1 ACPI ACPI support is enabled. 2 AGP AGP (Accelerated Graphics Port) is enabled. 3 ALSA ALSA sound support is enabled. 4 APIC APIC support is enabled. 5 APM Advanced Power Management support is enabled. 6 APPARMOR AppArmor support is enabled. 7 ARM ARM architecture is enabled. 8 ARM64 ARM64 architecture is enabled. 9 CLK Common clock infrastructure is enabled. 10 CMA Contiguous Memory Area support is enabled. 11 CPU_FREQ CPU frequency scaling is enabled. 12 CPU_IDLE CPU idle time management is enabled. 13 DRM Direct Rendering Management support is enabled. 14 DYNAMIC_DEBUG Build in debug messages and enable them at runtime 15 EARLY Parameter processed too early to be embedded in initrd. 16 EDD BIOS Enhanced Disk Drive Services (EDD) is enabled 17 EFI EFI Partitioning (GPT) is enabled 18 EVM Extended Verification Module 19 FB The frame buffer device is enabled. 20 FTRACE Function tracing enabled. 21 GCOV GCOV profiling is enabled. 22 HIBERNATION HIBERNATION is enabled. 23 HW Appropriate hardware is enabled. 24 HYPER_V HYPERV support is enabled. 25 IMA Integrity measurement architecture is enabled. 26 IP_PNP IP DHCP, BOOTP, or RARP is enabled. 27 IPV6 IPv6 support is enabled. 28 ISAPNP ISA PnP code is enabled. 29 ISOL CPU Isolation is enabled. 30 JOY Appropriate joystick support is enabled. 31 KGDB Kernel debugger support is enabled. 32 KVM Kernel Virtual Machine support is enabled. 33 LIBATA Libata driver is enabled 34 LOONGARCH LoongArch architecture is enabled. 35 LOOP Loopback device support is enabled. 36 LP Printer support is enabled. 37 M68k M68k architecture is enabled. 38 These options have more detailed description inside of 39 Documentation/arch/m68k/kernel-options.rst. 40 MIPS MIPS architecture is enabled. 41 MOUSE Appropriate mouse support is enabled. 42 MSI Message Signaled Interrupts (PCI). 43 MTD MTD (Memory Technology Device) support is enabled. 44 NET Appropriate network support is enabled. 45 NFS Appropriate NFS support is enabled. 46 NUMA NUMA support is enabled. 47 OF Devicetree is enabled. 48 PARISC The PA-RISC architecture is enabled. 49 PCI PCI bus support is enabled. 50 PCIE PCI Express support is enabled. 51 PCMCIA The PCMCIA subsystem is enabled. 52 PM Power Management support is enabled. 53 PNP Plug & Play support is enabled. 54 PPC PowerPC architecture is enabled. 55 PPT Parallel port support is enabled. 56 PS2 Appropriate PS/2 support is enabled. 57 PV_OPS A paravirtualized kernel is enabled. 58 RAM RAM disk support is enabled. 59 RDT Intel Resource Director Technology. 60 RISCV RISCV architecture is enabled. 61 S390 S390 architecture is enabled. 62 SCSI Appropriate SCSI support is enabled. 63 A lot of drivers have their options described inside 64 the Documentation/scsi/ sub-directory. 65 SDW SoundWire support is enabled. 66 SECURITY Different security models are enabled. 67 SELINUX SELinux support is enabled. 68 SERIAL Serial support is enabled. 69 SH SuperH architecture is enabled. 70 SMP The kernel is an SMP kernel. 71 SPARC Sparc architecture is enabled. 72 SUSPEND System suspend states are enabled. 73 SWSUSP Software suspend (hibernation) is enabled. 74 TPM TPM drivers are enabled. 75 UMS USB Mass Storage support is enabled. 76 USB USB support is enabled. 77 NVME NVMe support is enabled 78 USBHID USB Human Interface Device support is enabled. 79 V4L Video For Linux support is enabled. 80 VGA The VGA console has been enabled. 81 VMMIO Driver for memory mapped virtio devices is enabled. 82 VT Virtual terminal support is enabled. 83 WDT Watchdog support is enabled. 84 X86-32 X86-32, aka i386 architecture is enabled. 85 X86-64 X86-64 architecture is enabled. 86 X86 Either 32-bit or 64-bit x86 (same as X86-32+X86-64) 87 X86_UV SGI UV support is enabled. 88 XEN Xen support is enabled 89 XTENSA xtensa architecture is enabled. 90 91In addition, the following text indicates that the option 92 93 BOOT Is a boot loader parameter. 94 BUGS= Relates to possible processor bugs on the said processor. 95 KNL Is a kernel start-up parameter. 96 97 98Kernel parameters 99 100 accept_memory= [MM] 101 Format: { eager | lazy } 102 default: lazy 103 By default, unaccepted memory is accepted lazily to 104 avoid prolonged boot times. The lazy option will add 105 some runtime overhead until all memory is eventually 106 accepted. In most cases the overhead is negligible. 107 For some workloads or for debugging purposes 108 accept_memory=eager can be used to accept all memory 109 at once during boot. 110 111 acpi= [HW,ACPI,X86,ARM64,RISCV64,EARLY] 112 Advanced Configuration and Power Interface 113 Format: { force | on | off | strict | noirq | rsdt | 114 copy_dsdt | nospcr } 115 force -- enable ACPI if default was off 116 on -- enable ACPI but allow fallback to DT [arm64,riscv64] 117 off -- disable ACPI if default was on 118 noirq -- do not use ACPI for IRQ routing 119 strict -- Be less tolerant of platforms that are not 120 strictly ACPI specification compliant. 121 rsdt -- prefer RSDT over (default) XSDT 122 copy_dsdt -- copy DSDT to memory 123 nocmcff -- Disable firmware first mode for corrected 124 errors. This disables parsing the HEST CMC error 125 source to check if firmware has set the FF flag. This 126 may result in duplicate corrected error reports. 127 nospcr -- disable console in ACPI SPCR table as 128 default _serial_ console on ARM64 129 spcr -- enable console in ACPI SPCR table as 130 default _serial_ console on x86 131 For ARM64, ONLY "acpi=off", "acpi=on", "acpi=force" or 132 "acpi=nospcr" are available 133 For RISCV64, ONLY "acpi=off", "acpi=on" or "acpi=force" 134 are available 135 136 See also Documentation/power/runtime_pm.rst, pci=noacpi 137 138 acpi_apic_instance= [ACPI,IOAPIC,EARLY] 139 Format: <int> 140 2: use 2nd APIC table, if available 141 1,0: use 1st APIC table 142 default: 0 143 144 acpi_backlight= [HW,ACPI] 145 { vendor | video | native | none } 146 If set to vendor, prefer vendor-specific driver 147 (e.g. thinkpad_acpi, sony_acpi, etc.) instead 148 of the ACPI video.ko driver. 149 If set to video, use the ACPI video.ko driver. 150 If set to native, use the device's native backlight mode. 151 If set to none, disable the ACPI backlight interface. 152 153 acpi_force_32bit_fadt_addr [ACPI,EARLY] 154 force FADT to use 32 bit addresses rather than the 155 64 bit X_* addresses. Some firmware have broken 64 156 bit addresses for force ACPI ignore these and use 157 the older legacy 32 bit addresses. 158 159 acpica_no_return_repair [HW, ACPI] 160 Disable AML predefined validation mechanism 161 This mechanism can repair the evaluation result to make 162 the return objects more ACPI specification compliant. 163 This option is useful for developers to identify the 164 root cause of an AML interpreter issue when the issue 165 has something to do with the repair mechanism. 166 167 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG] 168 acpi.debug_level= [HW,ACPI,ACPI_DEBUG] 169 Format: <int> 170 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI 171 debug output. Bits in debug_layer correspond to a 172 _COMPONENT in an ACPI source file, e.g., 173 #define _COMPONENT ACPI_EVENTS 174 Bits in debug_level correspond to a level in 175 ACPI_DEBUG_PRINT statements, e.g., 176 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ... 177 The debug_level mask defaults to "info". See 178 Documentation/firmware-guide/acpi/debug.rst for more information about 179 debug layers and levels. 180 181 Enable processor driver info messages: 182 acpi.debug_layer=0x20000000 183 Enable AML "Debug" output, i.e., stores to the Debug 184 object while interpreting AML: 185 acpi.debug_layer=0xffffffff acpi.debug_level=0x2 186 Enable all messages related to ACPI hardware: 187 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff 188 189 Some values produce so much output that the system is 190 unusable. The "log_buf_len" parameter may be useful 191 if you need to capture more output. 192 193 acpi.poweroff_on_fatal= [ACPI] 194 {0 | 1} 195 Causes the system to poweroff when the ACPI bytecode signals 196 a fatal error. The default value of this setting is 1. 197 Overriding this value should only be done for diagnosing 198 ACPI firmware problems, as the system might behave erratically 199 after having encountered a fatal ACPI error. 200 201 acpi_enforce_resources= [ACPI] 202 { strict | lax | no } 203 Check for resource conflicts between native drivers 204 and ACPI OperationRegions (SystemIO and SystemMemory 205 only). IO ports and memory declared in ACPI might be 206 used by the ACPI subsystem in arbitrary AML code and 207 can interfere with legacy drivers. 208 strict (default): access to resources claimed by ACPI 209 is denied; legacy drivers trying to access reserved 210 resources will fail to bind to device using them. 211 lax: access to resources claimed by ACPI is allowed; 212 legacy drivers trying to access reserved resources 213 will bind successfully but a warning message is logged. 214 no: ACPI OperationRegions are not marked as reserved, 215 no further checks are performed. 216 217 acpi_force_table_verification [HW,ACPI,EARLY] 218 Enable table checksum verification during early stage. 219 By default, this is disabled due to x86 early mapping 220 size limitation. 