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: <byte> 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 bootconfig [KNL] 441 Extended command line options can be added to an initrd 442 and this will cause the kernel to look for it. 443 444 See Documentation/admin-guide/bootconfig.rst 445 446 bert_disable [ACPI] 447 Disable BERT OS support on buggy BIOSes. 448 449 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 450 bttv.radio= Most important insmod options are available as 451 kernel args too. 452 bttv.pll= See Documentation/media/v4l-drivers/bttv.rst 453 bttv.tuner= 454 455 bulk_remove=off [PPC] This parameter disables the use of the pSeries 456 firmware feature for flushing multiple hpte entries 457 at a time. 458 459 c101= [NET] Moxa C101 synchronous serial card 460 461 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 462 Sometimes CPU hardware bugs make them report the cache 463 size incorrectly. The kernel will attempt work arounds 464 to fix known problems, but for some CPUs it is not 465 possible to determine what the correct size should be. 466 This option provides an override for these situations. 467 468 carrier_timeout= 469 [NET] Specifies amount of time (in seconds) that 470 the kernel should wait for a network carrier. By default 471 it waits 120 seconds. 472 473 ca_keys= [KEYS] This parameter identifies a specific key(s) on 474 the system trusted keyring to be used for certificate 475 trust validation. 476 format: { id:<keyid> | builtin } 477 478 cca= [MIPS] Override the kernel pages' cache coherency 479 algorithm. Accepted values range from 0 to 7 480 inclusive. See arch/mips/include/asm/pgtable-bits.h 481 for platform specific values (SB1, Loongson3 and 482 others). 483 484 ccw_timeout_log [S390] 485 See Documentation/s390/common_io.rst for details. 486 487 cgroup_disable= [KNL] Disable a particular controller 488 Format: {name of the controller(s) to disable} 489 The effects of cgroup_disable=foo are: 490 - foo isn't auto-mounted if you mount all cgroups in 491 a single hierarchy 492 - foo isn't visible as an individually mountable 493 subsystem 494 {Currently only "memory" controller deal with this and 495 cut the overhead, others just disable the usage. So 496 only cgroup_disable=memory is actually worthy} 497 498 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 499 Format: { { controller | "all" | "named" } 500 [,{ controller | "all" | "named" }...] } 501 Like cgroup_disable, but only applies to cgroup v1; 502 the blacklisted controllers remain available in cgroup2. 503 "all" blacklists all controllers and "named" disables 504 named mounts. Specifying both "all" and "named" disables 505 all v1 hierarchies. 506 507 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 508 Format: <string> 509 nosocket -- Disable socket memory accounting. 510 nokmem -- Disable kernel memory accounting. 511 512 checkreqprot [SELINUX] Set initial checkreqprot flag value. 513 Format: { "0" | "1" } 514 See security/selinux/Kconfig help text. 515 0 -- check protection applied by kernel (includes 516 any implied execute protection). 517 1 -- check protection requested by application. 518 Default value is set via a kernel config option. 519 Value can be changed at runtime via 520 /selinux/checkreqprot. 521 522 cio_ignore= [S390] 523 See Documentation/s390/common_io.rst for details. 524 clk_ignore_unused 525 [CLK] 526 Prevents the clock framework from automatically gating 527 clocks that have not been explicitly enabled by a Linux 528 device driver but are enabled in hardware at reset or 529 by the bootloader/firmware. Note that this does not 530 force such clocks to be always-on nor does it reserve 531 those clocks in any way. This parameter is useful for 532 debug and development, but should not be needed on a 533 platform with proper driver support. For more 534 information, see Documentation/driver-api/clk.rst. 535 536 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 537 [Deprecated] 538 Forces specified clocksource (if available) to be used 539 when calculating gettimeofday(). If specified 540 clocksource is not available, it defaults to PIT. 541 Format: { pit | tsc | cyclone | pmtmr } 542 543 clocksource= Override the default clocksource 544 Format: <string> 545 Override the default clocksource and use the clocksource 546 with the name specified. 547 Some clocksource names to choose from, depending on 548 the platform: 549 [all] jiffies (this is the base, fallback clocksource) 550 [ACPI] acpi_pm 551 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 552 pxa_timer,timer3,32k_counter,timer0_1 553 [X86-32] pit,hpet,tsc; 554 scx200_hrt on Geode; cyclone on IBM x440 555 [MIPS] MIPS 556 [PARISC] cr16 557 [S390] tod 558 [SH] SuperH 559 [SPARC64] tick 560 [X86-64] hpet,tsc 561 562 clocksource.arm_arch_timer.evtstrm= 563 [ARM,ARM64] 564 Format: <bool> 565 Enable/disable the eventstream feature of the ARM 566 architected timer so that code using WFE-based polling 567 loops can be debugged more effectively on production 568 systems. 569 570 clearcpuid=BITNUM [X86] 571 Disable CPUID feature X for the kernel. See 572 arch/x86/include/asm/cpufeatures.h for the valid bit 573 numbers. Note the Linux specific bits are not necessarily 574 stable over kernel options, but the vendor specific 575 ones should be. 576 Also note that user programs calling CPUID directly 577 or using the feature without checking anything 578 will still see it. This just prevents it from 579 being used by the kernel or shown in /proc/cpuinfo. 580 Also note the kernel might malfunction if you disable 581 some critical bits. 582 583 cma=nn[MG]@[start[MG][-end[MG]]] 584 [ARM,X86,KNL] 585 Sets the size of kernel global memory area for 586 contiguous memory allocations and optionally the 587 placement constraint by the physical address range of 588 memory allocations. A value of 0 disables CMA 589 altogether. For more information, see 590 include/linux/dma-contiguous.h 591 592 cmo_free_hint= [PPC] Format: { yes | no } 593 Specify whether pages are marked as being inactive 594 when they are freed. This is used in CMO environments 595 to determine OS memory pressure for page stealing by 596 a hypervisor. 597 Default: yes 598 599 coherent_pool=nn[KMG] [ARM,KNL] 600 Sets the size of memory pool for coherent, atomic dma 601 allocations, by default set to 256K. 602 603 com20020= [HW,NET] ARCnet - COM20020 chipset 604 Format: 605 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 606 607 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 608 Format: <io>[,<irq>] 609 610 com90xx= [HW,NET] 611 ARCnet - COM90xx chipset (memory-mapped buffers) 612 Format: <io>[,<irq>[,<memstart>]] 613 614 condev= [HW,S390] console device 615 conmode= 616 617 console= [KNL] Output console device and options. 618 619 tty<n> Use the virtual console device <n>. 620 621 ttyS<n>[,options] 622 ttyUSB0[,options] 623 Use the specified serial port. The options are of 624 the form "bbbbpnf", where "bbbb" is the baud rate, 625 "p" is parity ("n", "o", or "e"), "n" is number of 626 bits, and "f" is flow control ("r" for RTS or 627 omit it). Default is "9600n8". 628 629 See Documentation/admin-guide/serial-console.rst for more 630 information. See 631 Documentation/networking/netconsole.txt for an 632 alternative. 633 634 uart[8250],io,<addr>[,options] 635 uart[8250],mmio,<addr>[,options] 636 uart[8250],mmio16,<addr>[,options] 637 uart[8250],mmio32,<addr>[,options] 638 uart[8250],0x<addr>[,options] 639 Start an early, polled-mode console on the 8250/16550 640 UART at the specified I/O port or MMIO address, 641 switching to the matching ttyS device later. 642 MMIO inter-register address stride is either 8-bit 643 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 644 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 645 to be equivalent to 'mmio'. 'options' are specified in 646 the same format described for ttyS above; if unspecified, 647 the h/w is not re-initialized. 648 649 hvc<n> Use the hypervisor console device <n>. This is for 650 both Xen and PowerPC hypervisors. 651 652 If the device connected to the port is not a TTY but a braille 653 device, prepend "brl," before the device type, for instance 654 console=brl,ttyS0 655 For now, only VisioBraille is supported. 656 657 console_msg_format= 658 [KNL] Change console messages format 659 default 660 By default we print messages on consoles in 661 "[time stamp] text\n" format (time stamp may not be 662 printed, depending on CONFIG_PRINTK_TIME or 663 `printk_time' param). 664 syslog 665 Switch to syslog format: "<%u>[time stamp] text\n" 666 IOW, each message will have a facility and loglevel 667 prefix. The format is similar to one used by syslog() 668 syscall, or to executing "dmesg -S --raw" or to reading 669 from /proc/kmsg. 670 671 consoleblank= [KNL] The console blank (screen saver) timeout in 672 seconds. A value of 0 disables the blank timer. 673 Defaults to 0. 674 675 coredump_filter= 676 [KNL] Change the default value for 677 /proc/<pid>/coredump_filter. 678 See also Documentation/filesystems/proc.txt. 679 680 coresight_cpu_debug.enable 681 [ARM,ARM64] 682 Format: <bool> 683 Enable/disable the CPU sampling based debugging. 684 0: default value, disable debugging 685 1: enable debugging at boot time 686 687 cpuidle.off=1 [CPU_IDLE] 688 disable the cpuidle sub-system 689 690 cpuidle.governor= 691 [CPU_IDLE] Name of the cpuidle governor to use. 692 693 cpufreq.off=1 [CPU_FREQ] 694 disable the cpufreq sub-system 695 696 cpu_init_udelay=N 697 [X86] Delay for N microsec between assert and de-assert 698 of APIC INIT to start processors. This delay occurs 699 on every CPU online, such as boot, and resume from suspend. 700 Default: 10000 701 702 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 703 Format: 704 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 705 706 crashkernel=size[KMG][@offset[KMG]] 707 [KNL] Using kexec, Linux can switch to a 'crash kernel' 708 upon panic. This parameter reserves the physical 709 memory region [offset, offset + size] for that kernel 710 image. If '@offset' is omitted, then a suitable offset 711 is selected automatically. 712 [KNL, x86_64] select a region under 4G first, and 713 fall back to reserve region above 4G when '@offset' 714 hasn't been specified. 715 See Documentation/admin-guide/kdump/kdump.rst for further details. 716 717 crashkernel=range1:size1[,range2:size2,...][@offset] 718 [KNL] Same as above, but depends on the memory 719 in the running system. The syntax of range is 720 start-[end] where start and end are both 721 a memory unit (amount[KMG]). See also 722 Documentation/admin-guide/kdump/kdump.rst for an example. 723 724 crashkernel=size[KMG],high 725 [KNL, x86_64] range could be above 4G. Allow kernel 726 to allocate physical memory region from top, so could 727 be above 4G if system have more than 4G ram installed. 728 Otherwise memory region will be allocated below 4G, if 729 available. 730 It will be ignored if crashkernel=X is specified. 731 crashkernel=size[KMG],low 732 [KNL, x86_64] range under 4G. When crashkernel=X,high 733 is passed, kernel could allocate physical memory region 734 above 4G, that cause second kernel crash on system 735 that require some amount of low memory, e.g. swiotlb 736 requires at least 64M+32K low memory, also enough extra 737 low memory is needed to make sure DMA buffers for 32-bit 738 devices won't run out. Kernel would try to allocate at 739 at least 256M below 4G automatically. 740 This one let user to specify own low range under 4G 741 for second kernel instead. 742 0: to disable low allocation. 743 It will be ignored when crashkernel=X,high is not used 744 or memory reserved is below 4G. 745 746 cryptomgr.notests 747 [KNL] Disable crypto self-tests 748 749 cs89x0_dma= [HW,NET] 750 Format: <dma> 751 752 cs89x0_media= [HW,NET] 753 Format: { rj45 | aui | bnc } 754 755 dasd= [HW,NET] 756 See header of drivers/s390/block/dasd_devmap.c. 757 758 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 759 (one device per port) 760 Format: <port#>,<type> 761 See also Documentation/input/devices/joystick-parport.rst 762 763 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot 764 time. See 765 Documentation/admin-guide/dynamic-debug-howto.rst for 766 details. Deprecated, see dyndbg. 767 768 debug [KNL] Enable kernel debugging (events log level). 769 770 debug_boot_weak_hash 771 [KNL] Enable printing [hashed] pointers early in the 772 boot sequence. If enabled, we use a weak hash instead 773 of siphash to hash pointers. Use this option if you are 774 seeing instances of '(___ptrval___)') and need to see a 775 value (hashed pointer) instead. Cryptographically 776 insecure, please do not use on production kernels. 777 778 debug_locks_verbose= 779 [KNL] verbose self-tests 780 Format=<0|1> 781 Print debugging info while doing the locking API 782 self-tests. 783 We default to 0 (no extra messages), setting it to 784 1 will print _a lot_ more information - normally 785 only useful to kernel developers. 786 787 debug_objects [KNL] Enable object debugging 788 789 no_debug_objects 790 [KNL] Disable object debugging 791 792 debug_guardpage_minorder= 793 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this 794 parameter allows control of the order of pages that will 795 be intentionally kept free (and hence protected) by the 796 buddy allocator. Bigger value increase the probability 797 of catching random memory corruption, but reduce the 798 amount of memory for normal system use. The maximum 799 possible value is MAX_ORDER/2. Setting this parameter 800 to 1 or 2 should be enough to identify most random 801 memory corruption problems caused by bugs in kernel or 802 driver code when a CPU writes to (or reads from) a 803 random memory location. Note that there exists a class 804 of memory corruptions problems caused by buggy H/W or 805 F/W or by drivers badly programing DMA (basically when 806 memory is written at bus level and the CPU MMU is 807 bypassed) which are not detectable by 808 CONFIG_DEBUG_PAGEALLOC, hence this option will not help 809 tracking down these problems. 810 811 debug_pagealloc= 812 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 813 enables the feature at boot time. By default, it is 814 disabled and the system will work mostly the same as a 815 kernel built without CONFIG_DEBUG_PAGEALLOC. 816 Note: to get most of debug_pagealloc error reports, it's 817 useful to also enable the page_owner functionality. 818 on: enable the feature 819 820 debugpat [X86] Enable PAT debugging 821 822 decnet.addr= [HW,NET] 823 Format: <area>[,<node>] 824 See also Documentation/networking/decnet.txt. 825 826 default_hugepagesz= 827 [same as hugepagesz=] The size of the default 828 HugeTLB page size. This is the size represented by 829 the legacy /proc/ hugepages APIs, used for SHM, and 830 default size when mounting hugetlbfs filesystems. 831 Defaults to the default architecture's huge page size 832 if not specified. 833 834 deferred_probe_timeout= 835 [KNL] Debugging option to set a timeout in seconds for 836 deferred probe to give up waiting on dependencies to 837 probe. Only specific dependencies (subsystems or 838 drivers) that have opted in will be ignored. A timeout of 0 839 will timeout at the end of initcalls. This option will also 840 dump out devices still on the deferred probe list after 841 retrying. 842 843 dhash_entries= [KNL] 844 Set number of hash buckets for dentry cache. 845 846 disable_1tb_segments [PPC] 847 Disables the use of 1TB hash page table segments. This 848 causes the kernel to fall back to 256MB segments which 849 can be useful when debugging issues that require an SLB 850 miss to occur. 851 852 disable= [IPV6] 853 See Documentation/networking/ipv6.txt. 854 855 hardened_usercopy= 856 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 857 hardening is enabled for this boot. Hardened 858 usercopy checking is used to protect the kernel 859 from reading or writing beyond known memory 860 allocation boundaries as a proactive defense 861 against bounds-checking flaws in the kernel's 862 copy_to_user()/copy_from_user() interface. 863 on Perform hardened usercopy checks (default). 864 off Disable hardened usercopy checks. 865 866 disable_radix [PPC] 867 Disable RADIX MMU mode on POWER9 868 869 disable_tlbie [PPC] 870 Disable TLBIE instruction. Currently does not work 871 with KVM, with HASH MMU, or with coherent accelerators. 872 873 disable_cpu_apicid= [X86,APIC,SMP] 874 Format: <int> 875 The number of initial APIC ID for the 876 corresponding CPU to be disabled at boot, 877 mostly used for the kdump 2nd kernel to 878 disable BSP to wake up multiple CPUs without 879 causing system reset or hang due to sending 880 INIT from AP to BSP. 881 882 perf_v4_pmi= [X86,INTEL] 883 Format: <bool> 884 Disable Intel PMU counter freezing feature. 885 The feature only exists starting from 886 Arch Perfmon v4 (Skylake and newer). 887 888 disable_ddw [PPC/PSERIES] 889 Disable Dynamic DMA Window support. Use this if 890 to workaround buggy firmware. 891 892 disable_ipv6= [IPV6] 893 See Documentation/networking/ipv6.txt. 894 895 disable_mtrr_cleanup [X86] 896 The kernel tries to adjust MTRR layout from continuous 897 to discrete, to make X server driver able to add WB 898 entry later. This parameter disables that. 899 900 disable_mtrr_trim [X86, Intel and AMD only] 901 By default the kernel will trim any uncacheable 902 memory out of your available memory pool based on 903 MTRR settings. This parameter disables that behavior, 904 possibly causing your machine to run very slowly. 905 906 disable_timer_pin_1 [X86] 907 Disable PIN 1 of APIC timer 908 Can be useful to work around chipset bugs. 909 910 dis_ucode_ldr [X86] Disable the microcode loader. 911 912 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 913 this option disables the debugging code at boot. 914 915 dma_debug_entries=<number> 916 This option allows to tune the number of preallocated 917 entries for DMA-API debugging code. One entry is 918 required per DMA-API allocation. Use this if the 919 DMA-API debugging code disables itself because the 920 architectural default is too low. 921 922 dma_debug_driver=<driver_name> 923 With this option the DMA-API debugging driver 924 filter feature can be enabled at boot time. Just 925 pass the driver to filter for as the parameter. 926 The filter can be disabled or changed to another 927 driver later using sysfs. 928 929 driver_async_probe= [KNL] 930 List of driver names to be probed asynchronously. 931 Format: <driver_name1>,<driver_name2>... 932 933 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 934 Broken monitors, graphic adapters, KVMs and EDIDless 935 panels may send no or incorrect EDID data sets. 936 This parameter allows to specify an EDID data sets 937 in the /lib/firmware directory that are used instead. 938 Generic built-in EDID data sets are used, if one of 939 edid/1024x768.bin, edid/1280x1024.bin, 940 edid/1680x1050.bin, or edid/1920x1080.bin is given 941 and no file with the same name exists. Details and 942 instructions how to build your own EDID data are 943 available in Documentation/driver-api/edid.rst. An EDID 944 data set will only be used for a particular connector, 945 if its name and a colon are prepended to the EDID 946 name. Each connector may use a unique EDID data 947 set by separating the files with a comma. An EDID 948 data set with no connector name will be used for 949 any connectors not explicitly specified. 950 951 dscc4.setup= [NET] 952 953 dt_cpu_ftrs= [PPC] 954 Format: {"off" | "known"} 955 Control how the dt_cpu_ftrs device-tree binding is 956 used for CPU feature discovery and setup (if it 957 exists). 958 off: Do not use it, fall back to legacy cpu table. 959 known: Do not pass through unknown features to guests 960 or userspace, only those that the kernel is aware of. 961 962 dump_apple_properties [X86] 963 Dump name and content of EFI device properties on 964 x86 Macs. Useful for driver authors to determine 965 what data is available or for reverse-engineering. 966 967 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 968 module.dyndbg[="val"] 969 Enable debug messages at boot time. See 970 Documentation/admin-guide/dynamic-debug-howto.rst 971 for details. 972 973 nompx [X86] Disables Intel Memory Protection Extensions. 974 See Documentation/x86/intel_mpx.rst for more 975 information about the feature. 976 977 nopku [X86] Disable Memory Protection Keys CPU feature found 978 in some Intel CPUs. 979 980 module.async_probe [KNL] 981 Enable asynchronous probe on this module. 982 983 early_ioremap_debug [KNL] 984 Enable debug messages in early_ioremap support. This 985 is useful for tracking down temporary early mappings 986 which are not unmapped. 987 988 earlycon= [KNL] Output early console device and options. 989 990 When used with no options, the early console is 991 determined by stdout-path property in device tree's 992 chosen node or the ACPI SPCR table if supported by 993 the platform. 994 995 cdns,<addr>[,options] 996 Start an early, polled-mode console on a Cadence 997 (xuartps) serial port at the specified address. Only 998 supported option is baud rate. If baud rate is not 999 specified, the serial port must already be setup and 1000 configured. 1001 1002 uart[8250],io,<addr>[,options] 1003 uart[8250],mmio,<addr>[,options] 1004 uart[8250],mmio32,<addr>[,options] 1005 uart[8250],mmio32be,<addr>[,options] 1006 uart[8250],0x<addr>[,options] 1007 Start an early, polled-mode console on the 8250/16550 1008 UART at the specified I/O port or MMIO address. 1009 MMIO inter-register address stride is either 8-bit 1010 (mmio) or 32-bit (mmio32 or mmio32be). 1011 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1012 to be equivalent to 'mmio'. 'options' are specified 1013 in the same format described for "console=ttyS<n>"; if 1014 unspecified, the h/w is not initialized. 1015 1016 pl011,<addr> 1017 pl011,mmio32,<addr> 1018 Start an early, polled-mode console on a pl011 serial 1019 port at the specified address. The pl011 serial port 1020 must already be setup and configured. Options are not 1021 yet supported. If 'mmio32' is specified, then only 1022 the driver will use only 32-bit accessors to read/write 1023 the device registers. 1024 1025 meson,<addr> 1026 Start an early, polled-mode console on a meson serial 1027 port at the specified address. The serial port must 1028 already be setup and configured. Options are not yet 1029 supported. 1030 1031 msm_serial,<addr> 1032 Start an early, polled-mode console on an msm serial 1033 port at the specified address. The serial port 1034 must already be setup and configured. Options are not 1035 yet supported. 1036 1037 msm_serial_dm,<addr> 1038 Start an early, polled-mode console on an msm serial 1039 dm port at the specified address. The serial port 1040 must already be setup and configured. Options are not 1041 yet supported. 1042 1043 owl,<addr> 1044 Start an early, polled-mode console on a serial port 1045 of an Actions Semi SoC, such as S500 or S900, at the 1046 specified address. The serial port must already be 1047 setup and configured. Options are not yet supported. 1048 1049 rda,<addr> 1050 Start an early, polled-mode console on a serial port 1051 of an RDA Micro SoC, such as RDA8810PL, at the 1052 specified address. The serial port must already be 1053 setup and configured. Options are not yet supported. 1054 1055 sbi 1056 Use RISC-V SBI (Supervisor Binary Interface) for early 1057 console. 1058 1059 smh Use ARM semihosting calls for early console. 1060 1061 s3c2410,<addr> 1062 s3c2412,<addr> 1063 s3c2440,<addr> 1064 s3c6400,<addr> 1065 s5pv210,<addr> 1066 exynos4210,<addr> 1067 Use early console provided by serial driver available 1068 on Samsung SoCs, requires selecting proper type and 1069 a correct base address of the selected UART port. The 1070 serial port must already be setup and configured. 