but may have a 1286 functionally usable PAE implementation. 1287 Warning: use of this parameter will taint the kernel 1288 and may cause unknown problems. 1289 1290 ftrace=[tracer] 1291 [FTRACE] will set and start the specified tracer 1292 as early as possible in order to facilitate early 1293 boot debugging. 1294 1295 ftrace_dump_on_oops[=orig_cpu] 1296 [FTRACE] will dump the trace buffers on oops. 1297 If no parameter is passed, ftrace will dump 1298 buffers of all CPUs, but if you pass orig_cpu, it will 1299 dump only the buffer of the CPU that triggered the 1300 oops. 1301 1302 ftrace_filter=[function-list] 1303 [FTRACE] Limit the functions traced by the function 1304 tracer at boot up. function-list is a comma separated 1305 list of functions. This list can be changed at run 1306 time by the set_ftrace_filter file in the debugfs 1307 tracing directory. 1308 1309 ftrace_notrace=[function-list] 1310 [FTRACE] Do not trace the functions specified in 1311 function-list. This list can be changed at run time 1312 by the set_ftrace_notrace file in the debugfs 1313 tracing directory. 1314 1315 ftrace_graph_filter=[function-list] 1316 [FTRACE] Limit the top level callers functions traced 1317 by the function graph tracer at boot up. 1318 function-list is a comma separated list of functions 1319 that can be changed at run time by the 1320 set_graph_function file in the debugfs tracing directory. 1321 1322 ftrace_graph_notrace=[function-list] 1323 [FTRACE] Do not trace from the functions specified in 1324 function-list. This list is a comma separated list of 1325 functions that can be changed at run time by the 1326 set_graph_notrace file in the debugfs tracing directory. 1327 1328 ftrace_graph_max_depth=<uint> 1329 [FTRACE] Used with the function graph tracer. This is 1330 the max depth it will trace into a function. This value 1331 can be changed at run time by the max_graph_depth file 1332 in the tracefs tracing directory. default: 0 (no limit) 1333 1334 gamecon.map[2|3]= 1335 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1336 support via parallel port (up to 5 devices per port) 1337 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1338 See also Documentation/input/devices/joystick-parport.rst 1339 1340 gamma= [HW,DRM] 1341 1342 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART 1343 Format: off | on 1344 default: on 1345 1346 gcov_persist= [GCOV] When non-zero (default), profiling data for 1347 kernel modules is saved and remains accessible via 1348 debugfs, even when the module is unloaded/reloaded. 1349 When zero, profiling data is discarded and associated 1350 debugfs files are removed at module unload time. 1351 1352 goldfish [X86] Enable the goldfish android emulator platform. 1353 Don't use this when you are not running on the 1354 android emulator 1355 1356 gpt [EFI] Forces disk with valid GPT signature but 1357 invalid Protective MBR to be treated as GPT. If the 1358 primary GPT is corrupted, it enables the backup/alternate 1359 GPT to be used instead. 1360 1361 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1362 the "Enable 0" bit of the configuration register. 1363 Format: 0 | 1 1364 Default: 0 1365 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1366 the "Enable 0" bit of the configuration register. 1367 Format: 0 | 1 1368 Default: 0 1369 grcan.select= [HW] Select which physical interface to use. 1370 Format: 0 | 1 1371 Default: 0 1372 grcan.txsize= [HW] Sets the size of the tx buffer. 1373 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1374 Default: 1024 1375 grcan.rxsize= [HW] Sets the size of the rx buffer. 1376 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1377 Default: 1024 1378 1379 gpio-mockup.gpio_mockup_ranges 1380 [HW] Sets the ranges of gpiochip of for this device. 1381 Format: <start1>,<end1>,<start2>,<end2>... 1382 1383 hardlockup_all_cpu_backtrace= 1384 [KNL] Should the hard-lockup detector generate 1385 backtraces on all cpus. 1386 Format: <integer> 1387 1388 hashdist= [KNL,NUMA] Large hashes allocated during boot 1389 are distributed across NUMA nodes. Defaults on 1390 for 64-bit NUMA, off otherwise. 1391 Format: 0 | 1 (for off | on) 1392 1393 hcl= [IA-64] SGI's Hardware Graph compatibility layer 1394 1395 hd= [EIDE] (E)IDE hard drive subsystem geometry 1396 Format: <cyl>,<head>,<sect> 1397 1398 hest_disable [ACPI] 1399 Disable Hardware Error Source Table (HEST) support; 1400 corresponding firmware-first mode error processing 1401 logic will be disabled. 1402 1403 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact 1404 size of <nn>. This works even on boxes that have no 1405 highmem otherwise. This also works to reduce highmem 1406 size on bigger boxes. 1407 1408 highres= [KNL] Enable/disable high resolution timer mode. 1409 Valid parameters: "on", "off" 1410 Default: "on" 1411 1412 hlt [BUGS=ARM,SH] 1413 1414 hpet= [X86-32,HPET] option to control HPET usage 1415 Format: { enable (default) | disable | force | 1416 verbose } 1417 disable: disable HPET and use PIT instead 1418 force: allow force enabled of undocumented chips (ICH4, 1419 VIA, nVidia) 1420 verbose: show contents of HPET registers during setup 1421 1422 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 1423 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 1424 1425 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot. 1426 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages. 1427 On x86-64 and powerpc, this option can be specified 1428 multiple times interleaved with hugepages= to reserve 1429 huge pages of different sizes. Valid pages sizes on 1430 x86-64 are 2M (when the CPU supports "pse") and 1G 1431 (when the CPU supports the "pdpe1gb" cpuinfo flag). 1432 1433 hung_task_panic= 1434 [KNL] Should the hung task detector generate panics. 1435 Format: <integer> 1436 1437 A nonzero value instructs the kernel to panic when a 1438 hung task is detected. The default value is controlled 1439 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time 1440 option. The value selected by this boot parameter can 1441 be changed later by the kernel.hung_task_panic sysctl. 1442 1443 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 1444 terminal devices. Valid values: 0..8 1445 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 1446 If specified, z/VM IUCV HVC accepts connections 1447 from listed z/VM user IDs only. 1448 1449 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations 1450 which allow the hypervisor to 'idle' the 1451 guest on lock contention. 1452 1453 keep_bootcon [KNL] 1454 Do not unregister boot console at start. This is only 1455 useful for debugging when something happens in the window 1456 between unregistering the boot console and initializing 1457 the real console. 1458 1459 i2c_bus= [HW] Override the default board specific I2C bus speed 1460 or register an additional I2C bus that is not 1461 registered from board initialization code. 1462 Format: 1463 <bus_id>,<clkrate> 1464 1465 i8042.debug [HW] Toggle i8042 debug mode 1466 i8042.unmask_kbd_data 1467 [HW] Enable printing of interrupt data from the KBD port 1468 (disabled by default, and as a pre-condition 1469 requires that i8042.debug=1 be enabled) 1470 i8042.direct [HW] Put keyboard port into non-translated mode 1471 i8042.dumbkbd [HW] Pretend that controller can only read data from 1472 keyboard and cannot control its state 1473 (Don't attempt to blink the leds) 1474 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 1475 i8042.nokbd [HW] Don't check/create keyboard port 1476 i8042.noloop [HW] Disable the AUX Loopback command while probing 1477 for the AUX port 1478 i8042.nomux [HW] Don't check presence of an active multiplexing 1479 controller 1480 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 1481 controllers 1482 i8042.notimeout [HW] Ignore timeout condition signalled by controller 1483 i8042.reset [HW] Reset the controller during init, cleanup and 1484 suspend-to-ram transitions, only during s2r 1485 transitions, or never reset 1486 Format: { 1 | Y | y | 0 | N | n } 1487 1, Y, y: always reset controller 1488 0, N, n: don't ever reset controller 1489 Default: only on s2r transitions on x86; most other 1490 architectures force reset to be always executed 1491 i8042.unlock [HW] Unlock (ignore) the keylock 1492 i8042.kbdreset [HW] Reset device connected to KBD port 1493 1494 i810= [HW,DRM] 1495 1496 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data 1497 indicates that the driver is running on unsupported 1498 hardware. 1499 i8k.force [HW] Activate i8k driver even if SMM BIOS signature 1500 does not match list of supported models. 1501 i8k.power_status 1502 [HW] Report power status in /proc/i8k 1503 (disabled by default) 1504 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN 1505 capability is set. 1506 1507 i915.invert_brightness= 1508 [DRM] Invert the sense of the variable that is used to 1509 set the brightness of the panel backlight. Normally a 1510 brightness value of 0 indicates backlight switched off, 1511 and the maximum of the brightness value sets the backlight 1512 to maximum brightness. If this parameter is set to 0 1513 (default) and the machine requires it, or this parameter 1514 is set to 1, a brightness value of 0 sets the backlight 1515 to maximum brightness, and the maximum of the brightness 1516 value switches the backlight off. 1517 -1 -- never invert brightness 1518 0 -- machine default 1519 1 -- force brightness inversion 1520 1521 icn= [HW,ISDN] 1522 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 1523 1524 ide-core.nodma= [HW] (E)IDE subsystem 1525 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc 1526 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr 1527 .cdrom .chs .ignore_cable are additional options 1528 See Documentation/ide/ide.rst. 1529 1530 ide-generic.probe-mask= [HW] (E)IDE subsystem 1531 Format: <int> 1532 Probe mask for legacy ISA IDE ports. Depending on 1533 platform up to 6 ports are supported, enabled by 1534 setting corresponding bits in the mask to 1. The 1535 default value is 0x0, which has a special meaning. 1536 On systems that have PCI, it triggers scanning the 1537 PCI bus for the first and the second port, which 1538 are then probed. On systems without PCI the value 1539 of 0x0 enables probing the two first ports as if it 1540 was 0x3. 1541 1542 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem 1543 Claim all unknown PCI IDE storage controllers. 1544 1545 idle= [X86] 1546 Format: idle=poll, idle=halt, idle=nomwait 1547 Poll forces a polling idle loop that can slightly 1548 improve the performance of waking up a idle CPU, but 1549 will use a lot of power and make the system run hot. 1550 Not recommended. 1551 idle=halt: Halt is forced to be used for CPU idle. 1552 In such case C2/C3 won't be used again. 1553 idle=nomwait: Disable mwait for CPU C-states 1554 1555 ieee754= [MIPS] Select IEEE Std 754 conformance mode 1556 Format: { strict | legacy | 2008 | relaxed } 1557 Default: strict 1558 1559 Choose which programs will be accepted for execution 1560 based on the IEEE 754 NaN encoding(s) supported by 1561 the FPU and the NaN encoding requested with the value 1562 of an ELF file header flag individually set by each 1563 binary. Hardware implementations are permitted to 1564 support either or both of the legacy and the 2008 NaN 1565 encoding mode. 1566 1567 Available settings are as follows: 1568 strict accept binaries that request a NaN encoding 1569 supported by the FPU 1570 legacy only accept legacy-NaN binaries, if supported 1571 by the FPU 1572 2008 only accept 2008-NaN binaries, if supported 1573 by the FPU 1574 relaxed accept any binaries regardless of whether 1575 supported by the FPU 1576 1577 The FPU emulator is always able to support both NaN 1578 encodings, so if no FPU hardware is present or it has 1579 been disabled with 'nofpu', then the settings of 1580 'legacy' and '2008' strap the emulator accordingly, 1581 'relaxed' straps the emulator for both legacy-NaN and 1582 2008-NaN, whereas 'strict' enables legacy-NaN only on 1583 legacy processors and both NaN encodings on MIPS32 or 1584 MIPS64 CPUs. 1585 1586 The setting for ABS.fmt/NEG.fmt instruction execution 1587 mode generally follows that for the NaN encoding, 1588 except where unsupported by hardware. 1589 1590 ignore_loglevel [KNL] 1591 Ignore loglevel setting - this will print /all/ 1592 kernel messages to the console. Useful for debugging. 1593 We also add it as printk module parameter, so users 1594 could change it dynamically, usually by 1595 /sys/module/printk/parameters/ignore_loglevel. 1596 1597 ignore_rlimit_data 1598 Ignore RLIMIT_DATA setting for data mappings, 1599 print warning at first misuse. Can be changed via 1600 /sys/module/kernel/parameters/ignore_rlimit_data. 1601 1602 ihash_entries= [KNL] 1603 Set number of hash buckets for inode cache. 1604 1605 ima_appraise= [IMA] appraise integrity measurements 1606 Format: { "off" | "enforce" | "fix" | "log" } 1607 default: "enforce" 1608 1609 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 1610 The builtin appraise policy appraises all files 1611 owned by uid=0. 1612 1613 ima_canonical_fmt [IMA] 1614 Use the canonical format for the binary runtime 1615 measurements, instead of host native format. 1616 1617 ima_hash= [IMA] 1618 Format: { md5 | sha1 | rmd160 | sha256 | sha384 1619 | sha512 | ... } 1620 default: "sha1" 1621 1622 The list of supported hash algorithms is defined 1623 in crypto/hash_info.h. 1624 1625 ima_policy= [IMA] 1626 The builtin policies to load during IMA setup. 1627 Format: "tcb | appraise_tcb | secure_boot | 1628 fail_securely" 1629 1630 The "tcb" policy measures all programs exec'd, files 1631 mmap'd for exec, and all files opened with the read 1632 mode bit set by either the effective uid (euid=0) or 1633 uid=0. 1634 1635 The "appraise_tcb" policy appraises the integrity of 1636 all files owned by root. 1637 1638 The "secure_boot" policy appraises the integrity 1639 of files (eg. kexec kernel image, kernel modules, 1640 firmware, policy, etc) based on file signatures. 1641 1642 The "fail_securely" policy forces file signature 1643 verification failure also on privileged mounted 1644 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 1645 flag. 1646 1647 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 1648 Load a policy which meets the needs of the Trusted 1649 Computing Base. This means IMA will measure all 1650 programs exec'd, files mmap'd for exec, and all files 1651 opened for read by uid=0. 1652 1653 ima_template= [IMA] 1654 Select one of defined IMA measurements template formats. 1655 Formats: { "ima" | "ima-ng" | "ima-sig" } 1656 Default: "ima-ng" 1657 1658 ima_template_fmt= 1659 [IMA] Define a custom template format. 1660 Format: { "field1|...|fieldN" } 1661 1662 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 1663 Format: <min_file_size> 1664 Set the minimal file size for using asynchronous hash. 1665 If left unspecified, ahash usage is disabled. 1666 1667 ahash performance varies for different data sizes on 1668 different crypto accelerators. This option can be used 1669 to achieve the best performance for a particular HW. 1670 1671 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size 1672 Format: <bufsize> 1673 Set hashing buffer size. Default: 4k. 1674 1675 ahash performance varies for different chunk sizes on 1676 different crypto accelerators. This option can be used 1677 to achieve best performance for particular HW. 1678 1679 init= [KNL] 1680 Format: <full_path> 1681 Run specified binary instead of /sbin/init as init 1682 process. 1683 1684 initcall_debug [KNL] Trace initcalls as they are executed. Useful 1685 for working out where the kernel is dying during 1686 startup. 1687 1688 initcall_blacklist= [KNL] Do not execute a comma-separated list of 1689 initcall functions. Useful for debugging built-in 1690 modules and initcalls. 1691 1692 initrd= [BOOT] Specify the location of the initial ramdisk 1693 1694 init_on_alloc= [MM] Fill newly allocated pages and heap objects with 1695 zeroes. 1696 Format: 0 | 1 1697 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 1698 1699 init_on_free= [MM] Fill freed pages and heap objects with zeroes. 1700 Format: 0 | 1 1701 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 1702 1703 init_pkru= [x86] Specify the default memory protection keys rights 1704 register contents for all processes. 0x55555554 by 1705 default (disallow access to all but pkey 0). Can 1706 override in debugfs after boot. 1707 1708 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 1709 Format: <irq> 1710 1711 int_pln_enable [x86] Enable power limit notification interrupt 1712 1713 integrity_audit=[IMA] 1714 Format: { "0" | "1" } 1715 0 -- basic integrity auditing messages. (Default) 1716 1 -- additional integrity auditing messages. 1717 1718 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 1719 on 1720 Enable intel iommu driver. 1721 off 1722 Disable intel iommu driver. 1723 igfx_off [Default Off] 1724 By default, gfx is mapped as normal device. If a gfx 1725 device has a dedicated DMAR unit, the DMAR unit is 1726 bypassed by not enabling DMAR with this option. In 1727 this case, gfx device will use physical address for 1728 DMA. 1729 forcedac [x86_64] 1730 With this option iommu will not optimize to look 1731 for io virtual address below 32-bit forcing dual 1732 address cycle on pci bus for cards supporting greater 1733 than 32-bit addressing. The default is to look 1734 for translation below 32-bit and if not available 1735 then look in the higher range. 1736 strict [Default Off] 1737 With this option on every unmap_single operation will 1738 result in a hardware IOTLB flush operation as opposed 1739 to batching them for performance. 1740 sp_off [Default Off] 1741 By default, super page will be supported if Intel IOMMU 1742 has the capability. With this option, super page will 1743 not be supported. 1744 sm_on [Default Off] 1745 By default, scalable mode will be disabled even if the 1746 hardware advertises that it has support for the scalable 1747 mode translation. With this option set, scalable mode 1748 will be used on hardware which claims to support it. 1749 tboot_noforce [Default Off] 1750 Do not force the Intel IOMMU enabled under tboot. 1751 By default, tboot will force Intel IOMMU on, which 1752 could harm performance of some high-throughput 1753 devices like 40GBit network cards, even if identity 1754 mapping is enabled. 1755 Note that using this option lowers the security 1756 provided by tboot because it makes the system 1757 vulnerable to DMA attacks. 1758 nobounce [Default off] 1759 Disable bounce buffer for unstrusted devices such as 1760 the Thunderbolt devices. This will treat the untrusted 1761 devices as the trusted ones, hence might expose security 1762 risks of DMA attacks. 1763 1764 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 1765 0 disables intel_idle and fall back on acpi_idle. 1766 1 to 9 specify maximum depth of C-state. 1767 1768 intel_pstate= [X86] 1769 disable 1770 Do not enable intel_pstate as the default 1771 scaling driver for the supported processors 1772 passive 1773 Use intel_pstate as a scaling driver, but configure it 1774 to work with generic cpufreq governors (instead of 1775 enabling its internal governor). This mode cannot be 1776 used along with the hardware-managed P-states (HWP) 1777 feature. 1778 force 1779 Enable intel_pstate on systems that prohibit it by default 1780 in favor of acpi-cpufreq. Forcing the intel_pstate driver 1781 instead of acpi-cpufreq may disable platform features, such 1782 as thermal controls and power capping, that rely on ACPI 1783 P-States information being indicated to OSPM and therefore 1784 should be used with caution. This option does not work with 1785 processors that aren't supported by the intel_pstate driver 1786 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 1787 no_hwp 1788 Do not enable hardware P state control (HWP) 1789 if available. 