221 222 acpi_irq_balance [HW,ACPI] 223 ACPI will balance active IRQs 224 default in APIC mode 225 226 acpi_irq_nobalance [HW,ACPI] 227 ACPI will not move active IRQs (default) 228 default in PIC mode 229 230 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 231 Format: <irq>,<irq>... 232 233 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 234 use by PCI 235 Format: <irq>,<irq>... 236 237 acpi_mask_gpe= [HW,ACPI] 238 Due to the existence of _Lxx/_Exx, some GPEs triggered 239 by unsupported hardware/firmware features can result in 240 GPE floodings that cannot be automatically disabled by 241 the GPE dispatcher. 242 This facility can be used to prevent such uncontrolled 243 GPE floodings. 244 Format: <byte> or <bitmap-list> 245 246 acpi_no_auto_serialize [HW,ACPI] 247 Disable auto-serialization of AML methods 248 AML control methods that contain the opcodes to create 249 named objects will be marked as "Serialized" by the 250 auto-serialization feature. 251 This feature is enabled by default. 252 This option allows to turn off the feature. 253 254 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 255 kernels. 256 257 acpi_no_static_ssdt [HW,ACPI,EARLY] 258 Disable installation of static SSDTs at early boot time 259 By default, SSDTs contained in the RSDT/XSDT will be 260 installed automatically and they will appear under 261 /sys/firmware/acpi/tables. 262 This option turns off this feature. 263 Note that specifying this option does not affect 264 dynamic table installation which will install SSDT 265 tables to /sys/firmware/acpi/tables/dynamic. 266 267 acpi_no_watchdog [HW,ACPI,WDT] 268 Ignore the ACPI-based watchdog interface (WDAT) and let 269 a native driver control the watchdog device instead. 270 271 acpi_rsdp= [ACPI,EFI,KEXEC,EARLY] 272 Pass the RSDP address to the kernel, mostly used 273 on machines running EFI runtime service to boot the 274 second kernel for kdump. 275 276 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 277 Format: To spoof as Windows 98: ="Microsoft Windows" 278 279 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 280 of 2 which is mandated by ACPI 6) as the supported ACPI 281 specification revision (when using this switch, it may 282 be necessary to carry out a cold reboot _twice_ in a 283 row to make it take effect on the platform firmware). 284 285 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 286 acpi_osi="string1" # add string1 287 acpi_osi="!string2" # remove string2 288 acpi_osi=!* # remove all strings 289 acpi_osi=! # disable all built-in OS vendor 290 strings 291 acpi_osi=!! # enable all built-in OS vendor 292 strings 293 acpi_osi= # disable all strings 294 295 'acpi_osi=!' can be used in combination with single or 296 multiple 'acpi_osi="string1"' to support specific OS 297 vendor string(s). Note that such command can only 298 affect the default state of the OS vendor strings, thus 299 it cannot affect the default state of the feature group 300 strings and the current state of the OS vendor strings, 301 specifying it multiple times through kernel command line 302 is meaningless. This command is useful when one do not 303 care about the state of the feature group strings which 304 should be controlled by the OSPM. 305 Examples: 306 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 307 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 308 can make '_OSI("Windows 2000")' TRUE. 309 310 'acpi_osi=' cannot be used in combination with other 311 'acpi_osi=' command lines, the _OSI method will not 312 exist in the ACPI namespace. NOTE that such command can 313 only affect the _OSI support state, thus specifying it 314 multiple times through kernel command line is also 315 meaningless. 316 Examples: 317 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 318 FALSE. 319 320 'acpi_osi=!*' can be used in combination with single or 321 multiple 'acpi_osi="string1"' to support specific 322 string(s). Note that such command can affect the 323 current state of both the OS vendor strings and the 324 feature group strings, thus specifying it multiple times 325 through kernel command line is meaningful. But it may 326 still not able to affect the final state of a string if 327 there are quirks related to this string. This command 328 is useful when one want to control the state of the 329 feature group strings to debug BIOS issues related to 330 the OSPM features. 331 Examples: 332 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 333 '_OSI("Module Device")' FALSE. 334 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 335 '_OSI("Module Device")' TRUE. 336 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 337 equivalent to 338 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 339 and 340 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 341 they all will make '_OSI("Windows 2000")' TRUE. 342 343 acpi_pm_good [X86] 344 Override the pmtimer bug detection: force the kernel 345 to assume that this machine's pmtimer latches its value 346 and always returns good values. 347 348 acpi_sci= [HW,ACPI,EARLY] ACPI System Control Interrupt trigger mode 349 Format: { level | edge | high | low } 350 351 acpi_skip_timer_override [HW,ACPI,EARLY] 352 Recognize and ignore IRQ0/pin2 Interrupt Override. 353 For broken nForce2 BIOS resulting in XT-PIC timer. 354 355 acpi_sleep= [HW,ACPI] Sleep options 356 Format: { s3_bios, s3_mode, s3_beep, s4_hwsig, 357 s4_nohwsig, old_ordering, nonvs, 358 sci_force_enable, nobl } 359 See Documentation/power/video.rst for information on 360 s3_bios and s3_mode. 361 s3_beep is for debugging; it makes the PC's speaker beep 362 as soon as the kernel's real-mode entry point is called. 363 s4_hwsig causes the kernel to check the ACPI hardware 364 signature during resume from hibernation, and gracefully 365 refuse to resume if it has changed. This complies with 366 the ACPI specification but not with reality, since 367 Windows does not do this and many laptops do change it 368 on docking. So the default behaviour is to allow resume 369 and simply warn when the signature changes, unless the 370 s4_hwsig option is enabled. 371 s4_nohwsig prevents ACPI hardware signature from being 372 used (or even warned about) during resume. 373 old_ordering causes the ACPI 1.0 ordering of the _PTS 374 control method, with respect to putting devices into 375 low power states, to be enforced (the ACPI 2.0 ordering 376 of _PTS is used by default). 377 nonvs prevents the kernel from saving/restoring the 378 ACPI NVS memory during suspend/hibernation and resume. 379 sci_force_enable causes the kernel to set SCI_EN directly 380 on resume from S1/S3 (which is against the ACPI spec, 381 but some broken systems don't work without it). 382 nobl causes the internal blacklist of systems known to 383 behave incorrectly in some ways with respect to system 384 suspend and resume to be ignored (use wisely). 385 386 acpi_use_timer_override [HW,ACPI,EARLY] 387 Use timer override. For some broken Nvidia NF5 boards 388 that require a timer override, but don't have HPET 389 390 add_efi_memmap [EFI,X86,EARLY] Include EFI memory map in 391 kernel's map of available physical RAM. 392 393 agp= [AGP] 394 { off | try_unsupported } 395 off: disable AGP support 396 try_unsupported: try to drive unsupported chipsets 397 (may crash computer or cause data corruption) 398 399 ALSA [HW,ALSA] 400 See Documentation/sound/alsa-configuration.rst 401 402 alignment= [KNL,ARM] 403 Allow the default userspace alignment fault handler 404 behaviour to be specified. Bit 0 enables warnings, 405 bit 1 enables fixups, and bit 2 sends a segfault. 406 407 align_va_addr= [X86-64] 408 Align virtual addresses by clearing slice [14:12] when 409 allocating a VMA at process creation time. This option 410 gives you up to 3% performance improvement on AMD F15h 411 machines (where it is enabled by default) for a 412 CPU-intensive style benchmark, and it can vary highly in 413 a microbenchmark depending on workload and compiler. 414 415 32: only for 32-bit processes 416 64: only for 64-bit processes 417 on: enable for both 32- and 64-bit processes 418 off: disable for both 32- and 64-bit processes 419 420 alloc_snapshot [FTRACE] 421 Allocate the ftrace snapshot buffer on boot up when the 422 main buffer is allocated. This is handy if debugging 423 and you need to use tracing_snapshot() on boot up, and 424 do not want to use tracing_snapshot_alloc() as it needs 425 to be done where GFP_KERNEL allocations are allowed. 426 427 allow_mismatched_32bit_el0 [ARM64,EARLY] 428 Allow execve() of 32-bit applications and setting of the 429 PER_LINUX32 personality on systems where only a strict 430 subset of the CPUs support 32-bit EL0. When this 431 parameter is present, the set of CPUs supporting 32-bit 432 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0 433 and hot-unplug operations may be restricted. 434 435 See Documentation/arch/arm64/asymmetric-32bit.rst for more 436 information. 437 438 amd_iommu= [HW,X86-64] 439 Pass parameters to the AMD IOMMU driver in the system. 