1071 Options are not yet supported. 1072 1073 lantiq,<addr> 1074 Start an early, polled-mode console on a lantiq serial 1075 (lqasc) port at the specified address. The serial port 1076 must already be setup and configured. Options are not 1077 yet supported. 1078 1079 lpuart,<addr> 1080 lpuart32,<addr> 1081 Use early console provided by Freescale LP UART driver 1082 found on Freescale Vybrid and QorIQ LS1021A processors. 1083 A valid base address must be provided, and the serial 1084 port must already be setup and configured. 1085 1086 ar3700_uart,<addr> 1087 Start an early, polled-mode console on the 1088 Armada 3700 serial port at the specified 1089 address. The serial port must already be setup 1090 and configured. Options are not yet supported. 1091 1092 qcom_geni,<addr> 1093 Start an early, polled-mode console on a Qualcomm 1094 Generic Interface (GENI) based serial port at the 1095 specified address. The serial port must already be 1096 setup and configured. Options are not yet supported. 1097 1098 efifb,[options] 1099 Start an early, unaccelerated console on the EFI 1100 memory mapped framebuffer (if available). On cache 1101 coherent non-x86 systems that use system memory for 1102 the framebuffer, pass the 'ram' option so that it is 1103 mapped with the correct attributes. 1104 1105 linflex,<addr> 1106 Use early console provided by Freescale LINFlexD UART 1107 serial driver for NXP S32V234 SoCs. A valid base 1108 address must be provided, and the serial port must 1109 already be setup and configured. 1110 1111 earlyprintk= [X86,SH,ARM,M68k,S390] 1112 earlyprintk=vga 1113 earlyprintk=sclp 1114 earlyprintk=xen 1115 earlyprintk=serial[,ttySn[,baudrate]] 1116 earlyprintk=serial[,0x...[,baudrate]] 1117 earlyprintk=ttySn[,baudrate] 1118 earlyprintk=dbgp[debugController#] 1119 earlyprintk=pciserial[,force],bus:device.function[,baudrate] 1120 earlyprintk=xdbc[xhciController#] 1121 1122 earlyprintk is useful when the kernel crashes before 1123 the normal console is initialized. It is not enabled by 1124 default because it has some cosmetic problems. 1125 1126 Append ",keep" to not disable it when the real console 1127 takes over. 1128 1129 Only one of vga, efi, serial, or usb debug port can 1130 be used at a time. 1131 1132 Currently only ttyS0 and ttyS1 may be specified by 1133 name. Other I/O ports may be explicitly specified 1134 on some architectures (x86 and arm at least) by 1135 replacing ttySn with an I/O port address, like this: 1136 earlyprintk=serial,0x1008,115200 1137 You can find the port for a given device in 1138 /proc/tty/driver/serial: 1139 2: uart:ST16650V2 port:00001008 irq:18 ... 1140 1141 Interaction with the standard serial driver is not 1142 very good. 1143 1144 The VGA and EFI output is eventually overwritten by 1145 the real console. 1146 1147 The xen output can only be used by Xen PV guests. 1148 1149 The sclp output can only be used on s390. 1150 1151 The optional "force" to "pciserial" enables use of a 1152 PCI device even when its classcode is not of the 1153 UART class. 1154 1155 edac_report= [HW,EDAC] Control how to report EDAC event 1156 Format: {"on" | "off" | "force"} 1157 on: enable EDAC to report H/W event. May be overridden 1158 by other higher priority error reporting module. 1159 off: disable H/W event reporting through EDAC. 1160 force: enforce the use of EDAC to report H/W event. 1161 default: on. 1162 1163 ekgdboc= [X86,KGDB] Allow early kernel console debugging 1164 ekgdboc=kbd 1165 1166 This is designed to be used in conjunction with 1167 the boot argument: earlyprintk=vga 1168 1169 edd= [EDD] 1170 Format: {"off" | "on" | "skip[mbr]"} 1171 1172 efi= [EFI] 1173 Format: { "old_map", "nochunk", "noruntime", "debug", 1174 "nosoftreserve" } 1175 old_map [X86-64]: switch to the old ioremap-based EFI 1176 runtime services mapping. 32-bit still uses this one by 1177 default. 1178 nochunk: disable reading files in "chunks" in the EFI 1179 boot stub, as chunking can cause problems with some 1180 firmware implementations. 1181 noruntime : disable EFI runtime services support 1182 debug: enable misc debug output 1183 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1184 attribute may cause the kernel to reserve the 1185 memory range for a memory mapping driver to 1186 claim. Specify efi=nosoftreserve to disable this 1187 reservation and treat the memory by its base type 1188 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1189 1190 efi_no_storage_paranoia [EFI; X86] 1191 Using this parameter you can use more than 50% of 1192 your efi variable storage. Use this parameter only if 1193 you are really sure that your UEFI does sane gc and 1194 fulfills the spec otherwise your board may brick. 1195 1196 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86] 1197 Add arbitrary attribute to specific memory range by 1198 updating original EFI memory map. 1199 Region of memory which aa attribute is added to is 1200 from ss to ss+nn. 1201 1202 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000 1203 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000) 1204 attribute is added to range 0x100000000-0x180000000 and 1205 0x10a0000000-0x1120000000. 1206 1207 If efi_fake_mem=8G@9G:0x40000 is specified, the 1208 EFI_MEMORY_SP(0x40000) attribute is added to 1209 range 0x240000000-0x43fffffff. 1210 1211 Using this parameter you can do debugging of EFI memmap 1212 related features. For example, you can do debugging of 1213 Address Range Mirroring feature even if your box 1214 doesn't support it, or mark specific memory as 1215 "soft reserved". 1216 1217 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1218 that is to be dynamically loaded by Linux. If there are 1219 multiple variables with the same name but with different 1220 vendor GUIDs, all of them will be loaded. See 1221 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1222 1223 1224 eisa_irq_edge= [PARISC,HW] 1225 See header of drivers/parisc/eisa.c. 1226 1227 elanfreq= [X86-32] 1228 See comment before function elanfreq_setup() in 1229 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1230 1231 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390] 1232 Specifies physical address of start of kernel core 1233 image elf header and optionally the size. Generally 1234 kexec loader will pass this option to capture kernel. 1235 See Documentation/admin-guide/kdump/kdump.rst for details. 1236 1237 enable_mtrr_cleanup [X86] 1238 The kernel tries to adjust MTRR layout from continuous 1239 to discrete, to make X server driver able to add WB 1240 entry later. This parameter enables that. 1241 1242 enable_timer_pin_1 [X86] 1243 Enable PIN 1 of APIC timer 1244 Can be useful to work around chipset bugs 1245 (in particular on some ATI chipsets). 1246 The kernel tries to set a reasonable default. 1247 1248 enforcing [SELINUX] Set initial enforcing status. 1249 Format: {"0" | "1"} 1250 See security/selinux/Kconfig help text. 1251 0 -- permissive (log only, no denials). 1252 1 -- enforcing (deny and log). 1253 Default value is 0. 1254 Value can be changed at runtime via /selinux/enforce. 1255 1256 erst_disable [ACPI] 1257 Disable Error Record Serialization Table (ERST) 1258 support. 1259 1260 ether= [HW,NET] Ethernet cards parameters 1261 This option is obsoleted by the "netdev=" option, which 1262 has equivalent usage. See its documentation for details. 1263 1264 evm= [EVM] 1265 Format: { "fix" } 1266 Permit 'security.evm' to be updated regardless of 1267 current integrity status. 1268 1269 failslab= 1270 fail_page_alloc= 1271 fail_make_request=[KNL] 1272 General fault injection mechanism. 1273 Format: <interval>,<probability>,<space>,<times> 1274 See also Documentation/fault-injection/. 1275 1276 floppy= [HW] 1277 See Documentation/admin-guide/blockdev/floppy.rst. 1278 1279 force_pal_cache_flush 1280 [IA-64] Avoid check_sal_cache_flush which may hang on 1281 buggy SAL_CACHE_FLUSH implementations. Using this 1282 parameter will force ia64_sal_cache_flush to call 1283 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH. 1284 1285 forcepae [X86-32] 1286 Forcefully enable Physical Address Extension (PAE). 1287 Many Pentium M systems disable PAE but may have a 1288 functionally usable PAE implementation. 1289 Warning: use of this parameter will taint the kernel 1290 and may cause unknown problems. 1291 1292 ftrace=[tracer] 1293 [FTRACE] will set and start the specified tracer 1294 as early as possible in order to facilitate early 1295 boot debugging. 1296 1297 ftrace_dump_on_oops[=orig_cpu] 1298 [FTRACE] will dump the trace buffers on oops. 1299 If no parameter is passed, ftrace will dump 1300 buffers of all CPUs, but if you pass orig_cpu, it will 1301 dump only the buffer of the CPU that triggered the 1302 oops. 1303 1304 ftrace_filter=[function-list] 1305 [FTRACE] Limit the functions traced by the function 1306 tracer at boot up. function-list is a comma separated 1307 list of functions. This list can be changed at run 1308 time by the set_ftrace_filter file in the debugfs 1309 tracing directory. 1310 1311 ftrace_notrace=[function-list] 1312 [FTRACE] Do not trace the functions specified in 1313 function-list. This list can be changed at run time 1314 by the set_ftrace_notrace file in the debugfs 1315 tracing directory. 1316 1317 ftrace_graph_filter=[function-list] 1318 [FTRACE] Limit the top level callers functions traced 1319 by the function graph tracer at boot up. 1320 function-list is a comma separated list of functions 1321 that can be changed at run time by the 1322 set_graph_function file in the debugfs tracing directory. 1323 1324 ftrace_graph_notrace=[function-list] 1325 [FTRACE] Do not trace from the functions specified in 1326 function-list. This list is a comma separated list of 1327 functions that can be changed at run time by the 1328 set_graph_notrace file in the debugfs tracing directory. 1329 1330 ftrace_graph_max_depth=<uint> 1331 [FTRACE] Used with the function graph tracer. This is 1332 the max depth it will trace into a function. This value 1333 can be changed at run time by the max_graph_depth file 1334 in the tracefs tracing directory. default: 0 (no limit) 1335 1336 gamecon.map[2|3]= 1337 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1338 support via parallel port (up to 5 devices per port) 1339 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1340 See also Documentation/input/devices/joystick-parport.rst 1341 1342 gamma= [HW,DRM] 1343 1344 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART 1345 Format: off | on 1346 default: on 1347 1348 gcov_persist= [GCOV] When non-zero (default), profiling data for 1349 kernel modules is saved and remains accessible via 1350 debugfs, even when the module is unloaded/reloaded. 1351 When zero, profiling data is discarded and associated 1352 debugfs files are removed at module unload time. 1353 1354 goldfish [X86] Enable the goldfish android emulator platform. 1355 Don't use this when you are not running on the 1356 android emulator 1357 1358 gpt [EFI] Forces disk with valid GPT signature but 1359 invalid Protective MBR to be treated as GPT. If the 1360 primary GPT is corrupted, it enables the backup/alternate 1361 GPT to be used instead. 1362 1363 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1364 the "Enable 0" bit of the configuration register. 1365 Format: 0 | 1 1366 Default: 0 1367 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1368 the "Enable 0" bit of the configuration register. 1369 Format: 0 | 1 1370 Default: 0 1371 grcan.select= [HW] Select which physical interface to use. 1372 Format: 0 | 1 1373 Default: 0 1374 grcan.txsize= [HW] Sets the size of the tx buffer. 1375 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1376 Default: 1024 1377 grcan.rxsize= [HW] Sets the size of the rx buffer. 1378 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1379 Default: 1024 1380 1381 gpio-mockup.gpio_mockup_ranges 1382 [HW] Sets the ranges of gpiochip of for this device. 1383 Format: <start1>,<end1>,<start2>,<end2>... 1384 1385 hardlockup_all_cpu_backtrace= 1386 [KNL] Should the hard-lockup detector generate 1387 backtraces on all cpus. 1388 Format: <integer> 1389 1390 hashdist= [KNL,NUMA] Large hashes allocated during boot 1391 are distributed across NUMA nodes. Defaults on 1392 for 64-bit NUMA, off otherwise. 1393 Format: 0 | 1 (for off | on) 1394 1395 hcl= [IA-64] SGI's Hardware Graph compatibility layer 1396 1397 hd= [EIDE] (E)IDE hard drive subsystem geometry 1398 Format: <cyl>,<head>,<sect> 1399 1400 hest_disable [ACPI] 1401 Disable Hardware Error Source Table (HEST) support; 1402 corresponding firmware-first mode error processing 1403 logic will be disabled. 1404 1405 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact 1406 size of <nn>. This works even on boxes that have no 1407 highmem otherwise. This also works to reduce highmem 1408 size on bigger boxes. 1409 1410 highres= [KNL] Enable/disable high resolution timer mode. 1411 Valid parameters: "on", "off" 1412 Default: "on" 1413 1414 hlt [BUGS=ARM,SH] 1415 1416 hpet= [X86-32,HPET] option to control HPET usage 1417 Format: { enable (default) | disable | force | 1418 verbose } 1419 disable: disable HPET and use PIT instead 1420 force: allow force enabled of undocumented chips (ICH4, 1421 VIA, nVidia) 1422 verbose: show contents of HPET registers during setup 1423 1424 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 1425 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 1426 1427 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot. 1428 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages. 1429 On x86-64 and powerpc, this option can be specified 1430 multiple times interleaved with hugepages= to reserve 1431 huge pages of different sizes. Valid pages sizes on 1432 x86-64 are 2M (when the CPU supports "pse") and 1G 1433 (when the CPU supports the "pdpe1gb" cpuinfo flag). 1434 1435 hung_task_panic= 1436 [KNL] Should the hung task detector generate panics. 1437 Format: <integer> 1438 1439 A nonzero value instructs the kernel to panic when a 1440 hung task is detected. The default value is controlled 1441 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time 1442 option. The value selected by this boot parameter can 1443 be changed later by the kernel.hung_task_panic sysctl. 1444 1445 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 1446 terminal devices. Valid values: 0..8 1447 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 1448 If specified, z/VM IUCV HVC accepts connections 1449 from listed z/VM user IDs only. 1450 1451 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations 1452 which allow the hypervisor to 'idle' the 1453 guest on lock contention. 1454 1455 keep_bootcon [KNL] 1456 Do not unregister boot console at start. This is only 1457 useful for debugging when something happens in the window 1458 between unregistering the boot console and initializing 1459 the real console. 1460 1461 i2c_bus= [HW] Override the default board specific I2C bus speed 1462 or register an additional I2C bus that is not 1463 registered from board initialization code. 1464 Format: 1465 <bus_id>,<clkrate> 1466 1467 i8042.debug [HW] Toggle i8042 debug mode 1468 i8042.unmask_kbd_data 1469 [HW] Enable printing of interrupt data from the KBD port 1470 (disabled by default, and as a pre-condition 1471 requires that i8042.debug=1 be enabled) 1472 i8042.direct [HW] Put keyboard port into non-translated mode 1473 i8042.dumbkbd [HW] Pretend that controller can only read data from 1474 keyboard and cannot control its state 1475 (Don't attempt to blink the leds) 1476 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 1477 i8042.nokbd [HW] Don't check/create keyboard port 1478 i8042.noloop [HW] Disable the AUX Loopback command while probing 1479 for the AUX port 1480 i8042.nomux [HW] Don't check presence of an active multiplexing 1481 controller 1482 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 1483 controllers 1484 i8042.notimeout [HW] Ignore timeout condition signalled by controller 1485 i8042.reset [HW] Reset the controller during init, cleanup and 1486 suspend-to-ram transitions, only during s2r 1487 transitions, or never reset 1488 Format: { 1 | Y | y | 0 | N | n } 1489 1, Y, y: always reset controller 1490 0, N, n: don't ever reset controller 1491 Default: only on s2r transitions on x86; most other 1492 architectures force reset to be always executed 1493 i8042.unlock [HW] Unlock (ignore) the keylock 1494 i8042.kbdreset [HW] Reset device connected to KBD port 1495 1496 i810= [HW,DRM] 1497 1498 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data 1499 indicates that the driver is running on unsupported 1500 hardware. 1501 i8k.force [HW] Activate i8k driver even if SMM BIOS signature 1502 does not match list of supported models. 1503 i8k.power_status 1504 [HW] Report power status in /proc/i8k 1505 (disabled by default) 1506 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN 1507 capability is set. 1508 1509 i915.invert_brightness= 1510 [DRM] Invert the sense of the variable that is used to 1511 set the brightness of the panel backlight. Normally a 1512 brightness value of 0 indicates backlight switched off, 1513 and the maximum of the brightness value sets the backlight 1514 to maximum brightness. If this parameter is set to 0 1515 (default) and the machine requires it, or this parameter 1516 is set to 1, a brightness value of 0 sets the backlight 1517 to maximum brightness, and the maximum of the brightness 1518 value switches the backlight off. 1519 -1 -- never invert brightness 1520 0 -- machine default 1521 1 -- force brightness inversion 1522 1523 icn= [HW,ISDN] 1524 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 1525 1526 ide-core.nodma= [HW] (E)IDE subsystem 1527 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc 1528 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr 1529 .cdrom .chs .ignore_cable are additional options 1530 See Documentation/ide/ide.rst. 1531 1532 ide-generic.probe-mask= [HW] (E)IDE subsystem 1533 Format: <int> 1534 Probe mask for legacy ISA IDE ports. Depending on 1535 platform up to 6 ports are supported, enabled by 1536 setting corresponding bits in the mask to 1. The 1537 default value is 0x0, which has a special meaning. 1538 On systems that have PCI, it triggers scanning the 1539 PCI bus for the first and the second port, which 1540 are then probed. On systems without PCI the value 1541 of 0x0 enables probing the two first ports as if it 1542 was 0x3. 1543 1544 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem 1545 Claim all unknown PCI IDE storage controllers. 1546 1547 idle= [X86] 1548 Format: idle=poll, idle=halt, idle=nomwait 1549 Poll forces a polling idle loop that can slightly 1550 improve the performance of waking up a idle CPU, but 1551 will use a lot of power and make the system run hot. 1552 Not recommended. 1553 idle=halt: Halt is forced to be used for CPU idle. 1554 In such case C2/C3 won't be used again. 1555 idle=nomwait: Disable mwait for CPU C-states 1556 1557 ieee754= [MIPS] Select IEEE Std 754 conformance mode 1558 Format: { strict | legacy | 2008 | relaxed } 1559 Default: strict 1560 1561 Choose which programs will be accepted for execution 1562 based on the IEEE 754 NaN encoding(s) supported by 1563 the FPU and the NaN encoding requested with the value 1564 of an ELF file header flag individually set by each 1565 binary. Hardware implementations are permitted to 1566 support either or both of the legacy and the 2008 NaN 1567 encoding mode. 1568 1569 Available settings are as follows: 1570 strict accept binaries that request a NaN encoding 1571 supported by the FPU 1572 legacy only accept legacy-NaN binaries, if supported 1573 by the FPU 1574 2008 only accept 2008-NaN binaries, if supported 1575 by the FPU 1576 relaxed accept any binaries regardless of whether 1577 supported by the FPU 1578 1579 The FPU emulator is always able to support both NaN 1580 encodings, so if no FPU hardware is present or it has 1581 been disabled with 'nofpu', then the settings of 1582 'legacy' and '2008' strap the emulator accordingly, 1583 'relaxed' straps the emulator for both legacy-NaN and 1584 2008-NaN, whereas 'strict' enables legacy-NaN only on 1585 legacy processors and both NaN encodings on MIPS32 or 1586 MIPS64 CPUs. 1587 1588 The setting for ABS.fmt/NEG.fmt instruction execution 1589 mode generally follows that for the NaN encoding, 1590 except where unsupported by hardware. 1591 1592 ignore_loglevel [KNL] 1593 Ignore loglevel setting - this will print /all/ 1594 kernel messages to the console. Useful for debugging. 1595 We also add it as printk module parameter, so users 1596 could change it dynamically, usually by 1597 /sys/module/printk/parameters/ignore_loglevel. 1598 1599 ignore_rlimit_data 1600 Ignore RLIMIT_DATA setting for data mappings, 1601 print warning at first misuse. Can be changed via 1602 /sys/module/kernel/parameters/ignore_rlimit_data. 1603 1604 ihash_entries= [KNL] 1605 Set number of hash buckets for inode cache. 1606 1607 ima_appraise= [IMA] appraise integrity measurements 1608 Format: { "off" | "enforce" | "fix" | "log" } 1609 default: "enforce" 1610 1611 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 1612 The builtin appraise policy appraises all files 1613 owned by uid=0. 1614 1615 ima_canonical_fmt [IMA] 1616 Use the canonical format for the binary runtime 1617 measurements, instead of host native format. 1618 1619 ima_hash= [IMA] 1620 Format: { md5 | sha1 | rmd160 | sha256 | sha384 1621 | sha512 | ... } 1622 default: "sha1" 1623 1624 The list of supported hash algorithms is defined 1625 in crypto/hash_info.h. 1626 1627 ima_policy= [IMA] 1628 The builtin policies to load during IMA setup. 1629 Format: "tcb | appraise_tcb | secure_boot | 1630 fail_securely" 1631 1632 The "tcb" policy measures all programs exec'd, files 1633 mmap'd for exec, and all files opened with the read 1634 mode bit set by either the effective uid (euid=0) or 1635 uid=0. 1636 1637 The "appraise_tcb" policy appraises the integrity of 1638 all files owned by root. 1639 1640 The "secure_boot" policy appraises the integrity 1641 of files (eg. kexec kernel image, kernel modules, 1642 firmware, policy, etc) based on file signatures. 1643 1644 The "fail_securely" policy forces file signature 1645 verification failure also on privileged mounted 1646 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 1647 flag. 1648 1649 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 1650 Load a policy which meets the needs of the Trusted 1651 Computing Base. This means IMA will measure all 1652 programs exec'd, files mmap'd for exec, and all files 1653 opened for read by uid=0. 1654 1655 ima_template= [IMA] 1656 Select one of defined IMA measurements template formats. 1657 Formats: { "ima" | "ima-ng" | "ima-sig" } 1658 Default: "ima-ng" 1659 1660 ima_template_fmt= 1661 [IMA] Define a custom template format. 1662 Format: { "field1|...|fieldN" } 1663 1664 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 1665 Format: <min_file_size> 1666 Set the minimal file size for using asynchronous hash. 1667 If left unspecified, ahash usage is disabled. 1668 1669 ahash performance varies for different data sizes on 1670 different crypto accelerators. This option can be used 1671 to achieve the best performance for a particular HW. 1672 1673 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size 1674 Format: <bufsize> 1675 Set hashing buffer size. Default: 4k. 1676 1677 ahash performance varies for different chunk sizes on 1678 different crypto accelerators. This option can be used 1679 to achieve best performance for particular HW. 1680 1681 init= [KNL] 1682 Format: <full_path> 1683 Run specified binary instead of /sbin/init as init 1684 process. 1685 1686 initcall_debug [KNL] Trace initcalls as they are executed. Useful 1687 for working out where the kernel is dying during 1688 startup. 