1790 hwp_only 1791 Only load intel_pstate on systems which support 1792 hardware P state control (HWP) if available. 1793 support_acpi_ppc 1794 Enforce ACPI _PPC performance limits. If the Fixed ACPI 1795 Description Table, specifies preferred power management 1796 profile as "Enterprise Server" or "Performance Server", 1797 then this feature is turned on by default. 1798 per_cpu_perf_limits 1799 Allow per-logical-CPU P-State performance control limits using 1800 cpufreq sysfs interface 1801 1802 intremap= [X86-64, Intel-IOMMU] 1803 on enable Interrupt Remapping (default) 1804 off disable Interrupt Remapping 1805 nosid disable Source ID checking 1806 no_x2apic_optout 1807 BIOS x2APIC opt-out request will be ignored 1808 nopost disable Interrupt Posting 1809 1810 iomem= Disable strict checking of access to MMIO memory 1811 strict regions from userspace. 1812 relaxed 1813 1814 iommu= [x86] 1815 off 1816 force 1817 noforce 1818 biomerge 1819 panic 1820 nopanic 1821 merge 1822 nomerge 1823 soft 1824 pt [x86] 1825 nopt [x86] 1826 nobypass [PPC/POWERNV] 1827 Disable IOMMU bypass, using IOMMU for PCI devices. 1828 1829 iommu.strict= [ARM64] Configure TLB invalidation behaviour 1830 Format: { "0" | "1" } 1831 0 - Lazy mode. 1832 Request that DMA unmap operations use deferred 1833 invalidation of hardware TLBs, for increased 1834 throughput at the cost of reduced device isolation. 1835 Will fall back to strict mode if not supported by 1836 the relevant IOMMU driver. 1837 1 - Strict mode (default). 1838 DMA unmap operations invalidate IOMMU hardware TLBs 1839 synchronously. 1840 1841 iommu.passthrough= 1842 [ARM64, X86] Configure DMA to bypass the IOMMU by default. 1843 Format: { "0" | "1" } 1844 0 - Use IOMMU translation for DMA. 1845 1 - Bypass the IOMMU for DMA. 1846 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 1847 1848 io7= [HW] IO7 for Marvel based alpha systems 1849 See comment before marvel_specify_io7 in 1850 arch/alpha/kernel/core_marvel.c. 1851 1852 io_delay= [X86] I/O delay method 1853 0x80 1854 Standard port 0x80 based delay 1855 0xed 1856 Alternate port 0xed based delay (needed on some systems) 1857 udelay 1858 Simple two microseconds delay 1859 none 1860 No delay 1861 1862 ip= [IP_PNP] 1863 See Documentation/filesystems/nfs/nfsroot.txt. 1864 1865 ipcmni_extend [KNL] Extend the maximum number of unique System V 1866 IPC identifiers from 32,768 to 16,777,216. 1867 1868 irqaffinity= [SMP] Set the default irq affinity mask 1869 The argument is a cpu list, as described above. 1870 1871 irqchip.gicv2_force_probe= 1872 [ARM, ARM64] 1873 Format: <bool> 1874 Force the kernel to look for the second 4kB page 1875 of a GICv2 controller even if the memory range 1876 exposed by the device tree is too small. 1877 1878 irqchip.gicv3_nolpi= 1879 [ARM, ARM64] 1880 Force the kernel to ignore the availability of 1881 LPIs (and by consequence ITSs). Intended for system 1882 that use the kernel as a bootloader, and thus want 1883 to let secondary kernels in charge of setting up 1884 LPIs. 1885 1886 irqchip.gicv3_pseudo_nmi= [ARM64] 1887 Enables support for pseudo-NMIs in the kernel. This 1888 requires the kernel to be built with 1889 CONFIG_ARM64_PSEUDO_NMI. 1890 1891 irqfixup [HW] 1892 When an interrupt is not handled search all handlers 1893 for it. Intended to get systems with badly broken 1894 firmware running. 1895 1896 irqpoll [HW] 1897 When an interrupt is not handled search all handlers 1898 for it. Also check all handlers each timer 1899 interrupt. Intended to get systems with badly broken 1900 firmware running. 1901 1902 isapnp= [ISAPNP] 1903 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 1904 1905 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 1906 [Deprecated - use cpusets instead] 1907 Format: [flag-list,]<cpu-list> 1908 1909 Specify one or more CPUs to isolate from disturbances 1910 specified in the flag list (default: domain): 1911 1912 nohz 1913 Disable the tick when a single task runs. 1914 1915 A residual 1Hz tick is offloaded to workqueues, which you 1916 need to affine to housekeeping through the global 1917 workqueue's affinity configured via the 1918 /sys/devices/virtual/workqueue/cpumask sysfs file, or 1919 by using the 'domain' flag described below. 1920 1921 NOTE: by default the global workqueue runs on all CPUs, 1922 so to protect individual CPUs the 'cpumask' file has to 1923 be configured manually after bootup. 1924 1925 domain 1926 Isolate from the general SMP balancing and scheduling 1927 algorithms. Note that performing domain isolation this way 1928 is irreversible: it's not possible to bring back a CPU to 1929 the domains once isolated through isolcpus. It's strongly 1930 advised to use cpusets instead to disable scheduler load 1931 balancing through the "cpuset.sched_load_balance" file. 1932 It offers a much more flexible interface where CPUs can 1933 move in and out of an isolated set anytime. 1934 1935 You can move a process onto or off an "isolated" CPU via 1936 the CPU affinity syscalls or cpuset. 1937 <cpu number> begins at 0 and the maximum value is 1938 "number of CPUs in system - 1". 1939 1940 The format of <cpu-list> is described above. 1941 1942 1943 1944 iucv= [HW,NET] 1945 1946 ivrs_ioapic [HW,X86_64] 1947 Provide an override to the IOAPIC-ID<->DEVICE-ID 1948 mapping provided in the IVRS ACPI table. For 1949 example, to map IOAPIC-ID decimal 10 to 1950 PCI device 00:14.0 write the parameter as: 1951 ivrs_ioapic[10]=00:14.0 1952 1953 ivrs_hpet [HW,X86_64] 1954 Provide an override to the HPET-ID<->DEVICE-ID 1955 mapping provided in the IVRS ACPI table. For 1956 example, to map HPET-ID decimal 0 to 1957 PCI device 00:14.0 write the parameter as: 1958 ivrs_hpet[0]=00:14.0 1959 1960 ivrs_acpihid [HW,X86_64] 1961 Provide an override to the ACPI-HID:UID<->DEVICE-ID 1962 mapping provided in the IVRS ACPI table. For 1963 example, to map UART-HID:UID AMD0020:0 to 1964 PCI device 00:14.5 write the parameter as: 1965 ivrs_acpihid[00:14.5]=AMD0020:0 1966 1967 js= [HW,JOY] Analog joystick 1968 See Documentation/input/joydev/joystick.rst. 1969 1970 nokaslr [KNL] 1971 When CONFIG_RANDOMIZE_BASE is set, this disables 1972 kernel and module base offset ASLR (Address Space 1973 Layout Randomization). 1974 1975 kasan_multi_shot 1976 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 1977 report on every invalid memory access. Without this 1978 parameter KASAN will print report only for the first 1979 invalid access. 1980 1981 keepinitrd [HW,ARM] 1982 1983 kernelcore= [KNL,X86,IA-64,PPC] 1984 Format: nn[KMGTPE] | nn% | "mirror" 1985 This parameter specifies the amount of memory usable by 1986 the kernel for non-movable allocations. The requested 1987 amount is spread evenly throughout all nodes in the 1988 system as ZONE_NORMAL. The remaining memory is used for 1989 movable memory in its own zone, ZONE_MOVABLE. In the 1990 event, a node is too small to have both ZONE_NORMAL and 1991 ZONE_MOVABLE, kernelcore memory will take priority and 1992 other nodes will have a larger ZONE_MOVABLE. 1993 1994 ZONE_MOVABLE is used for the allocation of pages that 1995 may be reclaimed or moved by the page migration 1996 subsystem. Note that allocations like PTEs-from-HighMem 1997 still use the HighMem zone if it exists, and the Normal 1998 zone if it does not. 1999 2000 It is possible to specify the exact amount of memory in 2001 the form of "nn[KMGTPE]", a percentage of total system 2002 memory in the form of "nn%", or "mirror". If "mirror" 2003 option is specified, mirrored (reliable) memory is used 2004 for non-movable allocations and remaining memory is used 2005 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2006 are exclusive, so you cannot specify multiple forms. 2007 2008 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port. 2009 Format: <Controller#>[,poll interval] 2010 The controller # is the number of the ehci usb debug 2011 port as it is probed via PCI. The poll interval is 2012 optional and is the number seconds in between 2013 each poll cycle to the debug port in case you need 2014 the functionality for interrupting the kernel with 2015 gdb or control-c on the dbgp connection. When 2016 not using this parameter you use sysrq-g to break into 2017 the kernel debugger. 2018 2019 kgdboc= [KGDB,HW] kgdb over consoles. 2020 Requires a tty driver that supports console polling, 2021 or a supported polling keyboard driver (non-usb). 2022 Serial only format: <serial_device>[,baud] 2023 keyboard only format: kbd 2024 keyboard and serial format: kbd,<serial_device>[,baud] 2025 Optional Kernel mode setting: 2026 kms, kbd format: kms,kbd 2027 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2028 2029 kgdbwait [KGDB] Stop kernel execution and enter the 2030 kernel debugger at the earliest opportunity. 2031 2032 kmac= [MIPS] korina ethernet MAC address. 2033 Configure the RouterBoard 532 series on-chip 2034 Ethernet adapter MAC address. 2035 2036 kmemleak= [KNL] Boot-time kmemleak enable/disable 2037 Valid arguments: on, off 2038 Default: on 2039 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 2040 the default is off. 2041 2042 kprobe_event=[probe-list] 2043 [FTRACE] Add kprobe events and enable at boot time. 2044 The probe-list is a semicolon delimited list of probe 2045 definitions. Each definition is same as kprobe_events 2046 interface, but the parameters are comma delimited. 2047 For example, to add a kprobe event on vfs_read with 2048 arg1 and arg2, add to the command line; 2049 2050 kprobe_event=p,vfs_read,$arg1,$arg2 2051 2052 See also Documentation/trace/kprobetrace.rst "Kernel 2053 Boot Parameter" section. 2054 2055 kpti= [ARM64] Control page table isolation of user 2056 and kernel address spaces. 2057 Default: enabled on cores which need mitigation. 2058 0: force disabled 2059 1: force enabled 2060 2061 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 2062 Default is 0 (don't ignore, but inject #GP) 2063 2064 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 2065 Default is false (don't support). 2066 2067 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit 2068 KVM MMU at runtime. 2069 Default is 0 (off) 2070 2071 kvm.nx_huge_pages= 2072 [KVM] Controls the software workaround for the 2073 X86_BUG_ITLB_MULTIHIT bug. 2074 force : Always deploy workaround. 2075 off : Never deploy workaround. 2076 auto : Deploy workaround based on the presence of 2077 X86_BUG_ITLB_MULTIHIT. 2078 2079 Default is 'auto'. 2080 2081 If the software workaround is enabled for the host, 2082 guests do need not to enable it for nested guests. 2083 2084 kvm.nx_huge_pages_recovery_ratio= 2085 [KVM] Controls how many 4KiB pages are periodically zapped 2086 back to huge pages. 0 disables the recovery, otherwise if 2087 the value is N KVM will zap 1/Nth of the 4KiB pages every 2088 minute. The default is 60. 2089 2090 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM. 2091 Default is 1 (enabled) 2092 2093 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU) 2094 for all guests. 2095 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode. 2096 2097 kvm-arm.vgic_v3_group0_trap= 2098 [KVM,ARM] Trap guest accesses to GICv3 group-0 2099 system registers 2100 2101 kvm-arm.vgic_v3_group1_trap= 2102 [KVM,ARM] Trap guest accesses to GICv3 group-1 2103 system registers 2104 2105 kvm-arm.vgic_v3_common_trap= 2106 [KVM,ARM] Trap guest accesses to GICv3 common 2107 system registers 2108 2109 kvm-arm.vgic_v4_enable= 2110 [KVM,ARM] Allow use of GICv4 for direct injection of 2111 LPIs. 2112 2113 kvm-intel.ept= [KVM,Intel] Disable extended page tables 2114 (virtualized MMU) support on capable Intel chips. 2115 Default is 1 (enabled) 2116 2117 kvm-intel.emulate_invalid_guest_state= 2118 [KVM,Intel] Enable emulation of invalid guest states 2119 Default is 0 (disabled) 2120 2121 kvm-intel.flexpriority= 2122 [KVM,Intel] Disable FlexPriority feature (TPR shadow). 2123 Default is 1 (enabled) 2124 2125 kvm-intel.nested= 2126 [KVM,Intel] Enable VMX nesting (nVMX). 2127 Default is 0 (disabled) 2128 2129 kvm-intel.unrestricted_guest= 2130 [KVM,Intel] Disable unrestricted guest feature 2131 (virtualized real and unpaged mode) on capable 2132 Intel chips. Default is 1 (enabled) 2133 2134 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 2135 CVE-2018-3620. 2136 2137 Valid arguments: never, cond, always 2138 2139 always: L1D cache flush on every VMENTER. 2140 cond: Flush L1D on VMENTER only when the code between 2141 VMEXIT and VMENTER can leak host memory. 2142 never: Disables the mitigation 2143 2144 Default is cond (do L1 cache flush in specific instances) 2145 2146 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification 2147 feature (tagged TLBs) on capable Intel chips. 2148 Default is 1 (enabled) 2149 2150 l1tf= [X86] Control mitigation of the L1TF vulnerability on 2151 affected CPUs 2152 2153 The kernel PTE inversion protection is unconditionally 2154 enabled and cannot be disabled. 2155 2156 full 2157 Provides all available mitigations for the 2158 L1TF vulnerability. Disables SMT and 2159 enables all mitigations in the 2160 hypervisors, i.e. unconditional L1D flush. 2161 2162 SMT control and L1D flush control via the 2163 sysfs interface is still possible after 2164 boot. Hypervisors will issue a warning 2165 when the first VM is started in a 2166 potentially insecure configuration, 2167 i.e. SMT enabled or L1D flush disabled. 2168 2169 full,force 2170 Same as 'full', but disables SMT and L1D 2171 flush runtime control. Implies the 2172 'nosmt=force' command line option. 2173 (i.e. sysfs control of SMT is disabled.) 2174 2175 flush 2176 Leaves SMT enabled and enables the default 2177 hypervisor mitigation, i.e. conditional 2178 L1D flush. 2179 2180 SMT control and L1D flush control via the 2181 sysfs interface is still possible after 2182 boot. Hypervisors will issue a warning 2183 when the first VM is started in a 2184 potentially insecure configuration, 2185 i.e. SMT enabled or L1D flush disabled. 2186 2187 flush,nosmt 2188 2189 Disables SMT and enables the default 2190 hypervisor mitigation. 2191 2192 SMT control and L1D flush control via the 2193 sysfs interface is still possible after 2194 boot. Hypervisors will issue a warning 2195 when the first VM is started in a 2196 potentially insecure configuration, 2197 i.e. SMT enabled or L1D flush disabled. 2198 2199 flush,nowarn 2200 Same as 'flush', but hypervisors will not 2201 warn when a VM is started in a potentially 2202 insecure configuration. 2203 2204 off 2205 Disables hypervisor mitigations and doesn't 2206 emit any warnings. 2207 It also drops the swap size and available 2208 RAM limit restriction on both hypervisor and 2209 bare metal. 2210 2211 Default is 'flush'. 2212 2213 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 2214 2215 l2cr= [PPC] 2216 2217 l3cr= [PPC] 2218 2219 lapic [X86-32,APIC] Enable the local APIC even if BIOS 2220 disabled it. 2221 2222 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline 2223 value for LAPIC timer one-shot implementation. Default 2224 back to the programmable timer unit in the LAPIC. 2225 2226 lapic_timer_c2_ok [X86,APIC] trust the local apic timer 2227 in C2 power state. 2228 2229 libata.dma= [LIBATA] DMA control 2230 libata.dma=0 Disable all PATA and SATA DMA 2231 libata.dma=1 PATA and SATA Disk DMA only 2232 libata.dma=2 ATAPI (CDROM) DMA only 2233 libata.dma=4 Compact Flash DMA only 2234 Combinations also work, so libata.dma=3 enables DMA 2235 for disks and CDROMs, but not CFs. 2236 2237 libata.ignore_hpa= [LIBATA] Ignore HPA limit 2238 libata.ignore_hpa=0 keep BIOS limits (default) 2239 libata.ignore_hpa=1 ignore limits, using full disk 2240 2241 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 2242 when set. 2243 Format: <int> 2244 2245 libata.force= [LIBATA] Force configurations. The format is comma 2246 separated list of "[ID:]VAL" where ID is 2247 PORT[.DEVICE]. PORT and DEVICE are decimal numbers 2248 matching port, link or device. Basically, it matches 2249 the ATA ID string printed on console by libata. If 2250 the whole ID part is omitted, the last PORT and DEVICE 2251 values are used. If ID hasn't been specified yet, the 2252 configuration applies to all ports, links and devices. 2253 2254 If only DEVICE is omitted, the parameter applies to 2255 the port and all links and devices behind it. DEVICE 2256 number of 0 either selects the first device or the 2257 first fan-out link behind PMP device. It does not 2258 select the host link. DEVICE number of 15 selects the 2259 host link and device attached to it. 2260 2261 The VAL specifies the configuration to force. As long 2262 as there's no ambiguity shortcut notation is allowed. 2263 For example, both 1.5 and 1.5G would work for 1.5Gbps. 2264 The following configurations can be forced. 2265 2266 * Cable type: 40c, 80c, short40c, unk, ign or sata. 2267 Any ID with matching PORT is used. 2268 2269 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 2270 2271 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 2272 udma[/][16,25,33,44,66,100,133] notation is also 2273 allowed. 2274 2275 * [no]ncq: Turn on or off NCQ. 2276 2277 * [no]ncqtrim: Turn off queued DSM TRIM. 2278 2279 * nohrst, nosrst, norst: suppress hard, soft 2280 and both resets. 2281 2282 * rstonce: only attempt one reset during 2283 hot-unplug link recovery 2284 2285 * dump_id: dump IDENTIFY data. 2286 2287 * atapi_dmadir: Enable ATAPI DMADIR bridge support 2288 2289 * disable: Disable this device. 2290 2291 If there are multiple matching configurations changing 2292 the same attribute, the last one is used. 2293 2294 memblock=debug [KNL] Enable memblock debug messages. 2295 2296 load_ramdisk= [RAM] List of ramdisks to load from floppy 2297 See Documentation/admin-guide/blockdev/ramdisk.rst. 2298 2299 lockd.nlm_grace_period=P [NFS] Assign grace period. 2300 Format: <integer> 2301 2302 lockd.nlm_tcpport=N [NFS] Assign TCP port. 2303 Format: <integer> 2304 2305 lockd.nlm_timeout=T [NFS] Assign timeout value. 2306 Format: <integer> 2307 2308 lockd.nlm_udpport=M [NFS] Assign UDP port. 2309 Format: <integer> 2310 2311 lockdown= [SECURITY] 2312 { integrity | confidentiality } 2313 Enable the kernel lockdown feature. If set to 2314 integrity, kernel features that allow userland to 2315 modify the running kernel are disabled. If set to 2316 confidentiality, kernel features that allow userland 2317 to extract confidential information from the kernel 2318 are also disabled. 2319 2320 locktorture.nreaders_stress= [KNL] 2321 Set the number of locking read-acquisition kthreads. 2322 Defaults to being automatically set based on the 2323 number of online CPUs. 2324 2325 locktorture.nwriters_stress= [KNL] 2326 Set the number of locking write-acquisition kthreads. 2327 2328 locktorture.onoff_holdoff= [KNL] 2329 Set time (s) after boot for CPU-hotplug testing. 2330 2331 locktorture.onoff_interval= [KNL] 2332 Set time (s) between CPU-hotplug operations, or 2333 zero to disable CPU-hotplug testing. 2334 2335 locktorture.shuffle_interval= [KNL] 2336 Set task-shuffle interval (jiffies). Shuffling 2337 tasks allows some CPUs to go into dyntick-idle 2338 mode during the locktorture test. 2339 2340 locktorture.shutdown_secs= [KNL] 2341 Set time (s) after boot system shutdown. This 2342 is useful for hands-off automated testing. 2343 2344 locktorture.stat_interval= [KNL] 2345 Time (s) between statistics printk()s. 2346 2347 locktorture.stutter= [KNL] 2348 Time (s) to stutter testing, for example, 2349 specifying five seconds causes the test to run for 2350 five seconds, wait for five seconds, and so on. 2351 This tests the locking primitive's ability to 2352 transition abruptly to and from idle. 2353 2354 locktorture.torture_type= [KNL] 2355 Specify the locking implementation to test. 2356 2357 locktorture.verbose= [KNL] 2358 Enable additional printk() statements. 