440 Possible values are: 441 fullflush - Deprecated, equivalent to iommu.strict=1 442 off - do not initialize any AMD IOMMU found in 443 the system 444 force_isolation - Force device isolation for all 445 devices. The IOMMU driver is not 446 allowed anymore to lift isolation 447 requirements as needed. This option 448 does not override iommu=pt 449 force_enable - Force enable the IOMMU on platforms known 450 to be buggy with IOMMU enabled. Use this 451 option with care. 452 pgtbl_v1 - Use v1 page table for DMA-API (Default). 453 pgtbl_v2 - Use v2 page table for DMA-API. 454 irtcachedis - Disable Interrupt Remapping Table (IRT) caching. 455 nohugepages - Limit page-sizes used for v1 page-tables 456 to 4 KiB. 457 v2_pgsizes_only - Limit page-sizes used for v1 page-tables 458 to 4KiB/2Mib/1GiB. 459 460 461 amd_iommu_dump= [HW,X86-64] 462 Enable AMD IOMMU driver option to dump the ACPI table 463 for AMD IOMMU. With this option enabled, AMD IOMMU 464 driver will print ACPI tables for AMD IOMMU during 465 IOMMU initialization. 466 467 amd_iommu_intr= [HW,X86-64] 468 Specifies one of the following AMD IOMMU interrupt 469 remapping modes: 470 legacy - Use legacy interrupt remapping mode. 471 vapic - Use virtual APIC mode, which allows IOMMU 472 to inject interrupts directly into guest. 473 This mode requires kvm-amd.avic=1. 474 (Default when IOMMU HW support is present.) 475 476 amd_pstate= [X86,EARLY] 477 disable 478 Do not enable amd_pstate as the default 479 scaling driver for the supported processors 480 passive 481 Use amd_pstate with passive mode as a scaling driver. 482 In this mode autonomous selection is disabled. 483 Driver requests a desired performance level and platform 484 tries to match the same performance level if it is 485 satisfied by guaranteed performance level. 486 active 487 Use amd_pstate_epp driver instance as the scaling driver, 488 driver provides a hint to the hardware if software wants 489 to bias toward performance (0x0) or energy efficiency (0xff) 490 to the CPPC firmware. then CPPC power algorithm will 491 calculate the runtime workload and adjust the realtime cores 492 frequency. 493 guided 494 Activate guided autonomous mode. Driver requests minimum and 495 maximum performance level and the platform autonomously 496 selects a performance level in this range and appropriate 497 to the current workload. 498 499 amd_prefcore= 500 [X86] 501 disable 502 Disable amd-pstate preferred core. 503 504 amijoy.map= [HW,JOY] Amiga joystick support 505 Map of devices attached to JOY0DAT and JOY1DAT 506 Format: <a>,<b> 507 See also Documentation/input/joydev/joystick.rst 508 509 analog.map= [HW,JOY] Analog joystick and gamepad support 510 Specifies type or capabilities of an analog joystick 511 connected to one of 16 gameports 512 Format: <type1>,<type2>,..<type16> 513 514 apc= [HW,SPARC] 515 Power management functions (SPARCstation-4/5 + deriv.) 516 Format: noidle 517 Disable APC CPU standby support. SPARCstation-Fox does 518 not play well with APC CPU idle - disable it if you have 519 APC and your system crashes randomly. 520 521 apic [APIC,X86-64] Use IO-APIC. Default. 522 523 apic= [APIC,X86,EARLY] Advanced Programmable Interrupt Controller 524 Change the output verbosity while booting 525 Format: { quiet (default) | verbose | debug } 526 Change the amount of debugging information output 527 when initialising the APIC and IO-APIC components. 528 529 apic_extnmi= [APIC,X86,EARLY] External NMI delivery setting 530 Format: { bsp (default) | all | none } 531 bsp: External NMI is delivered only to CPU 0 532 all: External NMIs are broadcast to all CPUs as a 533 backup of CPU 0 534 none: External NMI is masked for all CPUs. This is 535 useful so that a dump capture kernel won't be 536 shot down by NMI 537 538 apicpmtimer Do APIC timer calibration using the pmtimer. Implies 539 apicmaintimer. Useful when your PIT timer is totally 540 broken. 541 542 autoconf= [IPV6] 543 See Documentation/networking/ipv6.rst. 544 545 apm= [APM] Advanced Power Management 546 See header of arch/x86/kernel/apm_32.c. 547 548 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 549 Format: { "0" | "1" } 550 See security/apparmor/Kconfig help text 551 0 -- disable. 552 1 -- enable. 553 Default value is set via kernel config option. 554 555 arm64.no32bit_el0 [ARM64] Unconditionally disable the execution of 556 32 bit applications. 557 558 arm64.nobti [ARM64] Unconditionally disable Branch Target 559 Identification support 560 561 arm64.nogcs [ARM64] Unconditionally disable Guarded Control Stack 562 support 563 564 arm64.nohaft [ARM64] Unconditionally disable Hardware managed Access 565 Flag for Table descriptors support 566 567 arm64.nomops [ARM64] Unconditionally disable Memory Copy and Memory 568 Set instructions support 569 570 arm64.nompam [ARM64] Unconditionally disable Memory Partitioning And 571 Monitoring support. Only for a machine that does not 572 boot without it. See Documentation/arch/arm64/mpam.rst 573 574 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension 575 support 576 577 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 578 support 579 580 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix 581 Extension support 582 583 arm64.nosve [ARM64] Unconditionally disable Scalable Vector 584 Extension support 585 586 arm64.nowfxt [ARM64] Unconditionally disable Wait For Event with 587 Timeout and Wait For Interrupt with Timeout support 588 589 ataflop= [HW,M68k] 590 591 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 592 EzKey and similar keyboards 593 594 atkbd.reset= [HW] Reset keyboard during initialization 595 596 atkbd.set= [HW] Select keyboard code set 597 Format: <int> (2 = AT (default), 3 = PS/2) 598 599 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 600 keyboards 601 602 atkbd.softraw= [HW] Choose between synthetic and real raw mode 603 Format: <bool> (0 = real, 1 = synthetic (default)) 604 605 atkbd.softrepeat= [HW] 606 Use software keyboard repeat 607 608 audit= [KNL] Enable the audit sub-system 609 Format: { "0" | "1" | "off" | "on" } 610 0 | off - kernel audit is disabled and can not be 611 enabled until the next reboot 612 unset - kernel audit is initialized but disabled and 613 will be fully enabled by the userspace auditd. 614 1 | on - kernel audit is initialized and partially 615 enabled, storing at most audit_backlog_limit 616 messages in RAM until it is fully enabled by the 617 userspace auditd. 618 Default: unset 619 620 audit_backlog_limit= [KNL] Set the audit queue size limit. 621 Format: <int> (must be >=0) 622 Default: 64 623 624 bdev_allow_write_mounted= 625 Format: <bool> 626 Control the ability to open a mounted block device 627 for writing, i.e., allow / disallow writes that bypass 628 the FS. This was implemented as a means to prevent 629 fuzzers from crashing the kernel by overwriting the 630 metadata underneath a mounted FS without its awareness. 631 This also prevents destructive formatting of mounted 632 filesystems by naive storage tooling that don't use 633 O_EXCL. Default is Y and can be changed through the 634 Kconfig option CONFIG_BLK_DEV_WRITE_MOUNTED. 635 636 bert_disable [ACPI] 637 Disable BERT OS support on buggy BIOSes. 638 639 bgrt_disable [ACPI,X86,EARLY] 640 Disable BGRT to avoid flickering OEM logo. 641 642 blkdevparts= Manual partition parsing of block device(s) for 643 embedded devices based on command line input. 644 See Documentation/block/cmdline-partition.rst 645 646 boot_delay= [KNL,EARLY] 647 Milliseconds to delay each printk during boot. 648 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled, 649 and you may also have to specify "lpj=". Boot_delay 650 values larger than 10 seconds (10000) are assumed 651 erroneous and ignored. 652 Format: integer 653 654 bootconfig [KNL,EARLY] 655 Extended command line options can be added to an initrd 656 and this will cause the kernel to look for it. 657 658 See Documentation/admin-guide/bootconfig.rst 659 660 bpf.keyring_unsealed= 661 [BPF] When set to 1, leave the bpf keyring unsealed 662 after initialization so that userspace can provision 663 keys. Once the keyring is restricted it becomes active 664 and can be used for BPF program signature verification. 665 666 Setting this also means that the bpf keyring is the 667 only non-system keyring a loader may select for the 668 rest of the boot: caller-supplied user/session 669 keyrings are refused with -EPERM, whether or not 670 provisioning actually completed. The system keyrings 671 stay selectable. Leaving it unset keeps the prior 672 behaviour, where a caller-supplied keyring is allowed. 673 674 See Documentation/bpf/signing.rst 675 676 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 677 bttv.radio= Most important insmod options are available as 678 kernel args too. 679 bttv.pll= See Documentation/admin-guide/media/bttv.rst 680 bttv.tuner= 681 682 bulk_remove=off [PPC] This parameter disables the use of the pSeries 683 firmware feature for flushing multiple hpte entries 684 at a time. 685 686 c101= [NET] Moxa C101 synchronous serial card 687 688 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 689 Sometimes CPU hardware bugs make them report the cache 690 size incorrectly. The kernel will attempt work arounds 691 to fix known problems, but for some CPUs it is not 692 possible to determine what the correct size should be. 693 This option provides an override for these situations. 