1689 1690 initcall_blacklist= [KNL] Do not execute a comma-separated list of 1691 initcall functions. Useful for debugging built-in 1692 modules and initcalls. 1693 1694 initrd= [BOOT] Specify the location of the initial ramdisk 1695 1696 init_on_alloc= [MM] Fill newly allocated pages and heap objects with 1697 zeroes. 1698 Format: 0 | 1 1699 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 1700 1701 init_on_free= [MM] Fill freed pages and heap objects with zeroes. 1702 Format: 0 | 1 1703 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 1704 1705 init_pkru= [x86] Specify the default memory protection keys rights 1706 register contents for all processes. 0x55555554 by 1707 default (disallow access to all but pkey 0). Can 1708 override in debugfs after boot. 1709 1710 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 1711 Format: <irq> 1712 1713 int_pln_enable [x86] Enable power limit notification interrupt 1714 1715 integrity_audit=[IMA] 1716 Format: { "0" | "1" } 1717 0 -- basic integrity auditing messages. (Default) 1718 1 -- additional integrity auditing messages. 1719 1720 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 1721 on 1722 Enable intel iommu driver. 1723 off 1724 Disable intel iommu driver. 1725 igfx_off [Default Off] 1726 By default, gfx is mapped as normal device. If a gfx 1727 device has a dedicated DMAR unit, the DMAR unit is 1728 bypassed by not enabling DMAR with this option. In 1729 this case, gfx device will use physical address for 1730 DMA. 1731 forcedac [x86_64] 1732 With this option iommu will not optimize to look 1733 for io virtual address below 32-bit forcing dual 1734 address cycle on pci bus for cards supporting greater 1735 than 32-bit addressing. The default is to look 1736 for translation below 32-bit and if not available 1737 then look in the higher range. 1738 strict [Default Off] 1739 With this option on every unmap_single operation will 1740 result in a hardware IOTLB flush operation as opposed 1741 to batching them for performance. 1742 sp_off [Default Off] 1743 By default, super page will be supported if Intel IOMMU 1744 has the capability. With this option, super page will 1745 not be supported. 1746 sm_on [Default Off] 1747 By default, scalable mode will be disabled even if the 1748 hardware advertises that it has support for the scalable 1749 mode translation. With this option set, scalable mode 1750 will be used on hardware which claims to support it. 1751 tboot_noforce [Default Off] 1752 Do not force the Intel IOMMU enabled under tboot. 1753 By default, tboot will force Intel IOMMU on, which 1754 could harm performance of some high-throughput 1755 devices like 40GBit network cards, even if identity 1756 mapping is enabled. 1757 Note that using this option lowers the security 1758 provided by tboot because it makes the system 1759 vulnerable to DMA attacks. 1760 nobounce [Default off] 1761 Disable bounce buffer for unstrusted devices such as 1762 the Thunderbolt devices. This will treat the untrusted 1763 devices as the trusted ones, hence might expose security 1764 risks of DMA attacks. 1765 1766 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 1767 0 disables intel_idle and fall back on acpi_idle. 1768 1 to 9 specify maximum depth of C-state. 1769 1770 intel_pstate= [X86] 1771 disable 1772 Do not enable intel_pstate as the default 1773 scaling driver for the supported processors 1774 passive 1775 Use intel_pstate as a scaling driver, but configure it 1776 to work with generic cpufreq governors (instead of 1777 enabling its internal governor). This mode cannot be 1778 used along with the hardware-managed P-states (HWP) 1779 feature. 1780 force 1781 Enable intel_pstate on systems that prohibit it by default 1782 in favor of acpi-cpufreq. Forcing the intel_pstate driver 1783 instead of acpi-cpufreq may disable platform features, such 1784 as thermal controls and power capping, that rely on ACPI 1785 P-States information being indicated to OSPM and therefore 1786 should be used with caution. This option does not work with 1787 processors that aren't supported by the intel_pstate driver 1788 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 1789 no_hwp 1790 Do not enable hardware P state control (HWP) 1791 if available. 1792 hwp_only 1793 Only load intel_pstate on systems which support 1794 hardware P state control (HWP) if available. 1795 support_acpi_ppc 1796 Enforce ACPI _PPC performance limits. If the Fixed ACPI 1797 Description Table, specifies preferred power management 1798 profile as "Enterprise Server" or "Performance Server", 1799 then this feature is turned on by default. 1800 per_cpu_perf_limits 1801 Allow per-logical-CPU P-State performance control limits using 1802 cpufreq sysfs interface 1803 1804 intremap= [X86-64, Intel-IOMMU] 1805 on enable Interrupt Remapping (default) 1806 off disable Interrupt Remapping 1807 nosid disable Source ID checking 1808 no_x2apic_optout 1809 BIOS x2APIC opt-out request will be ignored 1810 nopost disable Interrupt Posting 1811 1812 iomem= Disable strict checking of access to MMIO memory 1813 strict regions from userspace. 1814 relaxed 1815 1816 iommu= [x86] 1817 off 1818 force 1819 noforce 1820 biomerge 1821 panic 1822 nopanic 1823 merge 1824 nomerge 1825 soft 1826 pt [x86] 1827 nopt [x86] 1828 nobypass [PPC/POWERNV] 1829 Disable IOMMU bypass, using IOMMU for PCI devices. 1830 1831 iommu.strict= [ARM64] Configure TLB invalidation behaviour 1832 Format: { "0" | "1" } 1833 0 - Lazy mode. 1834 Request that DMA unmap operations use deferred 1835 invalidation of hardware TLBs, for increased 1836 throughput at the cost of reduced device isolation. 1837 Will fall back to strict mode if not supported by 1838 the relevant IOMMU driver. 1839 1 - Strict mode (default). 1840 DMA unmap operations invalidate IOMMU hardware TLBs 1841 synchronously. 1842 1843 iommu.passthrough= 1844 [ARM64, X86] Configure DMA to bypass the IOMMU by default. 1845 Format: { "0" | "1" } 1846 0 - Use IOMMU translation for DMA. 1847 1 - Bypass the IOMMU for DMA. 1848 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 1849 1850 io7= [HW] IO7 for Marvel based alpha systems 1851 See comment before marvel_specify_io7 in 1852 arch/alpha/kernel/core_marvel.c. 1853 1854 io_delay= [X86] I/O delay method 1855 0x80 1856 Standard port 0x80 based delay 1857 0xed 1858 Alternate port 0xed based delay (needed on some systems) 1859 udelay 1860 Simple two microseconds delay 1861 none 1862 No delay 1863 1864 ip= [IP_PNP] 1865 See Documentation/filesystems/nfs/nfsroot.txt. 1866 1867 ipcmni_extend [KNL] Extend the maximum number of unique System V 1868 IPC identifiers from 32,768 to 16,777,216. 1869 1870 irqaffinity= [SMP] Set the default irq affinity mask 1871 The argument is a cpu list, as described above. 1872 1873 irqchip.gicv2_force_probe= 1874 [ARM, ARM64] 1875 Format: <bool> 1876 Force the kernel to look for the second 4kB page 1877 of a GICv2 controller even if the memory range 1878 exposed by the device tree is too small. 1879 1880 irqchip.gicv3_nolpi= 1881 [ARM, ARM64] 1882 Force the kernel to ignore the availability of 1883 LPIs (and by consequence ITSs). Intended for system 1884 that use the kernel as a bootloader, and thus want 1885 to let secondary kernels in charge of setting up 1886 LPIs. 1887 1888 irqchip.gicv3_pseudo_nmi= [ARM64] 1889 Enables support for pseudo-NMIs in the kernel. This 1890 requires the kernel to be built with 1891 CONFIG_ARM64_PSEUDO_NMI. 1892 1893 irqfixup [HW] 1894 When an interrupt is not handled search all handlers 1895 for it. Intended to get systems with badly broken 1896 firmware running. 1897 1898 irqpoll [HW] 1899 When an interrupt is not handled search all handlers 1900 for it. Also check all handlers each timer 1901 interrupt. Intended to get systems with badly broken 1902 firmware running. 1903 1904 isapnp= [ISAPNP] 1905 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 1906 1907 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 1908 [Deprecated - use cpusets instead] 1909 Format: [flag-list,]<cpu-list> 1910 1911 Specify one or more CPUs to isolate from disturbances 1912 specified in the flag list (default: domain): 1913 1914 nohz 1915 Disable the tick when a single task runs. 1916 1917 A residual 1Hz tick is offloaded to workqueues, which you 1918 need to affine to housekeeping through the global 1919 workqueue's affinity configured via the 1920 /sys/devices/virtual/workqueue/cpumask sysfs file, or 1921 by using the 'domain' flag described below. 1922 1923 NOTE: by default the global workqueue runs on all CPUs, 1924 so to protect individual CPUs the 'cpumask' file has to 1925 be configured manually after bootup. 1926 1927 domain 1928 Isolate from the general SMP balancing and scheduling 1929 algorithms. Note that performing domain isolation this way 1930 is irreversible: it's not possible to bring back a CPU to 1931 the domains once isolated through isolcpus. It's strongly 1932 advised to use cpusets instead to disable scheduler load 1933 balancing through the "cpuset.sched_load_balance" file. 1934 It offers a much more flexible interface where CPUs can 1935 move in and out of an isolated set anytime. 1936 1937 You can move a process onto or off an "isolated" CPU via 1938 the CPU affinity syscalls or cpuset. 1939 <cpu number> begins at 0 and the maximum value is 1940 "number of CPUs in system - 1". 1941 1942 The format of <cpu-list> is described above. 1943 1944 1945 1946 iucv= [HW,NET] 1947 1948 ivrs_ioapic [HW,X86_64] 1949 Provide an override to the IOAPIC-ID<->DEVICE-ID 1950 mapping provided in the IVRS ACPI table. For 1951 example, to map IOAPIC-ID decimal 10 to 1952 PCI device 00:14.0 write the parameter as: 1953 ivrs_ioapic[10]=00:14.0 1954 1955 ivrs_hpet [HW,X86_64] 1956 Provide an override to the HPET-ID<->DEVICE-ID 1957 mapping provided in the IVRS ACPI table. For 1958 example, to map HPET-ID decimal 0 to 1959 PCI device 00:14.0 write the parameter as: 1960 ivrs_hpet[0]=00:14.0 1961 1962 ivrs_acpihid [HW,X86_64] 1963 Provide an override to the ACPI-HID:UID<->DEVICE-ID 1964 mapping provided in the IVRS ACPI table. For 1965 example, to map UART-HID:UID AMD0020:0 to 1966 PCI device 00:14.5 write the parameter as: 1967 ivrs_acpihid[00:14.5]=AMD0020:0 1968 1969 js= [HW,JOY] Analog joystick 1970 See Documentation/input/joydev/joystick.rst. 1971 1972 nokaslr [KNL] 1973 When CONFIG_RANDOMIZE_BASE is set, this disables 1974 kernel and module base offset ASLR (Address Space 1975 Layout Randomization). 1976 1977 kasan_multi_shot 1978 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 1979 report on every invalid memory access. Without this 1980 parameter KASAN will print report only for the first 1981 invalid access. 1982 1983 keepinitrd [HW,ARM] 1984 1985 kernelcore= [KNL,X86,IA-64,PPC] 1986 Format: nn[KMGTPE] | nn% | "mirror" 1987 This parameter specifies the amount of memory usable by 1988 the kernel for non-movable allocations. The requested 1989 amount is spread evenly throughout all nodes in the 1990 system as ZONE_NORMAL. The remaining memory is used for 1991 movable memory in its own zone, ZONE_MOVABLE. In the 1992 event, a node is too small to have both ZONE_NORMAL and 1993 ZONE_MOVABLE, kernelcore memory will take priority and 1994 other nodes will have a larger ZONE_MOVABLE. 1995 1996 ZONE_MOVABLE is used for the allocation of pages that 1997 may be reclaimed or moved by the page migration 1998 subsystem. Note that allocations like PTEs-from-HighMem 1999 still use the HighMem zone if it exists, and the Normal 2000 zone if it does not. 2001 2002 It is possible to specify the exact amount of memory in 2003 the form of "nn[KMGTPE]", a percentage of total system 2004 memory in the form of "nn%", or "mirror". If "mirror" 2005 option is specified, mirrored (reliable) memory is used 2006 for non-movable allocations and remaining memory is used 2007 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2008 are exclusive, so you cannot specify multiple forms. 2009 2010 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port. 2011 Format: <Controller#>[,poll interval] 2012 The controller # is the number of the ehci usb debug 2013 port as it is probed via PCI. The poll interval is 2014 optional and is the number seconds in between 2015 each poll cycle to the debug port in case you need 2016 the functionality for interrupting the kernel with 2017 gdb or control-c on the dbgp connection. When 2018 not using this parameter you use sysrq-g to break into 2019 the kernel debugger. 2020 2021 kgdboc= [KGDB,HW] kgdb over consoles. 2022 Requires a tty driver that supports console polling, 2023 or a supported polling keyboard driver (non-usb). 2024 Serial only format: <serial_device>[,baud] 2025 keyboard only format: kbd 2026 keyboard and serial format: kbd,<serial_device>[,baud] 2027 Optional Kernel mode setting: 2028 kms, kbd format: kms,kbd 2029 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2030 2031 kgdbwait [KGDB] Stop kernel execution and enter the 2032 kernel debugger at the earliest opportunity. 2033 2034 kmac= [MIPS] korina ethernet MAC address. 2035 Configure the RouterBoard 532 series on-chip 2036 Ethernet adapter MAC address. 2037 2038 kmemleak= [KNL] Boot-time kmemleak enable/disable 2039 Valid arguments: on, off 2040 Default: on 2041 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 2042 the default is off. 2043 2044 kprobe_event=[probe-list] 2045 [FTRACE] Add kprobe events and enable at boot time. 2046 The probe-list is a semicolon delimited list of probe 2047 definitions. Each definition is same as kprobe_events 2048 interface, but the parameters are comma delimited. 2049 For example, to add a kprobe event on vfs_read with 2050 arg1 and arg2, add to the command line; 2051 2052 kprobe_event=p,vfs_read,$arg1,$arg2 2053 2054 See also Documentation/trace/kprobetrace.rst "Kernel 2055 Boot Parameter" section. 2056 2057 kpti= [ARM64] Control page table isolation of user 2058 and kernel address spaces. 2059 Default: enabled on cores which need mitigation. 2060 0: force disabled 2061 1: force enabled 2062 2063 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 2064 Default is 0 (don't ignore, but inject #GP) 2065 2066 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 2067 Default is false (don't support). 2068 2069 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit 2070 KVM MMU at runtime. 2071 Default is 0 (off) 2072 2073 kvm.nx_huge_pages= 2074 [KVM] Controls the software workaround for the 2075 X86_BUG_ITLB_MULTIHIT bug. 2076 force : Always deploy workaround. 2077 off : Never deploy workaround. 2078 auto : Deploy workaround based on the presence of 2079 X86_BUG_ITLB_MULTIHIT. 2080 2081 Default is 'auto'. 2082 2083 If the software workaround is enabled for the host, 2084 guests do need not to enable it for nested guests. 2085 2086 kvm.nx_huge_pages_recovery_ratio= 2087 [KVM] Controls how many 4KiB pages are periodically zapped 2088 back to huge pages. 0 disables the recovery, otherwise if 2089 the value is N KVM will zap 1/Nth of the 4KiB pages every 2090 minute. The default is 60. 2091 2092 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM. 2093 Default is 1 (enabled) 2094 2095 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU) 2096 for all guests. 2097 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode. 2098 2099 kvm-arm.vgic_v3_group0_trap= 2100 [KVM,ARM] Trap guest accesses to GICv3 group-0 2101 system registers 2102 2103 kvm-arm.vgic_v3_group1_trap= 2104 [KVM,ARM] Trap guest accesses to GICv3 group-1 2105 system registers 2106 2107 kvm-arm.vgic_v3_common_trap= 2108 [KVM,ARM] Trap guest accesses to GICv3 common 2109 system registers 2110 2111 kvm-arm.vgic_v4_enable= 2112 [KVM,ARM] Allow use of GICv4 for direct injection of 2113 LPIs. 2114 2115 kvm-intel.ept= [KVM,Intel] Disable extended page tables 2116 (virtualized MMU) support on capable Intel chips. 2117 Default is 1 (enabled) 2118 2119 kvm-intel.emulate_invalid_guest_state= 2120 [KVM,Intel] Enable emulation of invalid guest states 2121 Default is 0 (disabled) 2122 2123 kvm-intel.flexpriority= 2124 [KVM,Intel] Disable FlexPriority feature (TPR shadow). 2125 Default is 1 (enabled) 2126 2127 kvm-intel.nested= 2128 [KVM,Intel] Enable VMX nesting (nVMX). 2129 Default is 0 (disabled) 2130 2131 kvm-intel.unrestricted_guest= 2132 [KVM,Intel] Disable unrestricted guest feature 2133 (virtualized real and unpaged mode) on capable 2134 Intel chips. Default is 1 (enabled) 2135 2136 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 2137 CVE-2018-3620. 2138 2139 Valid arguments: never, cond, always 2140 2141 always: L1D cache flush on every VMENTER. 2142 cond: Flush L1D on VMENTER only when the code between 2143 VMEXIT and VMENTER can leak host memory. 2144 never: Disables the mitigation 2145 2146 Default is cond (do L1 cache flush in specific instances) 2147 2148 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification 2149 feature (tagged TLBs) on capable Intel chips. 2150 Default is 1 (enabled) 2151 2152 l1tf= [X86] Control mitigation of the L1TF vulnerability on 2153 affected CPUs 2154 2155 The kernel PTE inversion protection is unconditionally 2156 enabled and cannot be disabled. 2157 2158 full 2159 Provides all available mitigations for the 2160 L1TF vulnerability. Disables SMT and 2161 enables all mitigations in the 2162 hypervisors, i.e. unconditional L1D flush. 2163 2164 SMT control and L1D flush control via the 2165 sysfs interface is still possible after 2166 boot. Hypervisors will issue a warning 2167 when the first VM is started in a 2168 potentially insecure configuration, 2169 i.e. SMT enabled or L1D flush disabled. 2170 2171 full,force 2172 Same as 'full', but disables SMT and L1D 2173 flush runtime control. Implies the 2174 'nosmt=force' command line option. 2175 (i.e. sysfs control of SMT is disabled.) 2176 2177 flush 2178 Leaves SMT enabled and enables the default 2179 hypervisor mitigation, i.e. conditional 2180 L1D flush. 2181 2182 SMT control and L1D flush control via the 2183 sysfs interface is still possible after 2184 boot. Hypervisors will issue a warning 2185 when the first VM is started in a 2186 potentially insecure configuration, 2187 i.e. SMT enabled or L1D flush disabled. 2188 2189 flush,nosmt 2190 2191 Disables SMT and enables the default 2192 hypervisor mitigation. 2193 2194 SMT control and L1D flush control via the 2195 sysfs interface is still possible after 2196 boot. Hypervisors will issue a warning 2197 when the first VM is started in a 2198 potentially insecure configuration, 2199 i.e. SMT enabled or L1D flush disabled. 2200 2201 flush,nowarn 2202 Same as 'flush', but hypervisors will not 2203 warn when a VM is started in a potentially 2204 insecure configuration. 2205 2206 off 2207 Disables hypervisor mitigations and doesn't 2208 emit any warnings. 2209 It also drops the swap size and available 2210 RAM limit restriction on both hypervisor and 2211 bare metal. 2212 2213 Default is 'flush'. 2214 2215 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 2216 2217 l2cr= [PPC] 2218 2219 l3cr= [PPC] 2220 2221 lapic [X86-32,APIC] Enable the local APIC even if BIOS 2222 disabled it. 2223 2224 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline 2225 value for LAPIC timer one-shot implementation. Default 2226 back to the programmable timer unit in the LAPIC. 2227 2228 lapic_timer_c2_ok [X86,APIC] trust the local apic timer 2229 in C2 power state. 2230 2231 libata.dma= [LIBATA] DMA control 2232 libata.dma=0 Disable all PATA and SATA DMA 2233 libata.dma=1 PATA and SATA Disk DMA only 2234 libata.dma=2 ATAPI (CDROM) DMA only 2235 libata.dma=4 Compact Flash DMA only 2236 Combinations also work, so libata.dma=3 enables DMA 2237 for disks and CDROMs, but not CFs. 2238 2239 libata.ignore_hpa= [LIBATA] Ignore HPA limit 2240 libata.ignore_hpa=0 keep BIOS limits (default) 2241 libata.ignore_hpa=1 ignore limits, using full disk 2242 2243 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 2244 when set. 2245 Format: <int> 2246 2247 libata.force= [LIBATA] Force configurations. The format is comma 2248 separated list of "[ID:]VAL" where ID is 2249 PORT[.DEVICE]. PORT and DEVICE are decimal numbers 2250 matching port, link or device. Basically, it matches 2251 the ATA ID string printed on console by libata. If 2252 the whole ID part is omitted, the last PORT and DEVICE 2253 values are used. If ID hasn't been specified yet, the 2254 configuration applies to all ports, links and devices. 2255 2256 If only DEVICE is omitted, the parameter applies to 2257 the port and all links and devices behind it. DEVICE 2258 number of 0 either selects the first device or the 2259 first fan-out link behind PMP device. It does not 2260 select the host link. DEVICE number of 15 selects the 2261 host link and device attached to it. 2262 2263 The VAL specifies the configuration to force. As long 2264 as there's no ambiguity shortcut notation is allowed. 2265 For example, both 1.5 and 1.5G would work for 1.5Gbps. 2266 The following configurations can be forced. 2267 2268 * Cable type: 40c, 80c, short40c, unk, ign or sata. 2269 Any ID with matching PORT is used. 2270 2271 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 2272 2273 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 2274 udma[/][16,25,33,44,66,100,133] notation is also 2275 allowed. 2276 2277 * [no]ncq: Turn on or off NCQ. 2278 2279 * [no]ncqtrim: Turn off queued DSM TRIM. 2280 2281 * nohrst, nosrst, norst: suppress hard, soft 2282 and both resets. 2283 2284 * rstonce: only attempt one reset during 2285 hot-unplug link recovery 2286 2287 * dump_id: dump IDENTIFY data. 2288 2289 * atapi_dmadir: Enable ATAPI DMADIR bridge support 2290 2291 * disable: Disable this device. 2292 2293 If there are multiple matching configurations changing 2294 the same attribute, the last one is used. 2295 2296 memblock=debug [KNL] Enable memblock debug messages. 2297 2298 load_ramdisk= [RAM] List of ramdisks to load from floppy 2299 See Documentation/admin-guide/blockdev/ramdisk.rst. 2300 2301 lockd.nlm_grace_period=P [NFS] Assign grace period. 2302 Format: <integer> 2303 2304 lockd.nlm_tcpport=N [NFS] Assign TCP port. 2305 Format: <integer> 2306 2307 lockd.nlm_timeout=T [NFS] Assign timeout value. 2308 Format: <integer> 2309 2310 lockd.nlm_udpport=M [NFS] Assign UDP port. 2311 Format: <integer> 2312 2313 lockdown= [SECURITY] 2314 { integrity | confidentiality } 2315 Enable the kernel lockdown feature. If set to 2316 integrity, kernel features that allow userland to 2317 modify the running kernel are disabled. If set to 2318 confidentiality, kernel features that allow userland 2319 to extract confidential information from the kernel 2320 are also disabled. 2321 2322 locktorture.nreaders_stress= [KNL] 2323 Set the number of locking read-acquisition kthreads. 2324 Defaults to being automatically set based on the 2325 number of online CPUs. 2326 2327 locktorture.nwriters_stress= [KNL] 2328 Set the number of locking write-acquisition kthreads. 2329 2330 locktorture.onoff_holdoff= [KNL] 2331 Set time (s) after boot for CPU-hotplug testing. 2332 2333 locktorture.onoff_interval= [KNL] 2334 Set time (s) between CPU-hotplug operations, or 2335 zero to disable CPU-hotplug testing. 2336 2337 locktorture.shuffle_interval= [KNL] 2338 Set task-shuffle interval (jiffies). Shuffling 2339 tasks allows some CPUs to go into dyntick-idle 2340 mode during the locktorture test. 2341 2342 locktorture.shutdown_secs= [KNL] 2343 Set time (s) after boot system shutdown. This 2344 is useful for hands-off automated testing. 2345 2346 locktorture.stat_interval= [KNL] 2347 Time (s) between statistics printk()s. 2348 2349 locktorture.stutter= [KNL] 2350 Time (s) to stutter testing, for example, 2351 specifying five seconds causes the test to run for 2352 five seconds, wait for five seconds, and so on. 2353 This tests the locking primitive's ability to 2354 transition abruptly to and from idle. 2355 2356 locktorture.torture_type= [KNL] 2357 Specify the locking implementation to test. 2358 2359 locktorture.verbose= [KNL] 2360 Enable additional printk() statements. 