2359 2360 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 2361 Format: <irq> 2362 2363 loglevel= All Kernel Messages with a loglevel smaller than the 2364 console loglevel will be printed to the console. It can 2365 also be changed with klogd or other programs. The 2366 loglevels are defined as follows: 2367 2368 0 (KERN_EMERG) system is unusable 2369 1 (KERN_ALERT) action must be taken immediately 2370 2 (KERN_CRIT) critical conditions 2371 3 (KERN_ERR) error conditions 2372 4 (KERN_WARNING) warning conditions 2373 5 (KERN_NOTICE) normal but significant condition 2374 6 (KERN_INFO) informational 2375 7 (KERN_DEBUG) debug-level messages 2376 2377 log_buf_len=n[KMG] Sets the size of the printk ring buffer, 2378 in bytes. n must be a power of two and greater 2379 than the minimal size. The minimal size is defined 2380 by LOG_BUF_SHIFT kernel config parameter. There is 2381 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter 2382 that allows to increase the default size depending on 2383 the number of CPUs. See init/Kconfig for more details. 2384 2385 logo.nologo [FB] Disables display of the built-in Linux logo. 2386 This may be used to provide more screen space for 2387 kernel log messages and is useful when debugging 2388 kernel boot problems. 2389 2390 lp=0 [LP] Specify parallel ports to use, e.g, 2391 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 2392 lp=reset first parallel port). 'lp=0' disables the 2393 lp=auto printer driver. 'lp=reset' (which can be 2394 specified in addition to the ports) causes 2395 attached printers to be reset. Using 2396 lp=port1,port2,... specifies the parallel ports 2397 to associate lp devices with, starting with 2398 lp0. A port specification may be 'none' to skip 2399 that lp device, or a parport name such as 2400 'parport0'. Specifying 'lp=auto' instead of a 2401 port specification list means that device IDs 2402 from each port should be examined, to see if 2403 an IEEE 1284-compliant printer is attached; if 2404 so, the driver will manage that printer. 2405 See also header of drivers/char/lp.c. 2406 2407 lpj=n [KNL] 2408 Sets loops_per_jiffy to given constant, thus avoiding 2409 time-consuming boot-time autodetection (up to 250 ms per 2410 CPU). 0 enables autodetection (default). To determine 2411 the correct value for your kernel, boot with normal 2412 autodetection and see what value is printed. Note that 2413 on SMP systems the preset will be applied to all CPUs, 2414 which is likely to cause problems if your CPUs need 2415 significantly divergent settings. An incorrect value 2416 will cause delays in the kernel to be wrong, leading to 2417 unpredictable I/O errors and other breakage. Although 2418 unlikely, in the extreme case this might damage your 2419 hardware. 2420 2421 ltpc= [NET] 2422 Format: <io>,<irq>,<dma> 2423 2424 lsm.debug [SECURITY] Enable LSM initialization debugging output. 2425 2426 lsm=lsm1,...,lsmN 2427 [SECURITY] Choose order of LSM initialization. This 2428 overrides CONFIG_LSM, and the "security=" parameter. 2429 2430 machvec= [IA-64] Force the use of a particular machine-vector 2431 (machvec) in a generic kernel. 2432 Example: machvec=hpzx1 2433 2434 machtype= [Loongson] Share the same kernel image file between different 2435 yeeloong laptop. 2436 Example: machtype=lemote-yeeloong-2f-7inch 2437 2438 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater 2439 than or equal to this physical address is ignored. 2440 2441 maxcpus= [SMP] Maximum number of processors that an SMP kernel 2442 will bring up during bootup. maxcpus=n : n >= 0 limits 2443 the kernel to bring up 'n' processors. Surely after 2444 bootup you can bring up the other plugged cpu by executing 2445 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 2446 only takes effect during system bootup. 2447 While n=0 is a special case, it is equivalent to "nosmp", 2448 which also disables the IO APIC. 2449 2450 max_loop= [LOOP] The number of loop block devices that get 2451 (loop.max_loop) unconditionally pre-created at init time. The default 2452 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 2453 of statically allocating a predefined number, loop 2454 devices can be requested on-demand with the 2455 /dev/loop-control interface. 2456 2457 mce [X86-32] Machine Check Exception 2458 2459 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.rst 2460 2461 md= [HW] RAID subsystems devices and level 2462 See Documentation/admin-guide/md.rst. 2463 2464 mdacon= [MDA] 2465 Format: <first>,<last> 2466 Specifies range of consoles to be captured by the MDA. 2467 2468 mds= [X86,INTEL] 2469 Control mitigation for the Micro-architectural Data 2470 Sampling (MDS) vulnerability. 2471 2472 Certain CPUs are vulnerable to an exploit against CPU 2473 internal buffers which can forward information to a 2474 disclosure gadget under certain conditions. 2475 2476 In vulnerable processors, the speculatively 2477 forwarded data can be used in a cache side channel 2478 attack, to access data to which the attacker does 2479 not have direct access. 2480 2481 This parameter controls the MDS mitigation. The 2482 options are: 2483 2484 full - Enable MDS mitigation on vulnerable CPUs 2485 full,nosmt - Enable MDS mitigation and disable 2486 SMT on vulnerable CPUs 2487 off - Unconditionally disable MDS mitigation 2488 2489 On TAA-affected machines, mds=off can be prevented by 2490 an active TAA mitigation as both vulnerabilities are 2491 mitigated with the same mechanism so in order to disable 2492 this mitigation, you need to specify tsx_async_abort=off 2493 too. 2494 2495 Not specifying this option is equivalent to 2496 mds=full. 2497 2498 For details see: Documentation/admin-guide/hw-vuln/mds.rst 2499 2500 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory 2501 Amount of memory to be used when the kernel is not able 2502 to see the whole system memory or for test. 2503 [X86] Work as limiting max address. Use together 2504 with memmap= to avoid physical address space collisions. 2505 Without memmap= PCI devices could be placed at addresses 2506 belonging to unused RAM. 2507 2508 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 2509 memory. 2510 2511 memchunk=nn[KMG] 2512 [KNL,SH] Allow user to override the default size for 2513 per-device physically contiguous DMA buffers. 2514 2515 memhp_default_state=online/offline 2516 [KNL] Set the initial state for the memory hotplug 2517 onlining policy. If not specified, the default value is 2518 set according to the 2519 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config 2520 option. 2521 See Documentation/admin-guide/mm/memory-hotplug.rst. 2522 2523 memmap=exactmap [KNL,X86] Enable setting of an exact 2524 E820 memory map, as specified by the user. 2525 Such memmap=exactmap lines can be constructed based on 2526 BIOS output or other requirements. See the memmap=nn@ss 2527 option description. 2528 2529 memmap=nn[KMG]@ss[KMG] 2530 [KNL] Force usage of a specific region of memory. 2531 Region of memory to be used is from ss to ss+nn. 2532 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 2533 which limits max address to nn[KMG]. 2534 Multiple different regions can be specified, 2535 comma delimited. 2536 Example: 2537 memmap=100M@2G,100M#3G,1G!1024G 2538 2539 memmap=nn[KMG]#ss[KMG] 2540 [KNL,ACPI] Mark specific memory as ACPI data. 2541 Region of memory to be marked is from ss to ss+nn. 2542 2543 memmap=nn[KMG]$ss[KMG] 2544 [KNL,ACPI] Mark specific memory as reserved. 2545 Region of memory to be reserved is from ss to ss+nn. 2546 Example: Exclude memory from 0x18690000-0x1869ffff 2547 memmap=64K$0x18690000 2548 or 2549 memmap=0x10000$0x18690000 2550 Some bootloaders may need an escape character before '$', 2551 like Grub2, otherwise '$' and the following number 2552 will be eaten. 2553 2554 memmap=nn[KMG]!ss[KMG] 2555 [KNL,X86] Mark specific memory as protected. 2556 Region of memory to be used, from ss to ss+nn. 2557 The memory region may be marked as e820 type 12 (0xc) 2558 and is NVDIMM or ADR memory. 2559 2560 memmap=<size>%<offset>-<oldtype>+<newtype> 2561 [KNL,ACPI] Convert memory within the specified region 2562 from <oldtype> to <newtype>. If "-<oldtype>" is left 2563 out, the whole region will be marked as <newtype>, 2564 even if previously unavailable. If "+<newtype>" is left 2565 out, matching memory will be removed. Types are 2566 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 2567 3 = ACPI, 12 = PRAM. 2568 2569 memory_corruption_check=0/1 [X86] 2570 Some BIOSes seem to corrupt the first 64k of 2571 memory when doing things like suspend/resume. 2572 Setting this option will scan the memory 2573 looking for corruption. Enabling this will 2574 both detect corruption and prevent the kernel 2575 from using the memory being corrupted. 2576 However, its intended as a diagnostic tool; if 2577 repeatable BIOS-originated corruption always 2578 affects the same memory, you can use memmap= 2579 to prevent the kernel from using that memory. 2580 2581 memory_corruption_check_size=size [X86] 2582 By default it checks for corruption in the low 2583 64k, making this memory unavailable for normal 2584 use. Use this parameter to scan for 2585 corruption in more or less memory. 2586 2587 memory_corruption_check_period=seconds [X86] 2588 By default it checks for corruption every 60 2589 seconds. Use this parameter to check at some 2590 other rate. 0 disables periodic checking. 2591 2592 memtest= [KNL,X86,ARM,PPC] Enable memtest 2593 Format: <integer> 2594 default : 0 <disable> 2595 Specifies the number of memtest passes to be 2596 performed. Each pass selects another test 2597 pattern from a given set of patterns. Memtest 2598 fills the memory with this pattern, validates 2599 memory contents and reserves bad memory 2600 regions that are detected. 2601 2602 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 2603 Valid arguments: on, off 2604 Default (depends on kernel configuration option): 2605 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y) 2606 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n) 2607 mem_encrypt=on: Activate SME 2608 mem_encrypt=off: Do not activate SME 2609 2610 Refer to Documentation/virt/kvm/amd-memory-encryption.rst 2611 for details on when memory encryption can be activated. 2612 2613 mem_sleep_default= [SUSPEND] Default system suspend mode: 2614 s2idle - Suspend-To-Idle 2615 shallow - Power-On Suspend or equivalent (if supported) 2616 deep - Suspend-To-RAM or equivalent (if supported) 2617 See Documentation/admin-guide/pm/sleep-states.rst. 2618 2619 meye.*= [HW] Set MotionEye Camera parameters 2620 See Documentation/media/v4l-drivers/meye.rst. 2621 2622 mfgpt_irq= [IA-32] Specify the IRQ to use for the 2623 Multi-Function General Purpose Timers on AMD Geode 2624 platforms. 2625 2626 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 2627 the BIOS has incorrectly applied a workaround. TinyBIOS 2628 version 0.98 is known to be affected, 0.99 fixes the 2629 problem by letting the user disable the workaround. 2630 2631 mga= [HW,DRM] 2632 2633 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this 2634 physical address is ignored. 2635 2636 mini2440= [ARM,HW,KNL] 2637 Format:[0..2][b][c][t] 2638 Default: "0tb" 2639 MINI2440 configuration specification: 2640 0 - The attached screen is the 3.5" TFT 2641 1 - The attached screen is the 7" TFT 2642 2 - The VGA Shield is attached (1024x768) 2643 Leaving out the screen size parameter will not load 2644 the TFT driver, and the framebuffer will be left 2645 unconfigured. 2646 b - Enable backlight. The TFT backlight pin will be 2647 linked to the kernel VESA blanking code and a GPIO 2648 LED. This parameter is not necessary when using the 2649 VGA shield. 2650 c - Enable the s3c camera interface. 2651 t - Reserved for enabling touchscreen support. The 2652 touchscreen support is not enabled in the mainstream 2653 kernel as of 2.6.30, a preliminary port can be found 2654 in the "bleeding edge" mini2440 support kernel at 2655 http://repo.or.cz/w/linux-2.6/mini2440.git 2656 2657 mitigations= 2658 [X86,PPC,S390,ARM64] Control optional mitigations for 2659 CPU vulnerabilities. This is a set of curated, 2660 arch-independent options, each of which is an 2661 aggregation of existing arch-specific options. 2662 2663 off 2664 Disable all optional CPU mitigations. This 2665 improves system performance, but it may also 2666 expose users to several CPU vulnerabilities. 2667 Equivalent to: nopti [X86,PPC] 2668 kpti=0 [ARM64] 2669 nospectre_v1 [X86,PPC] 2670 nobp=0 [S390] 2671 nospectre_v2 [X86,PPC,S390,ARM64] 2672 spectre_v2_user=off [X86] 2673 spec_store_bypass_disable=off [X86,PPC] 2674 ssbd=force-off [ARM64] 2675 l1tf=off [X86] 2676 mds=off [X86] 2677 tsx_async_abort=off [X86] 2678 kvm.nx_huge_pages=off [X86] 2679 2680 Exceptions: 2681 This does not have any effect on 2682 kvm.nx_huge_pages when 2683 kvm.nx_huge_pages=force. 2684 2685 auto (default) 2686 Mitigate all CPU vulnerabilities, but leave SMT 2687 enabled, even if it's vulnerable. This is for 2688 users who don't want to be surprised by SMT 2689 getting disabled across kernel upgrades, or who 2690 have other ways of avoiding SMT-based attacks. 2691 Equivalent to: (default behavior) 2692 2693 auto,nosmt 2694 Mitigate all CPU vulnerabilities, disabling SMT 2695 if needed. This is for users who always want to 2696 be fully mitigated, even if it means losing SMT. 2697 Equivalent to: l1tf=flush,nosmt [X86] 2698 mds=full,nosmt [X86] 2699 tsx_async_abort=full,nosmt [X86] 2700 2701 mminit_loglevel= 2702 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this 2703 parameter allows control of the logging verbosity for 2704 the additional memory initialisation checks. A value 2705 of 0 disables mminit logging and a level of 4 will 2706 log everything. Information is printed at KERN_DEBUG 2707 so loglevel=8 may also need to be specified. 2708 2709 module.sig_enforce 2710 [KNL] When CONFIG_MODULE_SIG is set, this means that 2711 modules without (valid) signatures will fail to load. 2712 Note that if CONFIG_MODULE_SIG_FORCE is set, that 2713 is always true, so this option does nothing. 2714 2715 module_blacklist= [KNL] Do not load a comma-separated list of 2716 modules. Useful for debugging problem modules. 2717 2718 mousedev.tap_time= 2719 [MOUSE] Maximum time between finger touching and 2720 leaving touchpad surface for touch to be considered 2721 a tap and be reported as a left button click (for 2722 touchpads working in absolute mode only). 2723 Format: <msecs> 2724 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 2725 reporting absolute coordinates, such as tablets 2726 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 2727 reporting absolute coordinates, such as tablets 2728 2729 movablecore= [KNL,X86,IA-64,PPC] 2730 Format: nn[KMGTPE] | nn% 2731 This parameter is the complement to kernelcore=, it 2732 specifies the amount of memory used for migratable 2733 allocations. If both kernelcore and movablecore is 2734 specified, then kernelcore will be at *least* the 2735 specified value but may be more. If movablecore on its 2736 own is specified, the administrator must be careful 2737 that the amount of memory usable for all allocations 2738 is not too small. 2739 2740 movable_node [KNL] Boot-time switch to make hotplugable memory 2741 NUMA nodes to be movable. This means that the memory 2742 of such nodes will be usable only for movable 2743 allocations which rules out almost all kernel 2744 allocations. Use with caution! 2745 2746 MTD_Partition= [MTD] 2747 Format: <name>,<region-number>,<size>,<offset> 2748 2749 MTD_Region= [MTD] Format: 2750 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 2751 2752 mtdparts= [MTD] 2753 See drivers/mtd/cmdlinepart.c. 2754 2755 multitce=off [PPC] This parameter disables the use of the pSeries 2756 firmware feature for updating multiple TCE entries 2757 at a time. 2758 2759 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 2760 2761 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 2762 2763 boundary - index of last SLC block on Flex-OneNAND. 2764 The remaining blocks are configured as MLC blocks. 2765 lock - Configure if Flex-OneNAND boundary should be locked. 2766 Once locked, the boundary cannot be changed. 2767 1 indicates lock status, 0 indicates unlock status. 2768 2769 mtdset= [ARM] 2770 ARM/S3C2412 JIVE boot control 2771 2772 See arch/arm/mach-s3c2412/mach-jive.c 2773 2774 mtouchusb.raw_coordinates= 2775 [HW] Make the MicroTouch USB driver use raw coordinates 2776 ('y', default) or cooked coordinates ('n') 2777 2778 mtrr_chunk_size=nn[KMG] [X86] 2779 used for mtrr cleanup. It is largest continuous chunk 2780 that could hold holes aka. UC entries. 2781 2782 mtrr_gran_size=nn[KMG] [X86] 2783 Used for mtrr cleanup. It is granularity of mtrr block. 2784 Default is 1. 2785 Large value could prevent small alignment from 2786 using up MTRRs. 2787 2788 mtrr_spare_reg_nr=n [X86] 2789 Format: <integer> 2790 Range: 0,7 : spare reg number 2791 Default : 1 2792 Used for mtrr cleanup. It is spare mtrr entries number. 2793 Set to 2 or more if your graphical card needs more. 2794 2795 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 2796 2797 netdev= [NET] Network devices parameters 2798 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 2799 Note that mem_start is often overloaded to mean 2800 something different and driver-specific. 2801 This usage is only documented in each driver source 2802 file if at all. 2803 2804 nf_conntrack.acct= 2805 [NETFILTER] Enable connection tracking flow accounting 2806 0 to disable accounting 2807 1 to enable accounting 2808 Default value is 0. 2809 2810 nfsaddrs= [NFS] Deprecated. Use ip= instead. 2811 See Documentation/filesystems/nfs/nfsroot.txt. 2812 2813 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 2814 See Documentation/filesystems/nfs/nfsroot.txt. 2815 2816 nfsrootdebug [NFS] enable nfsroot debugging messages. 2817 See Documentation/filesystems/nfs/nfsroot.txt. 2818 2819 nfs.callback_nr_threads= 2820 [NFSv4] set the total number of threads that the 2821 NFS client will assign to service NFSv4 callback 2822 requests. 2823 2824 nfs.callback_tcpport= 2825 [NFS] set the TCP port on which the NFSv4 callback 2826 channel should listen. 2827 2828 nfs.cache_getent= 2829 [NFS] sets the pathname to the program which is used 2830 to update the NFS client cache entries. 2831 2832 nfs.cache_getent_timeout= 2833 [NFS] sets the timeout after which an attempt to 2834 update a cache entry is deemed to have failed. 2835 2836 nfs.idmap_cache_timeout= 2837 [NFS] set the maximum lifetime for idmapper cache 2838 entries. 2839 2840 nfs.enable_ino64= 2841 [NFS] enable 64-bit inode numbers. 2842 If zero, the NFS client will fake up a 32-bit inode 2843 number for the readdir() and stat() syscalls instead 2844 of returning the full 64-bit number. 2845 The default is to return 64-bit inode numbers. 2846 2847 nfs.max_session_cb_slots= 2848 [NFSv4.1] Sets the maximum number of session 2849 slots the client will assign to the callback 2850 channel. This determines the maximum number of 2851 callbacks the client will process in parallel for 2852 a particular server. 2853 2854 nfs.max_session_slots= 2855 [NFSv4.1] Sets the maximum number of session slots 2856 the client will attempt to negotiate with the server. 2857 This limits the number of simultaneous RPC requests 2858 that the client can send to the NFSv4.1 server. 2859 Note that there is little point in setting this 2860 value higher than the max_tcp_slot_table_limit. 