694 695 carrier_timeout= 696 [NET] Specifies amount of time (in seconds) that 697 the kernel should wait for a network carrier. By default 698 it waits 120 seconds. 699 700 ca_keys= [KEYS] This parameter identifies a specific key(s) on 701 the system trusted keyring to be used for certificate 702 trust validation. 703 format: { id:<keyid> | builtin } 704 705 cca= [MIPS,EARLY] Override the kernel pages' cache coherency 706 algorithm. Accepted values range from 0 to 7 707 inclusive. See arch/mips/include/asm/pgtable-bits.h 708 for platform specific values (SB1, Loongson3 and 709 others). 710 711 ccw_timeout_log [S390] 712 See Documentation/arch/s390/common_io.rst for details. 713 714 cfi= [X86-64] Set Control Flow Integrity checking features 715 when CONFIG_FINEIBT is enabled. 716 Format: feature[,feature...] 717 Default: auto 718 719 auto: Use FineIBT if IBT available, otherwise kCFI. 720 Under FineIBT, enable "paranoid" mode when 721 FRED is not available. 722 off: Turn off CFI checking. 723 kcfi: Use kCFI (disable FineIBT). 724 fineibt: Use FineIBT (even if IBT not available). 725 norand: Do not re-randomize CFI hashes. 726 paranoid: Add caller hash checking under FineIBT. 727 bhi: Enable register poisoning to stop speculation 728 across FineIBT. (Disabled by default.) 729 warn: Do not enforce CFI checking: warn only. 730 debug: Report CFI initialization details. 731 732 cgroup_disable= [KNL] Disable a particular controller or optional feature 733 Format: {name of the controller(s) or feature(s) to disable} 734 The effects of cgroup_disable=foo are: 735 - foo isn't auto-mounted if you mount all cgroups in 736 a single hierarchy 737 - foo isn't visible as an individually mountable 738 subsystem 739 - if foo is an optional feature then the feature is 740 disabled and corresponding cgroup files are not 741 created 742 {Currently only "memory" controller deal with this and 743 cut the overhead, others just disable the usage. So 744 only cgroup_disable=memory is actually worthy} 745 Specifying "pressure" disables per-cgroup pressure 746 stall information accounting feature 747 748 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 749 Format: { { controller | "all" | "named" } 750 [,{ controller | "all" | "named" }...] } 751 Like cgroup_disable, but only applies to cgroup v1; 752 the blacklisted controllers remain available in cgroup2. 753 "all" blacklists all controllers and "named" disables 754 named mounts. Specifying both "all" and "named" disables 755 all v1 hierarchies. 756 757 cgroup_v1_proc= [KNL] Show also missing controllers in /proc/cgroups 758 Format: { "true" | "false" } 759 /proc/cgroups lists only v1 controllers by default. 760 This compatibility option enables listing also v2 761 controllers (whose v1 code is not compiled!), so that 762 semi-legacy software can check this file to decide 763 about usage of v2 (sic) controllers. 764 765 cgroup_favordynmods= [KNL] Enable or Disable favordynmods. 766 Format: { "true" | "false" } 767 Defaults to the value of CONFIG_CGROUP_FAVOR_DYNMODS. 768 769 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 770 Format: <string> 771 nosocket -- Disable socket memory accounting. 772 nokmem -- Disable kernel memory accounting. 773 nobpf -- Disable BPF memory accounting. 774 775 check_pages= [MM,EARLY] Enable sanity checking of pages after 776 allocations / before freeing. This adds checks to catch 777 double-frees, use-after-frees, and other sources of 778 page corruption by inspecting page internals (flags, 779 mapcount/refcount, memcg_data, etc.). 780 Format: { "0" | "1" } 781 Default: 0 (1 if CONFIG_DEBUG_VM is set) 782 783 checkreqprot= [SELINUX] Set initial checkreqprot flag value. 784 Format: { "0" | "1" } 785 See security/selinux/Kconfig help text. 786 0 -- check protection applied by kernel (includes 787 any implied execute protection). 788 1 -- check protection requested by application. 789 Default value is set via a kernel config option. 790 Value can be changed at runtime via 791 /sys/fs/selinux/checkreqprot. 792 Setting checkreqprot to 1 is deprecated. 793 794 cio_ignore= [S390] 795 See Documentation/arch/s390/common_io.rst for details. 796 797 clk_ignore_unused 798 [CLK] 799 Prevents the clock framework from automatically gating 800 clocks that have not been explicitly enabled by a Linux 801 device driver but are enabled in hardware at reset or 802 by the bootloader/firmware. Note that this does not 803 force such clocks to be always-on nor does it reserve 804 those clocks in any way. This parameter is useful for 805 debug and development, but should not be needed on a 806 platform with proper driver support. For more 807 information, see Documentation/driver-api/clk.rst. 808 809 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 810 [Deprecated] 811 Forces specified clocksource (if available) to be used 812 when calculating gettimeofday(). If specified 813 clocksource is not available, it defaults to PIT. 814 Format: { pit | tsc | cyclone | pmtmr } 815 816 clocksource= Override the default clocksource 817 Format: <string> 818 Override the default clocksource and use the clocksource 819 with the name specified. 820 Some clocksource names to choose from, depending on 821 the platform: 822 [all] jiffies (this is the base, fallback clocksource) 823 [ACPI] acpi_pm 824 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 825 pxa_timer,timer3,32k_counter,timer0_1 826 [X86-32] pit,hpet,tsc; 827 scx200_hrt on Geode; cyclone on IBM x440 828 [MIPS] MIPS 829 [PARISC] cr16 830 [S390] tod 831 [SH] SuperH 832 [SPARC64] tick 833 [X86-64] hpet,tsc 834 835 clocksource.arm_arch_timer.evtstrm= 836 [ARM,ARM64,EARLY] 837 Format: <bool> 838 Enable/disable the eventstream feature of the ARM 839 architected timer so that code using WFE-based polling 840 loops can be debugged more effectively on production 841 systems. 842 843 clocksource.verify_n_cpus= [KNL] 844 Limit the number of CPUs checked for clocksources 845 marked with CLOCK_SOURCE_VERIFY_PERCPU that 846 are marked unstable due to excessive skew. 847 A negative value says to check all CPUs, while 848 zero says not to check any. Values larger than 849 nr_cpu_ids are silently truncated to nr_cpu_ids. 850 The actual CPUs are chosen randomly, with 851 no replacement if the same CPU is chosen twice. 852 853 clocksource-wdtest.holdoff= [KNL] 854 Set the time in seconds that the clocksource 855 watchdog test waits before commencing its tests. 856 Defaults to zero when built as a module and to 857 10 seconds when built into the kernel. 858 859 cma=nn[MG]@[start[MG][-end[MG]]] 860 [KNL,CMA,EARLY] 861 Sets the size of kernel global memory area for 862 contiguous memory allocations and optionally the 863 placement constraint by the physical address range of 864 memory allocations. A value of 0 disables CMA 865 altogether. For more information, see 866 kernel/dma/contiguous.c 867 868 cma_pernuma=nn[MG] 869 [KNL,CMA,EARLY] 870 Sets the size of kernel per-numa memory area for 871 contiguous memory allocations. A value of 0 disables 872 per-numa CMA altogether. And If this option is not 873 specified, the default value is 0. 874 With per-numa CMA enabled, DMA users on node nid will 875 first try to allocate buffer from the pernuma area 876 which is located in node nid, if the allocation fails, 877 they will fallback to the global default memory area. 878 879 numa_cma=<node>:nn[MG][,<node>:nn[MG]] 880 [KNL,CMA,EARLY] 881 Sets the size of kernel numa memory area for 882 contiguous memory allocations. It will reserve CMA 883 area for the specified node. 884 885 If it is setup together with upper 'cmd_pernuma=' 886 (unlikely), its size setting takes priority for the 887 specified numa nodes. 888 889 With numa CMA enabled, DMA users on node nid will 890 first try to allocate buffer from the numa area 891 which is located in node nid, if the allocation fails, 892 they will fallback to the global default memory area. 893 894 cmo_free_hint= [PPC] Format: { yes | no } 895 Specify whether pages are marked as being inactive 896 when they are freed. This is used in CMO environments 897 to determine OS memory pressure for page stealing by 898 a hypervisor. 899 Default: yes 900 901 coherent_pool=nn[KMG] [ARM,KNL,EARLY] 902 Sets the size of memory pool for coherent, atomic dma 903 allocations, by default set to 256K. 904 905 condev= [HW,S390] console device 906 conmode= 907 908 con3215_drop= [S390,EARLY] 3215 console drop mode. 909 Format: y|n|Y|N|1|0 910 When set to true, drop data on the 3215 console when 911 the console buffer is full. In this case the 912 operator using a 3270 terminal emulator (for example 913 x3270) does not have to enter the clear key for the 914 console output to advance and the kernel to continue. 915 This leads to a much faster boot time when a 3270 916 terminal emulator is active. If no 3270 terminal 917 emulator is used, this parameter has no effect. 918 919 console= [KNL] Output console device and options. 920 921 tty<n> Use the virtual console device <n>. 