2361 2362 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 2363 Format: <irq> 2364 2365 loglevel= All Kernel Messages with a loglevel smaller than the 2366 console loglevel will be printed to the console. It can 2367 also be changed with klogd or other programs. The 2368 loglevels are defined as follows: 2369 2370 0 (KERN_EMERG) system is unusable 2371 1 (KERN_ALERT) action must be taken immediately 2372 2 (KERN_CRIT) critical conditions 2373 3 (KERN_ERR) error conditions 2374 4 (KERN_WARNING) warning conditions 2375 5 (KERN_NOTICE) normal but significant condition 2376 6 (KERN_INFO) informational 2377 7 (KERN_DEBUG) debug-level messages 2378 2379 log_buf_len=n[KMG] Sets the size of the printk ring buffer, 2380 in bytes. n must be a power of two and greater 2381 than the minimal size. The minimal size is defined 2382 by LOG_BUF_SHIFT kernel config parameter. There is 2383 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter 2384 that allows to increase the default size depending on 2385 the number of CPUs. See init/Kconfig for more details. 2386 2387 logo.nologo [FB] Disables display of the built-in Linux logo. 2388 This may be used to provide more screen space for 2389 kernel log messages and is useful when debugging 2390 kernel boot problems. 2391 2392 lp=0 [LP] Specify parallel ports to use, e.g, 2393 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 2394 lp=reset first parallel port). 'lp=0' disables the 2395 lp=auto printer driver. 'lp=reset' (which can be 2396 specified in addition to the ports) causes 2397 attached printers to be reset. Using 2398 lp=port1,port2,... specifies the parallel ports 2399 to associate lp devices with, starting with 2400 lp0. A port specification may be 'none' to skip 2401 that lp device, or a parport name such as 2402 'parport0'. Specifying 'lp=auto' instead of a 2403 port specification list means that device IDs 2404 from each port should be examined, to see if 2405 an IEEE 1284-compliant printer is attached; if 2406 so, the driver will manage that printer. 2407 See also header of drivers/char/lp.c. 2408 2409 lpj=n [KNL] 2410 Sets loops_per_jiffy to given constant, thus avoiding 2411 time-consuming boot-time autodetection (up to 250 ms per 2412 CPU). 0 enables autodetection (default). To determine 2413 the correct value for your kernel, boot with normal 2414 autodetection and see what value is printed. Note that 2415 on SMP systems the preset will be applied to all CPUs, 2416 which is likely to cause problems if your CPUs need 2417 significantly divergent settings. An incorrect value 2418 will cause delays in the kernel to be wrong, leading to 2419 unpredictable I/O errors and other breakage. Although 2420 unlikely, in the extreme case this might damage your 2421 hardware. 2422 2423 ltpc= [NET] 2424 Format: <io>,<irq>,<dma> 2425 2426 lsm.debug [SECURITY] Enable LSM initialization debugging output. 2427 2428 lsm=lsm1,...,lsmN 2429 [SECURITY] Choose order of LSM initialization. This 2430 overrides CONFIG_LSM, and the "security=" parameter. 2431 2432 machvec= [IA-64] Force the use of a particular machine-vector 2433 (machvec) in a generic kernel. 2434 Example: machvec=hpzx1 2435 2436 machtype= [Loongson] Share the same kernel image file between different 2437 yeeloong laptop. 2438 Example: machtype=lemote-yeeloong-2f-7inch 2439 2440 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater 2441 than or equal to this physical address is ignored. 2442 2443 maxcpus= [SMP] Maximum number of processors that an SMP kernel 2444 will bring up during bootup. maxcpus=n : n >= 0 limits 2445 the kernel to bring up 'n' processors. Surely after 2446 bootup you can bring up the other plugged cpu by executing 2447 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 2448 only takes effect during system bootup. 2449 While n=0 is a special case, it is equivalent to "nosmp", 2450 which also disables the IO APIC. 2451 2452 max_loop= [LOOP] The number of loop block devices that get 2453 (loop.max_loop) unconditionally pre-created at init time. The default 2454 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 2455 of statically allocating a predefined number, loop 2456 devices can be requested on-demand with the 2457 /dev/loop-control interface. 2458 2459 mce [X86-32] Machine Check Exception 2460 2461 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.rst 2462 2463 md= [HW] RAID subsystems devices and level 2464 See Documentation/admin-guide/md.rst. 2465 2466 mdacon= [MDA] 2467 Format: <first>,<last> 2468 Specifies range of consoles to be captured by the MDA. 2469 2470 mds= [X86,INTEL] 2471 Control mitigation for the Micro-architectural Data 2472 Sampling (MDS) vulnerability. 2473 2474 Certain CPUs are vulnerable to an exploit against CPU 2475 internal buffers which can forward information to a 2476 disclosure gadget under certain conditions. 2477 2478 In vulnerable processors, the speculatively 2479 forwarded data can be used in a cache side channel 2480 attack, to access data to which the attacker does 2481 not have direct access. 2482 2483 This parameter controls the MDS mitigation. The 2484 options are: 2485 2486 full - Enable MDS mitigation on vulnerable CPUs 2487 full,nosmt - Enable MDS mitigation and disable 2488 SMT on vulnerable CPUs 2489 off - Unconditionally disable MDS mitigation 2490 2491 On TAA-affected machines, mds=off can be prevented by 2492 an active TAA mitigation as both vulnerabilities are 2493 mitigated with the same mechanism so in order to disable 2494 this mitigation, you need to specify tsx_async_abort=off 2495 too. 2496 2497 Not specifying this option is equivalent to 2498 mds=full. 2499 2500 For details see: Documentation/admin-guide/hw-vuln/mds.rst 2501 2502 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory 2503 Amount of memory to be used when the kernel is not able 2504 to see the whole system memory or for test. 2505 [X86] Work as limiting max address. Use together 2506 with memmap= to avoid physical address space collisions. 2507 Without memmap= PCI devices could be placed at addresses 2508 belonging to unused RAM. 2509 2510 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 2511 memory. 2512 2513 memchunk=nn[KMG] 2514 [KNL,SH] Allow user to override the default size for 2515 per-device physically contiguous DMA buffers. 2516 2517 memhp_default_state=online/offline 2518 [KNL] Set the initial state for the memory hotplug 2519 onlining policy. If not specified, the default value is 2520 set according to the 2521 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config 2522 option. 2523 See Documentation/admin-guide/mm/memory-hotplug.rst. 2524 2525 memmap=exactmap [KNL,X86] Enable setting of an exact 2526 E820 memory map, as specified by the user. 2527 Such memmap=exactmap lines can be constructed based on 2528 BIOS output or other requirements. See the memmap=nn@ss 2529 option description. 2530 2531 memmap=nn[KMG]@ss[KMG] 2532 [KNL] Force usage of a specific region of memory. 2533 Region of memory to be used is from ss to ss+nn. 2534 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 2535 which limits max address to nn[KMG]. 2536 Multiple different regions can be specified, 2537 comma delimited. 2538 Example: 2539 memmap=100M@2G,100M#3G,1G!1024G 2540 2541 memmap=nn[KMG]#ss[KMG] 2542 [KNL,ACPI] Mark specific memory as ACPI data. 2543 Region of memory to be marked is from ss to ss+nn. 2544 2545 memmap=nn[KMG]$ss[KMG] 2546 [KNL,ACPI] Mark specific memory as reserved. 2547 Region of memory to be reserved is from ss to ss+nn. 2548 Example: Exclude memory from 0x18690000-0x1869ffff 2549 memmap=64K$0x18690000 2550 or 2551 memmap=0x10000$0x18690000 2552 Some bootloaders may need an escape character before '$', 2553 like Grub2, otherwise '$' and the following number 2554 will be eaten. 2555 2556 memmap=nn[KMG]!ss[KMG] 2557 [KNL,X86] Mark specific memory as protected. 2558 Region of memory to be used, from ss to ss+nn. 2559 The memory region may be marked as e820 type 12 (0xc) 2560 and is NVDIMM or ADR memory. 2561 2562 memmap=<size>%<offset>-<oldtype>+<newtype> 2563 [KNL,ACPI] Convert memory within the specified region 2564 from <oldtype> to <newtype>. If "-<oldtype>" is left 2565 out, the whole region will be marked as <newtype>, 2566 even if previously unavailable. If "+<newtype>" is left 2567 out, matching memory will be removed. Types are 2568 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 2569 3 = ACPI, 12 = PRAM. 2570 2571 memory_corruption_check=0/1 [X86] 2572 Some BIOSes seem to corrupt the first 64k of 2573 memory when doing things like suspend/resume. 2574 Setting this option will scan the memory 2575 looking for corruption. Enabling this will 2576 both detect corruption and prevent the kernel 2577 from using the memory being corrupted. 2578 However, its intended as a diagnostic tool; if 2579 repeatable BIOS-originated corruption always 2580 affects the same memory, you can use memmap= 2581 to prevent the kernel from using that memory. 2582 2583 memory_corruption_check_size=size [X86] 2584 By default it checks for corruption in the low 2585 64k, making this memory unavailable for normal 2586 use. Use this parameter to scan for 2587 corruption in more or less memory. 2588 2589 memory_corruption_check_period=seconds [X86] 2590 By default it checks for corruption every 60 2591 seconds. Use this parameter to check at some 2592 other rate. 0 disables periodic checking. 2593 2594 memtest= [KNL,X86,ARM,PPC] Enable memtest 2595 Format: <integer> 2596 default : 0 <disable> 2597 Specifies the number of memtest passes to be 2598 performed. Each pass selects another test 2599 pattern from a given set of patterns. Memtest 2600 fills the memory with this pattern, validates 2601 memory contents and reserves bad memory 2602 regions that are detected. 2603 2604 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 2605 Valid arguments: on, off 2606 Default (depends on kernel configuration option): 2607 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y) 2608 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n) 2609 mem_encrypt=on: Activate SME 2610 mem_encrypt=off: Do not activate SME 2611 2612 Refer to Documentation/virt/kvm/amd-memory-encryption.rst 2613 for details on when memory encryption can be activated. 2614 2615 mem_sleep_default= [SUSPEND] Default system suspend mode: 2616 s2idle - Suspend-To-Idle 2617 shallow - Power-On Suspend or equivalent (if supported) 2618 deep - Suspend-To-RAM or equivalent (if supported) 2619 See Documentation/admin-guide/pm/sleep-states.rst. 2620 2621 meye.*= [HW] Set MotionEye Camera parameters 2622 See Documentation/media/v4l-drivers/meye.rst. 2623 2624 mfgpt_irq= [IA-32] Specify the IRQ to use for the 2625 Multi-Function General Purpose Timers on AMD Geode 2626 platforms. 2627 2628 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 2629 the BIOS has incorrectly applied a workaround. TinyBIOS 2630 version 0.98 is known to be affected, 0.99 fixes the 2631 problem by letting the user disable the workaround. 2632 2633 mga= [HW,DRM] 2634 2635 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this 2636 physical address is ignored. 2637 2638 mini2440= [ARM,HW,KNL] 2639 Format:[0..2][b][c][t] 2640 Default: "0tb" 2641 MINI2440 configuration specification: 2642 0 - The attached screen is the 3.5" TFT 2643 1 - The attached screen is the 7" TFT 2644 2 - The VGA Shield is attached (1024x768) 2645 Leaving out the screen size parameter will not load 2646 the TFT driver, and the framebuffer will be left 2647 unconfigured. 2648 b - Enable backlight. The TFT backlight pin will be 2649 linked to the kernel VESA blanking code and a GPIO 2650 LED. This parameter is not necessary when using the 2651 VGA shield. 2652 c - Enable the s3c camera interface. 2653 t - Reserved for enabling touchscreen support. The 2654 touchscreen support is not enabled in the mainstream 2655 kernel as of 2.6.30, a preliminary port can be found 2656 in the "bleeding edge" mini2440 support kernel at 2657 http://repo.or.cz/w/linux-2.6/mini2440.git 2658 2659 mitigations= 2660 [X86,PPC,S390,ARM64] Control optional mitigations for 2661 CPU vulnerabilities. This is a set of curated, 2662 arch-independent options, each of which is an 2663 aggregation of existing arch-specific options. 2664 2665 off 2666 Disable all optional CPU mitigations. This 2667 improves system performance, but it may also 2668 expose users to several CPU vulnerabilities. 2669 Equivalent to: nopti [X86,PPC] 2670 kpti=0 [ARM64] 2671 nospectre_v1 [X86,PPC] 2672 nobp=0 [S390] 2673 nospectre_v2 [X86,PPC,S390,ARM64] 2674 spectre_v2_user=off [X86] 2675 spec_store_bypass_disable=off [X86,PPC] 2676 ssbd=force-off [ARM64] 2677 l1tf=off [X86] 2678 mds=off [X86] 2679 tsx_async_abort=off [X86] 2680 kvm.nx_huge_pages=off [X86] 2681 2682 Exceptions: 2683 This does not have any effect on 2684 kvm.nx_huge_pages when 2685 kvm.nx_huge_pages=force. 2686 2687 auto (default) 2688 Mitigate all CPU vulnerabilities, but leave SMT 2689 enabled, even if it's vulnerable. This is for 2690 users who don't want to be surprised by SMT 2691 getting disabled across kernel upgrades, or who 2692 have other ways of avoiding SMT-based attacks. 2693 Equivalent to: (default behavior) 2694 2695 auto,nosmt 2696 Mitigate all CPU vulnerabilities, disabling SMT 2697 if needed. This is for users who always want to 2698 be fully mitigated, even if it means losing SMT. 2699 Equivalent to: l1tf=flush,nosmt [X86] 2700 mds=full,nosmt [X86] 2701 tsx_async_abort=full,nosmt [X86] 2702 2703 mminit_loglevel= 2704 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this 2705 parameter allows control of the logging verbosity for 2706 the additional memory initialisation checks. A value 2707 of 0 disables mminit logging and a level of 4 will 2708 log everything. Information is printed at KERN_DEBUG 2709 so loglevel=8 may also need to be specified. 2710 2711 module.sig_enforce 2712 [KNL] When CONFIG_MODULE_SIG is set, this means that 2713 modules without (valid) signatures will fail to load. 2714 Note that if CONFIG_MODULE_SIG_FORCE is set, that 2715 is always true, so this option does nothing. 2716 2717 module_blacklist= [KNL] Do not load a comma-separated list of 2718 modules. Useful for debugging problem modules. 2719 2720 mousedev.tap_time= 2721 [MOUSE] Maximum time between finger touching and 2722 leaving touchpad surface for touch to be considered 2723 a tap and be reported as a left button click (for 2724 touchpads working in absolute mode only). 2725 Format: <msecs> 2726 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 2727 reporting absolute coordinates, such as tablets 2728 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 2729 reporting absolute coordinates, such as tablets 2730 2731 movablecore= [KNL,X86,IA-64,PPC] 2732 Format: nn[KMGTPE] | nn% 2733 This parameter is the complement to kernelcore=, it 2734 specifies the amount of memory used for migratable 2735 allocations. If both kernelcore and movablecore is 2736 specified, then kernelcore will be at *least* the 2737 specified value but may be more. If movablecore on its 2738 own is specified, the administrator must be careful 2739 that the amount of memory usable for all allocations 2740 is not too small. 2741 2742 movable_node [KNL] Boot-time switch to make hotplugable memory 2743 NUMA nodes to be movable. This means that the memory 2744 of such nodes will be usable only for movable 2745 allocations which rules out almost all kernel 2746 allocations. Use with caution! 2747 2748 MTD_Partition= [MTD] 2749 Format: <name>,<region-number>,<size>,<offset> 2750 2751 MTD_Region= [MTD] Format: 2752 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 2753 2754 mtdparts= [MTD] 2755 See drivers/mtd/cmdlinepart.c. 2756 2757 multitce=off [PPC] This parameter disables the use of the pSeries 2758 firmware feature for updating multiple TCE entries 2759 at a time. 2760 2761 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 2762 2763 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 2764 2765 boundary - index of last SLC block on Flex-OneNAND. 2766 The remaining blocks are configured as MLC blocks. 2767 lock - Configure if Flex-OneNAND boundary should be locked. 2768 Once locked, the boundary cannot be changed. 2769 1 indicates lock status, 0 indicates unlock status. 2770 2771 mtdset= [ARM] 2772 ARM/S3C2412 JIVE boot control 2773 2774 See arch/arm/mach-s3c2412/mach-jive.c 2775 2776 mtouchusb.raw_coordinates= 2777 [HW] Make the MicroTouch USB driver use raw coordinates 2778 ('y', default) or cooked coordinates ('n') 2779 2780 mtrr_chunk_size=nn[KMG] [X86] 2781 used for mtrr cleanup. It is largest continuous chunk 2782 that could hold holes aka. UC entries. 2783 2784 mtrr_gran_size=nn[KMG] [X86] 2785 Used for mtrr cleanup. It is granularity of mtrr block. 2786 Default is 1. 2787 Large value could prevent small alignment from 2788 using up MTRRs. 2789 2790 mtrr_spare_reg_nr=n [X86] 2791 Format: <integer> 2792 Range: 0,7 : spare reg number 2793 Default : 1 2794 Used for mtrr cleanup. It is spare mtrr entries number. 2795 Set to 2 or more if your graphical card needs more. 2796 2797 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 2798 2799 netdev= [NET] Network devices parameters 2800 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 2801 Note that mem_start is often overloaded to mean 2802 something different and driver-specific. 2803 This usage is only documented in each driver source 2804 file if at all. 2805 2806 nf_conntrack.acct= 2807 [NETFILTER] Enable connection tracking flow accounting 2808 0 to disable accounting 2809 1 to enable accounting 2810 Default value is 0. 2811 2812 nfsaddrs= [NFS] Deprecated. Use ip= instead. 2813 See Documentation/filesystems/nfs/nfsroot.txt. 2814 2815 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 2816 See Documentation/filesystems/nfs/nfsroot.txt. 2817 2818 nfsrootdebug [NFS] enable nfsroot debugging messages. 2819 See Documentation/filesystems/nfs/nfsroot.txt. 2820 2821 nfs.callback_nr_threads= 2822 [NFSv4] set the total number of threads that the 2823 NFS client will assign to service NFSv4 callback 2824 requests. 2825 2826 nfs.callback_tcpport= 2827 [NFS] set the TCP port on which the NFSv4 callback 2828 channel should listen. 2829 2830 nfs.cache_getent= 2831 [NFS] sets the pathname to the program which is used 2832 to update the NFS client cache entries. 2833 2834 nfs.cache_getent_timeout= 2835 [NFS] sets the timeout after which an attempt to 2836 update a cache entry is deemed to have failed. 2837 2838 nfs.idmap_cache_timeout= 2839 [NFS] set the maximum lifetime for idmapper cache 2840 entries. 2841 2842 nfs.enable_ino64= 2843 [NFS] enable 64-bit inode numbers. 2844 If zero, the NFS client will fake up a 32-bit inode 2845 number for the readdir() and stat() syscalls instead 2846 of returning the full 64-bit number. 2847 The default is to return 64-bit inode numbers. 2848 2849 nfs.max_session_cb_slots= 2850 [NFSv4.1] Sets the maximum number of session 2851 slots the client will assign to the callback 2852 channel. This determines the maximum number of 2853 callbacks the client will process in parallel for 2854 a particular server. 2855 2856 nfs.max_session_slots= 2857 [NFSv4.1] Sets the maximum number of session slots 2858 the client will attempt to negotiate with the server. 2859 This limits the number of simultaneous RPC requests 2860 that the client can send to the NFSv4.1 server. 2861 Note that there is little point in setting this 2862 value higher than the max_tcp_slot_table_limit. 2863 2864 nfs.nfs4_disable_idmapping= 2865 [NFSv4] When set to the default of '1', this option 2866 ensures that both the RPC level authentication 2867 scheme and the NFS level operations agree to use 2868 numeric uids/gids if the mount is using the 2869 'sec=sys' security flavour. In effect it is 2870 disabling idmapping, which can make migration from 2871 legacy NFSv2/v3 systems to NFSv4 easier. 2872 Servers that do not support this mode of operation 2873 will be autodetected by the client, and it will fall 2874 back to using the idmapper. 2875 To turn off this behaviour, set the value to '0'. 2876 nfs.nfs4_unique_id= 2877 [NFS4] Specify an additional fixed unique ident- 2878 ification string that NFSv4 clients can insert into 2879 their nfs_client_id4 string. This is typically a 2880 UUID that is generated at system install time. 2881 2882 nfs.send_implementation_id = 2883 [NFSv4.1] Send client implementation identification 2884 information in exchange_id requests. 2885 If zero, no implementation identification information 2886 will be sent. 2887 The default is to send the implementation identification 2888 information. 2889 2890 nfs.recover_lost_locks = 2891 [NFSv4] Attempt to recover locks that were lost due 2892 to a lease timeout on the server. Please note that 2893 doing this risks data corruption, since there are 2894 no guarantees that the file will remain unchanged 2895 after the locks are lost. 2896 If you want to enable the kernel legacy behaviour of 2897 attempting to recover these locks, then set this 2898 parameter to '1'. 2899 The default parameter value of '0' causes the kernel 2900 not to attempt recovery of lost locks. 2901 2902 nfs4.layoutstats_timer = 2903 [NFSv4.2] Change the rate at which the kernel sends 2904 layoutstats to the pNFS metadata server. 2905 2906 Setting this to value to 0 causes the kernel to use 2907 whatever value is the default set by the layout 2908 driver. A non-zero value sets the minimum interval 2909 in seconds between layoutstats transmissions. 2910 2911 nfsd.nfs4_disable_idmapping= 2912 [NFSv4] When set to the default of '1', the NFSv4 2913 server will return only numeric uids and gids to 2914 clients using auth_sys, and will accept numeric uids 2915 and gids from such clients. This is intended to ease 2916 migration from NFSv2/v3. 2917 2918 nmi_debug= [KNL,SH] Specify one or more actions to take 2919 when a NMI is triggered. 2920 Format: [state][,regs][,debounce][,die] 2921 2922 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 2923 Format: [panic,][nopanic,][num] 2924 Valid num: 0 or 1 2925 0 - turn hardlockup detector in nmi_watchdog off 2926 1 - turn hardlockup detector in nmi_watchdog on 2927 When panic is specified, panic when an NMI watchdog 2928 timeout occurs (or 'nopanic' to not panic on an NMI 2929 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 2930 To disable both hard and soft lockup detectors, 2931 please see 'nowatchdog'. 2932 This is useful when you use a panic=... timeout and 2933 need the box quickly up again. 2934 2935 These settings can be accessed at runtime via 2936 the nmi_watchdog and hardlockup_panic sysctls. 2937 2938 netpoll.carrier_timeout= 2939 [NET] Specifies amount of time (in seconds) that 2940 netpoll should wait for a carrier. By default netpoll 2941 waits 4 seconds. 2942 2943 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 2944 emulation library even if a 387 maths coprocessor 2945 is present. 2946 2947 no5lvl [X86-64] Disable 5-level paging mode. Forces 2948 kernel to use 4-level paging instead. 2949 2950 no_console_suspend 2951 [HW] Never suspend the console 2952 Disable suspending of consoles during suspend and 2953 hibernate operations. Once disabled, debugging 2954 messages can reach various consoles while the rest 2955 of the system is being put to sleep (ie, while 2956 debugging driver suspend/resume hooks). This may 2957 not work reliably with all consoles, but is known 2958 to work with serial and VGA consoles. 2959 To facilitate more flexible debugging, we also add 2960 console_suspend, a printk module parameter to control 2961 it. Users could use console_suspend (usually 2962 /sys/module/printk/parameters/console_suspend) to 2963 turn on/off it dynamically. 