2861 2862 nfs.nfs4_disable_idmapping= 2863 [NFSv4] When set to the default of '1', this option 2864 ensures that both the RPC level authentication 2865 scheme and the NFS level operations agree to use 2866 numeric uids/gids if the mount is using the 2867 'sec=sys' security flavour. In effect it is 2868 disabling idmapping, which can make migration from 2869 legacy NFSv2/v3 systems to NFSv4 easier. 2870 Servers that do not support this mode of operation 2871 will be autodetected by the client, and it will fall 2872 back to using the idmapper. 2873 To turn off this behaviour, set the value to '0'. 2874 nfs.nfs4_unique_id= 2875 [NFS4] Specify an additional fixed unique ident- 2876 ification string that NFSv4 clients can insert into 2877 their nfs_client_id4 string. This is typically a 2878 UUID that is generated at system install time. 2879 2880 nfs.send_implementation_id = 2881 [NFSv4.1] Send client implementation identification 2882 information in exchange_id requests. 2883 If zero, no implementation identification information 2884 will be sent. 2885 The default is to send the implementation identification 2886 information. 2887 2888 nfs.recover_lost_locks = 2889 [NFSv4] Attempt to recover locks that were lost due 2890 to a lease timeout on the server. Please note that 2891 doing this risks data corruption, since there are 2892 no guarantees that the file will remain unchanged 2893 after the locks are lost. 2894 If you want to enable the kernel legacy behaviour of 2895 attempting to recover these locks, then set this 2896 parameter to '1'. 2897 The default parameter value of '0' causes the kernel 2898 not to attempt recovery of lost locks. 2899 2900 nfs4.layoutstats_timer = 2901 [NFSv4.2] Change the rate at which the kernel sends 2902 layoutstats to the pNFS metadata server. 2903 2904 Setting this to value to 0 causes the kernel to use 2905 whatever value is the default set by the layout 2906 driver. A non-zero value sets the minimum interval 2907 in seconds between layoutstats transmissions. 2908 2909 nfsd.nfs4_disable_idmapping= 2910 [NFSv4] When set to the default of '1', the NFSv4 2911 server will return only numeric uids and gids to 2912 clients using auth_sys, and will accept numeric uids 2913 and gids from such clients. This is intended to ease 2914 migration from NFSv2/v3. 2915 2916 nmi_debug= [KNL,SH] Specify one or more actions to take 2917 when a NMI is triggered. 2918 Format: [state][,regs][,debounce][,die] 2919 2920 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 2921 Format: [panic,][nopanic,][num] 2922 Valid num: 0 or 1 2923 0 - turn hardlockup detector in nmi_watchdog off 2924 1 - turn hardlockup detector in nmi_watchdog on 2925 When panic is specified, panic when an NMI watchdog 2926 timeout occurs (or 'nopanic' to not panic on an NMI 2927 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 2928 To disable both hard and soft lockup detectors, 2929 please see 'nowatchdog'. 2930 This is useful when you use a panic=... timeout and 2931 need the box quickly up again. 2932 2933 These settings can be accessed at runtime via 2934 the nmi_watchdog and hardlockup_panic sysctls. 2935 2936 netpoll.carrier_timeout= 2937 [NET] Specifies amount of time (in seconds) that 2938 netpoll should wait for a carrier. By default netpoll 2939 waits 4 seconds. 2940 2941 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 2942 emulation library even if a 387 maths coprocessor 2943 is present. 2944 2945 no5lvl [X86-64] Disable 5-level paging mode. Forces 2946 kernel to use 4-level paging instead. 2947 2948 no_console_suspend 2949 [HW] Never suspend the console 2950 Disable suspending of consoles during suspend and 2951 hibernate operations. Once disabled, debugging 2952 messages can reach various consoles while the rest 2953 of the system is being put to sleep (ie, while 2954 debugging driver suspend/resume hooks). This may 2955 not work reliably with all consoles, but is known 2956 to work with serial and VGA consoles. 2957 To facilitate more flexible debugging, we also add 2958 console_suspend, a printk module parameter to control 2959 it. Users could use console_suspend (usually 2960 /sys/module/printk/parameters/console_suspend) to 2961 turn on/off it dynamically. 2962 2963 novmcoredd [KNL,KDUMP] 2964 Disable device dump. Device dump allows drivers to 2965 append dump data to vmcore so you can collect driver 2966 specified debug info. Drivers can append the data 2967 without any limit and this data is stored in memory, 2968 so this may cause significant memory stress. Disabling 2969 device dump can help save memory but the driver debug 2970 data will be no longer available. This parameter 2971 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 2972 is set. 2973 2974 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien 2975 caches in the slab allocator. Saves per-node memory, 2976 but will impact performance. 2977 2978 noalign [KNL,ARM] 2979 2980 noaltinstr [S390] Disables alternative instructions patching 2981 (CPU alternatives feature). 2982 2983 noapic [SMP,APIC] Tells the kernel to not make use of any 2984 IOAPICs that may be present in the system. 2985 2986 noautogroup Disable scheduler automatic task group creation. 2987 2988 nobats [PPC] Do not use BATs for mapping kernel lowmem 2989 on "Classic" PPC cores. 2990 2991 nocache [ARM] 2992 2993 noclflush [BUGS=X86] Don't use the CLFLUSH instruction 2994 2995 nodelayacct [KNL] Disable per-task delay accounting 2996 2997 nodsp [SH] Disable hardware DSP at boot time. 2998 2999 noefi Disable EFI runtime services support. 3000 3001 noexec [IA-64] 3002 3003 noexec [X86] 3004 On X86-32 available only on PAE configured kernels. 3005 noexec=on: enable non-executable mappings (default) 3006 noexec=off: disable non-executable mappings 3007 3008 nosmap [X86,PPC] 3009 Disable SMAP (Supervisor Mode Access Prevention) 3010 even if it is supported by processor. 3011 3012 nosmep [X86,PPC] 3013 Disable SMEP (Supervisor Mode Execution Prevention) 3014 even if it is supported by processor. 3015 3016 noexec32 [X86-64] 3017 This affects only 32-bit executables. 3018 noexec32=on: enable non-executable mappings (default) 3019 read doesn't imply executable mappings 3020 noexec32=off: disable non-executable mappings 3021 read implies executable mappings 3022 3023 nofpu [MIPS,SH] Disable hardware FPU at boot time. 3024 3025 nofxsr [BUGS=X86-32] Disables x86 floating point extended 3026 register save and restore. The kernel will only save 3027 legacy floating-point registers on task switch. 3028 3029 nohugeiomap [KNL,x86,PPC] Disable kernel huge I/O mappings. 3030 3031 nosmt [KNL,S390] Disable symmetric multithreading (SMT). 3032 Equivalent to smt=1. 3033 3034 [KNL,x86] Disable symmetric multithreading (SMT). 3035 nosmt=force: Force disable SMT, cannot be undone 3036 via the sysfs control file. 3037 3038 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1 3039 (bounds check bypass). With this option data leaks are 3040 possible in the system. 3041 3042 nospectre_v2 [X86,PPC_FSL_BOOK3E,ARM64] Disable all mitigations for 3043 the Spectre variant 2 (indirect branch prediction) 3044 vulnerability. System may allow data leaks with this 3045 option. 3046 3047 nospec_store_bypass_disable 3048 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability 3049 3050 noxsave [BUGS=X86] Disables x86 extended register state save 3051 and restore using xsave. The kernel will fallback to 3052 enabling legacy floating-point and sse state. 3053 3054 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 3055 register states. The kernel will fall back to use 3056 xsave to save the states. By using this parameter, 3057 performance of saving the states is degraded because 3058 xsave doesn't support modified optimization while 3059 xsaveopt supports it on xsaveopt enabled systems. 3060 3061 noxsaves [X86] Disables xsaves and xrstors used in saving and 3062 restoring x86 extended register state in compacted 3063 form of xsave area. The kernel will fall back to use 3064 xsaveopt and xrstor to save and restore the states 3065 in standard form of xsave area. By using this 3066 parameter, xsave area per process might occupy more 3067 memory on xsaves enabled systems. 3068 3069 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or 3070 wfi(ARM) instruction doesn't work correctly and not to 3071 use it. This is also useful when using JTAG debugger. 3072 3073 no_file_caps Tells the kernel not to honor file capabilities. The 3074 only way then for a file to be executed with privilege 3075 is to be setuid root or executed by root. 3076 3077 nohalt [IA-64] Tells the kernel not to use the power saving 3078 function PAL_HALT_LIGHT when idle. This increases 3079 power-consumption. On the positive side, it reduces 3080 interrupt wake-up latency, which may improve performance 3081 in certain environments such as networked servers or 3082 real-time systems. 3083 3084 nohibernate [HIBERNATION] Disable hibernation and resume. 3085 3086 nohz= [KNL] Boottime enable/disable dynamic ticks 3087 Valid arguments: on, off 3088 Default: on 3089 3090 nohz_full= [KNL,BOOT,SMP,ISOL] 3091 The argument is a cpu list, as described above. 3092 In kernels built with CONFIG_NO_HZ_FULL=y, set 3093 the specified list of CPUs whose tick will be stopped 3094 whenever possible. The boot CPU will be forced outside 3095 the range to maintain the timekeeping. Any CPUs 3096 in this list will have their RCU callbacks offloaded, 3097 just as if they had also been called out in the 3098 rcu_nocbs= boot parameter. 3099 3100 noiotrap [SH] Disables trapped I/O port accesses. 3101 3102 noirqdebug [X86-32] Disables the code which attempts to detect and 3103 disable unhandled interrupt sources. 3104 3105 no_timer_check [X86,APIC] Disables the code which tests for 3106 broken timer IRQ sources. 3107 3108 noisapnp [ISAPNP] Disables ISA PnP code. 3109 3110 noinitrd [RAM] Tells the kernel not to load any configured 3111 initial RAM disk. 3112 3113 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt 3114 remapping. 3115 [Deprecated - use intremap=off] 3116 3117 nointroute [IA-64] 3118 3119 noinvpcid [X86] Disable the INVPCID cpu feature. 3120 3121 nojitter [IA-64] Disables jitter checking for ITC timers. 3122 3123 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver 3124 3125 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page 3126 fault handling. 3127 3128 no-vmw-sched-clock 3129 [X86,PV_OPS] Disable paravirtualized VMware scheduler 3130 clock and use the default one. 3131 3132 no-steal-acc [X86,KVM,ARM64] Disable paravirtualized steal time 3133 accounting. steal time is computed, but won't 3134 influence scheduler behaviour 3135 3136 nolapic [X86-32,APIC] Do not enable or use the local APIC. 3137 3138 nolapic_timer [X86-32,APIC] Do not use the local APIC timer. 3139 3140 noltlbs [PPC] Do not use large page/tlb entries for kernel 3141 lowmem mapping on PPC40x and PPC8xx 3142 3143 nomca [IA-64] Disable machine check abort handling 3144 3145 nomce [X86-32] Disable Machine Check Exception 3146 3147 nomfgpt [X86-32] Disable Multi-Function General Purpose 3148 Timer usage (for AMD Geode machines). 3149 3150 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 3151 shutdown the other cpus. Instead use the REBOOT_VECTOR 3152 irq. 3153 3154 nomodule Disable module load 3155 3156 nopat [X86] Disable PAT (page attribute table extension of 3157 pagetables) support. 3158 3159 nopcid [X86-64] Disable the PCID cpu feature. 3160 3161 norandmaps Don't use address space randomization. Equivalent to 3162 echo 0 > /proc/sys/kernel/randomize_va_space 3163 3164 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 3165 with UP alternatives 3166 3167 nordrand [X86] Disable kernel use of the RDRAND and 3168 RDSEED instructions even if they are supported 3169 by the processor. RDRAND and RDSEED are still 3170 available to user space applications. 3171 3172 noresume [SWSUSP] Disables resume and restores original swap 3173 space. 3174 3175 no-scroll [VGA] Disables scrollback. 3176 This is required for the Braillex ib80-piezo Braille 3177 reader made by F.H. Papenmeier (Germany). 3178 3179 nosbagart [IA-64] 3180 3181 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support. 3182 3183 nosmp [SMP] Tells an SMP kernel to act as a UP kernel, 3184 and disable the IO APIC. legacy for "maxcpus=0". 3185 3186 nosoftlockup [KNL] Disable the soft-lockup detector. 3187 3188 nosync [HW,M68K] Disables sync negotiation for all devices. 3189 3190 nowatchdog [KNL] Disable both lockup detectors, i.e. 3191 soft-lockup and NMI watchdog (hard-lockup). 3192 3193 nowb [ARM] 3194 3195 nox2apic [X86-64,APIC] Do not enable x2APIC mode. 3196 3197 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when 3198 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off. 3199 Some features depend on CPU0. Known dependencies are: 3200 1. Resume from suspend/hibernate depends on CPU0. 3201 Suspend/hibernate will fail if CPU0 is offline and you 3202 need to online CPU0 before suspend/hibernate. 3203 2. PIC interrupts also depend on CPU0. CPU0 can't be 3204 removed if a PIC interrupt is detected. 3205 It's said poweroff/reboot may depend on CPU0 on some 3206 machines although I haven't seen such issues so far 3207 after CPU0 is offline on a few tested machines. 3208 If the dependencies are under your control, you can 3209 turn on cpu0_hotplug. 3210 3211 nps_mtm_hs_ctr= [KNL,ARC] 3212 This parameter sets the maximum duration, in 3213 cycles, each HW thread of the CTOP can run 3214 without interruptions, before HW switches it. 3215 The actual maximum duration is 16 times this 3216 parameter's value. 3217 Format: integer between 1 and 255 3218 Default: 255 3219 3220 nptcg= [IA-64] Override max number of concurrent global TLB 3221 purges which is reported from either PAL_VM_SUMMARY or 3222 SAL PALO. 3223 3224 nr_cpus= [SMP] Maximum number of processors that an SMP kernel 3225 could support. nr_cpus=n : n >= 1 limits the kernel to 3226 support 'n' processors. It could be larger than the 3227 number of already plugged CPU during bootup, later in 3228 runtime you can physically add extra cpu until it reaches 3229 n. So during boot up some boot time memory for per-cpu 3230 variables need be pre-allocated for later physical cpu 3231 hot plugging. 3232 3233 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 3234 3235 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing. 3236 Allowed values are enable and disable 3237 3238 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 3239 'node', 'default' can be specified 3240 This can be set from sysctl after boot. 3241 See Documentation/admin-guide/sysctl/vm.rst for details. 3242 3243 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver. 3244 See Documentation/debugging-via-ohci1394.txt for more 3245 info. 3246 3247 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 3248 Rather than timing out after 20 ms if an EC 3249 command is not properly ACKed, override the length 3250 of the timeout. We have interrupts disabled while 3251 waiting for the ACK, so if this is set too high 3252 interrupts *may* be lost! 3253 3254 omap_mux= [OMAP] Override bootloader pin multiplexing. 3255 Format: <mux_mode0.mode_name=value>... 3256 For example, to override I2C bus2: 3257 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 3258 3259 oprofile.timer= [HW] 3260 Use timer interrupt instead of performance counters 3261 3262 oprofile.cpu_type= Force an oprofile cpu type 3263 This might be useful if you have an older oprofile 3264 userland or if you want common events. 3265 Format: { arch_perfmon } 3266 arch_perfmon: [X86] Force use of architectural 3267 perfmon on Intel CPUs instead of the 3268 CPU specific event set. 3269 timer: [X86] Force use of architectural NMI 3270 timer mode (see also oprofile.timer 3271 for generic hr timer mode) 3272 3273 oops=panic Always panic on oopses. Default is to just kill the 3274 process, but there is a small probability of 3275 deadlocking the machine. 3276 This will also cause panics on machine check exceptions. 3277 Useful together with panic=30 to trigger a reboot. 3278 3279 page_alloc.shuffle= 3280 [KNL] Boolean flag to control whether the page allocator 3281 should randomize its free lists. The randomization may 3282 be automatically enabled if the kernel detects it is 3283 running on a platform with a direct-mapped memory-side 3284 cache, and this parameter can be used to 3285 override/disable that behavior. The state of the flag 3286 can be read from sysfs at: 3287 /sys/module/page_alloc/parameters/shuffle. 3288 3289 page_owner= [KNL] Boot-time page_owner enabling option. 3290 Storage of the information about who allocated 3291 each page is disabled in default. With this switch, 3292 we can turn it on. 3293 on: enable the feature 3294 3295 page_poison= [KNL] Boot-time parameter changing the state of 3296 poisoning on the buddy allocator, available with 3297 CONFIG_PAGE_POISONING=y. 3298 off: turn off poisoning (default) 3299 on: turn on poisoning 3300 3301 panic= [KNL] Kernel behaviour on panic: delay <timeout> 3302 timeout > 0: seconds before rebooting 3303 timeout = 0: wait forever 3304 timeout < 0: reboot immediately 3305 Format: <timeout> 3306 3307 panic_print= Bitmask for printing system info when panic happens. 3308 User can chose combination of the following bits: 3309 bit 0: print all tasks info 3310 bit 1: print system memory info 3311 bit 2: print timer info 3312 bit 3: print locks info if CONFIG_LOCKDEP is on 3313 bit 4: print ftrace buffer 3314 bit 5: print all printk messages in buffer 3315 3316 panic_on_warn panic() instead of WARN(). Useful to cause kdump 3317 on a WARN(). 3318 3319 crash_kexec_post_notifiers 3320 Run kdump after running panic-notifiers and dumping 3321 kmsg. This only for the users who doubt kdump always 3322 succeeds in any situation. 3323 Note that this also increases risks of kdump failure, 3324 because some panic notifiers can make the crashed 3325 kernel more unstable. 3326 3327 parkbd.port= [HW] Parallel port number the keyboard adapter is 3328 connected to, default is 0. 3329 Format: <parport#> 3330 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 3331 0 for XT, 1 for AT (default is AT). 3332 Format: <mode> 3333 3334 parport= [HW,PPT] Specify parallel ports. 0 disables. 3335 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 3336 Use 'auto' to force the driver to use any 3337 IRQ/DMA settings detected (the default is to 3338 ignore detected IRQ/DMA settings because of 3339 possible conflicts). You can specify the base 3340 address, IRQ, and DMA settings; IRQ and DMA 3341 should be numbers, or 'auto' (for using detected 3342 settings on that particular port), or 'nofifo' 3343 (to avoid using a FIFO even if it is detected). 3344 Parallel ports are assigned in the order they 3345 are specified on the command line, starting 3346 with parport0. 3347 3348 parport_init_mode= [HW,PPT] 3349 Configure VIA parallel port to operate in 3350 a specific mode. This is necessary on Pegasos 3351 computer where firmware has no options for setting 3352 up parallel port mode and sets it to spp. 3353 Currently this function knows 686a and 8231 chips. 3354 Format: [spp|ps2|epp|ecp|ecpepp] 3355 3356 pause_on_oops= 3357 Halt all CPUs after the first oops has been printed for 3358 the specified number of seconds. This is to be used if 3359 your oopses keep scrolling off the screen. 3360 3361 pcbit= [HW,ISDN] 3362 3363 pcd. [PARIDE] 3364 See header of drivers/block/paride/pcd.c. 3365 See also Documentation/admin-guide/blockdev/paride.rst. 3366 3367 pci=option[,option...] [PCI] various PCI subsystem options. 