922 923 ttyS<n>[,options] 924 ttyUSB0[,options] 925 Use the specified serial port. The options are of 926 the form "bbbbpnf", where "bbbb" is the baud rate, 927 "p" is parity ("n", "o", or "e"), "n" is number of 928 bits, and "f" is flow control ("r" for RTS or 929 omit it). Default is "9600n8". 930 931 See Documentation/admin-guide/serial-console.rst for more 932 information. See 933 Documentation/networking/netconsole.rst for an 934 alternative. 935 936 <DEVNAME>:<n>.<n>[,options] 937 Use the specified serial port on the serial core bus. 938 The addressing uses DEVNAME of the physical serial port 939 device, followed by the serial core controller instance, 940 and the serial port instance. The options are the same 941 as documented for the ttyS addressing above. 942 943 The mapping of the serial ports to the tty instances 944 can be viewed with: 945 946 $ ls -d /sys/bus/serial-base/devices/*:*.*/tty/* 947 /sys/bus/serial-base/devices/00:04:0.0/tty/ttyS0 948 949 In the above example, the console can be addressed with 950 console=00:04:0.0. Note that a console addressed this 951 way will only get added when the related device driver 952 is ready. The use of an earlycon parameter in addition to 953 the console may be desired for console output early on. 954 955 uart[8250],io,<addr>[,options] 956 uart[8250],mmio,<addr>[,options] 957 uart[8250],mmio16,<addr>[,options] 958 uart[8250],mmio32,<addr>[,options] 959 uart[8250],0x<addr>[,options] 960 Start an early, polled-mode console on the 8250/16550 961 UART at the specified I/O port or MMIO address, 962 switching to the matching ttyS device later. 963 MMIO inter-register address stride is either 8-bit 964 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 965 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 966 to be equivalent to 'mmio'. 'options' are specified in 967 the same format described for ttyS above; if unspecified, 968 the h/w is not re-initialized. 969 970 hvc<n> Use the hypervisor console device <n>. This is for 971 both Xen and PowerPC hypervisors. 972 973 { null | "" } 974 Use to disable console output, i.e., to have kernel 975 console messages discarded. 976 This must be the only console= parameter used on the 977 kernel command line. 978 979 If the device connected to the port is not a TTY but a braille 980 device, prepend "brl," before the device type, for instance 981 console=brl,ttyS0 982 For now, only VisioBraille is supported. 983 984 console_msg_format= 985 [KNL] Change console messages format 986 default 987 By default we print messages on consoles in 988 "[time stamp] text\n" format (time stamp may not be 989 printed, depending on CONFIG_PRINTK_TIME or 990 `printk_time' param). 991 syslog 992 Switch to syslog format: "<%u>[time stamp] text\n" 993 IOW, each message will have a facility and loglevel 994 prefix. The format is similar to one used by syslog() 995 syscall, or to executing "dmesg -S --raw" or to reading 996 from /proc/kmsg. 997 998 consoleblank= [KNL] The console blank (screen saver) timeout in 999 seconds. A value of 0 disables the blank timer. 1000 Defaults to 0. 1001 1002 coredump_filter= 1003 [KNL] Change the default value for 1004 /proc/<pid>/coredump_filter. 1005 See also Documentation/filesystems/proc.rst. 1006 1007 coresight_cpu_debug.enable 1008 [ARM,ARM64] 1009 Format: <bool> 1010 Enable/disable the CPU sampling based debugging. 1011 0: default value, disable debugging 1012 1: enable debugging at boot time 1013 1014 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 1015 Format: 1016 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 1017 1018 cpuidle.off=1 [CPU_IDLE] 1019 disable the cpuidle sub-system 1020 1021 cpuidle.governor= 1022 [CPU_IDLE] Name of the cpuidle governor to use. 1023 1024 cpufreq.off=1 [CPU_FREQ] 1025 disable the cpufreq sub-system 1026 1027 cpufreq.default_governor= 1028 [CPU_FREQ] Name of the default cpufreq governor or 1029 policy to use. This governor must be registered in the 1030 kernel before the cpufreq driver probes. 1031 1032 cpu_init_udelay=N 1033 [X86,EARLY] Delay for N microsec between assert and de-assert 1034 of APIC INIT to start processors. This delay occurs 1035 on every CPU online, such as boot, and resume from suspend. 1036 Default: 10000 1037 1038 cpuhp.parallel= 1039 [SMP] Enable/disable parallel bringup of secondary CPUs 1040 Format: <bool> 1041 Default is enabled if CONFIG_HOTPLUG_PARALLEL=y. Otherwise 1042 the parameter has no effect. 1043 1044 crash_kexec_post_notifiers 1045 Only jump to kdump kernel after running the panic 1046 notifiers and dumping kmsg. This option increases 1047 the risks of a kdump failure, since some panic 1048 notifiers can make the crashed kernel more unstable. 1049 In configurations where kdump may not be reliable, 1050 running the panic notifiers could allow collecting 1051 more data on dmesg, like stack traces from other CPUS 1052 or extra data dumped by panic_print. Note that some 1053 configurations enable this option unconditionally, 1054 like Hyper-V, PowerPC (fadump) and AMD SEV-SNP. 1055 1056 crashkernel=size[KMG][@offset[KMG]] 1057 [KNL,EARLY] Using kexec, Linux can switch to a 'crash kernel' 1058 upon panic. This parameter reserves the physical 1059 memory region [offset, offset + size] for that kernel 1060 image. If '@offset' is omitted, then a suitable offset 1061 is selected automatically. 1062 [KNL, X86-64, ARM64, RISCV, LoongArch] Select a region 1063 under 4G first, and fall back to reserve region above 1064 4G when '@offset' hasn't been specified. 1065 See Documentation/admin-guide/kdump/kdump.rst for further details. 1066 1067 crashkernel=range1:size1[,range2:size2,...][@offset] 1068 [KNL] Same as above, but depends on the memory 1069 in the running system. The syntax of range is 1070 start-[end] where start and end are both 1071 a memory unit (amount[KMG]). See also 1072 Documentation/admin-guide/kdump/kdump.rst for an example. 1073 1074 crashkernel=size[KMG],high 1075 [KNL, X86-64, ARM64, RISCV, LoongArch] range could be 1076 above 4G. 1077 Allow kernel to allocate physical memory region from top, 1078 so could be above 4G if system have more than 4G ram 1079 installed. Otherwise memory region will be allocated 1080 below 4G, if available. 1081 It will be ignored if crashkernel=X is specified. 1082 crashkernel=size[KMG],low 1083 [KNL, X86-64, ARM64, RISCV, LoongArch] range under 4G. 1084 When crashkernel=X,high is passed, kernel could allocate 1085 physical memory region above 4G, that cause second kernel 1086 crash on system that require some amount of low memory, 1087 e.g. swiotlb requires at least 64M+32K low memory, also 1088 enough extra low memory is needed to make sure DMA buffers 1089 for 32-bit devices won't run out. Kernel would try to allocate 1090 default size of memory below 4G automatically. The default 1091 size is platform dependent. 1092 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB) 1093 --> arm64: 128MiB 1094 --> riscv: 128MiB 1095 --> loongarch: 128MiB 1096 This one lets the user specify own low range under 4G 1097 for second kernel instead. 1098 0: to disable low allocation. 1099 It will be ignored when crashkernel=X,high is not used 1100 or memory reserved is below 4G. 1101 crashkernel=size[KMG],cma 1102 [KNL, X86, ARM64, RISCV, PPC] Reserve additional crash 1103 kernel memory from CMA. This reservation is usable by 1104 the first system's userspace memory and kernel movable 1105 allocations (memory balloon, zswap). Pages allocated 1106 from this memory range will not be included in the vmcore 1107 so this should not be used if dumping of userspace memory 1108 is intended and it has to be expected that some movable 1109 kernel pages may be missing from the dump. 1110 1111 A standard crashkernel reservation, as described above, 1112 is still needed to hold the crash kernel and initrd. 1113 1114 This option increases the risk of a kdump failure: DMA 1115 transfers configured by the first kernel may end up 1116 corrupting the second kernel's memory. 1117 1118 This reservation method is intended for systems that 1119 can't afford to sacrifice enough memory for standard 1120 crashkernel reservation and where less reliable and 1121 possibly incomplete kdump is preferable to no kdump at 1122 all. 1123 1124 cryptomgr.notests 1125 [KNL] Disable crypto self-tests 1126 1127 cs89x0_dma= [HW,NET] 1128 Format: <dma> 1129 1130 cs89x0_media= [HW,NET] 1131 Format: { rj45 | aui | bnc } 1132 1133 csdlock_debug= [KNL] Enable or disable debug add-ons of cross-CPU 1134 function call handling. When switched on, 1135 additional debug data is printed to the console 1136 in case a hanging CPU is detected, and that 1137 CPU is pinged again in order to try to resolve 1138 the hang situation. The default value of this 1139 option depends on the CSD_LOCK_WAIT_DEBUG_DEFAULT 1140 Kconfig option. 1141 1142 dasd= [HW,NET] 1143 See header of drivers/s390/block/dasd_devmap.c. 1144 1145 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 1146 (one device per port) 1147 Format: <port#>,<type> 1148 See also Documentation/input/devices/joystick-parport.rst 1149 1150 debug [KNL,EARLY] Enable kernel debugging (events log level). 