2964 2965 novmcoredd [KNL,KDUMP] 2966 Disable device dump. Device dump allows drivers to 2967 append dump data to vmcore so you can collect driver 2968 specified debug info. Drivers can append the data 2969 without any limit and this data is stored in memory, 2970 so this may cause significant memory stress. Disabling 2971 device dump can help save memory but the driver debug 2972 data will be no longer available. This parameter 2973 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 2974 is set. 2975 2976 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien 2977 caches in the slab allocator. Saves per-node memory, 2978 but will impact performance. 2979 2980 noalign [KNL,ARM] 2981 2982 noaltinstr [S390] Disables alternative instructions patching 2983 (CPU alternatives feature). 2984 2985 noapic [SMP,APIC] Tells the kernel to not make use of any 2986 IOAPICs that may be present in the system. 2987 2988 noautogroup Disable scheduler automatic task group creation. 2989 2990 nobats [PPC] Do not use BATs for mapping kernel lowmem 2991 on "Classic" PPC cores. 2992 2993 nocache [ARM] 2994 2995 noclflush [BUGS=X86] Don't use the CLFLUSH instruction 2996 2997 nodelayacct [KNL] Disable per-task delay accounting 2998 2999 nodsp [SH] Disable hardware DSP at boot time. 3000 3001 noefi Disable EFI runtime services support. 3002 3003 noexec [IA-64] 3004 3005 noexec [X86] 3006 On X86-32 available only on PAE configured kernels. 3007 noexec=on: enable non-executable mappings (default) 3008 noexec=off: disable non-executable mappings 3009 3010 nosmap [X86,PPC] 3011 Disable SMAP (Supervisor Mode Access Prevention) 3012 even if it is supported by processor. 3013 3014 nosmep [X86,PPC] 3015 Disable SMEP (Supervisor Mode Execution Prevention) 3016 even if it is supported by processor. 3017 3018 noexec32 [X86-64] 3019 This affects only 32-bit executables. 3020 noexec32=on: enable non-executable mappings (default) 3021 read doesn't imply executable mappings 3022 noexec32=off: disable non-executable mappings 3023 read implies executable mappings 3024 3025 nofpu [MIPS,SH] Disable hardware FPU at boot time. 3026 3027 nofxsr [BUGS=X86-32] Disables x86 floating point extended 3028 register save and restore. The kernel will only save 3029 legacy floating-point registers on task switch. 3030 3031 nohugeiomap [KNL,x86,PPC] Disable kernel huge I/O mappings. 3032 3033 nosmt [KNL,S390] Disable symmetric multithreading (SMT). 3034 Equivalent to smt=1. 3035 3036 [KNL,x86] Disable symmetric multithreading (SMT). 3037 nosmt=force: Force disable SMT, cannot be undone 3038 via the sysfs control file. 3039 3040 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1 3041 (bounds check bypass). With this option data leaks are 3042 possible in the system. 3043 3044 nospectre_v2 [X86,PPC_FSL_BOOK3E,ARM64] Disable all mitigations for 3045 the Spectre variant 2 (indirect branch prediction) 3046 vulnerability. System may allow data leaks with this 3047 option. 3048 3049 nospec_store_bypass_disable 3050 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability 3051 3052 noxsave [BUGS=X86] Disables x86 extended register state save 3053 and restore using xsave. The kernel will fallback to 3054 enabling legacy floating-point and sse state. 3055 3056 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 3057 register states. The kernel will fall back to use 3058 xsave to save the states. By using this parameter, 3059 performance of saving the states is degraded because 3060 xsave doesn't support modified optimization while 3061 xsaveopt supports it on xsaveopt enabled systems. 3062 3063 noxsaves [X86] Disables xsaves and xrstors used in saving and 3064 restoring x86 extended register state in compacted 3065 form of xsave area. The kernel will fall back to use 3066 xsaveopt and xrstor to save and restore the states 3067 in standard form of xsave area. By using this 3068 parameter, xsave area per process might occupy more 3069 memory on xsaves enabled systems. 3070 3071 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or 3072 wfi(ARM) instruction doesn't work correctly and not to 3073 use it. This is also useful when using JTAG debugger. 3074 3075 no_file_caps Tells the kernel not to honor file capabilities. The 3076 only way then for a file to be executed with privilege 3077 is to be setuid root or executed by root. 3078 3079 nohalt [IA-64] Tells the kernel not to use the power saving 3080 function PAL_HALT_LIGHT when idle. This increases 3081 power-consumption. On the positive side, it reduces 3082 interrupt wake-up latency, which may improve performance 3083 in certain environments such as networked servers or 3084 real-time systems. 3085 3086 nohibernate [HIBERNATION] Disable hibernation and resume. 3087 3088 nohz= [KNL] Boottime enable/disable dynamic ticks 3089 Valid arguments: on, off 3090 Default: on 3091 3092 nohz_full= [KNL,BOOT,SMP,ISOL] 3093 The argument is a cpu list, as described above. 3094 In kernels built with CONFIG_NO_HZ_FULL=y, set 3095 the specified list of CPUs whose tick will be stopped 3096 whenever possible. The boot CPU will be forced outside 3097 the range to maintain the timekeeping. Any CPUs 3098 in this list will have their RCU callbacks offloaded, 3099 just as if they had also been called out in the 3100 rcu_nocbs= boot parameter. 3101 3102 noiotrap [SH] Disables trapped I/O port accesses. 3103 3104 noirqdebug [X86-32] Disables the code which attempts to detect and 3105 disable unhandled interrupt sources. 3106 3107 no_timer_check [X86,APIC] Disables the code which tests for 3108 broken timer IRQ sources. 3109 3110 noisapnp [ISAPNP] Disables ISA PnP code. 3111 3112 noinitrd [RAM] Tells the kernel not to load any configured 3113 initial RAM disk. 3114 3115 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt 3116 remapping. 3117 [Deprecated - use intremap=off] 3118 3119 nointroute [IA-64] 3120 3121 noinvpcid [X86] Disable the INVPCID cpu feature. 3122 3123 nojitter [IA-64] Disables jitter checking for ITC timers. 3124 3125 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver 3126 3127 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page 3128 fault handling. 3129 3130 no-vmw-sched-clock 3131 [X86,PV_OPS] Disable paravirtualized VMware scheduler 3132 clock and use the default one. 3133 3134 no-steal-acc [X86,KVM,ARM64] Disable paravirtualized steal time 3135 accounting. steal time is computed, but won't 3136 influence scheduler behaviour 3137 3138 nolapic [X86-32,APIC] Do not enable or use the local APIC. 3139 3140 nolapic_timer [X86-32,APIC] Do not use the local APIC timer. 3141 3142 noltlbs [PPC] Do not use large page/tlb entries for kernel 3143 lowmem mapping on PPC40x and PPC8xx 3144 3145 nomca [IA-64] Disable machine check abort handling 3146 3147 nomce [X86-32] Disable Machine Check Exception 3148 3149 nomfgpt [X86-32] Disable Multi-Function General Purpose 3150 Timer usage (for AMD Geode machines). 3151 3152 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 3153 shutdown the other cpus. Instead use the REBOOT_VECTOR 3154 irq. 3155 3156 nomodule Disable module load 3157 3158 nopat [X86] Disable PAT (page attribute table extension of 3159 pagetables) support. 3160 3161 nopcid [X86-64] Disable the PCID cpu feature. 3162 3163 norandmaps Don't use address space randomization. Equivalent to 3164 echo 0 > /proc/sys/kernel/randomize_va_space 3165 3166 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 3167 with UP alternatives 3168 3169 nordrand [X86] Disable kernel use of the RDRAND and 3170 RDSEED instructions even if they are supported 3171 by the processor. RDRAND and RDSEED are still 3172 available to user space applications. 3173 3174 noresume [SWSUSP] Disables resume and restores original swap 3175 space. 3176 3177 no-scroll [VGA] Disables scrollback. 3178 This is required for the Braillex ib80-piezo Braille 3179 reader made by F.H. Papenmeier (Germany). 3180 3181 nosbagart [IA-64] 3182 3183 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support. 3184 3185 nosmp [SMP] Tells an SMP kernel to act as a UP kernel, 3186 and disable the IO APIC. legacy for "maxcpus=0". 3187 3188 nosoftlockup [KNL] Disable the soft-lockup detector. 3189 3190 nosync [HW,M68K] Disables sync negotiation for all devices. 3191 3192 nowatchdog [KNL] Disable both lockup detectors, i.e. 3193 soft-lockup and NMI watchdog (hard-lockup). 3194 3195 nowb [ARM] 3196 3197 nox2apic [X86-64,APIC] Do not enable x2APIC mode. 3198 3199 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when 3200 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off. 3201 Some features depend on CPU0. Known dependencies are: 3202 1. Resume from suspend/hibernate depends on CPU0. 3203 Suspend/hibernate will fail if CPU0 is offline and you 3204 need to online CPU0 before suspend/hibernate. 3205 2. PIC interrupts also depend on CPU0. CPU0 can't be 3206 removed if a PIC interrupt is detected. 3207 It's said poweroff/reboot may depend on CPU0 on some 3208 machines although I haven't seen such issues so far 3209 after CPU0 is offline on a few tested machines. 3210 If the dependencies are under your control, you can 3211 turn on cpu0_hotplug. 3212 3213 nps_mtm_hs_ctr= [KNL,ARC] 3214 This parameter sets the maximum duration, in 3215 cycles, each HW thread of the CTOP can run 3216 without interruptions, before HW switches it. 3217 The actual maximum duration is 16 times this 3218 parameter's value. 3219 Format: integer between 1 and 255 3220 Default: 255 3221 3222 nptcg= [IA-64] Override max number of concurrent global TLB 3223 purges which is reported from either PAL_VM_SUMMARY or 3224 SAL PALO. 3225 3226 nr_cpus= [SMP] Maximum number of processors that an SMP kernel 3227 could support. nr_cpus=n : n >= 1 limits the kernel to 3228 support 'n' processors. It could be larger than the 3229 number of already plugged CPU during bootup, later in 3230 runtime you can physically add extra cpu until it reaches 3231 n. So during boot up some boot time memory for per-cpu 3232 variables need be pre-allocated for later physical cpu 3233 hot plugging. 3234 3235 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 3236 3237 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing. 3238 Allowed values are enable and disable 3239 3240 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 3241 'node', 'default' can be specified 3242 This can be set from sysctl after boot. 3243 See Documentation/admin-guide/sysctl/vm.rst for details. 3244 3245 of_devlink [OF, KNL] Create device links between consumer and 3246 supplier devices by scanning the devictree to infer the 3247 consumer/supplier relationships. A consumer device 3248 will not be probed until all the supplier devices have 3249 probed successfully. 3250 3251 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver. 3252 See Documentation/debugging-via-ohci1394.txt for more 3253 info. 3254 3255 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 3256 Rather than timing out after 20 ms if an EC 3257 command is not properly ACKed, override the length 3258 of the timeout. We have interrupts disabled while 3259 waiting for the ACK, so if this is set too high 3260 interrupts *may* be lost! 3261 3262 omap_mux= [OMAP] Override bootloader pin multiplexing. 3263 Format: <mux_mode0.mode_name=value>... 3264 For example, to override I2C bus2: 3265 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 3266 3267 oprofile.timer= [HW] 3268 Use timer interrupt instead of performance counters 3269 3270 oprofile.cpu_type= Force an oprofile cpu type 3271 This might be useful if you have an older oprofile 3272 userland or if you want common events. 3273 Format: { arch_perfmon } 3274 arch_perfmon: [X86] Force use of architectural 3275 perfmon on Intel CPUs instead of the 3276 CPU specific event set. 3277 timer: [X86] Force use of architectural NMI 3278 timer mode (see also oprofile.timer 3279 for generic hr timer mode) 3280 3281 oops=panic Always panic on oopses. Default is to just kill the 3282 process, but there is a small probability of 3283 deadlocking the machine. 3284 This will also cause panics on machine check exceptions. 3285 Useful together with panic=30 to trigger a reboot. 3286 3287 page_alloc.shuffle= 3288 [KNL] Boolean flag to control whether the page allocator 3289 should randomize its free lists. The randomization may 3290 be automatically enabled if the kernel detects it is 3291 running on a platform with a direct-mapped memory-side 3292 cache, and this parameter can be used to 3293 override/disable that behavior. The state of the flag 3294 can be read from sysfs at: 3295 /sys/module/page_alloc/parameters/shuffle. 3296 3297 page_owner= [KNL] Boot-time page_owner enabling option. 3298 Storage of the information about who allocated 3299 each page is disabled in default. With this switch, 3300 we can turn it on. 3301 on: enable the feature 3302 3303 page_poison= [KNL] Boot-time parameter changing the state of 3304 poisoning on the buddy allocator, available with 3305 CONFIG_PAGE_POISONING=y. 3306 off: turn off poisoning (default) 3307 on: turn on poisoning 3308 3309 panic= [KNL] Kernel behaviour on panic: delay <timeout> 3310 timeout > 0: seconds before rebooting 3311 timeout = 0: wait forever 3312 timeout < 0: reboot immediately 3313 Format: <timeout> 3314 3315 panic_print= Bitmask for printing system info when panic happens. 3316 User can chose combination of the following bits: 3317 bit 0: print all tasks info 3318 bit 1: print system memory info 3319 bit 2: print timer info 3320 bit 3: print locks info if CONFIG_LOCKDEP is on 3321 bit 4: print ftrace buffer 3322 bit 5: print all printk messages in buffer 3323 3324 panic_on_warn panic() instead of WARN(). Useful to cause kdump 3325 on a WARN(). 3326 3327 crash_kexec_post_notifiers 3328 Run kdump after running panic-notifiers and dumping 3329 kmsg. This only for the users who doubt kdump always 3330 succeeds in any situation. 3331 Note that this also increases risks of kdump failure, 3332 because some panic notifiers can make the crashed 3333 kernel more unstable. 3334 3335 parkbd.port= [HW] Parallel port number the keyboard adapter is 3336 connected to, default is 0. 3337 Format: <parport#> 3338 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 3339 0 for XT, 1 for AT (default is AT). 3340 Format: <mode> 3341 3342 parport= [HW,PPT] Specify parallel ports. 0 disables. 3343 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 3344 Use 'auto' to force the driver to use any 3345 IRQ/DMA settings detected (the default is to 3346 ignore detected IRQ/DMA settings because of 3347 possible conflicts). You can specify the base 3348 address, IRQ, and DMA settings; IRQ and DMA 3349 should be numbers, or 'auto' (for using detected 3350 settings on that particular port), or 'nofifo' 3351 (to avoid using a FIFO even if it is detected). 3352 Parallel ports are assigned in the order they 3353 are specified on the command line, starting 3354 with parport0. 3355 3356 parport_init_mode= [HW,PPT] 3357 Configure VIA parallel port to operate in 3358 a specific mode. This is necessary on Pegasos 3359 computer where firmware has no options for setting 3360 up parallel port mode and sets it to spp. 3361 Currently this function knows 686a and 8231 chips. 3362 Format: [spp|ps2|epp|ecp|ecpepp] 3363 3364 pause_on_oops= 3365 Halt all CPUs after the first oops has been printed for 3366 the specified number of seconds. This is to be used if 3367 your oopses keep scrolling off the screen. 3368 3369 pcbit= [HW,ISDN] 3370 3371 pcd. [PARIDE] 3372 See header of drivers/block/paride/pcd.c. 3373 See also Documentation/admin-guide/blockdev/paride.rst. 3374 3375 pci=option[,option...] [PCI] various PCI subsystem options. 3376 3377 Some options herein operate on a specific device 3378 or a set of devices (<pci_dev>). These are 3379 specified in one of the following formats: 3380 3381 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 3382 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 3383 3384 Note: the first format specifies a PCI 3385 bus/device/function address which may change 3386 if new hardware is inserted, if motherboard 3387 firmware changes, or due to changes caused 3388 by other kernel parameters. If the 3389 domain is left unspecified, it is 3390 taken to be zero. Optionally, a path 3391 to a device through multiple device/function 3392 addresses can be specified after the base 3393 address (this is more robust against 3394 renumbering issues). The second format 3395 selects devices using IDs from the 3396 configuration space which may match multiple 3397 devices in the system. 3398 3399 earlydump dump PCI config space before the kernel 3400 changes anything 3401 off [X86] don't probe for the PCI bus 3402 bios [X86-32] force use of PCI BIOS, don't access 3403 the hardware directly. Use this if your machine 3404 has a non-standard PCI host bridge. 3405 nobios [X86-32] disallow use of PCI BIOS, only direct 3406 hardware access methods are allowed. Use this 3407 if you experience crashes upon bootup and you 3408 suspect they are caused by the BIOS. 3409 conf1 [X86] Force use of PCI Configuration Access 3410 Mechanism 1 (config address in IO port 0xCF8, 3411 data in IO port 0xCFC, both 32-bit). 3412 conf2 [X86] Force use of PCI Configuration Access 3413 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 3414 the function, IO port 0xCFA, also 8-bit, sets 3415 bus number. The config space is then accessed 3416 through ports 0xC000-0xCFFF). 3417 See http://wiki.osdev.org/PCI for more info 3418 on the configuration access mechanisms. 3419 noaer [PCIE] If the PCIEAER kernel config parameter is 3420 enabled, this kernel boot option can be used to 3421 disable the use of PCIE advanced error reporting. 3422 nodomains [PCI] Disable support for multiple PCI 3423 root domains (aka PCI segments, in ACPI-speak). 3424 nommconf [X86] Disable use of MMCONFIG for PCI 3425 Configuration 3426 check_enable_amd_mmconf [X86] check for and enable 3427 properly configured MMIO access to PCI 3428 config space on AMD family 10h CPU 3429 nomsi [MSI] If the PCI_MSI kernel config parameter is 3430 enabled, this kernel boot option can be used to 3431 disable the use of MSI interrupts system-wide. 3432 noioapicquirk [APIC] Disable all boot interrupt quirks. 3433 Safety option to keep boot IRQs enabled. This 3434 should never be necessary. 3435 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 3436 primary IO-APIC for bridges that cannot disable 3437 boot IRQs. This fixes a source of spurious IRQs 3438 when the system masks IRQs. 3439 noioapicreroute [APIC] Disable workaround that uses the 3440 boot IRQ equivalent of an IRQ that connects to 3441 a chipset where boot IRQs cannot be disabled. 3442 The opposite of ioapicreroute. 3443 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 3444 routing table. These calls are known to be buggy 3445 on several machines and they hang the machine 3446 when used, but on other computers it's the only 3447 way to get the interrupt routing table. Try 3448 this option if the kernel is unable to allocate 3449 IRQs or discover secondary PCI buses on your 3450 motherboard. 3451 rom [X86] Assign address space to expansion ROMs. 3452 Use with caution as certain devices share 3453 address decoders between ROMs and other 3454 resources. 3455 norom [X86] Do not assign address space to 3456 expansion ROMs that do not already have 3457 BIOS assigned address ranges. 3458 nobar [X86] Do not assign address space to the 3459 BARs that weren't assigned by the BIOS. 3460 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 3461 assigned automatically to PCI devices. You can 3462 make the kernel exclude IRQs of your ISA cards 3463 this way. 3464 pirqaddr=0xAAAAA [X86] Specify the physical address 3465 of the PIRQ table (normally generated 3466 by the BIOS) if it is outside the 3467 F0000h-100000h range. 3468 lastbus=N [X86] Scan all buses thru bus #N. Can be 3469 useful if the kernel is unable to find your 3470 secondary buses and you want to tell it 3471 explicitly which ones they are. 3472 assign-busses [X86] Always assign all PCI bus 3473 numbers ourselves, overriding 3474 whatever the firmware may have done. 3475 usepirqmask [X86] Honor the possible IRQ mask stored 3476 in the BIOS $PIR table. This is needed on 3477 some systems with broken BIOSes, notably 3478 some HP Pavilion N5400 and Omnibook XE3 3479 notebooks. This will have no effect if ACPI 3480 IRQ routing is enabled. 3481 noacpi [X86] Do not use ACPI for IRQ routing 3482 or for PCI scanning. 3483 use_crs [X86] Use PCI host bridge window information 3484 from ACPI. On BIOSes from 2008 or later, this 3485 is enabled by default. If you need to use this, 3486 please report a bug. 3487 nocrs [X86] Ignore PCI host bridge windows from ACPI. 3488 If you need to use this, please report a bug. 3489 routeirq Do IRQ routing for all PCI devices. 3490 This is normally done in pci_enable_device(), 3491 so this option is a temporary workaround 3492 for broken drivers that don't call it. 3493 skip_isa_align [X86] do not align io start addr, so can 3494 handle more pci cards 3495 noearly [X86] Don't do any early type 1 scanning. 3496 This might help on some broken boards which 3497 machine check when some devices' config space 3498 is read. But various workarounds are disabled 3499 and some IOMMU drivers will not work. 3500 bfsort Sort PCI devices into breadth-first order. 3501 This sorting is done to get a device 3502 order compatible with older (<= 2.4) kernels. 3503 nobfsort Don't sort PCI devices into breadth-first order. 3504 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 3505 tuning and use the BIOS-configured MPS defaults. 3506 pcie_bus_safe Set every device's MPS to the largest value 3507 supported by all devices below the root complex. 3508 pcie_bus_perf Set device MPS to the largest allowable MPS 3509 based on its parent bus. Also set MRRS (Max 3510 Read Request Size) to the largest supported 3511 value (no larger than the MPS that the device 3512 or bus can support) for best performance. 3513 pcie_bus_peer2peer Set every device's MPS to 128B, which 3514 every device is guaranteed to support. This 3515 configuration allows peer-to-peer DMA between 3516 any pair of devices, possibly at the cost of 3517 reduced performance. This also guarantees 3518 that hot-added devices will work. 3519 cbiosize=nn[KMG] The fixed amount of bus space which is 3520 reserved for the CardBus bridge's IO window. 3521 The default value is 256 bytes. 3522 cbmemsize=nn[KMG] The fixed amount of bus space which is 3523 reserved for the CardBus bridge's memory 3524 window. The default value is 64 megabytes. 3525 resource_alignment= 3526 Format: 3527 [<order of align>@]<pci_dev>[; ...] 3528 Specifies alignment and device to reassign 3529 aligned memory resources. How to 3530 specify the device is described above. 3531 If <order of align> is not specified, 3532 PAGE_SIZE is used as alignment. 3533 A PCI-PCI bridge can be specified if resource 3534 windows need to be expanded. 3535 To specify the alignment for several 3536 instances of a device, the PCI vendor, 3537 device, subvendor, and subdevice may be 3538 specified, e.g., 12@pci:8086:9c22:103c:198f 3539 for 4096-byte alignment. 3540 ecrc= Enable/disable PCIe ECRC (transaction layer 3541 end-to-end CRC checking). 3542 bios: Use BIOS/firmware settings. This is the 3543 the default. 3544 off: Turn ECRC off 3545 on: Turn ECRC on. 3546 hpiosize=nn[KMG] The fixed amount of bus space which is 3547 reserved for hotplug bridge's IO window. 3548 Default size is 256 bytes. 3549 hpmmiosize=nn[KMG] The fixed amount of bus space which is 3550 reserved for hotplug bridge's MMIO window. 3551 Default size is 2 megabytes. 3552 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 3553 reserved for hotplug bridge's MMIO_PREF window. 3554 Default size is 2 megabytes. 3555 hpmemsize=nn[KMG] The fixed amount of bus space which is 3556 reserved for hotplug bridge's MMIO and 3557 MMIO_PREF window. 