3368 3369 Some options herein operate on a specific device 3370 or a set of devices (<pci_dev>). These are 3371 specified in one of the following formats: 3372 3373 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 3374 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 3375 3376 Note: the first format specifies a PCI 3377 bus/device/function address which may change 3378 if new hardware is inserted, if motherboard 3379 firmware changes, or due to changes caused 3380 by other kernel parameters. If the 3381 domain is left unspecified, it is 3382 taken to be zero. Optionally, a path 3383 to a device through multiple device/function 3384 addresses can be specified after the base 3385 address (this is more robust against 3386 renumbering issues). The second format 3387 selects devices using IDs from the 3388 configuration space which may match multiple 3389 devices in the system. 3390 3391 earlydump dump PCI config space before the kernel 3392 changes anything 3393 off [X86] don't probe for the PCI bus 3394 bios [X86-32] force use of PCI BIOS, don't access 3395 the hardware directly. Use this if your machine 3396 has a non-standard PCI host bridge. 3397 nobios [X86-32] disallow use of PCI BIOS, only direct 3398 hardware access methods are allowed. Use this 3399 if you experience crashes upon bootup and you 3400 suspect they are caused by the BIOS. 3401 conf1 [X86] Force use of PCI Configuration Access 3402 Mechanism 1 (config address in IO port 0xCF8, 3403 data in IO port 0xCFC, both 32-bit). 3404 conf2 [X86] Force use of PCI Configuration Access 3405 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 3406 the function, IO port 0xCFA, also 8-bit, sets 3407 bus number. The config space is then accessed 3408 through ports 0xC000-0xCFFF). 3409 See http://wiki.osdev.org/PCI for more info 3410 on the configuration access mechanisms. 3411 noaer [PCIE] If the PCIEAER kernel config parameter is 3412 enabled, this kernel boot option can be used to 3413 disable the use of PCIE advanced error reporting. 3414 nodomains [PCI] Disable support for multiple PCI 3415 root domains (aka PCI segments, in ACPI-speak). 3416 nommconf [X86] Disable use of MMCONFIG for PCI 3417 Configuration 3418 check_enable_amd_mmconf [X86] check for and enable 3419 properly configured MMIO access to PCI 3420 config space on AMD family 10h CPU 3421 nomsi [MSI] If the PCI_MSI kernel config parameter is 3422 enabled, this kernel boot option can be used to 3423 disable the use of MSI interrupts system-wide. 3424 noioapicquirk [APIC] Disable all boot interrupt quirks. 3425 Safety option to keep boot IRQs enabled. This 3426 should never be necessary. 3427 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 3428 primary IO-APIC for bridges that cannot disable 3429 boot IRQs. This fixes a source of spurious IRQs 3430 when the system masks IRQs. 3431 noioapicreroute [APIC] Disable workaround that uses the 3432 boot IRQ equivalent of an IRQ that connects to 3433 a chipset where boot IRQs cannot be disabled. 3434 The opposite of ioapicreroute. 3435 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 3436 routing table. These calls are known to be buggy 3437 on several machines and they hang the machine 3438 when used, but on other computers it's the only 3439 way to get the interrupt routing table. Try 3440 this option if the kernel is unable to allocate 3441 IRQs or discover secondary PCI buses on your 3442 motherboard. 3443 rom [X86] Assign address space to expansion ROMs. 3444 Use with caution as certain devices share 3445 address decoders between ROMs and other 3446 resources. 3447 norom [X86] Do not assign address space to 3448 expansion ROMs that do not already have 3449 BIOS assigned address ranges. 3450 nobar [X86] Do not assign address space to the 3451 BARs that weren't assigned by the BIOS. 3452 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 3453 assigned automatically to PCI devices. You can 3454 make the kernel exclude IRQs of your ISA cards 3455 this way. 3456 pirqaddr=0xAAAAA [X86] Specify the physical address 3457 of the PIRQ table (normally generated 3458 by the BIOS) if it is outside the 3459 F0000h-100000h range. 3460 lastbus=N [X86] Scan all buses thru bus #N. Can be 3461 useful if the kernel is unable to find your 3462 secondary buses and you want to tell it 3463 explicitly which ones they are. 3464 assign-busses [X86] Always assign all PCI bus 3465 numbers ourselves, overriding 3466 whatever the firmware may have done. 3467 usepirqmask [X86] Honor the possible IRQ mask stored 3468 in the BIOS $PIR table. This is needed on 3469 some systems with broken BIOSes, notably 3470 some HP Pavilion N5400 and Omnibook XE3 3471 notebooks. This will have no effect if ACPI 3472 IRQ routing is enabled. 3473 noacpi [X86] Do not use ACPI for IRQ routing 3474 or for PCI scanning. 3475 use_crs [X86] Use PCI host bridge window information 3476 from ACPI. On BIOSes from 2008 or later, this 3477 is enabled by default. If you need to use this, 3478 please report a bug. 3479 nocrs [X86] Ignore PCI host bridge windows from ACPI. 3480 If you need to use this, please report a bug. 3481 routeirq Do IRQ routing for all PCI devices. 3482 This is normally done in pci_enable_device(), 3483 so this option is a temporary workaround 3484 for broken drivers that don't call it. 3485 skip_isa_align [X86] do not align io start addr, so can 3486 handle more pci cards 3487 noearly [X86] Don't do any early type 1 scanning. 3488 This might help on some broken boards which 3489 machine check when some devices' config space 3490 is read. But various workarounds are disabled 3491 and some IOMMU drivers will not work. 3492 bfsort Sort PCI devices into breadth-first order. 3493 This sorting is done to get a device 3494 order compatible with older (<= 2.4) kernels. 3495 nobfsort Don't sort PCI devices into breadth-first order. 3496 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 3497 tuning and use the BIOS-configured MPS defaults. 3498 pcie_bus_safe Set every device's MPS to the largest value 3499 supported by all devices below the root complex. 3500 pcie_bus_perf Set device MPS to the largest allowable MPS 3501 based on its parent bus. Also set MRRS (Max 3502 Read Request Size) to the largest supported 3503 value (no larger than the MPS that the device 3504 or bus can support) for best performance. 3505 pcie_bus_peer2peer Set every device's MPS to 128B, which 3506 every device is guaranteed to support. This 3507 configuration allows peer-to-peer DMA between 3508 any pair of devices, possibly at the cost of 3509 reduced performance. This also guarantees 3510 that hot-added devices will work. 3511 cbiosize=nn[KMG] The fixed amount of bus space which is 3512 reserved for the CardBus bridge's IO window. 3513 The default value is 256 bytes. 3514 cbmemsize=nn[KMG] The fixed amount of bus space which is 3515 reserved for the CardBus bridge's memory 3516 window. The default value is 64 megabytes. 3517 resource_alignment= 3518 Format: 3519 [<order of align>@]<pci_dev>[; ...] 3520 Specifies alignment and device to reassign 3521 aligned memory resources. How to 3522 specify the device is described above. 3523 If <order of align> is not specified, 3524 PAGE_SIZE is used as alignment. 3525 A PCI-PCI bridge can be specified if resource 3526 windows need to be expanded. 3527 To specify the alignment for several 3528 instances of a device, the PCI vendor, 3529 device, subvendor, and subdevice may be 3530 specified, e.g., 12@pci:8086:9c22:103c:198f 3531 for 4096-byte alignment. 3532 ecrc= Enable/disable PCIe ECRC (transaction layer 3533 end-to-end CRC checking). 3534 bios: Use BIOS/firmware settings. This is the 3535 the default. 3536 off: Turn ECRC off 3537 on: Turn ECRC on. 3538 hpiosize=nn[KMG] The fixed amount of bus space which is 3539 reserved for hotplug bridge's IO window. 3540 Default size is 256 bytes. 3541 hpmemsize=nn[KMG] The fixed amount of bus space which is 3542 reserved for hotplug bridge's memory window. 3543 Default size is 2 megabytes. 3544 hpbussize=nn The minimum amount of additional bus numbers 3545 reserved for buses below a hotplug bridge. 3546 Default is 1. 3547 realloc= Enable/disable reallocating PCI bridge resources 3548 if allocations done by BIOS are too small to 3549 accommodate resources required by all child 3550 devices. 3551 off: Turn realloc off 3552 on: Turn realloc on 3553 realloc same as realloc=on 3554 noari do not use PCIe ARI. 3555 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 3556 do not use PCIe ATS (and IOMMU device IOTLB). 3557 pcie_scan_all Scan all possible PCIe devices. Otherwise we 3558 only look for one device below a PCIe downstream 3559 port. 3560 big_root_window Try to add a big 64bit memory window to the PCIe 3561 root complex on AMD CPUs. Some GFX hardware 3562 can resize a BAR to allow access to all VRAM. 3563 Adding the window is slightly risky (it may 3564 conflict with unreported devices), so this 3565 taints the kernel. 3566 disable_acs_redir=<pci_dev>[; ...] 3567 Specify one or more PCI devices (in the format 3568 specified above) separated by semicolons. 3569 Each device specified will have the PCI ACS 3570 redirect capabilities forced off which will 3571 allow P2P traffic between devices through 3572 bridges without forcing it upstream. Note: 3573 this removes isolation between devices and 3574 may put more devices in an IOMMU group. 3575 force_floating [S390] Force usage of floating interrupts. 3576 nomio [S390] Do not use MIO instructions. 3577 3578 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power 3579 Management. 3580 off Disable ASPM. 3581 force Enable ASPM even on devices that claim not to support it. 3582 WARNING: Forcing ASPM on may cause system lockups. 3583 3584 pcie_ports= [PCIE] PCIe port services handling: 3585 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 3586 even if the platform doesn't give the OS permission to 3587 use them. This may cause conflicts if the platform 3588 also tries to use these services. 3589 compat Disable native PCIe services (PME, AER, DPC, PCIe 3590 hotplug). 3591 3592 pcie_port_pm= [PCIE] PCIe port power management handling: 3593 off Disable power management of all PCIe ports 3594 force Forcibly enable power management of all PCIe ports 3595 3596 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 3597 nomsi Do not use MSI for native PCIe PME signaling (this makes 3598 all PCIe root ports use INTx for all services). 3599 3600 pcmv= [HW,PCMCIA] BadgePAD 4 3601 3602 pd_ignore_unused 3603 [PM] 3604 Keep all power-domains already enabled by bootloader on, 3605 even if no driver has claimed them. This is useful 3606 for debug and development, but should not be 3607 needed on a platform with proper driver support. 3608 3609 pd. [PARIDE] 3610 See Documentation/admin-guide/blockdev/paride.rst. 3611 3612 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 3613 boot time. 3614 Format: { 0 | 1 } 3615 See arch/parisc/kernel/pdc_chassis.c 3616 3617 percpu_alloc= Select which percpu first chunk allocator to use. 3618 Currently supported values are "embed" and "page". 3619 Archs may support subset or none of the selections. 3620 See comments in mm/percpu.c for details on each 3621 allocator. This parameter is primarily for debugging 3622 and performance comparison. 3623 3624 pf. [PARIDE] 3625 See Documentation/admin-guide/blockdev/paride.rst. 3626 3627 pg. [PARIDE] 3628 See Documentation/admin-guide/blockdev/paride.rst. 3629 3630 pirq= [SMP,APIC] Manual mp-table setup 3631 See Documentation/x86/i386/IO-APIC.rst. 3632 3633 plip= [PPT,NET] Parallel port network link 3634 Format: { parport<nr> | timid | 0 } 3635 See also Documentation/admin-guide/parport.rst. 3636 3637 pmtmr= [X86] Manual setup of pmtmr I/O Port. 3638 Override pmtimer IOPort with a hex value. 3639 e.g. pmtmr=0x508 3640 3641 pnp.debug=1 [PNP] 3642 Enable PNP debug messages (depends on the 3643 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 3644 via /sys/module/pnp/parameters/debug. We always show 3645 current resource usage; turning this on also shows 3646 possible settings and some assignment information. 3647 3648 pnpacpi= [ACPI] 3649 { off } 3650 3651 pnpbios= [ISAPNP] 3652 { on | off | curr | res | no-curr | no-res } 3653 3654 pnp_reserve_irq= 3655 [ISAPNP] Exclude IRQs for the autoconfiguration 3656 3657 pnp_reserve_dma= 3658 [ISAPNP] Exclude DMAs for the autoconfiguration 3659 3660 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 3661 Ranges are in pairs (I/O port base and size). 3662 3663 pnp_reserve_mem= 3664 [ISAPNP] Exclude memory regions for the 3665 autoconfiguration. 3666 Ranges are in pairs (memory base and size). 3667 3668 ports= [IP_VS_FTP] IPVS ftp helper module 3669 Default is 21. 3670 Up to 8 (IP_VS_APP_MAX_PORTS) ports 3671 may be specified. 3672 Format: <port>,<port>.... 3673 3674 powersave=off [PPC] This option disables power saving features. 3675 It specifically disables cpuidle and sets the 3676 platform machine description specific power_save 3677 function to NULL. On Idle the CPU just reduces 3678 execution priority. 3679 3680 ppc_strict_facility_enable 3681 [PPC] This option catches any kernel floating point, 3682 Altivec, VSX and SPE outside of regions specifically 3683 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 3684 There is some performance impact when enabling this. 3685 3686 ppc_tm= [PPC] 3687 Format: {"off"} 3688 Disable Hardware Transactional Memory 3689 3690 print-fatal-signals= 3691 [KNL] debug: print fatal signals 3692 3693 If enabled, warn about various signal handling 3694 related application anomalies: too many signals, 3695 too many POSIX.1 timers, fatal signals causing a 3696 coredump - etc. 3697 3698 If you hit the warning due to signal overflow, 3699 you might want to try "ulimit -i unlimited". 3700 3701 default: off. 3702 3703 printk.always_kmsg_dump= 3704 Trigger kmsg_dump for cases other than kernel oops or 3705 panics 3706 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 3707 default: disabled 3708 3709 printk.devkmsg={on,off,ratelimit} 3710 Control writing to /dev/kmsg. 3711 on - unlimited logging to /dev/kmsg from userspace 3712 off - logging to /dev/kmsg disabled 3713 ratelimit - ratelimit the logging 3714 Default: ratelimit 3715 3716 printk.time= Show timing data prefixed to each printk message line 3717 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 3718 3719 processor.max_cstate= [HW,ACPI] 3720 Limit processor to maximum C-state 3721 max_cstate=9 overrides any DMI blacklist limit. 3722 3723 processor.nocst [HW,ACPI] 3724 Ignore the _CST method to determine C-states, 3725 instead using the legacy FADT method 3726 3727 profile= [KNL] Enable kernel profiling via /proc/profile 3728 Format: [<profiletype>,]<number> 3729 Param: <profiletype>: "schedule", "sleep", or "kvm" 3730 [defaults to kernel profiling] 3731 Param: "schedule" - profile schedule points. 3732 Param: "sleep" - profile D-state sleeping (millisecs). 3733 Requires CONFIG_SCHEDSTATS 3734 Param: "kvm" - profile VM exits. 3735 Param: <number> - step/bucket size as a power of 2 for 3736 statistical time based profiling. 3737 3738 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk 3739 before loading. 3740 See Documentation/admin-guide/blockdev/ramdisk.rst. 3741 3742 psi= [KNL] Enable or disable pressure stall information 3743 tracking. 3744 Format: <bool> 3745 3746 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 3747 probe for; one of (bare|imps|exps|lifebook|any). 3748 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 3749 per second. 3750 psmouse.resetafter= [HW,MOUSE] 3751 Try to reset the device after so many bad packets 3752 (0 = never). 3753 psmouse.resolution= 3754 [HW,MOUSE] Set desired mouse resolution, in dpi. 3755 psmouse.smartscroll= 3756 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 3757 0 = disabled, 1 = enabled (default). 3758 3759 pstore.backend= Specify the name of the pstore backend to use 3760 3761 pt. [PARIDE] 3762 See Documentation/admin-guide/blockdev/paride.rst. 3763 3764 pti= [X86_64] Control Page Table Isolation of user and 3765 kernel address spaces. Disabling this feature 3766 removes hardening, but improves performance of 3767 system calls and interrupts. 3768 3769 on - unconditionally enable 3770 off - unconditionally disable 3771 auto - kernel detects whether your CPU model is 3772 vulnerable to issues that PTI mitigates 3773 3774 Not specifying this option is equivalent to pti=auto. 3775 3776 nopti [X86_64] 3777 Equivalent to pti=off 3778 3779 pty.legacy_count= 3780 [KNL] Number of legacy pty's. Overwrites compiled-in 3781 default number. 3782 3783 quiet [KNL] Disable most log messages 3784 3785 r128= [HW,DRM] 3786 3787 raid= [HW,RAID] 3788 See Documentation/admin-guide/md.rst. 3789 3790 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 3791 See Documentation/admin-guide/blockdev/ramdisk.rst. 3792 3793 random.trust_cpu={on,off} 3794 [KNL] Enable or disable trusting the use of the 3795 CPU's random number generator (if available) to 3796 fully seed the kernel's CRNG. Default is controlled 3797 by CONFIG_RANDOM_TRUST_CPU. 3798 3799 ras=option[,option,...] [KNL] RAS-specific options 3800 3801 cec_disable [X86] 3802 Disable the Correctable Errors Collector, 3803 see CONFIG_RAS_CEC help text. 3804 3805 rcu_nocbs= [KNL] 3806 The argument is a cpu list, as described above, 3807 except that the string "all" can be used to 3808 specify every CPU on the system. 3809 3810 In kernels built with CONFIG_RCU_NOCB_CPU=y, set 3811 the specified list of CPUs to be no-callback CPUs. 3812 Invocation of these CPUs' RCU callbacks will be 3813 offloaded to "rcuox/N" kthreads created for that 3814 purpose, where "x" is "p" for RCU-preempt, and 3815 "s" for RCU-sched, and "N" is the CPU number. 3816 This reduces OS jitter on the offloaded CPUs, 3817 which can be useful for HPC and real-time 3818 workloads. It can also improve energy efficiency 3819 for asymmetric multiprocessors. 3820 3821 rcu_nocb_poll [KNL] 3822 Rather than requiring that offloaded CPUs 3823 (specified by rcu_nocbs= above) explicitly 3824 awaken the corresponding "rcuoN" kthreads, 3825 make these kthreads poll for callbacks. 3826 This improves the real-time response for the 3827 offloaded CPUs by relieving them of the need to 3828 wake up the corresponding kthread, but degrades 3829 energy efficiency by requiring that the kthreads 3830 periodically wake up to do the polling. 3831 3832 rcutree.blimit= [KNL] 3833 Set maximum number of finished RCU callbacks to 3834 process in one batch. 3835 3836 rcutree.dump_tree= [KNL] 3837 Dump the structure of the rcu_node combining tree 3838 out at early boot. This is used for diagnostic 3839 purposes, to verify correct tree setup. 3840 3841 rcutree.gp_cleanup_delay= [KNL] 3842 Set the number of jiffies to delay each step of 3843 RCU grace-period cleanup. 3844 3845 rcutree.gp_init_delay= [KNL] 3846 Set the number of jiffies to delay each step of 3847 RCU grace-period initialization. 3848 3849 rcutree.gp_preinit_delay= [KNL] 3850 Set the number of jiffies to delay each step of 3851 RCU grace-period pre-initialization, that is, 3852 the propagation of recent CPU-hotplug changes up 3853 the rcu_node combining tree. 3854 3855 rcutree.use_softirq= [KNL] 3856 If set to zero, move all RCU_SOFTIRQ processing to 3857 per-CPU rcuc kthreads. Defaults to a non-zero 3858 value, meaning that RCU_SOFTIRQ is used by default. 3859 Specify rcutree.use_softirq=0 to use rcuc kthreads. 3860 3861 rcutree.rcu_fanout_exact= [KNL] 3862 Disable autobalancing of the rcu_node combining 3863 tree. This is used by rcutorture, and might 3864 possibly be useful for architectures having high 3865 cache-to-cache transfer latencies. 3866 3867 rcutree.rcu_fanout_leaf= [KNL] 3868 Change the number of CPUs assigned to each 3869 leaf rcu_node structure. Useful for very 3870 large systems, which will choose the value 64, 3871 and for NUMA systems with large remote-access 3872 latencies, which will choose a value aligned 3873 with the appropriate hardware boundaries. 