1151 1152 debug_boot_weak_hash 1153 [KNL,EARLY] Enable printing [hashed] pointers early in the 1154 boot sequence. If enabled, we use a weak hash instead 1155 of siphash to hash pointers. Use this option if you are 1156 seeing instances of '(___ptrval___)') and need to see a 1157 value (hashed pointer) instead. Cryptographically 1158 insecure, please do not use on production kernels. 1159 1160 debug_locks_verbose= 1161 [KNL] verbose locking self-tests 1162 Format: <int> 1163 Print debugging info while doing the locking API 1164 self-tests. 1165 Bitmask for the various LOCKTYPE_ tests. Defaults to 0 1166 (no extra messages), setting it to -1 (all bits set) 1167 will print _a_lot_ more information - normally only 1168 useful to lockdep developers. 1169 1170 debug_objects [KNL,EARLY] Enable object debugging 1171 1172 debug_guardpage_minorder= 1173 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this 1174 parameter allows control of the order of pages that will 1175 be intentionally kept free (and hence protected) by the 1176 buddy allocator. Bigger value increase the probability 1177 of catching random memory corruption, but reduce the 1178 amount of memory for normal system use. The maximum 1179 possible value is MAX_PAGE_ORDER/2. Setting this 1180 parameter to 1 or 2 should be enough to identify most 1181 random memory corruption problems caused by bugs in 1182 kernel or driver code when a CPU writes to (or reads 1183 from) a random memory location. Note that there exists 1184 a class of memory corruptions problems caused by buggy 1185 H/W or F/W or by drivers badly programming DMA 1186 (basically when memory is written at bus level and the 1187 CPU MMU is bypassed) which are not detectable by 1188 CONFIG_DEBUG_PAGEALLOC, hence this option will not 1189 help tracking down these problems. 1190 1191 debug_pagealloc= 1192 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 1193 enables the feature at boot time. By default, it is 1194 disabled and the system will work mostly the same as a 1195 kernel built without CONFIG_DEBUG_PAGEALLOC. 1196 Note: to get most of debug_pagealloc error reports, it's 1197 useful to also enable the page_owner functionality. 1198 on: enable the feature 1199 1200 debugfs= [KNL,EARLY] This parameter enables what is exposed to 1201 userspace and debugfs internal clients. 1202 Format: { on, off } 1203 on: All functions are enabled. 1204 off: Filesystem is not registered and clients 1205 get a -EPERM as result when trying to register files 1206 or directories within debugfs. 1207 This is equivalent of the runtime functionality if 1208 debugfs was not enabled in the kernel at all. 1209 Default value is set in build-time with a kernel configuration. 1210 1211 debugpat [X86] Enable PAT debugging 1212 1213 default_hugepagesz= 1214 [HW] The size of the default HugeTLB page. This is 1215 the size represented by the legacy /proc/ hugepages 1216 APIs. In addition, this is the default hugetlb size 1217 used for shmget(), mmap() and mounting hugetlbfs 1218 filesystems. If not specified, defaults to the 1219 architecture's default huge page size. Huge page 1220 sizes are architecture dependent. See also 1221 Documentation/admin-guide/mm/hugetlbpage.rst. 1222 Format: size[KMG] 1223 1224 deferred_probe_timeout= 1225 [KNL] Debugging option to set a timeout in seconds for 1226 deferred probe to give up waiting on dependencies to 1227 probe. Only specific dependencies (subsystems or 1228 drivers) that have opted in will be ignored. A timeout 1229 of 0 will timeout at the end of initcalls; a negative 1230 value is treated as an infinite timeout value. If the 1231 timeout hasn't expired, it'll be restarted by each 1232 successful driver registration. This option will also 1233 dump out devices still on the deferred probe list after 1234 retrying. 1235 1236 delayacct [KNL] Enable per-task delay accounting 1237 1238 dell_smm_hwmon.ignore_dmi= 1239 [HW] Continue probing hardware even if DMI data 1240 indicates that the driver is running on unsupported 1241 hardware. 1242 1243 dell_smm_hwmon.force= 1244 [HW] Activate driver even if SMM BIOS signature does 1245 not match list of supported models and enable otherwise 1246 blacklisted features. 1247 1248 dell_smm_hwmon.power_status= 1249 [HW] Report power status in /proc/i8k 1250 (disabled by default). 1251 1252 dell_smm_hwmon.restricted= 1253 [HW] Allow controlling fans only if SYS_ADMIN 1254 capability is set. 1255 1256 dell_smm_hwmon.fan_mult= 1257 [HW] Factor to multiply fan speed with. 1258 1259 dell_smm_hwmon.fan_max= 1260 [HW] Maximum configurable fan speed. 1261 1262 dfltcc= [HW,S390] 1263 Format: { on | off | def_only | inf_only | always } 1264 on: s390 zlib hardware support for compression on 1265 level 1 and decompression (default) 1266 off: No s390 zlib hardware support 1267 def_only: s390 zlib hardware support for deflate 1268 only (compression on level 1) 1269 inf_only: s390 zlib hardware support for inflate 1270 only (decompression) 1271 always: Same as 'on' but ignores the selected compression 1272 level always using hardware support (used for debugging) 1273 1274 dhash_entries= [KNL] 1275 Set number of hash buckets for dentry cache. 1276 1277 disable_1tb_segments [PPC,EARLY] 1278 Disables the use of 1TB hash page table segments. This 1279 causes the kernel to fall back to 256MB segments which 1280 can be useful when debugging issues that require an SLB 1281 miss to occur. 1282 1283 disable= [IPV6] 1284 See Documentation/networking/ipv6.rst. 1285 1286 disable_radix [PPC,EARLY] 1287 Disable RADIX MMU mode on POWER9 1288 1289 disable_tlbie [PPC] 1290 Disable TLBIE instruction. Currently does not work 1291 with KVM, with HASH MMU, or with coherent accelerators. 1292 1293 disable_ddw [PPC/PSERIES,EARLY] 1294 Disable Dynamic DMA Window support. Use this 1295 to workaround buggy firmware. 1296 1297 disable_ipv6= [IPV6] 1298 See Documentation/networking/ipv6.rst. 1299 1300 disable_mtrr_cleanup [X86,EARLY] 1301 The kernel tries to adjust MTRR layout from continuous 1302 to discrete, to make X server driver able to add WB 1303 entry later. This parameter disables that. 1304 1305 disable_mtrr_trim [X86, Intel and AMD only,EARLY] 1306 By default the kernel will trim any uncacheable 1307 memory out of your available memory pool based on 1308 MTRR settings. This parameter disables that behavior, 1309 possibly causing your machine to run very slowly. 1310 1311 disable_timer_pin_1 [X86,EARLY] 1312 Disable PIN 1 of APIC timer 1313 Can be useful to work around chipset bugs. 1314 1315 dis_ucode_ldr [X86] Disable the microcode loader. 1316 1317 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 1318 this option disables the debugging code at boot. 1319 1320 dma_debug_entries=<number> 1321 This option allows to tune the number of preallocated 1322 entries for DMA-API debugging code. One entry is 1323 required per DMA-API allocation. Use this if the 1324 DMA-API debugging code disables itself because the 1325 architectural default is too low. 1326 1327 dma_debug_driver=<driver_name> 1328 With this option the DMA-API debugging driver 1329 filter feature can be enabled at boot time. Just 1330 pass the driver to filter for as the parameter. 1331 The filter can be disabled or changed to another 1332 driver later using sysfs. 1333 1334 reg_file_data_sampling= 1335 [X86] Controls mitigation for Register File Data 1336 Sampling (RFDS) vulnerability. RFDS is a CPU 1337 vulnerability which may allow userspace to infer 1338 kernel data values previously stored in floating point 1339 registers, vector registers, or integer registers. 1340 RFDS only affects Intel Atom processors. 1341 1342 on: Turns ON the mitigation. 1343 off: Turns OFF the mitigation. 1344 1345 This parameter overrides the compile time default set 1346 by CONFIG_MITIGATION_RFDS. Mitigation cannot be 1347 disabled when other VERW based mitigations (like MDS) 1348 are enabled. In order to disable RFDS mitigation all 1349 VERW based mitigations need to be disabled. 1350 1351 For details see: 1352 Documentation/admin-guide/hw-vuln/reg-file-data-sampling.rst 1353 1354 dm_verity.keyring_unsealed= 1355 [KNL] When set to 1, leave the dm-verity keyring 1356 unsealed after initialization so userspace can 1357 provision keys. Once the keyring is restricted 1358 it becomes active and is searched during signature 1359 verification. 1360 1361 driver_async_probe= [KNL] 1362 List of driver names to be probed asynchronously. * 1363 matches with all driver names. If * is specified, the 1364 rest of the listed driver names are those that will NOT 1365 match the *. 1366 Format: <driver_name1>,<driver_name2>... 1367 1368 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1369 Broken monitors, graphic adapters, KVMs and EDIDless 1370 panels may send no or incorrect EDID data sets. 1371 This parameter allows to specify an EDID data sets 1372 in the /lib/firmware directory that are used instead. 1373 An EDID data set will only be used for a particular 1374 connector, if its name and a colon are prepended to 1375 the EDID name. Each connector may use a unique EDID 1376 data set by separating the files with a comma. An EDID 1377 data set with no connector name will be used for 1378 any connectors not explicitly specified. 