3558 Default size is 2 megabytes. 3559 hpbussize=nn The minimum amount of additional bus numbers 3560 reserved for buses below a hotplug bridge. 3561 Default is 1. 3562 realloc= Enable/disable reallocating PCI bridge resources 3563 if allocations done by BIOS are too small to 3564 accommodate resources required by all child 3565 devices. 3566 off: Turn realloc off 3567 on: Turn realloc on 3568 realloc same as realloc=on 3569 noari do not use PCIe ARI. 3570 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 3571 do not use PCIe ATS (and IOMMU device IOTLB). 3572 pcie_scan_all Scan all possible PCIe devices. Otherwise we 3573 only look for one device below a PCIe downstream 3574 port. 3575 big_root_window Try to add a big 64bit memory window to the PCIe 3576 root complex on AMD CPUs. Some GFX hardware 3577 can resize a BAR to allow access to all VRAM. 3578 Adding the window is slightly risky (it may 3579 conflict with unreported devices), so this 3580 taints the kernel. 3581 disable_acs_redir=<pci_dev>[; ...] 3582 Specify one or more PCI devices (in the format 3583 specified above) separated by semicolons. 3584 Each device specified will have the PCI ACS 3585 redirect capabilities forced off which will 3586 allow P2P traffic between devices through 3587 bridges without forcing it upstream. Note: 3588 this removes isolation between devices and 3589 may put more devices in an IOMMU group. 3590 force_floating [S390] Force usage of floating interrupts. 3591 nomio [S390] Do not use MIO instructions. 3592 3593 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power 3594 Management. 3595 off Disable ASPM. 3596 force Enable ASPM even on devices that claim not to support it. 3597 WARNING: Forcing ASPM on may cause system lockups. 3598 3599 pcie_ports= [PCIE] PCIe port services handling: 3600 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 3601 even if the platform doesn't give the OS permission to 3602 use them. This may cause conflicts if the platform 3603 also tries to use these services. 3604 dpc-native Use native PCIe service for DPC only. May 3605 cause conflicts if firmware uses AER or DPC. 3606 compat Disable native PCIe services (PME, AER, DPC, PCIe 3607 hotplug). 3608 3609 pcie_port_pm= [PCIE] PCIe port power management handling: 3610 off Disable power management of all PCIe ports 3611 force Forcibly enable power management of all PCIe ports 3612 3613 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 3614 nomsi Do not use MSI for native PCIe PME signaling (this makes 3615 all PCIe root ports use INTx for all services). 3616 3617 pcmv= [HW,PCMCIA] BadgePAD 4 3618 3619 pd_ignore_unused 3620 [PM] 3621 Keep all power-domains already enabled by bootloader on, 3622 even if no driver has claimed them. This is useful 3623 for debug and development, but should not be 3624 needed on a platform with proper driver support. 3625 3626 pd. [PARIDE] 3627 See Documentation/admin-guide/blockdev/paride.rst. 3628 3629 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 3630 boot time. 3631 Format: { 0 | 1 } 3632 See arch/parisc/kernel/pdc_chassis.c 3633 3634 percpu_alloc= Select which percpu first chunk allocator to use. 3635 Currently supported values are "embed" and "page". 3636 Archs may support subset or none of the selections. 3637 See comments in mm/percpu.c for details on each 3638 allocator. This parameter is primarily for debugging 3639 and performance comparison. 3640 3641 pf. [PARIDE] 3642 See Documentation/admin-guide/blockdev/paride.rst. 3643 3644 pg. [PARIDE] 3645 See Documentation/admin-guide/blockdev/paride.rst. 3646 3647 pirq= [SMP,APIC] Manual mp-table setup 3648 See Documentation/x86/i386/IO-APIC.rst. 3649 3650 plip= [PPT,NET] Parallel port network link 3651 Format: { parport<nr> | timid | 0 } 3652 See also Documentation/admin-guide/parport.rst. 3653 3654 pmtmr= [X86] Manual setup of pmtmr I/O Port. 3655 Override pmtimer IOPort with a hex value. 3656 e.g. pmtmr=0x508 3657 3658 pnp.debug=1 [PNP] 3659 Enable PNP debug messages (depends on the 3660 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 3661 via /sys/module/pnp/parameters/debug. We always show 3662 current resource usage; turning this on also shows 3663 possible settings and some assignment information. 3664 3665 pnpacpi= [ACPI] 3666 { off } 3667 3668 pnpbios= [ISAPNP] 3669 { on | off | curr | res | no-curr | no-res } 3670 3671 pnp_reserve_irq= 3672 [ISAPNP] Exclude IRQs for the autoconfiguration 3673 3674 pnp_reserve_dma= 3675 [ISAPNP] Exclude DMAs for the autoconfiguration 3676 3677 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 3678 Ranges are in pairs (I/O port base and size). 3679 3680 pnp_reserve_mem= 3681 [ISAPNP] Exclude memory regions for the 3682 autoconfiguration. 3683 Ranges are in pairs (memory base and size). 3684 3685 ports= [IP_VS_FTP] IPVS ftp helper module 3686 Default is 21. 3687 Up to 8 (IP_VS_APP_MAX_PORTS) ports 3688 may be specified. 3689 Format: <port>,<port>.... 3690 3691 powersave=off [PPC] This option disables power saving features. 3692 It specifically disables cpuidle and sets the 3693 platform machine description specific power_save 3694 function to NULL. On Idle the CPU just reduces 3695 execution priority. 3696 3697 ppc_strict_facility_enable 3698 [PPC] This option catches any kernel floating point, 3699 Altivec, VSX and SPE outside of regions specifically 3700 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 3701 There is some performance impact when enabling this. 3702 3703 ppc_tm= [PPC] 3704 Format: {"off"} 3705 Disable Hardware Transactional Memory 3706 3707 print-fatal-signals= 3708 [KNL] debug: print fatal signals 3709 3710 If enabled, warn about various signal handling 3711 related application anomalies: too many signals, 3712 too many POSIX.1 timers, fatal signals causing a 3713 coredump - etc. 3714 3715 If you hit the warning due to signal overflow, 3716 you might want to try "ulimit -i unlimited". 3717 3718 default: off. 3719 3720 printk.always_kmsg_dump= 3721 Trigger kmsg_dump for cases other than kernel oops or 3722 panics 3723 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 3724 default: disabled 3725 3726 printk.devkmsg={on,off,ratelimit} 3727 Control writing to /dev/kmsg. 3728 on - unlimited logging to /dev/kmsg from userspace 3729 off - logging to /dev/kmsg disabled 3730 ratelimit - ratelimit the logging 3731 Default: ratelimit 3732 3733 printk.time= Show timing data prefixed to each printk message line 3734 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 3735 3736 processor.max_cstate= [HW,ACPI] 3737 Limit processor to maximum C-state 3738 max_cstate=9 overrides any DMI blacklist limit. 3739 3740 processor.nocst [HW,ACPI] 3741 Ignore the _CST method to determine C-states, 3742 instead using the legacy FADT method 3743 3744 profile= [KNL] Enable kernel profiling via /proc/profile 3745 Format: [<profiletype>,]<number> 3746 Param: <profiletype>: "schedule", "sleep", or "kvm" 3747 [defaults to kernel profiling] 3748 Param: "schedule" - profile schedule points. 3749 Param: "sleep" - profile D-state sleeping (millisecs). 3750 Requires CONFIG_SCHEDSTATS 3751 Param: "kvm" - profile VM exits. 3752 Param: <number> - step/bucket size as a power of 2 for 3753 statistical time based profiling. 3754 3755 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk 3756 before loading. 3757 See Documentation/admin-guide/blockdev/ramdisk.rst. 3758 3759 psi= [KNL] Enable or disable pressure stall information 3760 tracking. 3761 Format: <bool> 3762 3763 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 3764 probe for; one of (bare|imps|exps|lifebook|any). 3765 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 3766 per second. 3767 psmouse.resetafter= [HW,MOUSE] 3768 Try to reset the device after so many bad packets 3769 (0 = never). 3770 psmouse.resolution= 3771 [HW,MOUSE] Set desired mouse resolution, in dpi. 3772 psmouse.smartscroll= 3773 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 3774 0 = disabled, 1 = enabled (default). 3775 3776 pstore.backend= Specify the name of the pstore backend to use 3777 3778 pt. [PARIDE] 3779 See Documentation/admin-guide/blockdev/paride.rst. 3780 3781 pti= [X86_64] Control Page Table Isolation of user and 3782 kernel address spaces. Disabling this feature 3783 removes hardening, but improves performance of 3784 system calls and interrupts. 3785 3786 on - unconditionally enable 3787 off - unconditionally disable 3788 auto - kernel detects whether your CPU model is 3789 vulnerable to issues that PTI mitigates 3790 3791 Not specifying this option is equivalent to pti=auto. 3792 3793 nopti [X86_64] 3794 Equivalent to pti=off 3795 3796 pty.legacy_count= 3797 [KNL] Number of legacy pty's. Overwrites compiled-in 3798 default number. 3799 3800 quiet [KNL] Disable most log messages 3801 3802 r128= [HW,DRM] 3803 3804 raid= [HW,RAID] 3805 See Documentation/admin-guide/md.rst. 3806 3807 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 3808 See Documentation/admin-guide/blockdev/ramdisk.rst. 3809 3810 random.trust_cpu={on,off} 3811 [KNL] Enable or disable trusting the use of the 3812 CPU's random number generator (if available) to 3813 fully seed the kernel's CRNG. Default is controlled 3814 by CONFIG_RANDOM_TRUST_CPU. 3815 3816 ras=option[,option,...] [KNL] RAS-specific options 3817 3818 cec_disable [X86] 3819 Disable the Correctable Errors Collector, 3820 see CONFIG_RAS_CEC help text. 3821 3822 rcu_nocbs= [KNL] 3823 The argument is a cpu list, as described above, 3824 except that the string "all" can be used to 3825 specify every CPU on the system. 3826 3827 In kernels built with CONFIG_RCU_NOCB_CPU=y, set 3828 the specified list of CPUs to be no-callback CPUs. 3829 Invocation of these CPUs' RCU callbacks will be 3830 offloaded to "rcuox/N" kthreads created for that 3831 purpose, where "x" is "p" for RCU-preempt, and 3832 "s" for RCU-sched, and "N" is the CPU number. 3833 This reduces OS jitter on the offloaded CPUs, 3834 which can be useful for HPC and real-time 3835 workloads. It can also improve energy efficiency 3836 for asymmetric multiprocessors. 3837 3838 rcu_nocb_poll [KNL] 3839 Rather than requiring that offloaded CPUs 3840 (specified by rcu_nocbs= above) explicitly 3841 awaken the corresponding "rcuoN" kthreads, 3842 make these kthreads poll for callbacks. 3843 This improves the real-time response for the 3844 offloaded CPUs by relieving them of the need to 3845 wake up the corresponding kthread, but degrades 3846 energy efficiency by requiring that the kthreads 3847 periodically wake up to do the polling. 3848 3849 rcutree.blimit= [KNL] 3850 Set maximum number of finished RCU callbacks to 3851 process in one batch. 3852 3853 rcutree.dump_tree= [KNL] 3854 Dump the structure of the rcu_node combining tree 3855 out at early boot. This is used for diagnostic 3856 purposes, to verify correct tree setup. 3857 3858 rcutree.gp_cleanup_delay= [KNL] 3859 Set the number of jiffies to delay each step of 3860 RCU grace-period cleanup. 3861 3862 rcutree.gp_init_delay= [KNL] 3863 Set the number of jiffies to delay each step of 3864 RCU grace-period initialization. 3865 3866 rcutree.gp_preinit_delay= [KNL] 3867 Set the number of jiffies to delay each step of 3868 RCU grace-period pre-initialization, that is, 3869 the propagation of recent CPU-hotplug changes up 3870 the rcu_node combining tree. 3871 3872 rcutree.use_softirq= [KNL] 3873 If set to zero, move all RCU_SOFTIRQ processing to 3874 per-CPU rcuc kthreads. Defaults to a non-zero 3875 value, meaning that RCU_SOFTIRQ is used by default. 3876 Specify rcutree.use_softirq=0 to use rcuc kthreads. 3877 3878 rcutree.rcu_fanout_exact= [KNL] 3879 Disable autobalancing of the rcu_node combining 3880 tree. This is used by rcutorture, and might 3881 possibly be useful for architectures having high 3882 cache-to-cache transfer latencies. 3883 3884 rcutree.rcu_fanout_leaf= [KNL] 3885 Change the number of CPUs assigned to each 3886 leaf rcu_node structure. Useful for very 3887 large systems, which will choose the value 64, 3888 and for NUMA systems with large remote-access 3889 latencies, which will choose a value aligned 3890 with the appropriate hardware boundaries. 3891 3892 rcutree.jiffies_till_first_fqs= [KNL] 3893 Set delay from grace-period initialization to 3894 first attempt to force quiescent states. 3895 Units are jiffies, minimum value is zero, 3896 and maximum value is HZ. 3897 3898 rcutree.jiffies_till_next_fqs= [KNL] 3899 Set delay between subsequent attempts to force 3900 quiescent states. Units are jiffies, minimum 3901 value is one, and maximum value is HZ. 3902 3903 rcutree.jiffies_till_sched_qs= [KNL] 3904 Set required age in jiffies for a 3905 given grace period before RCU starts 3906 soliciting quiescent-state help from 3907 rcu_note_context_switch() and cond_resched(). 3908 If not specified, the kernel will calculate 3909 a value based on the most recent settings 3910 of rcutree.jiffies_till_first_fqs 3911 and rcutree.jiffies_till_next_fqs. 3912 This calculated value may be viewed in 3913 rcutree.jiffies_to_sched_qs. Any attempt to set 3914 rcutree.jiffies_to_sched_qs will be cheerfully 3915 overwritten. 3916 3917 rcutree.kthread_prio= [KNL,BOOT] 3918 Set the SCHED_FIFO priority of the RCU per-CPU 3919 kthreads (rcuc/N). This value is also used for 3920 the priority of the RCU boost threads (rcub/N) 3921 and for the RCU grace-period kthreads (rcu_bh, 3922 rcu_preempt, and rcu_sched). If RCU_BOOST is 3923 set, valid values are 1-99 and the default is 1 3924 (the least-favored priority). Otherwise, when 3925 RCU_BOOST is not set, valid values are 0-99 and 3926 the default is zero (non-realtime operation). 3927 3928 rcutree.rcu_nocb_gp_stride= [KNL] 3929 Set the number of NOCB callback kthreads in 3930 each group, which defaults to the square root 3931 of the number of CPUs. Larger numbers reduce 3932 the wakeup overhead on the global grace-period 3933 kthread, but increases that same overhead on 3934 each group's NOCB grace-period kthread. 3935 3936 rcutree.qhimark= [KNL] 3937 Set threshold of queued RCU callbacks beyond which 3938 batch limiting is disabled. 3939 3940 rcutree.qlowmark= [KNL] 3941 Set threshold of queued RCU callbacks below which 3942 batch limiting is re-enabled. 3943 3944 rcutree.rcu_idle_gp_delay= [KNL] 3945 Set wakeup interval for idle CPUs that have 3946 RCU callbacks (RCU_FAST_NO_HZ=y). 3947 3948 rcutree.rcu_idle_lazy_gp_delay= [KNL] 3949 Set wakeup interval for idle CPUs that have 3950 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y). 3951 Lazy RCU callbacks are those which RCU can 3952 prove do nothing more than free memory. 3953 3954 rcutree.rcu_kick_kthreads= [KNL] 3955 Cause the grace-period kthread to get an extra 3956 wake_up() if it sleeps three times longer than 3957 it should at force-quiescent-state time. 3958 This wake_up() will be accompanied by a 3959 WARN_ONCE() splat and an ftrace_dump(). 3960 3961 rcutree.sysrq_rcu= [KNL] 3962 Commandeer a sysrq key to dump out Tree RCU's 3963 rcu_node tree with an eye towards determining 3964 why a new grace period has not yet started. 3965 3966 rcuperf.gp_async= [KNL] 3967 Measure performance of asynchronous 3968 grace-period primitives such as call_rcu(). 3969 3970 rcuperf.gp_async_max= [KNL] 3971 Specify the maximum number of outstanding 3972 callbacks per writer thread. When a writer 3973 thread exceeds this limit, it invokes the 3974 corresponding flavor of rcu_barrier() to allow 3975 previously posted callbacks to drain. 3976 3977 rcuperf.gp_exp= [KNL] 3978 Measure performance of expedited synchronous 3979 grace-period primitives. 3980 3981 rcuperf.holdoff= [KNL] 3982 Set test-start holdoff period. The purpose of 3983 this parameter is to delay the start of the 3984 test until boot completes in order to avoid 3985 interference. 3986 3987 rcuperf.nreaders= [KNL] 3988 Set number of RCU readers. The value -1 selects 3989 N, where N is the number of CPUs. A value 3990 "n" less than -1 selects N-n+1, where N is again 3991 the number of CPUs. For example, -2 selects N 3992 (the number of CPUs), -3 selects N+1, and so on. 3993 A value of "n" less than or equal to -N selects 3994 a single reader. 3995 3996 rcuperf.nwriters= [KNL] 3997 Set number of RCU writers. The values operate 3998 the same as for rcuperf.nreaders. 3999 N, where N is the number of CPUs 4000 4001 rcuperf.perf_type= [KNL] 4002 Specify the RCU implementation to test. 4003 4004 rcuperf.shutdown= [KNL] 4005 Shut the system down after performance tests 4006 complete. This is useful for hands-off automated 4007 testing. 4008 4009 rcuperf.verbose= [KNL] 4010 Enable additional printk() statements. 4011 4012 rcuperf.writer_holdoff= [KNL] 4013 Write-side holdoff between grace periods, 4014 in microseconds. The default of zero says 4015 no holdoff. 4016 4017 rcutorture.fqs_duration= [KNL] 4018 Set duration of force_quiescent_state bursts 4019 in microseconds. 4020 4021 rcutorture.fqs_holdoff= [KNL] 4022 Set holdoff time within force_quiescent_state bursts 4023 in microseconds. 4024 4025 rcutorture.fqs_stutter= [KNL] 4026 Set wait time between force_quiescent_state bursts 4027 in seconds. 4028 4029 rcutorture.fwd_progress= [KNL] 4030 Enable RCU grace-period forward-progress testing 4031 for the types of RCU supporting this notion. 4032 4033 rcutorture.fwd_progress_div= [KNL] 4034 Specify the fraction of a CPU-stall-warning 4035 period to do tight-loop forward-progress testing. 4036 4037 rcutorture.fwd_progress_holdoff= [KNL] 4038 Number of seconds to wait between successive 4039 forward-progress tests. 4040 4041 rcutorture.fwd_progress_need_resched= [KNL] 4042 Enclose cond_resched() calls within checks for 4043 need_resched() during tight-loop forward-progress 4044 testing. 4045 4046 rcutorture.gp_cond= [KNL] 4047 Use conditional/asynchronous update-side 4048 primitives, if available. 4049 4050 rcutorture.gp_exp= [KNL] 4051 Use expedited update-side primitives, if available. 4052 4053 rcutorture.gp_normal= [KNL] 4054 Use normal (non-expedited) asynchronous 4055 update-side primitives, if available. 4056 4057 rcutorture.gp_sync= [KNL] 4058 Use normal (non-expedited) synchronous 4059 update-side primitives, if available. If all 4060 of rcutorture.gp_cond=, rcutorture.gp_exp=, 4061 rcutorture.gp_normal=, and rcutorture.gp_sync= 4062 are zero, rcutorture acts as if is interpreted 4063 they are all non-zero. 4064 4065 rcutorture.n_barrier_cbs= [KNL] 4066 Set callbacks/threads for rcu_barrier() testing. 4067 4068 rcutorture.nfakewriters= [KNL] 4069 Set number of concurrent RCU writers. These just 4070 stress RCU, they don't participate in the actual 4071 test, hence the "fake". 4072 4073 rcutorture.nreaders= [KNL] 4074 Set number of RCU readers. The value -1 selects 4075 N-1, where N is the number of CPUs. A value 4076 "n" less than -1 selects N-n-2, where N is again 4077 the number of CPUs. For example, -2 selects N 4078 (the number of CPUs), -3 selects N+1, and so on. 4079 4080 rcutorture.object_debug= [KNL] 4081 Enable debug-object double-call_rcu() testing. 4082 4083 rcutorture.onoff_holdoff= [KNL] 4084 Set time (s) after boot for CPU-hotplug testing. 4085 4086 rcutorture.onoff_interval= [KNL] 4087 Set time (jiffies) between CPU-hotplug operations, 4088 or zero to disable CPU-hotplug testing. 4089 4090 rcutorture.shuffle_interval= [KNL] 4091 Set task-shuffle interval (s). Shuffling tasks 4092 allows some CPUs to go into dyntick-idle mode 4093 during the rcutorture test. 4094 4095 rcutorture.shutdown_secs= [KNL] 4096 Set time (s) after boot system shutdown. This 4097 is useful for hands-off automated testing. 4098 4099 rcutorture.stall_cpu= [KNL] 4100 Duration of CPU stall (s) to test RCU CPU stall 4101 warnings, zero to disable. 4102 4103 rcutorture.stall_cpu_holdoff= [KNL] 4104 Time to wait (s) after boot before inducing stall. 4105 4106 rcutorture.stall_cpu_irqsoff= [KNL] 4107 Disable interrupts while stalling if set. 4108 4109 rcutorture.stat_interval= [KNL] 4110 Time (s) between statistics printk()s. 4111 4112 rcutorture.stutter= [KNL] 4113 Time (s) to stutter testing, for example, specifying 4114 five seconds causes the test to run for five seconds, 4115 wait for five seconds, and so on. This tests RCU's 4116 ability to transition abruptly to and from idle. 4117 4118 rcutorture.test_boost= [KNL] 4119 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 4120 "Maybe" means test if the RCU implementation 4121 under test support RCU priority boosting. 4122 4123 rcutorture.test_boost_duration= [KNL] 4124 Duration (s) of each individual boost test. 4125 4126 rcutorture.test_boost_interval= [KNL] 4127 Interval (s) between each boost test. 4128 4129 rcutorture.test_no_idle_hz= [KNL] 4130 Test RCU's dyntick-idle handling. See also the 4131 rcutorture.shuffle_interval parameter. 4132 4133 rcutorture.torture_type= [KNL] 4134 Specify the RCU implementation to test. 4135 4136 rcutorture.verbose= [KNL] 4137 Enable additional printk() statements. 4138 4139 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 4140 Dump ftrace buffer after reporting RCU CPU 4141 stall warning. 4142 4143 rcupdate.rcu_cpu_stall_suppress= [KNL] 4144 Suppress RCU CPU stall warning messages. 4145 4146 rcupdate.rcu_cpu_stall_timeout= [KNL] 4147 Set timeout for RCU CPU stall warning messages. 4148 4149 rcupdate.rcu_expedited= [KNL] 4150 Use expedited grace-period primitives, for 4151 example, synchronize_rcu_expedited() instead 4152 of synchronize_rcu(). This reduces latency, 4153 but can increase CPU utilization, degrade 4154 real-time latency, and degrade energy efficiency. 4155 No effect on CONFIG_TINY_RCU kernels. 4156 4157 rcupdate.rcu_normal= [KNL] 4158 Use only normal grace-period primitives, 4159 for example, synchronize_rcu() instead of 4160 synchronize_rcu_expedited(). This improves 4161 real-time latency, CPU utilization, and 4162 energy efficiency, but can expose users to 4163 increased grace-period latency. This parameter 4164 overrides rcupdate.rcu_expedited. No effect on 4165 CONFIG_TINY_RCU kernels. 4166 4167 rcupdate.rcu_normal_after_boot= [KNL] 4168 Once boot has completed (that is, after 4169 rcu_end_inkernel_boot() has been invoked), use 4170 only normal grace-period primitives. No effect 4171 on CONFIG_TINY_RCU kernels. 4172 4173 rcupdate.rcu_task_stall_timeout= [KNL] 4174 Set timeout in jiffies for RCU task stall warning 4175 messages. Disable with a value less than or equal 4176 to zero. 4177 4178 rcupdate.rcu_self_test= [KNL] 4179 Run the RCU early boot self tests 4180 4181 rdinit= [KNL] 4182 Format: <full_path> 4183 Run specified binary instead of /init from the ramdisk, 4184 used for early userspace startup. See initrd. 4185 4186 rdrand= [X86] 4187 force - Override the decision by the kernel to hide the 4188 advertisement of RDRAND support (this affects 4189 certain AMD processors because of buggy BIOS 4190 support, specifically around the suspend/resume 4191 path). 4192 4193 rdt= [HW,X86,RDT] 4194 Turn on/off individual RDT features. List is: 4195 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 4196 mba. 4197 E.g. to turn on cmt and turn off mba use: 4198 rdt=cmt,!mba 4199 4200 reboot= [KNL] 4201 Format (x86 or x86_64): 4202 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \ 4203 [[,]s[mp]#### \ 4204 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 4205 [[,]f[orce] 4206 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 4207 (prefix with 'panic_' to set mode for panic 4208 reboot only), 4209 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 4210 reboot_force is either force or not specified, 4211 reboot_cpu is s[mp]#### with #### being the processor 4212 to be used for rebooting. 4213 4214 relax_domain_level= 4215 [KNL, SMP] Set scheduler's default relax_domain_level. 4216 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 4217 4218 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 4219 Format: <base1>,<size1>[,<base2>,<size2>,...] 