3874 3875 rcutree.jiffies_till_first_fqs= [KNL] 3876 Set delay from grace-period initialization to 3877 first attempt to force quiescent states. 3878 Units are jiffies, minimum value is zero, 3879 and maximum value is HZ. 3880 3881 rcutree.jiffies_till_next_fqs= [KNL] 3882 Set delay between subsequent attempts to force 3883 quiescent states. Units are jiffies, minimum 3884 value is one, and maximum value is HZ. 3885 3886 rcutree.jiffies_till_sched_qs= [KNL] 3887 Set required age in jiffies for a 3888 given grace period before RCU starts 3889 soliciting quiescent-state help from 3890 rcu_note_context_switch() and cond_resched(). 3891 If not specified, the kernel will calculate 3892 a value based on the most recent settings 3893 of rcutree.jiffies_till_first_fqs 3894 and rcutree.jiffies_till_next_fqs. 3895 This calculated value may be viewed in 3896 rcutree.jiffies_to_sched_qs. Any attempt to set 3897 rcutree.jiffies_to_sched_qs will be cheerfully 3898 overwritten. 3899 3900 rcutree.kthread_prio= [KNL,BOOT] 3901 Set the SCHED_FIFO priority of the RCU per-CPU 3902 kthreads (rcuc/N). This value is also used for 3903 the priority of the RCU boost threads (rcub/N) 3904 and for the RCU grace-period kthreads (rcu_bh, 3905 rcu_preempt, and rcu_sched). If RCU_BOOST is 3906 set, valid values are 1-99 and the default is 1 3907 (the least-favored priority). Otherwise, when 3908 RCU_BOOST is not set, valid values are 0-99 and 3909 the default is zero (non-realtime operation). 3910 3911 rcutree.rcu_nocb_gp_stride= [KNL] 3912 Set the number of NOCB callback kthreads in 3913 each group, which defaults to the square root 3914 of the number of CPUs. Larger numbers reduce 3915 the wakeup overhead on the global grace-period 3916 kthread, but increases that same overhead on 3917 each group's NOCB grace-period kthread. 3918 3919 rcutree.qhimark= [KNL] 3920 Set threshold of queued RCU callbacks beyond which 3921 batch limiting is disabled. 3922 3923 rcutree.qlowmark= [KNL] 3924 Set threshold of queued RCU callbacks below which 3925 batch limiting is re-enabled. 3926 3927 rcutree.rcu_idle_gp_delay= [KNL] 3928 Set wakeup interval for idle CPUs that have 3929 RCU callbacks (RCU_FAST_NO_HZ=y). 3930 3931 rcutree.rcu_idle_lazy_gp_delay= [KNL] 3932 Set wakeup interval for idle CPUs that have 3933 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y). 3934 Lazy RCU callbacks are those which RCU can 3935 prove do nothing more than free memory. 3936 3937 rcutree.rcu_kick_kthreads= [KNL] 3938 Cause the grace-period kthread to get an extra 3939 wake_up() if it sleeps three times longer than 3940 it should at force-quiescent-state time. 3941 This wake_up() will be accompanied by a 3942 WARN_ONCE() splat and an ftrace_dump(). 3943 3944 rcutree.sysrq_rcu= [KNL] 3945 Commandeer a sysrq key to dump out Tree RCU's 3946 rcu_node tree with an eye towards determining 3947 why a new grace period has not yet started. 3948 3949 rcuperf.gp_async= [KNL] 3950 Measure performance of asynchronous 3951 grace-period primitives such as call_rcu(). 3952 3953 rcuperf.gp_async_max= [KNL] 3954 Specify the maximum number of outstanding 3955 callbacks per writer thread. When a writer 3956 thread exceeds this limit, it invokes the 3957 corresponding flavor of rcu_barrier() to allow 3958 previously posted callbacks to drain. 3959 3960 rcuperf.gp_exp= [KNL] 3961 Measure performance of expedited synchronous 3962 grace-period primitives. 3963 3964 rcuperf.holdoff= [KNL] 3965 Set test-start holdoff period. The purpose of 3966 this parameter is to delay the start of the 3967 test until boot completes in order to avoid 3968 interference. 3969 3970 rcuperf.nreaders= [KNL] 3971 Set number of RCU readers. The value -1 selects 3972 N, where N is the number of CPUs. A value 3973 "n" less than -1 selects N-n+1, where N is again 3974 the number of CPUs. For example, -2 selects N 3975 (the number of CPUs), -3 selects N+1, and so on. 3976 A value of "n" less than or equal to -N selects 3977 a single reader. 3978 3979 rcuperf.nwriters= [KNL] 3980 Set number of RCU writers. The values operate 3981 the same as for rcuperf.nreaders. 3982 N, where N is the number of CPUs 3983 3984 rcuperf.perf_type= [KNL] 3985 Specify the RCU implementation to test. 3986 3987 rcuperf.shutdown= [KNL] 3988 Shut the system down after performance tests 3989 complete. This is useful for hands-off automated 3990 testing. 3991 3992 rcuperf.verbose= [KNL] 3993 Enable additional printk() statements. 3994 3995 rcuperf.writer_holdoff= [KNL] 3996 Write-side holdoff between grace periods, 3997 in microseconds. The default of zero says 3998 no holdoff. 3999 4000 rcutorture.fqs_duration= [KNL] 4001 Set duration of force_quiescent_state bursts 4002 in microseconds. 4003 4004 rcutorture.fqs_holdoff= [KNL] 4005 Set holdoff time within force_quiescent_state bursts 4006 in microseconds. 4007 4008 rcutorture.fqs_stutter= [KNL] 4009 Set wait time between force_quiescent_state bursts 4010 in seconds. 4011 4012 rcutorture.fwd_progress= [KNL] 4013 Enable RCU grace-period forward-progress testing 4014 for the types of RCU supporting this notion. 4015 4016 rcutorture.fwd_progress_div= [KNL] 4017 Specify the fraction of a CPU-stall-warning 4018 period to do tight-loop forward-progress testing. 4019 4020 rcutorture.fwd_progress_holdoff= [KNL] 4021 Number of seconds to wait between successive 4022 forward-progress tests. 4023 4024 rcutorture.fwd_progress_need_resched= [KNL] 4025 Enclose cond_resched() calls within checks for 4026 need_resched() during tight-loop forward-progress 4027 testing. 4028 4029 rcutorture.gp_cond= [KNL] 4030 Use conditional/asynchronous update-side 4031 primitives, if available. 4032 4033 rcutorture.gp_exp= [KNL] 4034 Use expedited update-side primitives, if available. 4035 4036 rcutorture.gp_normal= [KNL] 4037 Use normal (non-expedited) asynchronous 4038 update-side primitives, if available. 4039 4040 rcutorture.gp_sync= [KNL] 4041 Use normal (non-expedited) synchronous 4042 update-side primitives, if available. If all 4043 of rcutorture.gp_cond=, rcutorture.gp_exp=, 4044 rcutorture.gp_normal=, and rcutorture.gp_sync= 4045 are zero, rcutorture acts as if is interpreted 4046 they are all non-zero. 4047 4048 rcutorture.n_barrier_cbs= [KNL] 4049 Set callbacks/threads for rcu_barrier() testing. 4050 4051 rcutorture.nfakewriters= [KNL] 4052 Set number of concurrent RCU writers. These just 4053 stress RCU, they don't participate in the actual 4054 test, hence the "fake". 4055 4056 rcutorture.nreaders= [KNL] 4057 Set number of RCU readers. The value -1 selects 4058 N-1, where N is the number of CPUs. A value 4059 "n" less than -1 selects N-n-2, where N is again 4060 the number of CPUs. For example, -2 selects N 4061 (the number of CPUs), -3 selects N+1, and so on. 4062 4063 rcutorture.object_debug= [KNL] 4064 Enable debug-object double-call_rcu() testing. 4065 4066 rcutorture.onoff_holdoff= [KNL] 4067 Set time (s) after boot for CPU-hotplug testing. 4068 4069 rcutorture.onoff_interval= [KNL] 4070 Set time (jiffies) between CPU-hotplug operations, 4071 or zero to disable CPU-hotplug testing. 4072 4073 rcutorture.shuffle_interval= [KNL] 4074 Set task-shuffle interval (s). Shuffling tasks 4075 allows some CPUs to go into dyntick-idle mode 4076 during the rcutorture test. 4077 4078 rcutorture.shutdown_secs= [KNL] 4079 Set time (s) after boot system shutdown. This 4080 is useful for hands-off automated testing. 4081 4082 rcutorture.stall_cpu= [KNL] 4083 Duration of CPU stall (s) to test RCU CPU stall 4084 warnings, zero to disable. 4085 4086 rcutorture.stall_cpu_holdoff= [KNL] 4087 Time to wait (s) after boot before inducing stall. 4088 4089 rcutorture.stall_cpu_irqsoff= [KNL] 4090 Disable interrupts while stalling if set. 4091 4092 rcutorture.stat_interval= [KNL] 4093 Time (s) between statistics printk()s. 4094 4095 rcutorture.stutter= [KNL] 4096 Time (s) to stutter testing, for example, specifying 4097 five seconds causes the test to run for five seconds, 4098 wait for five seconds, and so on. This tests RCU's 4099 ability to transition abruptly to and from idle. 4100 4101 rcutorture.test_boost= [KNL] 4102 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 4103 "Maybe" means test if the RCU implementation 4104 under test support RCU priority boosting. 4105 4106 rcutorture.test_boost_duration= [KNL] 4107 Duration (s) of each individual boost test. 4108 4109 rcutorture.test_boost_interval= [KNL] 4110 Interval (s) between each boost test. 4111 4112 rcutorture.test_no_idle_hz= [KNL] 4113 Test RCU's dyntick-idle handling. See also the 4114 rcutorture.shuffle_interval parameter. 4115 4116 rcutorture.torture_type= [KNL] 4117 Specify the RCU implementation to test. 4118 4119 rcutorture.verbose= [KNL] 4120 Enable additional printk() statements. 4121 4122 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 4123 Dump ftrace buffer after reporting RCU CPU 4124 stall warning. 4125 4126 rcupdate.rcu_cpu_stall_suppress= [KNL] 4127 Suppress RCU CPU stall warning messages. 4128 4129 rcupdate.rcu_cpu_stall_timeout= [KNL] 4130 Set timeout for RCU CPU stall warning messages. 4131 4132 rcupdate.rcu_expedited= [KNL] 4133 Use expedited grace-period primitives, for 4134 example, synchronize_rcu_expedited() instead 4135 of synchronize_rcu(). This reduces latency, 4136 but can increase CPU utilization, degrade 4137 real-time latency, and degrade energy efficiency. 4138 No effect on CONFIG_TINY_RCU kernels. 4139 4140 rcupdate.rcu_normal= [KNL] 4141 Use only normal grace-period primitives, 4142 for example, synchronize_rcu() instead of 4143 synchronize_rcu_expedited(). This improves 4144 real-time latency, CPU utilization, and 4145 energy efficiency, but can expose users to 4146 increased grace-period latency. This parameter 4147 overrides rcupdate.rcu_expedited. No effect on 4148 CONFIG_TINY_RCU kernels. 4149 4150 rcupdate.rcu_normal_after_boot= [KNL] 4151 Once boot has completed (that is, after 4152 rcu_end_inkernel_boot() has been invoked), use 4153 only normal grace-period primitives. No effect 4154 on CONFIG_TINY_RCU kernels. 4155 4156 rcupdate.rcu_task_stall_timeout= [KNL] 4157 Set timeout in jiffies for RCU task stall warning 4158 messages. Disable with a value less than or equal 4159 to zero. 4160 4161 rcupdate.rcu_self_test= [KNL] 4162 Run the RCU early boot self tests 4163 4164 rdinit= [KNL] 4165 Format: <full_path> 4166 Run specified binary instead of /init from the ramdisk, 4167 used for early userspace startup. See initrd. 4168 4169 rdrand= [X86] 4170 force - Override the decision by the kernel to hide the 4171 advertisement of RDRAND support (this affects 4172 certain AMD processors because of buggy BIOS 4173 support, specifically around the suspend/resume 4174 path). 4175 4176 rdt= [HW,X86,RDT] 4177 Turn on/off individual RDT features. List is: 4178 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 4179 mba. 4180 E.g. to turn on cmt and turn off mba use: 4181 rdt=cmt,!mba 4182 4183 reboot= [KNL] 4184 Format (x86 or x86_64): 4185 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \ 4186 [[,]s[mp]#### \ 4187 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 4188 [[,]f[orce] 4189 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 4190 (prefix with 'panic_' to set mode for panic 4191 reboot only), 4192 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 4193 reboot_force is either force or not specified, 4194 reboot_cpu is s[mp]#### with #### being the processor 4195 to be used for rebooting. 4196 4197 relax_domain_level= 4198 [KNL, SMP] Set scheduler's default relax_domain_level. 4199 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 4200 4201 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 4202 Format: <base1>,<size1>[,<base2>,<size2>,...] 4203 Reserve I/O ports or memory so the kernel won't use 4204 them. If <base> is less than 0x10000, the region 4205 is assumed to be I/O ports; otherwise it is memory. 4206 4207 reservetop= [X86-32] 4208 Format: nn[KMG] 4209 Reserves a hole at the top of the kernel virtual 4210 address space. 4211 4212 reservelow= [X86] 4213 Format: nn[K] 4214 Set the amount of memory to reserve for BIOS at 4215 the bottom of the address space. 4216 4217 reset_devices [KNL] Force drivers to reset the underlying device 4218 during initialization. 4219 4220 resume= [SWSUSP] 4221 Specify the partition device for software suspend 4222 Format: 4223 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 4224 4225 resume_offset= [SWSUSP] 4226 Specify the offset from the beginning of the partition 4227 given by "resume=" at which the swap header is located, 4228 in <PAGE_SIZE> units (needed only for swap files). 4229 See Documentation/power/swsusp-and-swap-files.rst 4230 4231 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 4232 read the resume files 4233 4234 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 4235 Useful for devices that are detected asynchronously 4236 (e.g. USB and MMC devices). 4237 4238 hibernate= [HIBERNATION] 4239 noresume Don't check if there's a hibernation image 4240 present during boot. 4241 nocompress Don't compress/decompress hibernation images. 4242 no Disable hibernation and resume. 4243 protect_image Turn on image protection during restoration 4244 (that will set all pages holding image data 4245 during restoration read-only). 4246 4247 retain_initrd [RAM] Keep initrd memory after extraction 4248 4249 rfkill.default_state= 4250 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 4251 etc. communication is blocked by default. 4252 1 Unblocked. 4253 4254 rfkill.master_switch_mode= 4255 0 The "airplane mode" button does nothing. 4256 1 The "airplane mode" button toggles between everything 4257 blocked and the previous configuration. 4258 2 The "airplane mode" button toggles between everything 4259 blocked and everything unblocked. 4260 4261 rhash_entries= [KNL,NET] 4262 Set number of hash buckets for route cache 4263 4264 ring3mwait=disable 4265 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 4266 CPUs. 4267 4268 ro [KNL] Mount root device read-only on boot 4269 4270 rodata= [KNL] 4271 on Mark read-only kernel memory as read-only (default). 4272 off Leave read-only kernel memory writable for debugging. 4273 4274 rockchip.usb_uart 4275 Enable the uart passthrough on the designated usb port 4276 on Rockchip SoCs. When active, the signals of the 4277 debug-uart get routed to the D+ and D- pins of the usb 4278 port and the regular usb controller gets disabled. 4279 4280 root= [KNL] Root filesystem 4281 See name_to_dev_t comment in init/do_mounts.c. 4282 4283 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 4284 mount the root filesystem 4285 4286 rootflags= [KNL] Set root filesystem mount option string 4287 4288 rootfstype= [KNL] Set root filesystem type 4289 4290 rootwait [KNL] Wait (indefinitely) for root device to show up. 4291 Useful for devices that are detected asynchronously 4292 (e.g. USB and MMC devices). 4293 4294 rproc_mem=nn[KMG][@address] 4295 [KNL,ARM,CMA] Remoteproc physical memory block. 4296 Memory area to be used by remote processor image, 4297 managed by CMA. 4298 4299 rw [KNL] Mount root device read-write on boot 4300 4301 S [KNL] Run init in single mode 4302 4303 s390_iommu= [HW,S390] 4304 Set s390 IOTLB flushing mode 4305 strict 4306 With strict flushing every unmap operation will result in 4307 an IOTLB flush. Default is lazy flushing before reuse, 4308 which is faster. 4309 4310 sa1100ir [NET] 4311 See drivers/net/irda/sa1100_ir.c. 4312 4313 sbni= [NET] Granch SBNI12 leased line adapter 4314 4315 sched_debug [KNL] Enables verbose scheduler debug messages. 4316 4317 schedstats= [KNL,X86] Enable or disable scheduled statistics. 4318 Allowed values are enable and disable. This feature 4319 incurs a small amount of overhead in the scheduler 4320 but is useful for debugging and performance tuning. 4321 4322 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate 4323 xtime_lock contention on larger systems, and/or RCU lock 4324 contention on all systems with CONFIG_MAXSMP set. 4325 Format: { "0" | "1" } 4326 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 4327 1 -- enable. 4328 Note: increases power consumption, thus should only be 4329 enabled if running jitter sensitive (HPC/RT) workloads. 4330 4331 security= [SECURITY] Choose a legacy "major" security module to 4332 enable at boot. This has been deprecated by the 4333 "lsm=" parameter. 4334 4335 selinux= [SELINUX] Disable or enable SELinux at boot time. 4336 Format: { "0" | "1" } 4337 See security/selinux/Kconfig help text. 4338 0 -- disable. 4339 1 -- enable. 4340 Default value is set via kernel config option. 4341 If enabled at boot time, /selinux/disable can be used 4342 later to disable prior to initial policy load. 4343 4344 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 4345 Format: { "0" | "1" } 4346 See security/apparmor/Kconfig help text 4347 0 -- disable. 4348 1 -- enable. 4349 Default value is set via kernel config option. 4350 4351 serialnumber [BUGS=X86-32] 4352 4353 shapers= [NET] 4354 Maximal number of shapers. 4355 4356 simeth= [IA-64] 4357 simscsi= 4358 4359 slram= [HW,MTD] 4360 4361 slab_nomerge [MM] 4362 Disable merging of slabs with similar size. May be 4363 necessary if there is some reason to distinguish 4364 allocs to different slabs, especially in hardened 4365 environments where the risk of heap overflows and 4366 layout control by attackers can usually be 4367 frustrated by disabling merging. This will reduce 4368 most of the exposure of a heap attack to a single 4369 cache (risks via metadata attacks are mostly 4370 unchanged). Debug options disable merging on their 4371 own. 4372 For more information see Documentation/vm/slub.rst. 4373 4374 slab_max_order= [MM, SLAB] 4375 Determines the maximum allowed order for slabs. 4376 A high setting may cause OOMs due to memory 4377 fragmentation. Defaults to 1 for systems with 4378 more than 32MB of RAM, 0 otherwise. 4379 4380 slub_debug[=options[,slabs]] [MM, SLUB] 4381 Enabling slub_debug allows one to determine the 4382 culprit if slab objects become corrupted. Enabling 4383 slub_debug can create guard zones around objects and 4384 may poison objects when not in use. Also tracks the 4385 last alloc / free. For more information see 4386 Documentation/vm/slub.rst. 4387 4388 slub_memcg_sysfs= [MM, SLUB] 4389 Determines whether to enable sysfs directories for 4390 memory cgroup sub-caches. 1 to enable, 0 to disable. 4391 The default is determined by CONFIG_SLUB_MEMCG_SYSFS_ON. 4392 Enabling this can lead to a very high number of debug 4393 directories and files being created under 4394 /sys/kernel/slub. 4395 4396 slub_max_order= [MM, SLUB] 4397 Determines the maximum allowed order for slabs. 4398 A high setting may cause OOMs due to memory 4399 fragmentation. For more information see 4400 Documentation/vm/slub.rst. 4401 4402 slub_min_objects= [MM, SLUB] 4403 The minimum number of objects per slab. SLUB will 4404 increase the slab order up to slub_max_order to 4405 generate a sufficiently large slab able to contain 4406 the number of objects indicated. The higher the number 4407 of objects the smaller the overhead of tracking slabs 4408 and the less frequently locks need to be acquired. 4409 For more information see Documentation/vm/slub.rst. 4410 4411 slub_min_order= [MM, SLUB] 4412 Determines the minimum page order for slabs. Must be 4413 lower than slub_max_order. 4414 For more information see Documentation/vm/slub.rst. 4415 4416 slub_nomerge [MM, SLUB] 4417 Same with slab_nomerge. This is supported for legacy. 4418 See slab_nomerge for more information. 