1379 1380 dscc4.setup= [NET] 1381 1382 dt_cpu_ftrs= [PPC,EARLY] 1383 Format: {"off" | "known"} 1384 Control how the dt_cpu_ftrs device-tree binding is 1385 used for CPU feature discovery and setup (if it 1386 exists). 1387 off: Do not use it, fall back to legacy cpu table. 1388 known: Do not pass through unknown features to guests 1389 or userspace, only those that the kernel is aware of. 1390 1391 dump_apple_properties [X86] 1392 Dump name and content of EFI device properties on 1393 x86 Macs. Useful for driver authors to determine 1394 what data is available or for reverse-engineering. 1395 1396 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1397 <module>.dyndbg[="val"] 1398 Enable debug messages at boot time. See 1399 Documentation/admin-guide/dynamic-debug-howto.rst 1400 for details. 1401 1402 early_ioremap_debug [KNL,EARLY] 1403 Enable debug messages in early_ioremap support. This 1404 is useful for tracking down temporary early mappings 1405 which are not unmapped. 1406 1407 earlycon= [KNL,EARLY] Output early console device and options. 1408 1409 When used with no options, the early console is 1410 determined by stdout-path property in device tree's 1411 chosen node or the ACPI SPCR table if supported by 1412 the platform. 1413 1414 cdns,<addr>[,options] 1415 Start an early, polled-mode console on a Cadence 1416 (xuartps) serial port at the specified address. Only 1417 supported option is baud rate. If baud rate is not 1418 specified, the serial port must already be setup and 1419 configured. 1420 1421 uart[8250],io,<addr>[,options[,uartclk]] 1422 uart[8250],mmio,<addr>[,options[,uartclk]] 1423 uart[8250],mmio32,<addr>[,options[,uartclk]] 1424 uart[8250],mmio32be,<addr>[,options[,uartclk]] 1425 uart[8250],0x<addr>[,options] 1426 Start an early, polled-mode console on the 8250/16550 1427 UART at the specified I/O port or MMIO address. 1428 MMIO inter-register address stride is either 8-bit 1429 (mmio) or 32-bit (mmio32 or mmio32be). 1430 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1431 to be equivalent to 'mmio'. 'options' are specified 1432 in the same format described for "console=ttyS<n>"; if 1433 unspecified, the h/w is not initialized. 'uartclk' is 1434 the uart clock frequency; if unspecified, it is set 1435 to 'BASE_BAUD' * 16. 1436 1437 pl011,<addr> 1438 pl011,mmio32,<addr> 1439 Start an early, polled-mode console on a pl011 serial 1440 port at the specified address. The pl011 serial port 1441 must already be setup and configured. Options are not 1442 yet supported. If 'mmio32' is specified, then only 1443 the driver will use only 32-bit accessors to read/write 1444 the device registers. 1445 1446 liteuart,<addr> 1447 Start an early console on a litex serial port at the 1448 specified address. The serial port must already be 1449 setup and configured. Options are not yet supported. 1450 1451 meson,<addr> 1452 Start an early, polled-mode console on a meson serial 1453 port at the specified address. The serial port must 1454 already be setup and configured. Options are not yet 1455 supported. 1456 1457 msm_serial,<addr> 1458 Start an early, polled-mode console on an msm serial 1459 port at the specified address. The serial port 1460 must already be setup and configured. Options are not 1461 yet supported. 1462 1463 msm_serial_dm,<addr> 1464 Start an early, polled-mode console on an msm serial 1465 dm port at the specified address. The serial port 1466 must already be setup and configured. Options are not 1467 yet supported. 1468 1469 owl,<addr> 1470 Start an early, polled-mode console on a serial port 1471 of an Actions Semi SoC, such as S500 or S900, at the 1472 specified address. The serial port must already be 1473 setup and configured. Options are not yet supported. 1474 1475 rda,<addr> 1476 Start an early, polled-mode console on a serial port 1477 of an RDA Micro SoC, such as RDA8810PL, at the 1478 specified address. The serial port must already be 1479 setup and configured. Options are not yet supported. 1480 1481 sbi 1482 Use RISC-V SBI (Supervisor Binary Interface) for early 1483 console. 1484 1485 smh Use ARM semihosting calls for early console. 1486 1487 s3c2410,<addr> 1488 s3c2412,<addr> 1489 s3c2440,<addr> 1490 s3c6400,<addr> 1491 s5pv210,<addr> 1492 exynos4210,<addr> 1493 Use early console provided by serial driver available 1494 on Samsung SoCs, requires selecting proper type and 1495 a correct base address of the selected UART port. The 1496 serial port must already be setup and configured. 1497 Options are not yet supported. 1498 1499 lantiq,<addr> 1500 Start an early, polled-mode console on a lantiq serial 1501 (lqasc) port at the specified address. The serial port 1502 must already be setup and configured. Options are not 1503 yet supported. 1504 1505 lpuart,<addr> 1506 lpuart32,<addr> 1507 Use early console provided by Freescale LP UART driver 1508 found on Freescale Vybrid and QorIQ LS1021A processors. 1509 A valid base address must be provided, and the serial 1510 port must already be setup and configured. 1511 1512 ec_imx21,<addr> 1513 ec_imx6q,<addr> 1514 Start an early, polled-mode, output-only console on the 1515 Freescale i.MX UART at the specified address. The UART 1516 must already be setup and configured. 1517 1518 ar3700_uart,<addr> 1519 Start an early, polled-mode console on the 1520 Armada 3700 serial port at the specified 1521 address. The serial port must already be setup 1522 and configured. Options are not yet supported. 1523 1524 qcom_geni,<addr> 1525 Start an early, polled-mode console on a Qualcomm 1526 Generic Interface (GENI) based serial port at the 1527 specified address. The serial port must already be 1528 setup and configured. Options are not yet supported. 1529 1530 efifb,[options] 1531 Start an early, unaccelerated console on the EFI 1532 memory mapped framebuffer (if available). On cache 1533 coherent non-x86 systems that use system memory for 1534 the framebuffer, pass the 'ram' option so that it is 1535 mapped with the correct attributes. 1536 1537 linflex,<addr> 1538 Use early console provided by Freescale LINFlexD UART 1539 serial driver for NXP S32V234 SoCs. A valid base 1540 address must be provided, and the serial port must 1541 already be setup and configured. 1542 1543 earlyprintk= [X86,SH,ARM,M68k,S390,UM,EARLY] 1544 earlyprintk=vga 1545 earlyprintk=sclp 1546 earlyprintk=xen 1547 earlyprintk=serial[,ttySn[,baudrate]] 1548 earlyprintk=serial[,0x...[,baudrate]] 1549 earlyprintk=ttySn[,baudrate] 1550 earlyprintk=dbgp[debugController#] 1551 earlyprintk=mmio32,membase[,{nocfg|baudrate}] 1552 earlyprintk=pciserial[,force],bus:device.function[,{nocfg|baudrate}] 1553 earlyprintk=xdbc[xhciController#] 1554 earlyprintk=bios 1555 1556 earlyprintk is useful when the kernel crashes before 1557 the normal console is initialized. It is not enabled by 1558 default because it has some cosmetic problems. 1559 1560 Use "nocfg" to skip UART configuration, assume 1561 BIOS/firmware has configured UART correctly. 1562 1563 Append ",keep" to not disable it when the real console 1564 takes over. 1565 1566 Only one of vga, serial, or usb debug port can 1567 be used at a time. 1568 1569 Currently only ttyS0 and ttyS1 may be specified by 1570 name. Other I/O ports may be explicitly specified 1571 on some architectures (x86 and arm at least) by 1572 replacing ttySn with an I/O port address, like this: 1573 earlyprintk=serial,0x1008,115200 1574 You can find the port for a given device in 1575 /proc/tty/driver/serial: 1576 2: uart:ST16650V2 port:00001008 irq:18 ... 1577 1578 Interaction with the standard serial driver is not 1579 very good. 1580 1581 The VGA output is eventually overwritten by 1582 the real console. 1583 1584 The xen option can only be used in Xen domains. 1585 1586 The sclp output can only be used on s390. 1587 1588 The bios output can only be used on SuperH. 1589 1590 The optional "force" to "pciserial" enables use of a 1591 PCI device even when its classcode is not of the 1592 UART class. 1593 1594 edd= [EDD] 1595 Format: {"off" | "on" | "skip[mbr]"} 1596 1597 efi= [EFI,EARLY] 1598 Format: { "debug", "disable_early_pci_dma", 1599 "nochunk", "noruntime", "nosoftreserve", 1600 "novamap", "no_disable_early_pci_dma" } 1601 debug: enable misc debug output. 1602 disable_early_pci_dma: disable the busmaster bit on all 1603 PCI bridges while in the EFI boot stub. 1604 nochunk: disable reading files in "chunks" in the EFI 1605 boot stub, as chunking can cause problems with some 1606 firmware implementations. 1607 noruntime : disable EFI runtime services support 1608 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1609 attribute may cause the kernel to reserve the 1610 memory range for a memory mapping driver to 1611 claim. Specify efi=nosoftreserve to disable this 1612 reservation and treat the memory by its base type 1613 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1614 novamap: do not call SetVirtualAddressMap(). 1615 no_disable_early_pci_dma: Leave the busmaster bit set 1616 on all PCI bridges while in the EFI boot stub 1617 1618 efi_no_storage_paranoia [EFI,X86,EARLY] 1619 The kernel reserves 5KB of EFI variable storage for 1620 safety, because some UEFI implementation may fail to 1621 boot if there's insufficient space in the EFI variable 1622 storage. 