4220 Reserve I/O ports or memory so the kernel won't use 4221 them. If <base> is less than 0x10000, the region 4222 is assumed to be I/O ports; otherwise it is memory. 4223 4224 reservetop= [X86-32] 4225 Format: nn[KMG] 4226 Reserves a hole at the top of the kernel virtual 4227 address space. 4228 4229 reservelow= [X86] 4230 Format: nn[K] 4231 Set the amount of memory to reserve for BIOS at 4232 the bottom of the address space. 4233 4234 reset_devices [KNL] Force drivers to reset the underlying device 4235 during initialization. 4236 4237 resume= [SWSUSP] 4238 Specify the partition device for software suspend 4239 Format: 4240 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 4241 4242 resume_offset= [SWSUSP] 4243 Specify the offset from the beginning of the partition 4244 given by "resume=" at which the swap header is located, 4245 in <PAGE_SIZE> units (needed only for swap files). 4246 See Documentation/power/swsusp-and-swap-files.rst 4247 4248 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 4249 read the resume files 4250 4251 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 4252 Useful for devices that are detected asynchronously 4253 (e.g. USB and MMC devices). 4254 4255 hibernate= [HIBERNATION] 4256 noresume Don't check if there's a hibernation image 4257 present during boot. 4258 nocompress Don't compress/decompress hibernation images. 4259 no Disable hibernation and resume. 4260 protect_image Turn on image protection during restoration 4261 (that will set all pages holding image data 4262 during restoration read-only). 4263 4264 retain_initrd [RAM] Keep initrd memory after extraction 4265 4266 rfkill.default_state= 4267 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 4268 etc. communication is blocked by default. 4269 1 Unblocked. 4270 4271 rfkill.master_switch_mode= 4272 0 The "airplane mode" button does nothing. 4273 1 The "airplane mode" button toggles between everything 4274 blocked and the previous configuration. 4275 2 The "airplane mode" button toggles between everything 4276 blocked and everything unblocked. 4277 4278 rhash_entries= [KNL,NET] 4279 Set number of hash buckets for route cache 4280 4281 ring3mwait=disable 4282 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 4283 CPUs. 4284 4285 ro [KNL] Mount root device read-only on boot 4286 4287 rodata= [KNL] 4288 on Mark read-only kernel memory as read-only (default). 4289 off Leave read-only kernel memory writable for debugging. 4290 4291 rockchip.usb_uart 4292 Enable the uart passthrough on the designated usb port 4293 on Rockchip SoCs. When active, the signals of the 4294 debug-uart get routed to the D+ and D- pins of the usb 4295 port and the regular usb controller gets disabled. 4296 4297 root= [KNL] Root filesystem 4298 See name_to_dev_t comment in init/do_mounts.c. 4299 4300 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 4301 mount the root filesystem 4302 4303 rootflags= [KNL] Set root filesystem mount option string 4304 4305 rootfstype= [KNL] Set root filesystem type 4306 4307 rootwait [KNL] Wait (indefinitely) for root device to show up. 4308 Useful for devices that are detected asynchronously 4309 (e.g. USB and MMC devices). 4310 4311 rproc_mem=nn[KMG][@address] 4312 [KNL,ARM,CMA] Remoteproc physical memory block. 4313 Memory area to be used by remote processor image, 4314 managed by CMA. 4315 4316 rw [KNL] Mount root device read-write on boot 4317 4318 S [KNL] Run init in single mode 4319 4320 s390_iommu= [HW,S390] 4321 Set s390 IOTLB flushing mode 4322 strict 4323 With strict flushing every unmap operation will result in 4324 an IOTLB flush. Default is lazy flushing before reuse, 4325 which is faster. 4326 4327 sa1100ir [NET] 4328 See drivers/net/irda/sa1100_ir.c. 4329 4330 sbni= [NET] Granch SBNI12 leased line adapter 4331 4332 sched_debug [KNL] Enables verbose scheduler debug messages. 4333 4334 schedstats= [KNL,X86] Enable or disable scheduled statistics. 4335 Allowed values are enable and disable. This feature 4336 incurs a small amount of overhead in the scheduler 4337 but is useful for debugging and performance tuning. 4338 4339 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate 4340 xtime_lock contention on larger systems, and/or RCU lock 4341 contention on all systems with CONFIG_MAXSMP set. 4342 Format: { "0" | "1" } 4343 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 4344 1 -- enable. 4345 Note: increases power consumption, thus should only be 4346 enabled if running jitter sensitive (HPC/RT) workloads. 4347 4348 security= [SECURITY] Choose a legacy "major" security module to 4349 enable at boot. This has been deprecated by the 4350 "lsm=" parameter. 4351 4352 selinux= [SELINUX] Disable or enable SELinux at boot time. 4353 Format: { "0" | "1" } 4354 See security/selinux/Kconfig help text. 4355 0 -- disable. 4356 1 -- enable. 4357 Default value is set via kernel config option. 4358 If enabled at boot time, /selinux/disable can be used 4359 later to disable prior to initial policy load. 4360 4361 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 4362 Format: { "0" | "1" } 4363 See security/apparmor/Kconfig help text 4364 0 -- disable. 4365 1 -- enable. 4366 Default value is set via kernel config option. 4367 4368 serialnumber [BUGS=X86-32] 4369 4370 shapers= [NET] 4371 Maximal number of shapers. 4372 4373 simeth= [IA-64] 4374 simscsi= 4375 4376 slram= [HW,MTD] 4377 4378 slab_nomerge [MM] 4379 Disable merging of slabs with similar size. May be 4380 necessary if there is some reason to distinguish 4381 allocs to different slabs, especially in hardened 4382 environments where the risk of heap overflows and 4383 layout control by attackers can usually be 4384 frustrated by disabling merging. This will reduce 4385 most of the exposure of a heap attack to a single 4386 cache (risks via metadata attacks are mostly 4387 unchanged). Debug options disable merging on their 4388 own. 4389 For more information see Documentation/vm/slub.rst. 4390 4391 slab_max_order= [MM, SLAB] 4392 Determines the maximum allowed order for slabs. 4393 A high setting may cause OOMs due to memory 4394 fragmentation. Defaults to 1 for systems with 4395 more than 32MB of RAM, 0 otherwise. 4396 4397 slub_debug[=options[,slabs]] [MM, SLUB] 4398 Enabling slub_debug allows one to determine the 4399 culprit if slab objects become corrupted. Enabling 4400 slub_debug can create guard zones around objects and 4401 may poison objects when not in use. Also tracks the 4402 last alloc / free. For more information see 4403 Documentation/vm/slub.rst. 4404 4405 slub_memcg_sysfs= [MM, SLUB] 4406 Determines whether to enable sysfs directories for 4407 memory cgroup sub-caches. 1 to enable, 0 to disable. 4408 The default is determined by CONFIG_SLUB_MEMCG_SYSFS_ON. 4409 Enabling this can lead to a very high number of debug 4410 directories and files being created under 4411 /sys/kernel/slub. 4412 4413 slub_max_order= [MM, SLUB] 4414 Determines the maximum allowed order for slabs. 4415 A high setting may cause OOMs due to memory 4416 fragmentation. For more information see 4417 Documentation/vm/slub.rst. 4418 4419 slub_min_objects= [MM, SLUB] 4420 The minimum number of objects per slab. SLUB will 4421 increase the slab order up to slub_max_order to 4422 generate a sufficiently large slab able to contain 4423 the number of objects indicated. The higher the number 4424 of objects the smaller the overhead of tracking slabs 4425 and the less frequently locks need to be acquired. 4426 For more information see Documentation/vm/slub.rst. 4427 4428 slub_min_order= [MM, SLUB] 4429 Determines the minimum page order for slabs. Must be 4430 lower than slub_max_order. 4431 For more information see Documentation/vm/slub.rst. 4432 4433 slub_nomerge [MM, SLUB] 4434 Same with slab_nomerge. This is supported for legacy. 4435 See slab_nomerge for more information. 4436 4437 smart2= [HW] 4438 Format: <io1>[,<io2>[,...,<io8>]] 4439 4440 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 4441 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 4442 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 4443 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 4444 smsc-ircc2.ircc_irq= [HW] IRQ line 4445 smsc-ircc2.ircc_dma= [HW] DMA channel 4446 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 4447 0: Toshiba Satellite 1800 (GP data pin select) 4448 1: Fast pin select (default) 4449 2: ATC IRMode 4450 4451 smt [KNL,S390] Set the maximum number of threads (logical 4452 CPUs) to use per physical CPU on systems capable of 4453 symmetric multithreading (SMT). Will be capped to the 4454 actual hardware limit. 4455 Format: <integer> 4456 Default: -1 (no limit) 4457 4458 softlockup_panic= 4459 [KNL] Should the soft-lockup detector generate panics. 4460 Format: <integer> 4461 4462 A nonzero value instructs the soft-lockup detector 4463 to panic the machine when a soft-lockup occurs. This 4464 is also controlled by CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC 4465 which is the respective build-time switch to that 4466 functionality. 4467 4468 softlockup_all_cpu_backtrace= 4469 [KNL] Should the soft-lockup detector generate 4470 backtraces on all cpus. 4471 Format: <integer> 4472 4473 sonypi.*= [HW] Sony Programmable I/O Control Device driver 4474 See Documentation/admin-guide/laptops/sonypi.rst 4475 4476 spectre_v2= [X86] Control mitigation of Spectre variant 2 4477 (indirect branch speculation) vulnerability. 4478 The default operation protects the kernel from 4479 user space attacks. 4480 4481 on - unconditionally enable, implies 4482 spectre_v2_user=on 4483 off - unconditionally disable, implies 4484 spectre_v2_user=off 4485 auto - kernel detects whether your CPU model is 4486 vulnerable 4487 4488 Selecting 'on' will, and 'auto' may, choose a 4489 mitigation method at run time according to the 4490 CPU, the available microcode, the setting of the 4491 CONFIG_RETPOLINE configuration option, and the 4492 compiler with which the kernel was built. 4493 4494 Selecting 'on' will also enable the mitigation 4495 against user space to user space task attacks. 4496 4497 Selecting 'off' will disable both the kernel and 4498 the user space protections. 4499 4500 Specific mitigations can also be selected manually: 4501 4502 retpoline - replace indirect branches 4503 retpoline,generic - google's original retpoline 4504 retpoline,amd - AMD-specific minimal thunk 4505 4506 Not specifying this option is equivalent to 4507 spectre_v2=auto. 4508 4509 spectre_v2_user= 4510 [X86] Control mitigation of Spectre variant 2 4511 (indirect branch speculation) vulnerability between 4512 user space tasks 4513 4514 on - Unconditionally enable mitigations. Is 4515 enforced by spectre_v2=on 4516 4517 off - Unconditionally disable mitigations. Is 4518 enforced by spectre_v2=off 4519 4520 prctl - Indirect branch speculation is enabled, 4521 but mitigation can be enabled via prctl 4522 per thread. The mitigation control state 4523 is inherited on fork. 4524 4525 prctl,ibpb 4526 - Like "prctl" above, but only STIBP is 4527 controlled per thread. IBPB is issued 4528 always when switching between different user 4529 space processes. 4530 4531 seccomp 4532 - Same as "prctl" above, but all seccomp 4533 threads will enable the mitigation unless 4534 they explicitly opt out. 4535 4536 seccomp,ibpb 4537 - Like "seccomp" above, but only STIBP is 4538 controlled per thread. IBPB is issued 4539 always when switching between different 4540 user space processes. 4541 4542 auto - Kernel selects the mitigation depending on 4543 the available CPU features and vulnerability. 4544 4545 Default mitigation: 4546 If CONFIG_SECCOMP=y then "seccomp", otherwise "prctl" 4547 4548 Not specifying this option is equivalent to 4549 spectre_v2_user=auto. 4550 4551 spec_store_bypass_disable= 4552 [HW] Control Speculative Store Bypass (SSB) Disable mitigation 4553 (Speculative Store Bypass vulnerability) 4554 4555 Certain CPUs are vulnerable to an exploit against a 4556 a common industry wide performance optimization known 4557 as "Speculative Store Bypass" in which recent stores 4558 to the same memory location may not be observed by 4559 later loads during speculative execution. The idea 4560 is that such stores are unlikely and that they can 4561 be detected prior to instruction retirement at the 4562 end of a particular speculation execution window. 4563 4564 In vulnerable processors, the speculatively forwarded 4565 store can be used in a cache side channel attack, for 4566 example to read memory to which the attacker does not 4567 directly have access (e.g. inside sandboxed code). 4568 4569 This parameter controls whether the Speculative Store 4570 Bypass optimization is used. 4571 4572 On x86 the options are: 4573 4574 on - Unconditionally disable Speculative Store Bypass 4575 off - Unconditionally enable Speculative Store Bypass 4576 auto - Kernel detects whether the CPU model contains an 4577 implementation of Speculative Store Bypass and 4578 picks the most appropriate mitigation. If the 4579 CPU is not vulnerable, "off" is selected. If the 4580 CPU is vulnerable the default mitigation is 4581 architecture and Kconfig dependent. See below. 4582 prctl - Control Speculative Store Bypass per thread 4583 via prctl. Speculative Store Bypass is enabled 4584 for a process by default. The state of the control 4585 is inherited on fork. 4586 seccomp - Same as "prctl" above, but all seccomp threads 4587 will disable SSB unless they explicitly opt out. 4588 4589 Default mitigations: 4590 X86: If CONFIG_SECCOMP=y "seccomp", otherwise "prctl" 4591 4592 On powerpc the options are: 4593 4594 on,auto - On Power8 and Power9 insert a store-forwarding 4595 barrier on kernel entry and exit. On Power7 4596 perform a software flush on kernel entry and 4597 exit. 4598 off - No action. 4599 4600 Not specifying this option is equivalent to 4601 spec_store_bypass_disable=auto. 4602 4603 spia_io_base= [HW,MTD] 4604 spia_fio_base= 4605 spia_pedr= 4606 spia_peddr= 4607 4608 srcutree.counter_wrap_check [KNL] 4609 Specifies how frequently to check for 4610 grace-period sequence counter wrap for the 4611 srcu_data structure's ->srcu_gp_seq_needed field. 4612 The greater the number of bits set in this kernel 4613 parameter, the less frequently counter wrap will 4614 be checked for. Note that the bottom two bits 4615 are ignored. 4616 4617 srcutree.exp_holdoff [KNL] 4618 Specifies how many nanoseconds must elapse 4619 since the end of the last SRCU grace period for 4620 a given srcu_struct until the next normal SRCU 4621 grace period will be considered for automatic 4622 expediting. Set to zero to disable automatic 4623 expediting. 4624 4625 ssbd= [ARM64,HW] 4626 Speculative Store Bypass Disable control 4627 4628 On CPUs that are vulnerable to the Speculative 4629 Store Bypass vulnerability and offer a 4630 firmware based mitigation, this parameter 4631 indicates how the mitigation should be used: 4632 4633 force-on: Unconditionally enable mitigation for 4634 for both kernel and userspace 4635 force-off: Unconditionally disable mitigation for 4636 for both kernel and userspace 4637 kernel: Always enable mitigation in the 4638 kernel, and offer a prctl interface 4639 to allow userspace to register its 4640 interest in being mitigated too. 4641 4642 stack_guard_gap= [MM] 4643 override the default stack gap protection. The value 4644 is in page units and it defines how many pages prior 4645 to (for stacks growing down) resp. after (for stacks 4646 growing up) the main stack are reserved for no other 4647 mapping. Default value is 256 pages. 4648 4649 stacktrace [FTRACE] 4650 Enabled the stack tracer on boot up. 4651 4652 stacktrace_filter=[function-list] 4653 [FTRACE] Limit the functions that the stack tracer 4654 will trace at boot up. function-list is a comma separated 4655 list of functions. This list can be changed at run 4656 time by the stack_trace_filter file in the debugfs 4657 tracing directory. Note, this enables stack tracing 4658 and the stacktrace above is not needed. 4659 4660 sti= [PARISC,HW] 4661 Format: <num> 4662 Set the STI (builtin display/keyboard on the HP-PARISC 4663 machines) console (graphic card) which should be used 4664 as the initial boot-console. 4665 See also comment in drivers/video/console/sticore.c. 4666 4667 sti_font= [HW] 4668 See comment in drivers/video/console/sticore.c. 4669 4670 stifb= [HW] 4671 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 4672 4673 sunrpc.min_resvport= 4674 sunrpc.max_resvport= 4675 [NFS,SUNRPC] 4676 SunRPC servers often require that client requests 4677 originate from a privileged port (i.e. a port in the 4678 range 0 < portnr < 1024). 4679 An administrator who wishes to reserve some of these 4680 ports for other uses may adjust the range that the 4681 kernel's sunrpc client considers to be privileged 4682 using these two parameters to set the minimum and 4683 maximum port values. 4684 4685 sunrpc.svc_rpc_per_connection_limit= 4686 [NFS,SUNRPC] 4687 Limit the number of requests that the server will 4688 process in parallel from a single connection. 4689 The default value is 0 (no limit). 4690 4691 sunrpc.pool_mode= 4692 [NFS] 4693 Control how the NFS server code allocates CPUs to 4694 service thread pools. Depending on how many NICs 4695 you have and where their interrupts are bound, this 4696 option will affect which CPUs will do NFS serving. 4697 Note: this parameter cannot be changed while the 4698 NFS server is running. 4699 4700 auto the server chooses an appropriate mode 4701 automatically using heuristics 4702 global a single global pool contains all CPUs 4703 percpu one pool for each CPU 4704 pernode one pool for each NUMA node (equivalent 4705 to global on non-NUMA machines) 4706 4707 sunrpc.tcp_slot_table_entries= 4708 sunrpc.udp_slot_table_entries= 4709 [NFS,SUNRPC] 4710 Sets the upper limit on the number of simultaneous 4711 RPC calls that can be sent from the client to a 4712 server. Increasing these values may allow you to 4713 improve throughput, but will also increase the 4714 amount of memory reserved for use by the client. 4715 4716 suspend.pm_test_delay= 4717 [SUSPEND] 4718 Sets the number of seconds to remain in a suspend test 4719 mode before resuming the system (see 4720 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 4721 is set. Default value is 5. 4722 4723 svm= [PPC] 4724 Format: { on | off | y | n | 1 | 0 } 4725 This parameter controls use of the Protected 4726 Execution Facility on pSeries. 4727 4728 swapaccount=[0|1] 4729 [KNL] Enable accounting of swap in memory resource 4730 controller if no parameter or 1 is given or disable 4731 it if 0 is given (See Documentation/admin-guide/cgroup-v1/memory.rst) 4732 4733 swiotlb= [ARM,IA-64,PPC,MIPS,X86] 4734 Format: { <int> | force | noforce } 4735 <int> -- Number of I/O TLB slabs 4736 force -- force using of bounce buffers even if they 4737 wouldn't be automatically used by the kernel 4738 noforce -- Never use bounce buffers (for debugging) 4739 4740 switches= [HW,M68k] 4741 4742 sysfs.deprecated=0|1 [KNL] 4743 Enable/disable old style sysfs layout for old udev 4744 on older distributions. When this option is enabled 4745 very new udev will not work anymore. When this option 4746 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled) 4747 in older udev will not work anymore. 4748 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in 4749 the kernel configuration. 4750 4751 sysrq_always_enabled 4752 [KNL] 4753 Ignore sysrq setting - this boot parameter will 4754 neutralize any effect of /proc/sys/kernel/sysrq. 4755 Useful for debugging. 4756 4757 tcpmhash_entries= [KNL,NET] 4758 Set the number of tcp_metrics_hash slots. 4759 Default value is 8192 or 16384 depending on total 4760 ram pages. This is used to specify the TCP metrics 4761 cache size. See Documentation/networking/ip-sysctl.txt 4762 "tcp_no_metrics_save" section for more details. 4763 4764 tdfx= [HW,DRM] 4765 4766 test_suspend= [SUSPEND][,N] 4767 Specify "mem" (for Suspend-to-RAM) or "standby" (for 4768 standby suspend) or "freeze" (for suspend type freeze) 4769 as the system sleep state during system startup with 4770 the optional capability to repeat N number of times. 4771 The system is woken from this state using a 4772 wakeup-capable RTC alarm. 4773 4774 thash_entries= [KNL,NET] 4775 Set number of hash buckets for TCP connection 4776 4777 thermal.act= [HW,ACPI] 4778 -1: disable all active trip points in all thermal zones 4779 <degrees C>: override all lowest active trip points 4780 4781 thermal.crt= [HW,ACPI] 4782 -1: disable all critical trip points in all thermal zones 4783 <degrees C>: override all critical trip points 4784 4785 thermal.nocrt= [HW,ACPI] 4786 Set to disable actions on ACPI thermal zone 4787 critical and hot trip points. 4788 4789 thermal.off= [HW,ACPI] 4790 1: disable ACPI thermal control 4791 4792 thermal.psv= [HW,ACPI] 4793 -1: disable all passive trip points 4794 <degrees C>: override all passive trip points to this 4795 value 4796 4797 thermal.tzp= [HW,ACPI] 4798 Specify global default ACPI thermal zone polling rate 4799 <deci-seconds>: poll all this frequency 4800 0: no polling (default) 4801 4802 threadirqs [KNL] 4803 Force threading of all interrupt handlers except those 4804 marked explicitly IRQF_NO_THREAD. 4805 4806 topology= [S390] 4807 Format: {off | on} 4808 Specify if the kernel should make use of the cpu 4809 topology information if the hardware supports this. 4810 The scheduler will make use of this information and 4811 e.g. base its process migration decisions on it. 4812 Default is on. 4813 4814 topology_updates= [KNL, PPC, NUMA] 4815 Format: {off} 4816 Specify if the kernel should ignore (off) 4817 topology updates sent by the hypervisor to this 4818 LPAR. 4819 4820 tp720= [HW,PS2] 4821 4822 tpm_suspend_pcr=[HW,TPM] 4823 Format: integer pcr id 4824 Specify that at suspend time, the tpm driver 4825 should extend the specified pcr with zeros, 4826 as a workaround for some chips which fail to 4827 flush the last written pcr on TPM_SaveState. 4828 This will guarantee that all the other pcrs 4829 are saved. 4830 4831 trace_buf_size=nn[KMG] 4832 [FTRACE] will set tracing buffer size on each cpu. 4833 4834 trace_event=[event-list] 4835 [FTRACE] Set and start specified trace events in order 4836 to facilitate early boot debugging. The event-list is a 4837 comma separated list of trace events to enable. See 4838 also Documentation/trace/events.rst 4839 4840 trace_options=[option-list] 4841 [FTRACE] Enable or disable tracer options at boot. 4842 The option-list is a comma delimited list of options 4843 that can be enabled or disabled just as if you were 4844 to echo the option name into 4845 4846 /sys/kernel/debug/tracing/trace_options 4847 4848 For example, to enable stacktrace option (to dump the 4849 stack trace of each event), add to the command line: 4850 4851 trace_options=stacktrace 4852 4853 See also Documentation/trace/ftrace.rst "trace options" 4854 section. 4855 4856 tp_printk[FTRACE] 4857 Have the tracepoints sent to printk as well as the 4858 tracing ring buffer. This is useful for early boot up 4859 where the system hangs or reboots and does not give the 4860 option for reading the tracing buffer or performing a 4861 ftrace_dump_on_oops. 4862 4863 To turn off having tracepoints sent to printk, 4864 echo 0 > /proc/sys/kernel/tracepoint_printk 4865 Note, echoing 1 into this file without the 4866 tracepoint_printk kernel cmdline option has no effect. 4867 4868 ** CAUTION ** 4869 4870 Having tracepoints sent to printk() and activating high 4871 frequency tracepoints such as irq or sched, can cause 4872 the system to live lock. 4873 4874 traceoff_on_warning 4875 [FTRACE] enable this option to disable tracing when a 4876 warning is hit. This turns off "tracing_on". Tracing can 4877 be enabled again by echoing '1' into the "tracing_on" 4878 file located in /sys/kernel/debug/tracing/ 4879 4880 This option is useful, as it disables the trace before 4881 the WARNING dump is called, which prevents the trace to 4882 be filled with content caused by the warning output. 