4419 4420 smart2= [HW] 4421 Format: <io1>[,<io2>[,...,<io8>]] 4422 4423 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 4424 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 4425 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 4426 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 4427 smsc-ircc2.ircc_irq= [HW] IRQ line 4428 smsc-ircc2.ircc_dma= [HW] DMA channel 4429 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 4430 0: Toshiba Satellite 1800 (GP data pin select) 4431 1: Fast pin select (default) 4432 2: ATC IRMode 4433 4434 smt [KNL,S390] Set the maximum number of threads (logical 4435 CPUs) to use per physical CPU on systems capable of 4436 symmetric multithreading (SMT). Will be capped to the 4437 actual hardware limit. 4438 Format: <integer> 4439 Default: -1 (no limit) 4440 4441 softlockup_panic= 4442 [KNL] Should the soft-lockup detector generate panics. 4443 Format: <integer> 4444 4445 A nonzero value instructs the soft-lockup detector 4446 to panic the machine when a soft-lockup occurs. This 4447 is also controlled by CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC 4448 which is the respective build-time switch to that 4449 functionality. 4450 4451 softlockup_all_cpu_backtrace= 4452 [KNL] Should the soft-lockup detector generate 4453 backtraces on all cpus. 4454 Format: <integer> 4455 4456 sonypi.*= [HW] Sony Programmable I/O Control Device driver 4457 See Documentation/admin-guide/laptops/sonypi.rst 4458 4459 spectre_v2= [X86] Control mitigation of Spectre variant 2 4460 (indirect branch speculation) vulnerability. 4461 The default operation protects the kernel from 4462 user space attacks. 4463 4464 on - unconditionally enable, implies 4465 spectre_v2_user=on 4466 off - unconditionally disable, implies 4467 spectre_v2_user=off 4468 auto - kernel detects whether your CPU model is 4469 vulnerable 4470 4471 Selecting 'on' will, and 'auto' may, choose a 4472 mitigation method at run time according to the 4473 CPU, the available microcode, the setting of the 4474 CONFIG_RETPOLINE configuration option, and the 4475 compiler with which the kernel was built. 4476 4477 Selecting 'on' will also enable the mitigation 4478 against user space to user space task attacks. 4479 4480 Selecting 'off' will disable both the kernel and 4481 the user space protections. 4482 4483 Specific mitigations can also be selected manually: 4484 4485 retpoline - replace indirect branches 4486 retpoline,generic - google's original retpoline 4487 retpoline,amd - AMD-specific minimal thunk 4488 4489 Not specifying this option is equivalent to 4490 spectre_v2=auto. 4491 4492 spectre_v2_user= 4493 [X86] Control mitigation of Spectre variant 2 4494 (indirect branch speculation) vulnerability between 4495 user space tasks 4496 4497 on - Unconditionally enable mitigations. Is 4498 enforced by spectre_v2=on 4499 4500 off - Unconditionally disable mitigations. Is 4501 enforced by spectre_v2=off 4502 4503 prctl - Indirect branch speculation is enabled, 4504 but mitigation can be enabled via prctl 4505 per thread. The mitigation control state 4506 is inherited on fork. 4507 4508 prctl,ibpb 4509 - Like "prctl" above, but only STIBP is 4510 controlled per thread. IBPB is issued 4511 always when switching between different user 4512 space processes. 4513 4514 seccomp 4515 - Same as "prctl" above, but all seccomp 4516 threads will enable the mitigation unless 4517 they explicitly opt out. 4518 4519 seccomp,ibpb 4520 - Like "seccomp" above, but only STIBP is 4521 controlled per thread. IBPB is issued 4522 always when switching between different 4523 user space processes. 4524 4525 auto - Kernel selects the mitigation depending on 4526 the available CPU features and vulnerability. 4527 4528 Default mitigation: 4529 If CONFIG_SECCOMP=y then "seccomp", otherwise "prctl" 4530 4531 Not specifying this option is equivalent to 4532 spectre_v2_user=auto. 4533 4534 spec_store_bypass_disable= 4535 [HW] Control Speculative Store Bypass (SSB) Disable mitigation 4536 (Speculative Store Bypass vulnerability) 4537 4538 Certain CPUs are vulnerable to an exploit against a 4539 a common industry wide performance optimization known 4540 as "Speculative Store Bypass" in which recent stores 4541 to the same memory location may not be observed by 4542 later loads during speculative execution. The idea 4543 is that such stores are unlikely and that they can 4544 be detected prior to instruction retirement at the 4545 end of a particular speculation execution window. 4546 4547 In vulnerable processors, the speculatively forwarded 4548 store can be used in a cache side channel attack, for 4549 example to read memory to which the attacker does not 4550 directly have access (e.g. inside sandboxed code). 4551 4552 This parameter controls whether the Speculative Store 4553 Bypass optimization is used. 4554 4555 On x86 the options are: 4556 4557 on - Unconditionally disable Speculative Store Bypass 4558 off - Unconditionally enable Speculative Store Bypass 4559 auto - Kernel detects whether the CPU model contains an 4560 implementation of Speculative Store Bypass and 4561 picks the most appropriate mitigation. If the 4562 CPU is not vulnerable, "off" is selected. If the 4563 CPU is vulnerable the default mitigation is 4564 architecture and Kconfig dependent. See below. 4565 prctl - Control Speculative Store Bypass per thread 4566 via prctl. Speculative Store Bypass is enabled 4567 for a process by default. The state of the control 4568 is inherited on fork. 4569 seccomp - Same as "prctl" above, but all seccomp threads 4570 will disable SSB unless they explicitly opt out. 4571 4572 Default mitigations: 4573 X86: If CONFIG_SECCOMP=y "seccomp", otherwise "prctl" 4574 4575 On powerpc the options are: 4576 4577 on,auto - On Power8 and Power9 insert a store-forwarding 4578 barrier on kernel entry and exit. On Power7 4579 perform a software flush on kernel entry and 4580 exit. 4581 off - No action. 4582 4583 Not specifying this option is equivalent to 4584 spec_store_bypass_disable=auto. 4585 4586 spia_io_base= [HW,MTD] 4587 spia_fio_base= 4588 spia_pedr= 4589 spia_peddr= 4590 4591 srcutree.counter_wrap_check [KNL] 4592 Specifies how frequently to check for 4593 grace-period sequence counter wrap for the 4594 srcu_data structure's ->srcu_gp_seq_needed field. 4595 The greater the number of bits set in this kernel 4596 parameter, the less frequently counter wrap will 4597 be checked for. Note that the bottom two bits 4598 are ignored. 4599 4600 srcutree.exp_holdoff [KNL] 4601 Specifies how many nanoseconds must elapse 4602 since the end of the last SRCU grace period for 4603 a given srcu_struct until the next normal SRCU 4604 grace period will be considered for automatic 4605 expediting. Set to zero to disable automatic 4606 expediting. 4607 4608 ssbd= [ARM64,HW] 4609 Speculative Store Bypass Disable control 4610 4611 On CPUs that are vulnerable to the Speculative 4612 Store Bypass vulnerability and offer a 4613 firmware based mitigation, this parameter 4614 indicates how the mitigation should be used: 4615 4616 force-on: Unconditionally enable mitigation for 4617 for both kernel and userspace 4618 force-off: Unconditionally disable mitigation for 4619 for both kernel and userspace 4620 kernel: Always enable mitigation in the 4621 kernel, and offer a prctl interface 4622 to allow userspace to register its 4623 interest in being mitigated too. 4624 4625 stack_guard_gap= [MM] 4626 override the default stack gap protection. The value 4627 is in page units and it defines how many pages prior 4628 to (for stacks growing down) resp. after (for stacks 4629 growing up) the main stack are reserved for no other 4630 mapping. Default value is 256 pages. 4631 4632 stacktrace [FTRACE] 4633 Enabled the stack tracer on boot up. 4634 4635 stacktrace_filter=[function-list] 4636 [FTRACE] Limit the functions that the stack tracer 4637 will trace at boot up. function-list is a comma separated 4638 list of functions. This list can be changed at run 4639 time by the stack_trace_filter file in the debugfs 4640 tracing directory. Note, this enables stack tracing 4641 and the stacktrace above is not needed. 4642 4643 sti= [PARISC,HW] 4644 Format: <num> 4645 Set the STI (builtin display/keyboard on the HP-PARISC 4646 machines) console (graphic card) which should be used 4647 as the initial boot-console. 4648 See also comment in drivers/video/console/sticore.c. 4649 4650 sti_font= [HW] 4651 See comment in drivers/video/console/sticore.c. 4652 4653 stifb= [HW] 4654 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 4655 4656 sunrpc.min_resvport= 4657 sunrpc.max_resvport= 4658 [NFS,SUNRPC] 4659 SunRPC servers often require that client requests 4660 originate from a privileged port (i.e. a port in the 4661 range 0 < portnr < 1024). 4662 An administrator who wishes to reserve some of these 4663 ports for other uses may adjust the range that the 4664 kernel's sunrpc client considers to be privileged 4665 using these two parameters to set the minimum and 4666 maximum port values. 4667 4668 sunrpc.svc_rpc_per_connection_limit= 4669 [NFS,SUNRPC] 4670 Limit the number of requests that the server will 4671 process in parallel from a single connection. 4672 The default value is 0 (no limit). 4673 4674 sunrpc.pool_mode= 4675 [NFS] 4676 Control how the NFS server code allocates CPUs to 4677 service thread pools. Depending on how many NICs 4678 you have and where their interrupts are bound, this 4679 option will affect which CPUs will do NFS serving. 4680 Note: this parameter cannot be changed while the 4681 NFS server is running. 4682 4683 auto the server chooses an appropriate mode 4684 automatically using heuristics 4685 global a single global pool contains all CPUs 4686 percpu one pool for each CPU 4687 pernode one pool for each NUMA node (equivalent 4688 to global on non-NUMA machines) 4689 4690 sunrpc.tcp_slot_table_entries= 4691 sunrpc.udp_slot_table_entries= 4692 [NFS,SUNRPC] 4693 Sets the upper limit on the number of simultaneous 4694 RPC calls that can be sent from the client to a 4695 server. Increasing these values may allow you to 4696 improve throughput, but will also increase the 4697 amount of memory reserved for use by the client. 4698 4699 suspend.pm_test_delay= 4700 [SUSPEND] 4701 Sets the number of seconds to remain in a suspend test 4702 mode before resuming the system (see 4703 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 4704 is set. Default value is 5. 4705 4706 svm= [PPC] 4707 Format: { on | off | y | n | 1 | 0 } 4708 This parameter controls use of the Protected 4709 Execution Facility on pSeries. 4710 4711 swapaccount=[0|1] 4712 [KNL] Enable accounting of swap in memory resource 4713 controller if no parameter or 1 is given or disable 4714 it if 0 is given (See Documentation/admin-guide/cgroup-v1/memory.rst) 4715 4716 swiotlb= [ARM,IA-64,PPC,MIPS,X86] 4717 Format: { <int> | force | noforce } 4718 <int> -- Number of I/O TLB slabs 4719 force -- force using of bounce buffers even if they 4720 wouldn't be automatically used by the kernel 4721 noforce -- Never use bounce buffers (for debugging) 4722 4723 switches= [HW,M68k] 4724 4725 sysfs.deprecated=0|1 [KNL] 4726 Enable/disable old style sysfs layout for old udev 4727 on older distributions. When this option is enabled 4728 very new udev will not work anymore. When this option 4729 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled) 4730 in older udev will not work anymore. 4731 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in 4732 the kernel configuration. 4733 4734 sysrq_always_enabled 4735 [KNL] 4736 Ignore sysrq setting - this boot parameter will 4737 neutralize any effect of /proc/sys/kernel/sysrq. 4738 Useful for debugging. 4739 4740 tcpmhash_entries= [KNL,NET] 4741 Set the number of tcp_metrics_hash slots. 4742 Default value is 8192 or 16384 depending on total 4743 ram pages. This is used to specify the TCP metrics 4744 cache size. See Documentation/networking/ip-sysctl.txt 4745 "tcp_no_metrics_save" section for more details. 4746 4747 tdfx= [HW,DRM] 4748 4749 test_suspend= [SUSPEND][,N] 4750 Specify "mem" (for Suspend-to-RAM) or "standby" (for 4751 standby suspend) or "freeze" (for suspend type freeze) 4752 as the system sleep state during system startup with 4753 the optional capability to repeat N number of times. 4754 The system is woken from this state using a 4755 wakeup-capable RTC alarm. 4756 4757 thash_entries= [KNL,NET] 4758 Set number of hash buckets for TCP connection 4759 4760 thermal.act= [HW,ACPI] 4761 -1: disable all active trip points in all thermal zones 4762 <degrees C>: override all lowest active trip points 4763 4764 thermal.crt= [HW,ACPI] 4765 -1: disable all critical trip points in all thermal zones 4766 <degrees C>: override all critical trip points 4767 4768 thermal.nocrt= [HW,ACPI] 4769 Set to disable actions on ACPI thermal zone 4770 critical and hot trip points. 4771 4772 thermal.off= [HW,ACPI] 4773 1: disable ACPI thermal control 4774 4775 thermal.psv= [HW,ACPI] 4776 -1: disable all passive trip points 4777 <degrees C>: override all passive trip points to this 4778 value 4779 4780 thermal.tzp= [HW,ACPI] 4781 Specify global default ACPI thermal zone polling rate 4782 <deci-seconds>: poll all this frequency 4783 0: no polling (default) 4784 4785 threadirqs [KNL] 4786 Force threading of all interrupt handlers except those 4787 marked explicitly IRQF_NO_THREAD. 4788 4789 topology= [S390] 4790 Format: {off | on} 4791 Specify if the kernel should make use of the cpu 4792 topology information if the hardware supports this. 4793 The scheduler will make use of this information and 4794 e.g. base its process migration decisions on it. 4795 Default is on. 4796 4797 topology_updates= [KNL, PPC, NUMA] 4798 Format: {off} 4799 Specify if the kernel should ignore (off) 4800 topology updates sent by the hypervisor to this 4801 LPAR. 4802 4803 tp720= [HW,PS2] 4804 4805 tpm_suspend_pcr=[HW,TPM] 4806 Format: integer pcr id 4807 Specify that at suspend time, the tpm driver 4808 should extend the specified pcr with zeros, 4809 as a workaround for some chips which fail to 4810 flush the last written pcr on TPM_SaveState. 4811 This will guarantee that all the other pcrs 4812 are saved. 4813 4814 trace_buf_size=nn[KMG] 4815 [FTRACE] will set tracing buffer size on each cpu. 4816 4817 trace_event=[event-list] 4818 [FTRACE] Set and start specified trace events in order 4819 to facilitate early boot debugging. The event-list is a 4820 comma separated list of trace events to enable. See 4821 also Documentation/trace/events.rst 4822 4823 trace_options=[option-list] 4824 [FTRACE] Enable or disable tracer options at boot. 4825 The option-list is a comma delimited list of options 4826 that can be enabled or disabled just as if you were 4827 to echo the option name into 4828 4829 /sys/kernel/debug/tracing/trace_options 4830 4831 For example, to enable stacktrace option (to dump the 4832 stack trace of each event), add to the command line: 4833 4834 trace_options=stacktrace 4835 4836 See also Documentation/trace/ftrace.rst "trace options" 4837 section. 4838 4839 tp_printk[FTRACE] 4840 Have the tracepoints sent to printk as well as the 4841 tracing ring buffer. This is useful for early boot up 4842 where the system hangs or reboots and does not give the 4843 option for reading the tracing buffer or performing a 4844 ftrace_dump_on_oops. 4845 4846 To turn off having tracepoints sent to printk, 4847 echo 0 > /proc/sys/kernel/tracepoint_printk 4848 Note, echoing 1 into this file without the 4849 tracepoint_printk kernel cmdline option has no effect. 4850 4851 ** CAUTION ** 4852 4853 Having tracepoints sent to printk() and activating high 4854 frequency tracepoints such as irq or sched, can cause 4855 the system to live lock. 4856 4857 traceoff_on_warning 4858 [FTRACE] enable this option to disable tracing when a 4859 warning is hit. This turns off "tracing_on". Tracing can 4860 be enabled again by echoing '1' into the "tracing_on" 4861 file located in /sys/kernel/debug/tracing/ 4862 4863 This option is useful, as it disables the trace before 4864 the WARNING dump is called, which prevents the trace to 4865 be filled with content caused by the warning output. 4866 4867 This option can also be set at run time via the sysctl 4868 option: kernel/traceoff_on_warning 4869 4870 transparent_hugepage= 4871 [KNL] 4872 Format: [always|madvise|never] 4873 Can be used to control the default behavior of the system 4874 with respect to transparent hugepages. 4875 See Documentation/admin-guide/mm/transhuge.rst 4876 for more details. 4877 4878 tsc= Disable clocksource stability checks for TSC. 4879 Format: <string> 4880 [x86] reliable: mark tsc clocksource as reliable, this 4881 disables clocksource verification at runtime, as well 4882 as the stability checks done at bootup. Used to enable 4883 high-resolution timer mode on older hardware, and in 4884 virtualized environment. 4885 [x86] noirqtime: Do not use TSC to do irq accounting. 4886 Used to run time disable IRQ_TIME_ACCOUNTING on any 4887 platforms where RDTSC is slow and this accounting 4888 can add overhead. 4889 [x86] unstable: mark the TSC clocksource as unstable, this 4890 marks the TSC unconditionally unstable at bootup and 4891 avoids any further wobbles once the TSC watchdog notices. 4892 [x86] nowatchdog: disable clocksource watchdog. Used 4893 in situations with strict latency requirements (where 4894 interruptions from clocksource watchdog are not 4895 acceptable). 4896 4897 tsx= [X86] Control Transactional Synchronization 4898 Extensions (TSX) feature in Intel processors that 4899 support TSX control. 4900 4901 This parameter controls the TSX feature. The options are: 4902 4903 on - Enable TSX on the system. Although there are 4904 mitigations for all known security vulnerabilities, 4905 TSX has been known to be an accelerator for 4906 several previous speculation-related CVEs, and 4907 so there may be unknown security risks associated 4908 with leaving it enabled. 4909 4910 off - Disable TSX on the system. (Note that this 4911 option takes effect only on newer CPUs which are 4912 not vulnerable to MDS, i.e., have 4913 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 4914 the new IA32_TSX_CTRL MSR through a microcode 4915 update. This new MSR allows for the reliable 4916 deactivation of the TSX functionality.) 4917 4918 auto - Disable TSX if X86_BUG_TAA is present, 4919 otherwise enable TSX on the system. 4920 4921 Not specifying this option is equivalent to tsx=off. 4922 4923 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 4924 for more details. 4925 4926 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async 4927 Abort (TAA) vulnerability. 4928 4929 Similar to Micro-architectural Data Sampling (MDS) 4930 certain CPUs that support Transactional 4931 Synchronization Extensions (TSX) are vulnerable to an 4932 exploit against CPU internal buffers which can forward 4933 information to a disclosure gadget under certain 4934 conditions. 4935 4936 In vulnerable processors, the speculatively forwarded 4937 data can be used in a cache side channel attack, to 4938 access data to which the attacker does not have direct 4939 access. 4940 4941 This parameter controls the TAA mitigation. The 4942 options are: 4943 4944 full - Enable TAA mitigation on vulnerable CPUs 4945 if TSX is enabled. 4946 4947 full,nosmt - Enable TAA mitigation and disable SMT on 4948 vulnerable CPUs. If TSX is disabled, SMT 4949 is not disabled because CPU is not 4950 vulnerable to cross-thread TAA attacks. 4951 off - Unconditionally disable TAA mitigation 4952 4953 On MDS-affected machines, tsx_async_abort=off can be 4954 prevented by an active MDS mitigation as both vulnerabilities 4955 are mitigated with the same mechanism so in order to disable 4956 this mitigation, you need to specify mds=off too. 4957 4958 Not specifying this option is equivalent to 4959 tsx_async_abort=full. On CPUs which are MDS affected 4960 and deploy MDS mitigation, TAA mitigation is not 4961 required and doesn't provide any additional 4962 mitigation. 4963 4964 For details see: 4965 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 4966 4967 turbografx.map[2|3]= [HW,JOY] 4968 TurboGraFX parallel port interface 4969 Format: 4970 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 4971 See also Documentation/input/devices/joystick-parport.rst 4972 4973 udbg-immortal [PPC] When debugging early kernel crashes that 4974 happen after console_init() and before a proper 4975 console driver takes over, this boot options might 4976 help "seeing" what's going on. 4977 4978 uhash_entries= [KNL,NET] 4979 Set number of hash buckets for UDP/UDP-Lite connections 4980 4981 uhci-hcd.ignore_oc= 4982 [USB] Ignore overcurrent events (default N). 4983 Some badly-designed motherboards generate lots of 4984 bogus events, for ports that aren't wired to 4985 anything. Set this parameter to avoid log spamming. 4986 Note that genuine overcurrent events won't be 4987 reported either. 4988 4989 unknown_nmi_panic 4990 [X86] Cause panic on unknown NMI. 4991 4992 usbcore.authorized_default= 4993 [USB] Default USB device authorization: 4994 (default -1 = authorized except for wireless USB, 4995 0 = not authorized, 1 = authorized, 2 = authorized 4996 if device connected to internal port) 4997 4998 usbcore.autosuspend= 4999 [USB] The autosuspend time delay (in seconds) used 5000 for newly-detected USB devices (default 2). This 5001 is the time required before an idle device will be 5002 autosuspended. Devices for which the delay is set 5003 to a negative value won't be autosuspended at all. 5004 5005 usbcore.usbfs_snoop= 5006 [USB] Set to log all usbfs traffic (default 0 = off). 5007 5008 usbcore.usbfs_snoop_max= 5009 [USB] Maximum number of bytes to snoop in each URB 5010 (default = 65536). 5011 5012 usbcore.blinkenlights= 5013 [USB] Set to cycle leds on hubs (default 0 = off). 5014 5015 usbcore.old_scheme_first= 5016 [USB] Start with the old device initialization 5017 scheme, applies only to low and full-speed devices 5018 (default 0 = off). 5019 5020 usbcore.usbfs_memory_mb= 5021 [USB] Memory limit (in MB) for buffers allocated by 5022 usbfs (default = 16, 0 = max = 2047). 5023 5024 usbcore.use_both_schemes= 5025 [USB] Try the other device initialization scheme 5026 if the first one fails (default 1 = enabled). 5027 5028 usbcore.initial_descriptor_timeout= 5029 [USB] Specifies timeout for the initial 64-byte 5030 USB_REQ_GET_DESCRIPTOR request in milliseconds 5031 (default 5000 = 5.0 seconds). 5032 5033 usbcore.nousb [USB] Disable the USB subsystem 5034 5035 usbcore.quirks= 5036 [USB] A list of quirk entries to augment the built-in 5037 usb core quirk list. List entries are separated by 5038 commas. Each entry has the form 5039 VendorID:ProductID:Flags. The IDs are 4-digit hex 5040 numbers and Flags is a set of letters. Each letter 5041 will change the built-in quirk; setting it if it is 5042 clear and clearing it if it is set. The letters have 5043 the following meanings: 5044 a = USB_QUIRK_STRING_FETCH_255 (string 5045 descriptors must not be fetched using 5046 a 255-byte read); 5047 b = USB_QUIRK_RESET_RESUME (device can't resume 5048 correctly so reset it instead); 5049 c = USB_QUIRK_NO_SET_INTF (device can't handle 5050 Set-Interface requests); 5051 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 5052 handle its Configuration or Interface 5053 strings); 5054 e = USB_QUIRK_RESET (device can't be reset 5055 (e.g morph devices), don't use reset); 5056 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 5057 more interface descriptions than the 5058 bNumInterfaces count, and can't handle 5059 talking to these interfaces); 5060 g = USB_QUIRK_DELAY_INIT (device needs a pause 5061 during initialization, after we read 5062 the device descriptor); 5063 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 5064 high speed and super speed interrupt 5065 endpoints, the USB 2.0 and USB 3.0 spec 5066 require the interval in microframes (1 5067 microframe = 125 microseconds) to be 5068 calculated as interval = 2 ^ 5069 (bInterval-1). 5070 Devices with this quirk report their 5071 bInterval as the result of this 5072 calculation instead of the exponent 5073 variable used in the calculation); 5074 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 5075 handle device_qualifier descriptor 5076 requests); 5077 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 5078 generates spurious wakeup, ignore 5079 remote wakeup capability); 5080 k = USB_QUIRK_NO_LPM (device can't handle Link 5081 Power Management); 5082 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 5083 (Device reports its bInterval as linear 5084 frames instead of the USB 2.0 5085 calculation); 5086 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 5087 to be disconnected before suspend to 5088 prevent spurious wakeup); 5089 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 5090 pause after every control message); 5091 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 5092 delay after resetting its port); 5093 Example: quirks=0781:5580:bk,0a5c:5834:gij 5094 5095 usbhid.mousepoll= 5096 [USBHID] The interval which mice are to be polled at. 5097 5098 usbhid.jspoll= 5099 [USBHID] The interval which joysticks are to be polled at. 5100 5101 usbhid.kbpoll= 5102 [USBHID] The interval which keyboards are to be polled at. 5103 5104 usb-storage.delay_use= 5105 [UMS] The delay in seconds before a new device is 5106 scanned for Logical Units (default 1). 5107 5108 usb-storage.quirks= 5109 [UMS] A list of quirks entries to supplement or 5110 override the built-in unusual_devs list. List 5111 entries are separated by commas. Each entry has 5112 the form VID:PID:Flags where VID and PID are Vendor 5113 and Product ID values (4-digit hex numbers) and 5114 Flags is a set of characters, each corresponding 5115 to a common usb-storage quirk flag as follows: 5116 a = SANE_SENSE (collect more than 18 bytes 5117 of sense data); 5118 b = BAD_SENSE (don't collect more than 18 5119 bytes of sense data); 5120 c = FIX_CAPACITY (decrease the reported 5121 device capacity by one sector); 5122 d = NO_READ_DISC_INFO (don't use 5123 READ_DISC_INFO command); 5124 e = NO_READ_CAPACITY_16 (don't use 5125 READ_CAPACITY_16 command); 5126 f = NO_REPORT_OPCODES (don't use report opcodes 5127 command, uas only); 5128 g = MAX_SECTORS_240 (don't transfer more than 5129 240 sectors at a time, uas only); 5130 h = CAPACITY_HEURISTICS (decrease the 5131 reported device capacity by one 5132 sector if the number is odd); 5133 i = IGNORE_DEVICE (don't bind to this 5134 device); 5135 j = NO_REPORT_LUNS (don't use report luns 5136 command, uas only); 5137 l = NOT_LOCKABLE (don't try to lock and 5138 unlock ejectable media); 5139 m = MAX_SECTORS_64 (don't transfer more 5140 than 64 sectors = 32 KB at a time); 5141 n = INITIAL_READ10 (force a retry of the 5142 initial READ(10) command); 5143 o = CAPACITY_OK (accept the capacity 5144 reported by the device); 5145 p = WRITE_CACHE (the device cache is ON 5146 by default); 5147 r = IGNORE_RESIDUE (the device reports 5148 bogus residue values); 5149 s = SINGLE_LUN (the device has only one 5150 Logical Unit); 5151 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 5152 commands, uas only); 5153 u = IGNORE_UAS (don't bind to the uas driver); 5154 w = NO_WP_DETECT (don't test whether the 5155 medium is write-protected). 5156 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 5157 even if the device claims no cache) 5158 Example: quirks=0419:aaf5:rl,0421:0433:rc 5159 5160 user_debug= [KNL,ARM] 5161 Format: <int> 5162 See arch/arm/Kconfig.debug help text. 5163 1 - undefined instruction events 5164 2 - system calls 5165 4 - invalid data aborts 5166 8 - SIGSEGV faults 5167 16 - SIGBUS faults 5168 Example: user_debug=31 5169 5170 userpte= 5171 [X86] Flags controlling user PTE allocations. 5172 5173 nohigh = do not allocate PTE pages in 5174 HIGHMEM regardless of setting 5175 of CONFIG_HIGHPTE. 5176 5177 vdso= [X86,SH] 5178 On X86_32, this is an alias for vdso32=. Otherwise: 5179 5180 vdso=1: enable VDSO (the default) 5181 vdso=0: disable VDSO mapping 5182 5183 vdso32= [X86] Control the 32-bit vDSO 5184 vdso32=1: enable 32-bit VDSO 5185 vdso32=0 or vdso32=2: disable 32-bit VDSO 5186 5187 See the help text for CONFIG_COMPAT_VDSO for more 5188 details. If CONFIG_COMPAT_VDSO is set, the default is 5189 vdso32=0; otherwise, the default is vdso32=1. 5190 5191 For compatibility with older kernels, vdso32=2 is an 5192 alias for vdso32=0. 5193 5194 Try vdso32=0 if you encounter an error that says: 5195 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 5196 5197 vector= [IA-64,SMP] 5198 vector=percpu: enable percpu vector domain 5199 5200 video= [FB] Frame buffer configuration 5201 See Documentation/fb/modedb.rst. 5202 5203 video.brightness_switch_enabled= [0,1] 5204 If set to 1, on receiving an ACPI notify event 5205 generated by hotkey, video driver will adjust brightness 5206 level and then send out the event to user space through 5207 the allocated input device; If set to 0, video driver 5208 will only send out the event without touching backlight 5209 brightness level. 5210 default: 1 5211 5212 virtio_mmio.device= 5213 [VMMIO] Memory mapped virtio (platform) device. 5214 5215 <size>@<baseaddr>:<irq>[:<id>] 5216 where: 5217 <size> := size (can use standard suffixes 5218 like K, M and G) 5219 <baseaddr> := physical base address 5220 <irq> := interrupt number (as passed to 5221 request_irq()) 5222 <id> := (optional) platform device id 5223 example: 5224 virtio_mmio.device=1K@0x100b0000:48:7 5225 5226 Can be used multiple times for multiple devices. 5227 5228 vga= [BOOT,X86-32] Select a particular video mode 5229 See Documentation/x86/boot.rst and 5230 Documentation/admin-guide/svga.rst. 5231 Use vga=ask for menu. 5232 This is actually a boot loader parameter; the value is 5233 passed to the kernel using a special protocol. 5234 5235 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 5236 May slow down system boot speed, especially when 5237 enabled on systems with a large amount of memory. 5238 All options are enabled by default, and this 5239 interface is meant to allow for selectively 5240 enabling or disabling specific virtual memory 5241 debugging features. 5242 5243 Available options are: 5244 P Enable page structure init time poisoning 5245 - Disable all of the above options 5246 5247 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact 5248 size of <nn>. This can be used to increase the 5249 minimum size (128MB on x86). It can also be used to 5250 decrease the size and leave more room for directly 5251 mapped kernel RAM. 5252 5253 vmcp_cma=nn[MG] [KNL,S390] 5254 Sets the memory size reserved for contiguous memory 5255 allocations for the vmcp device driver. 5256 5257 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 5258 Format: <command> 5259 5260 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 5261 Format: <command> 5262 5263 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 5264 Format: <command> 5265 5266 vsyscall= [X86-64] 5267 Controls the behavior of vsyscalls (i.e. calls to 5268 fixed addresses of 0xffffffffff600x00 from legacy 5269 code). Most statically-linked binaries and older 5270 versions of glibc use these calls. Because these 5271 functions are at fixed addresses, they make nice 5272 targets for exploits that can control RIP. 5273 5274 emulate [default] Vsyscalls turn into traps and are 5275 emulated reasonably safely. The vsyscall 5276 page is readable. 5277 5278 xonly Vsyscalls turn into traps and are 5279 emulated reasonably safely. The vsyscall 5280 page is not readable. 5281 5282 none Vsyscalls don't work at all. This makes 5283 them quite hard to use for exploits but 5284 might break your system. 5285 5286 vt.color= [VT] Default text color. 5287 Format: 0xYX, X = foreground, Y = background. 5288 Default: 0x07 = light gray on black. 5289 5290 vt.cur_default= [VT] Default cursor shape. 5291 Format: 0xCCBBAA, where AA, BB, and CC are the same as 5292 the parameters of the <Esc>[?A;B;Cc escape sequence; 5293 see VGA-softcursor.txt. Default: 2 = underline. 5294 5295 vt.default_blu= [VT] 5296 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 5297 Change the default blue palette of the console. 5298 This is a 16-member array composed of values 5299 ranging from 0-255. 5300 5301 vt.default_grn= [VT] 5302 Format: <green0>,<green1>,<green2>,...,<green15> 5303 Change the default green palette of the console. 5304 This is a 16-member array composed of values 5305 ranging from 0-255. 5306 5307 vt.default_red= [VT] 5308 Format: <red0>,<red1>,<red2>,...,<red15> 5309 Change the default red palette of the console. 5310 This is a 16-member array composed of values 5311 ranging from 0-255. 5312 5313 vt.default_utf8= 5314 [VT] 5315 Format=<0|1> 5316 Set system-wide default UTF-8 mode for all tty's. 5317 Default is 1, i.e. UTF-8 mode is enabled for all 5318 newly opened terminals. 5319 5320 vt.global_cursor_default= 5321 [VT] 5322 Format=<-1|0|1> 5323 Set system-wide default for whether a cursor 5324 is shown on new VTs. Default is -1, 5325 i.e. cursors will be created by default unless 5326 overridden by individual drivers. 0 will hide 5327 cursors, 1 will display them. 5328 5329 vt.italic= [VT] Default color for italic text; 0-15. 5330 Default: 2 = green. 5331 5332 vt.underline= [VT] Default color for underlined text; 0-15. 5333 Default: 3 = cyan. 5334 5335 watchdog timers [HW,WDT] For information on watchdog timers, 5336 see Documentation/watchdog/watchdog-parameters.rst 5337 or other driver-specific files in the 5338 Documentation/watchdog/ directory. 5339 5340 watchdog_thresh= 5341 [KNL] 5342 Set the hard lockup detector stall duration 5343 threshold in seconds. The soft lockup detector 5344 threshold is set to twice the value. A value of 0 5345 disables both lockup detectors. Default is 10 5346 seconds. 5347 5348 workqueue.watchdog_thresh= 5349 If CONFIG_WQ_WATCHDOG is configured, workqueue can 5350 warn stall conditions and dump internal state to 5351 help debugging. 0 disables workqueue stall 5352 detection; otherwise, it's the stall threshold 5353 duration in seconds. The default value is 30 and 5354 it can be updated at runtime by writing to the 5355 corresponding sysfs file. 5356 5357 workqueue.disable_numa 5358 By default, all work items queued to unbound 5359 workqueues are affine to the NUMA nodes they're 5360 issued on, which results in better behavior in 5361 general. If NUMA affinity needs to be disabled for 5362 whatever reason, this option can be used. Note 5363 that this also can be controlled per-workqueue for 5364 workqueues visible under /sys/bus/workqueue/. 5365 5366 workqueue.power_efficient 5367 Per-cpu workqueues are generally preferred because 5368 they show better performance thanks to cache 5369 locality; unfortunately, per-cpu workqueues tend to 5370 be more power hungry than unbound workqueues. 5371 5372 Enabling this makes the per-cpu workqueues which 5373 were observed to contribute significantly to power 5374 consumption unbound, leading to measurably lower 5375 power usage at the cost of small performance 5376 overhead. 5377 5378 The default value of this parameter is determined by 5379 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 5380 5381 workqueue.debug_force_rr_cpu 5382 Workqueue used to implicitly guarantee that work 5383 items queued without explicit CPU specified are put 5384 on the local CPU. This guarantee is no longer true 5385 and while local CPU is still preferred work items 5386 may be put on foreign CPUs. This debug option 5387 forces round-robin CPU selection to flush out 5388 usages which depend on the now broken guarantee. 5389 When enabled, memory and cache locality will be 5390 impacted. 5391 5392 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of 5393 default x2apic cluster mode on platforms 5394 supporting x2apic. 5395 5396 x86_intel_mid_timer= [X86-32,APBT] 5397 Choose timer option for x86 Intel MID platform. 5398 Two valid options are apbt timer only and lapic timer 5399 plus one apbt timer for broadcast timer. 5400 x86_intel_mid_timer=apbt_only | lapic_and_apbt 5401 5402 xen_512gb_limit [KNL,X86-64,XEN] 5403 Restricts the kernel running paravirtualized under Xen 5404 to use only up to 512 GB of RAM. The reason to do so is 5405 crash analysis tools and Xen tools for doing domain 5406 save/restore/migration must be enabled to handle larger 5407 domains. 5408 5409 xen_emul_unplug= [HW,X86,XEN] 5410 Unplug Xen emulated devices 5411 Format: [unplug0,][unplug1] 5412 ide-disks -- unplug primary master IDE devices 5413 aux-ide-disks -- unplug non-primary-master IDE devices 5414 nics -- unplug network devices 5415 all -- unplug all emulated devices (NICs and IDE disks) 5416 unnecessary -- unplugging emulated devices is 5417 unnecessary even if the host did not respond to 5418 the unplug protocol 5419 never -- do not unplug even if version check succeeds 5420 5421 xen_legacy_crash [X86,XEN] 5422 Crash from Xen panic notifier, without executing late 5423 panic() code such as dumping handler. 5424 5425 xen_nopvspin [X86,XEN] 5426 Disables the ticketlock slowpath using Xen PV 5427 optimizations. 5428 5429 xen_nopv [X86] 5430 Disables the PV optimizations forcing the HVM guest to 5431 run as generic HVM guest with no PV drivers. 5432 This option is obsoleted by the "nopv" option, which 5433 has equivalent effect for XEN platform. 5434 5435 xen_scrub_pages= [XEN] 5436 Boolean option to control scrubbing pages before giving them back 5437 to Xen, for use by other domains. Can be also changed at runtime 5438 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 5439 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 5440 5441 xen_timer_slop= [X86-64,XEN] 5442 Set the timer slop (in nanoseconds) for the virtual Xen 5443 timers (default is 100000). This adjusts the minimum 5444 delta of virtualized Xen timers, where lower values 5445 improve timer resolution at the expense of processing 5446 more timer interrupts. 5447 5448 nopv= [X86,XEN,KVM,HYPER_V,VMWARE] 5449 Disables the PV optimizations forcing the guest to run 5450 as generic guest with no PV drivers. Currently support 5451 XEN HVM, KVM, HYPER_V and VMWARE guest. 5452 5453 xirc2ps_cs= [NET,PCMCIA] 5454 Format: 5455 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 5456 5457 xive= [PPC] 5458 By default on POWER9 and above, the kernel will 5459 natively use the XIVE interrupt controller. This option 5460 allows the fallback firmware mode to be used: 5461 5462 off Fallback to firmware control of XIVE interrupt 5463 controller on both pseries and powernv 5464 platforms. Only useful on POWER9 and above. 5465 5466 xhci-hcd.quirks [USB,KNL] 5467 A hex value specifying bitmask with supplemental xhci 5468 host controller quirks. Meaning of each bit can be 5469 consulted in header drivers/usb/host/xhci.h. 5470 5471 xmon [PPC] 5472 Format: { early | on | rw | ro | off } 5473 Controls if xmon debugger is enabled. Default is off. 5474 Passing only "xmon" is equivalent to "xmon=early". 5475 early Call xmon as early as possible on boot; xmon 5476 debugger is called from setup_arch(). 5477 on xmon debugger hooks will be installed so xmon 5478 is only called on a kernel crash. Default mode, 5479 i.e. either "ro" or "rw" mode, is controlled 5480 with CONFIG_XMON_DEFAULT_RO_MODE. 5481 rw xmon debugger hooks will be installed so xmon 5482 is called only on a kernel crash, mode is write, 5483 meaning SPR registers, memory and, other data 5484 can be written using xmon commands. 5485 ro same as "rw" option above but SPR registers, 5486 memory, and other data can't be written using 5487 xmon commands. 5488 off xmon is disabled. 5489