4883 4884 This option can also be set at run time via the sysctl 4885 option: kernel/traceoff_on_warning 4886 4887 transparent_hugepage= 4888 [KNL] 4889 Format: [always|madvise|never] 4890 Can be used to control the default behavior of the system 4891 with respect to transparent hugepages. 4892 See Documentation/admin-guide/mm/transhuge.rst 4893 for more details. 4894 4895 tsc= Disable clocksource stability checks for TSC. 4896 Format: <string> 4897 [x86] reliable: mark tsc clocksource as reliable, this 4898 disables clocksource verification at runtime, as well 4899 as the stability checks done at bootup. Used to enable 4900 high-resolution timer mode on older hardware, and in 4901 virtualized environment. 4902 [x86] noirqtime: Do not use TSC to do irq accounting. 4903 Used to run time disable IRQ_TIME_ACCOUNTING on any 4904 platforms where RDTSC is slow and this accounting 4905 can add overhead. 4906 [x86] unstable: mark the TSC clocksource as unstable, this 4907 marks the TSC unconditionally unstable at bootup and 4908 avoids any further wobbles once the TSC watchdog notices. 4909 [x86] nowatchdog: disable clocksource watchdog. Used 4910 in situations with strict latency requirements (where 4911 interruptions from clocksource watchdog are not 4912 acceptable). 4913 4914 tsx= [X86] Control Transactional Synchronization 4915 Extensions (TSX) feature in Intel processors that 4916 support TSX control. 4917 4918 This parameter controls the TSX feature. The options are: 4919 4920 on - Enable TSX on the system. Although there are 4921 mitigations for all known security vulnerabilities, 4922 TSX has been known to be an accelerator for 4923 several previous speculation-related CVEs, and 4924 so there may be unknown security risks associated 4925 with leaving it enabled. 4926 4927 off - Disable TSX on the system. (Note that this 4928 option takes effect only on newer CPUs which are 4929 not vulnerable to MDS, i.e., have 4930 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 4931 the new IA32_TSX_CTRL MSR through a microcode 4932 update. This new MSR allows for the reliable 4933 deactivation of the TSX functionality.) 4934 4935 auto - Disable TSX if X86_BUG_TAA is present, 4936 otherwise enable TSX on the system. 4937 4938 Not specifying this option is equivalent to tsx=off. 4939 4940 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 4941 for more details. 4942 4943 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async 4944 Abort (TAA) vulnerability. 4945 4946 Similar to Micro-architectural Data Sampling (MDS) 4947 certain CPUs that support Transactional 4948 Synchronization Extensions (TSX) are vulnerable to an 4949 exploit against CPU internal buffers which can forward 4950 information to a disclosure gadget under certain 4951 conditions. 4952 4953 In vulnerable processors, the speculatively forwarded 4954 data can be used in a cache side channel attack, to 4955 access data to which the attacker does not have direct 4956 access. 4957 4958 This parameter controls the TAA mitigation. The 4959 options are: 4960 4961 full - Enable TAA mitigation on vulnerable CPUs 4962 if TSX is enabled. 4963 4964 full,nosmt - Enable TAA mitigation and disable SMT on 4965 vulnerable CPUs. If TSX is disabled, SMT 4966 is not disabled because CPU is not 4967 vulnerable to cross-thread TAA attacks. 4968 off - Unconditionally disable TAA mitigation 4969 4970 On MDS-affected machines, tsx_async_abort=off can be 4971 prevented by an active MDS mitigation as both vulnerabilities 4972 are mitigated with the same mechanism so in order to disable 4973 this mitigation, you need to specify mds=off too. 4974 4975 Not specifying this option is equivalent to 4976 tsx_async_abort=full. On CPUs which are MDS affected 4977 and deploy MDS mitigation, TAA mitigation is not 4978 required and doesn't provide any additional 4979 mitigation. 4980 4981 For details see: 4982 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 4983 4984 turbografx.map[2|3]= [HW,JOY] 4985 TurboGraFX parallel port interface 4986 Format: 4987 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 4988 See also Documentation/input/devices/joystick-parport.rst 4989 4990 udbg-immortal [PPC] When debugging early kernel crashes that 4991 happen after console_init() and before a proper 4992 console driver takes over, this boot options might 4993 help "seeing" what's going on. 4994 4995 uhash_entries= [KNL,NET] 4996 Set number of hash buckets for UDP/UDP-Lite connections 4997 4998 uhci-hcd.ignore_oc= 4999 [USB] Ignore overcurrent events (default N). 5000 Some badly-designed motherboards generate lots of 5001 bogus events, for ports that aren't wired to 5002 anything. Set this parameter to avoid log spamming. 5003 Note that genuine overcurrent events won't be 5004 reported either. 5005 5006 unknown_nmi_panic 5007 [X86] Cause panic on unknown NMI. 5008 5009 usbcore.authorized_default= 5010 [USB] Default USB device authorization: 5011 (default -1 = authorized except for wireless USB, 5012 0 = not authorized, 1 = authorized, 2 = authorized 5013 if device connected to internal port) 5014 5015 usbcore.autosuspend= 5016 [USB] The autosuspend time delay (in seconds) used 5017 for newly-detected USB devices (default 2). This 5018 is the time required before an idle device will be 5019 autosuspended. Devices for which the delay is set 5020 to a negative value won't be autosuspended at all. 5021 5022 usbcore.usbfs_snoop= 5023 [USB] Set to log all usbfs traffic (default 0 = off). 5024 5025 usbcore.usbfs_snoop_max= 5026 [USB] Maximum number of bytes to snoop in each URB 5027 (default = 65536). 5028 5029 usbcore.blinkenlights= 5030 [USB] Set to cycle leds on hubs (default 0 = off). 5031 5032 usbcore.old_scheme_first= 5033 [USB] Start with the old device initialization 5034 scheme, applies only to low and full-speed devices 5035 (default 0 = off). 5036 5037 usbcore.usbfs_memory_mb= 5038 [USB] Memory limit (in MB) for buffers allocated by 5039 usbfs (default = 16, 0 = max = 2047). 5040 5041 usbcore.use_both_schemes= 5042 [USB] Try the other device initialization scheme 5043 if the first one fails (default 1 = enabled). 5044 5045 usbcore.initial_descriptor_timeout= 5046 [USB] Specifies timeout for the initial 64-byte 5047 USB_REQ_GET_DESCRIPTOR request in milliseconds 5048 (default 5000 = 5.0 seconds). 5049 5050 usbcore.nousb [USB] Disable the USB subsystem 5051 5052 usbcore.quirks= 5053 [USB] A list of quirk entries to augment the built-in 5054 usb core quirk list. List entries are separated by 5055 commas. Each entry has the form 5056 VendorID:ProductID:Flags. The IDs are 4-digit hex 5057 numbers and Flags is a set of letters. Each letter 5058 will change the built-in quirk; setting it if it is 5059 clear and clearing it if it is set. The letters have 5060 the following meanings: 5061 a = USB_QUIRK_STRING_FETCH_255 (string 5062 descriptors must not be fetched using 5063 a 255-byte read); 5064 b = USB_QUIRK_RESET_RESUME (device can't resume 5065 correctly so reset it instead); 5066 c = USB_QUIRK_NO_SET_INTF (device can't handle 5067 Set-Interface requests); 5068 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 5069 handle its Configuration or Interface 5070 strings); 5071 e = USB_QUIRK_RESET (device can't be reset 5072 (e.g morph devices), don't use reset); 5073 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 5074 more interface descriptions than the 5075 bNumInterfaces count, and can't handle 5076 talking to these interfaces); 5077 g = USB_QUIRK_DELAY_INIT (device needs a pause 5078 during initialization, after we read 5079 the device descriptor); 5080 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 5081 high speed and super speed interrupt 5082 endpoints, the USB 2.0 and USB 3.0 spec 5083 require the interval in microframes (1 5084 microframe = 125 microseconds) to be 5085 calculated as interval = 2 ^ 5086 (bInterval-1). 5087 Devices with this quirk report their 5088 bInterval as the result of this 5089 calculation instead of the exponent 5090 variable used in the calculation); 5091 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 5092 handle device_qualifier descriptor 5093 requests); 5094 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 5095 generates spurious wakeup, ignore 5096 remote wakeup capability); 5097 k = USB_QUIRK_NO_LPM (device can't handle Link 5098 Power Management); 5099 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 5100 (Device reports its bInterval as linear 5101 frames instead of the USB 2.0 5102 calculation); 5103 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 5104 to be disconnected before suspend to 5105 prevent spurious wakeup); 5106 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 5107 pause after every control message); 5108 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 5109 delay after resetting its port); 5110 Example: quirks=0781:5580:bk,0a5c:5834:gij 5111 5112 usbhid.mousepoll= 5113 [USBHID] The interval which mice are to be polled at. 5114 5115 usbhid.jspoll= 5116 [USBHID] The interval which joysticks are to be polled at. 5117 5118 usbhid.kbpoll= 5119 [USBHID] The interval which keyboards are to be polled at. 5120 5121 usb-storage.delay_use= 5122 [UMS] The delay in seconds before a new device is 5123 scanned for Logical Units (default 1). 5124 5125 usb-storage.quirks= 5126 [UMS] A list of quirks entries to supplement or 5127 override the built-in unusual_devs list. List 5128 entries are separated by commas. Each entry has 5129 the form VID:PID:Flags where VID and PID are Vendor 5130 and Product ID values (4-digit hex numbers) and 5131 Flags is a set of characters, each corresponding 5132 to a common usb-storage quirk flag as follows: 5133 a = SANE_SENSE (collect more than 18 bytes 5134 of sense data, not on uas); 5135 b = BAD_SENSE (don't collect more than 18 5136 bytes of sense data, not on uas); 5137 c = FIX_CAPACITY (decrease the reported 5138 device capacity by one sector); 5139 d = NO_READ_DISC_INFO (don't use 5140 READ_DISC_INFO command, not on uas); 5141 e = NO_READ_CAPACITY_16 (don't use 5142 READ_CAPACITY_16 command); 5143 f = NO_REPORT_OPCODES (don't use report opcodes 5144 command, uas only); 5145 g = MAX_SECTORS_240 (don't transfer more than 5146 240 sectors at a time, uas only); 5147 h = CAPACITY_HEURISTICS (decrease the 5148 reported device capacity by one 5149 sector if the number is odd); 5150 i = IGNORE_DEVICE (don't bind to this 5151 device); 5152 j = NO_REPORT_LUNS (don't use report luns 5153 command, uas only); 5154 l = NOT_LOCKABLE (don't try to lock and 5155 unlock ejectable media, not on uas); 5156 m = MAX_SECTORS_64 (don't transfer more 5157 than 64 sectors = 32 KB at a time, 5158 not on uas); 5159 n = INITIAL_READ10 (force a retry of the 5160 initial READ(10) command, not on uas); 5161 o = CAPACITY_OK (accept the capacity 5162 reported by the device, not on uas); 5163 p = WRITE_CACHE (the device cache is ON 5164 by default, not on uas); 5165 r = IGNORE_RESIDUE (the device reports 5166 bogus residue values, not on uas); 5167 s = SINGLE_LUN (the device has only one 5168 Logical Unit); 5169 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 5170 commands, uas only); 5171 u = IGNORE_UAS (don't bind to the uas driver); 5172 w = NO_WP_DETECT (don't test whether the 5173 medium is write-protected). 5174 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 5175 even if the device claims no cache, 5176 not on uas) 5177 Example: quirks=0419:aaf5:rl,0421:0433:rc 5178 5179 user_debug= [KNL,ARM] 5180 Format: <int> 5181 See arch/arm/Kconfig.debug help text. 5182 1 - undefined instruction events 5183 2 - system calls 5184 4 - invalid data aborts 5185 8 - SIGSEGV faults 5186 16 - SIGBUS faults 5187 Example: user_debug=31 5188 5189 userpte= 5190 [X86] Flags controlling user PTE allocations. 5191 5192 nohigh = do not allocate PTE pages in 5193 HIGHMEM regardless of setting 5194 of CONFIG_HIGHPTE. 5195 5196 vdso= [X86,SH] 5197 On X86_32, this is an alias for vdso32=. Otherwise: 5198 5199 vdso=1: enable VDSO (the default) 5200 vdso=0: disable VDSO mapping 5201 5202 vdso32= [X86] Control the 32-bit vDSO 5203 vdso32=1: enable 32-bit VDSO 5204 vdso32=0 or vdso32=2: disable 32-bit VDSO 5205 5206 See the help text for CONFIG_COMPAT_VDSO for more 5207 details. If CONFIG_COMPAT_VDSO is set, the default is 5208 vdso32=0; otherwise, the default is vdso32=1. 5209 5210 For compatibility with older kernels, vdso32=2 is an 5211 alias for vdso32=0. 5212 5213 Try vdso32=0 if you encounter an error that says: 5214 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 5215 5216 vector= [IA-64,SMP] 5217 vector=percpu: enable percpu vector domain 5218 5219 video= [FB] Frame buffer configuration 5220 See Documentation/fb/modedb.rst. 5221 5222 video.brightness_switch_enabled= [0,1] 5223 If set to 1, on receiving an ACPI notify event 5224 generated by hotkey, video driver will adjust brightness 5225 level and then send out the event to user space through 5226 the allocated input device; If set to 0, video driver 5227 will only send out the event without touching backlight 5228 brightness level. 5229 default: 1 5230 5231 virtio_mmio.device= 5232 [VMMIO] Memory mapped virtio (platform) device. 5233 5234 <size>@<baseaddr>:<irq>[:<id>] 5235 where: 5236 <size> := size (can use standard suffixes 5237 like K, M and G) 5238 <baseaddr> := physical base address 5239 <irq> := interrupt number (as passed to 5240 request_irq()) 5241 <id> := (optional) platform device id 5242 example: 5243 virtio_mmio.device=1K@0x100b0000:48:7 5244 5245 Can be used multiple times for multiple devices. 5246 5247 vga= [BOOT,X86-32] Select a particular video mode 5248 See Documentation/x86/boot.rst and 5249 Documentation/admin-guide/svga.rst. 5250 Use vga=ask for menu. 5251 This is actually a boot loader parameter; the value is 5252 passed to the kernel using a special protocol. 5253 5254 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 5255 May slow down system boot speed, especially when 5256 enabled on systems with a large amount of memory. 5257 All options are enabled by default, and this 5258 interface is meant to allow for selectively 5259 enabling or disabling specific virtual memory 5260 debugging features. 5261 5262 Available options are: 5263 P Enable page structure init time poisoning 5264 - Disable all of the above options 5265 5266 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact 5267 size of <nn>. This can be used to increase the 5268 minimum size (128MB on x86). It can also be used to 5269 decrease the size and leave more room for directly 5270 mapped kernel RAM. 5271 5272 vmcp_cma=nn[MG] [KNL,S390] 5273 Sets the memory size reserved for contiguous memory 5274 allocations for the vmcp device driver. 5275 5276 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 5277 Format: <command> 5278 5279 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 5280 Format: <command> 5281 5282 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 5283 Format: <command> 5284 5285 vsyscall= [X86-64] 5286 Controls the behavior of vsyscalls (i.e. calls to 5287 fixed addresses of 0xffffffffff600x00 from legacy 5288 code). Most statically-linked binaries and older 5289 versions of glibc use these calls. Because these 5290 functions are at fixed addresses, they make nice 5291 targets for exploits that can control RIP. 5292 5293 emulate [default] Vsyscalls turn into traps and are 5294 emulated reasonably safely. The vsyscall 5295 page is readable. 5296 5297 xonly Vsyscalls turn into traps and are 5298 emulated reasonably safely. The vsyscall 5299 page is not readable. 5300 5301 none Vsyscalls don't work at all. This makes 5302 them quite hard to use for exploits but 5303 might break your system. 5304 5305 vt.color= [VT] Default text color. 5306 Format: 0xYX, X = foreground, Y = background. 5307 Default: 0x07 = light gray on black. 5308 5309 vt.cur_default= [VT] Default cursor shape. 5310 Format: 0xCCBBAA, where AA, BB, and CC are the same as 5311 the parameters of the <Esc>[?A;B;Cc escape sequence; 5312 see VGA-softcursor.txt. Default: 2 = underline. 5313 5314 vt.default_blu= [VT] 5315 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 5316 Change the default blue palette of the console. 5317 This is a 16-member array composed of values 5318 ranging from 0-255. 5319 5320 vt.default_grn= [VT] 5321 Format: <green0>,<green1>,<green2>,...,<green15> 5322 Change the default green palette of the console. 5323 This is a 16-member array composed of values 5324 ranging from 0-255. 5325 5326 vt.default_red= [VT] 5327 Format: <red0>,<red1>,<red2>,...,<red15> 5328 Change the default red palette of the console. 5329 This is a 16-member array composed of values 5330 ranging from 0-255. 5331 5332 vt.default_utf8= 5333 [VT] 5334 Format=<0|1> 5335 Set system-wide default UTF-8 mode for all tty's. 5336 Default is 1, i.e. UTF-8 mode is enabled for all 5337 newly opened terminals. 5338 5339 vt.global_cursor_default= 5340 [VT] 5341 Format=<-1|0|1> 5342 Set system-wide default for whether a cursor 5343 is shown on new VTs. Default is -1, 5344 i.e. cursors will be created by default unless 5345 overridden by individual drivers. 0 will hide 5346 cursors, 1 will display them. 5347 5348 vt.italic= [VT] Default color for italic text; 0-15. 5349 Default: 2 = green. 5350 5351 vt.underline= [VT] Default color for underlined text; 0-15. 5352 Default: 3 = cyan. 5353 5354 watchdog timers [HW,WDT] For information on watchdog timers, 5355 see Documentation/watchdog/watchdog-parameters.rst 5356 or other driver-specific files in the 5357 Documentation/watchdog/ directory. 5358 5359 watchdog_thresh= 5360 [KNL] 5361 Set the hard lockup detector stall duration 5362 threshold in seconds. The soft lockup detector 5363 threshold is set to twice the value. A value of 0 5364 disables both lockup detectors. Default is 10 5365 seconds. 5366 5367 workqueue.watchdog_thresh= 5368 If CONFIG_WQ_WATCHDOG is configured, workqueue can 5369 warn stall conditions and dump internal state to 5370 help debugging. 0 disables workqueue stall 5371 detection; otherwise, it's the stall threshold 5372 duration in seconds. The default value is 30 and 5373 it can be updated at runtime by writing to the 5374 corresponding sysfs file. 5375 5376 workqueue.disable_numa 5377 By default, all work items queued to unbound 5378 workqueues are affine to the NUMA nodes they're 5379 issued on, which results in better behavior in 5380 general. If NUMA affinity needs to be disabled for 5381 whatever reason, this option can be used. Note 5382 that this also can be controlled per-workqueue for 5383 workqueues visible under /sys/bus/workqueue/. 5384 5385 workqueue.power_efficient 5386 Per-cpu workqueues are generally preferred because 5387 they show better performance thanks to cache 5388 locality; unfortunately, per-cpu workqueues tend to 5389 be more power hungry than unbound workqueues. 5390 5391 Enabling this makes the per-cpu workqueues which 5392 were observed to contribute significantly to power 5393 consumption unbound, leading to measurably lower 5394 power usage at the cost of small performance 5395 overhead. 5396 5397 The default value of this parameter is determined by 5398 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 5399 5400 workqueue.debug_force_rr_cpu 5401 Workqueue used to implicitly guarantee that work 5402 items queued without explicit CPU specified are put 5403 on the local CPU. This guarantee is no longer true 5404 and while local CPU is still preferred work items 5405 may be put on foreign CPUs. This debug option 5406 forces round-robin CPU selection to flush out 5407 usages which depend on the now broken guarantee. 5408 When enabled, memory and cache locality will be 5409 impacted. 5410 5411 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of 5412 default x2apic cluster mode on platforms 5413 supporting x2apic. 5414 5415 x86_intel_mid_timer= [X86-32,APBT] 5416 Choose timer option for x86 Intel MID platform. 5417 Two valid options are apbt timer only and lapic timer 5418 plus one apbt timer for broadcast timer. 5419 x86_intel_mid_timer=apbt_only | lapic_and_apbt 5420 5421 xen_512gb_limit [KNL,X86-64,XEN] 5422 Restricts the kernel running paravirtualized under Xen 5423 to use only up to 512 GB of RAM. The reason to do so is 5424 crash analysis tools and Xen tools for doing domain 5425 save/restore/migration must be enabled to handle larger 5426 domains. 5427 5428 xen_emul_unplug= [HW,X86,XEN] 5429 Unplug Xen emulated devices 5430 Format: [unplug0,][unplug1] 5431 ide-disks -- unplug primary master IDE devices 5432 aux-ide-disks -- unplug non-primary-master IDE devices 5433 nics -- unplug network devices 5434 all -- unplug all emulated devices (NICs and IDE disks) 5435 unnecessary -- unplugging emulated devices is 5436 unnecessary even if the host did not respond to 5437 the unplug protocol 5438 never -- do not unplug even if version check succeeds 5439 5440 xen_legacy_crash [X86,XEN] 5441 Crash from Xen panic notifier, without executing late 5442 panic() code such as dumping handler. 5443 5444 xen_nopvspin [X86,XEN] 5445 Disables the ticketlock slowpath using Xen PV 5446 optimizations. 5447 5448 xen_nopv [X86] 5449 Disables the PV optimizations forcing the HVM guest to 5450 run as generic HVM guest with no PV drivers. 5451 This option is obsoleted by the "nopv" option, which 5452 has equivalent effect for XEN platform. 5453 5454 xen_scrub_pages= [XEN] 5455 Boolean option to control scrubbing pages before giving them back 5456 to Xen, for use by other domains. Can be also changed at runtime 5457 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 5458 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 5459 5460 xen_timer_slop= [X86-64,XEN] 5461 Set the timer slop (in nanoseconds) for the virtual Xen 5462 timers (default is 100000). This adjusts the minimum 5463 delta of virtualized Xen timers, where lower values 5464 improve timer resolution at the expense of processing 5465 more timer interrupts. 5466 5467 nopv= [X86,XEN,KVM,HYPER_V,VMWARE] 5468 Disables the PV optimizations forcing the guest to run 5469 as generic guest with no PV drivers. Currently support 5470 XEN HVM, KVM, HYPER_V and VMWARE guest. 5471 5472 xirc2ps_cs= [NET,PCMCIA] 5473 Format: 5474 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 5475 5476 xive= [PPC] 5477 By default on POWER9 and above, the kernel will 5478 natively use the XIVE interrupt controller. This option 5479 allows the fallback firmware mode to be used: 5480 5481 off Fallback to firmware control of XIVE interrupt 5482 controller on both pseries and powernv 5483 platforms. Only useful on POWER9 and above. 5484 5485 xhci-hcd.quirks [USB,KNL] 5486 A hex value specifying bitmask with supplemental xhci 5487 host controller quirks. Meaning of each bit can be 5488 consulted in header drivers/usb/host/xhci.h. 5489 5490 xmon [PPC] 5491 Format: { early | on | rw | ro | off } 5492 Controls if xmon debugger is enabled. Default is off. 5493 Passing only "xmon" is equivalent to "xmon=early". 5494 early Call xmon as early as possible on boot; xmon 5495 debugger is called from setup_arch(). 5496 on xmon debugger hooks will be installed so xmon 5497 is only called on a kernel crash. Default mode, 5498 i.e. either "ro" or "rw" mode, is controlled 5499 with CONFIG_XMON_DEFAULT_RO_MODE. 5500 rw xmon debugger hooks will be installed so xmon 5501 is called only on a kernel crash, mode is write, 5502 meaning SPR registers, memory and, other data 5503 can be written using xmon commands. 5504 ro same as "rw" option above but SPR registers, 5505 memory, and other data can't be written using 5506 xmon commands. 5507 off xmon is disabled. 5508