1623 1624 Using this parameter, you can use the 5KB reservation 1625 in the EFI variable storage. 1626 1627 However, Use this parameter only if you are really 1628 sure that your UEFI does sane gc and fulfills the spec 1629 otherwise your board may brick. 1630 1631 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1632 that is to be dynamically loaded by Linux. If there are 1633 multiple variables with the same name but with different 1634 vendor GUIDs, all of them will be loaded. See 1635 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1636 1637 1638 eisa_irq_edge= [PARISC,HW] 1639 See header of drivers/parisc/eisa.c. 1640 1641 ekgdboc= [X86,KGDB,EARLY] Allow early kernel console debugging 1642 Format: ekgdboc=kbd 1643 1644 This is designed to be used in conjunction with 1645 the boot argument: earlyprintk=vga 1646 1647 This parameter works in place of the kgdboc parameter 1648 but can only be used if the backing tty is available 1649 very early in the boot process. For early debugging 1650 via a serial port see kgdboc_earlycon instead. 1651 1652 elfcorehdr=[size[KMG]@]offset[KMG] [PPC,SH,X86,S390,EARLY] 1653 Specifies physical address of start of kernel core 1654 image elf header and optionally the size. Generally 1655 kexec loader will pass this option to capture kernel. 1656 See Documentation/admin-guide/kdump/kdump.rst for details. 1657 1658 enable_mtrr_cleanup [X86,EARLY] 1659 The kernel tries to adjust MTRR layout from continuous 1660 to discrete, to make X server driver able to add WB 1661 entry later. This parameter enables that. 1662 1663 enforcing= [SELINUX] Set initial enforcing status. 1664 Format: {"0" | "1"} 1665 See security/selinux/Kconfig help text. 1666 0 -- permissive (log only, no denials). 1667 1 -- enforcing (deny and log). 1668 Default value is 0. 1669 Value can be changed at runtime via 1670 /sys/fs/selinux/enforce. 1671 1672 erst_disable [ACPI] 1673 Disable Error Record Serialization Table (ERST) 1674 support. 1675 1676 ether= [HW,NET] Ethernet cards parameters 1677 This option is obsoleted by the "netdev=" option, which 1678 has equivalent usage. See its documentation for details. 1679 1680 evm= [EVM] 1681 Format: { "fix" } 1682 Permit 'security.evm' to be updated regardless of 1683 current integrity status. 1684 1685 early_page_ext [KNL,EARLY] Enforces page_ext initialization to earlier 1686 stages so cover more early boot allocations. 1687 Please note that as side effect some optimizations 1688 might be disabled to achieve that (e.g. parallelized 1689 memory initialization is disabled) so the boot process 1690 might take longer, especially on systems with a lot of 1691 memory. Available with CONFIG_PAGE_EXTENSION=y. 1692 1693 failslab= 1694 fail_usercopy= 1695 fail_page_alloc= 1696 fail_skb_realloc= 1697 fail_make_request=[KNL] 1698 General fault injection mechanism. 1699 Format: <interval>,<probability>,<space>,<times> 1700 See also Documentation/fault-injection/. 1701 1702 fb_tunnels= [NET] 1703 Format: { initns | none } 1704 See Documentation/admin-guide/sysctl/net.rst for 1705 fb_tunnels_only_for_init_ns 1706 1707 floppy= [HW] 1708 See Documentation/admin-guide/blockdev/floppy.rst. 1709 1710 forcepae [X86-32] 1711 Forcefully enable Physical Address Extension (PAE). 1712 Many Pentium M systems disable PAE but may have a 1713 functionally usable PAE implementation. 1714 Warning: use of this parameter will taint the kernel 1715 and may cause unknown problems. 1716 1717 fred= [X86-64] 1718 Enable/disable Flexible Return and Event Delivery. 1719 Format: { on | off } 1720 on: enable FRED when it's present, the default setting. 1721 off: disable FRED. 1722 1723 ftrace=[tracer] 1724 [FTRACE] will set and start the specified tracer 1725 as early as possible in order to facilitate early 1726 boot debugging. 1727 1728 ftrace_boot_snapshot 1729 [FTRACE] On boot up, a snapshot will be taken of the 1730 ftrace ring buffer that can be read at: 1731 /sys/kernel/tracing/snapshot. 1732 This is useful if you need tracing information from kernel 1733 boot up that is likely to be overridden by user space 1734 start up functionality. 1735 1736 Optionally, the snapshot can also be defined for a tracing 1737 instance that was created by the trace_instance= command 1738 line parameter. 1739 1740 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo 1741 1742 The above will cause the "foo" tracing instance to trigger 1743 a snapshot at the end of boot up. 1744 1745 ftrace_dump_on_oops[=2(orig_cpu) | =<instance>][,<instance> | 1746 ,<instance>=2(orig_cpu)] 1747 [FTRACE] will dump the trace buffers on oops. 1748 If no parameter is passed, ftrace will dump global 1749 buffers of all CPUs, if you pass 2 or orig_cpu, it 1750 will dump only the buffer of the CPU that triggered 1751 the oops, or the specific instance will be dumped if 1752 its name is passed. Multiple instance dump is also 1753 supported, and instances are separated by commas. Each 1754 instance supports only dump on CPU that triggered the 1755 oops by passing 2 or orig_cpu to it. 1756 1757 ftrace_dump_on_oops=foo=orig_cpu 1758 1759 The above will dump only the buffer of "foo" instance 1760 on CPU that triggered the oops. 1761 1762 ftrace_dump_on_oops,foo,bar=orig_cpu 1763 1764 The above will dump global buffer on all CPUs, the 1765 buffer of "foo" instance on all CPUs and the buffer 1766 of "bar" instance on CPU that triggered the oops. 1767 1768 ftrace_filter=[function-list] 1769 [FTRACE] Limit the functions traced by the function 1770 tracer at boot up. function-list is a comma-separated 1771 list of functions. This list can be changed at run 1772 time by the set_ftrace_filter file in the debugfs 1773 tracing directory. 1774 1775 ftrace_notrace=[function-list] 1776 [FTRACE] Do not trace the functions specified in 1777 function-list. This list can be changed at run time 1778 by the set_ftrace_notrace file in the debugfs 1779 tracing directory. 1780 1781 ftrace_graph_filter=[function-list] 1782 [FTRACE] Limit the top level callers functions traced 1783 by the function graph tracer at boot up. 1784 function-list is a comma-separated list of functions 1785 that can be changed at run time by the 1786 set_graph_function file in the debugfs tracing directory. 1787 1788 ftrace_graph_notrace=[function-list] 1789 [FTRACE] Do not trace from the functions specified in 1790 function-list. This list is a comma-separated list of 1791 functions that can be changed at run time by the 1792 set_graph_notrace file in the debugfs tracing directory. 1793 1794 ftrace_graph_max_depth=<uint> 1795 [FTRACE] Used with the function graph tracer. This is 1796 the max depth it will trace into a function. This value 1797 can be changed at run time by the max_graph_depth file 1798 in the tracefs tracing directory. default: 0 (no limit) 1799 1800 fw_devlink= [KNL,EARLY] Create device links between consumer and supplier 1801 devices by scanning the firmware to infer the 1802 consumer/supplier relationships. This feature is 1803 especially useful when drivers are loaded as modules as 1804 it ensures proper ordering of tasks like device probing 1805 (suppliers first, then consumers), supplier boot state 1806 clean up (only after all consumers have probed), 1807 suspend/resume & runtime PM (consumers first, then 1808 suppliers). 1809 Format: { off | permissive | on | rpm } 1810 off -- Don't create device links from firmware info. 1811 permissive -- Create device links from firmware info 1812 but use it only for ordering boot state clean 1813 up (sync_state() calls). 1814 on -- Create device links from firmware info and use it 1815 to enforce probe and suspend/resume ordering. 1816 rpm -- Like "on", but also use to order runtime PM. 1817 1818 fw_devlink.strict=<bool> 1819 [KNL,EARLY] Treat all inferred dependencies as mandatory 1820 dependencies. This only applies for fw_devlink=on|rpm. 1821 Format: <bool> 1822 1823 fw_devlink.sync_state = 1824 [KNL,EARLY] When all devices that could probe have finished 1825 probing, this parameter controls what to do with 1826 devices that haven't yet received their sync_state() 1827 calls. 1828 Format: { strict | timeout } 1829 strict -- Default. Continue waiting on consumers to 1830 probe successfully. 1831 timeout -- Give up waiting on consumers and call 1832 sync_state() on any devices that haven't yet 1833 received their sync_state() calls after 1834 deferred_probe_timeout has expired or by 1835 late_initcall() if !CONFIG_MODULES. 1836 1837 gamecon.map[2|3]= 1838 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1839 support via parallel port (up to 5 devices per port) 1840 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1841 See also Documentation/input/devices/joystick-parport.rst 1842 1843 gamma= [HW,DRM] 1844 1845 gart_fix_e820= [X86-64,EARLY] disable the fix e820 for K8 GART 1846 Format: off | on 1847 default: on 1848 1849 gather_data_sampling= 1850 [X86,INTEL,EARLY] Control the Gather Data Sampling (GDS) 1851 mitigation. 1852 1853 Gather Data Sampling is a hardware vulnerability which 1854 allows unprivileged speculative access to data which was 1855 previously stored in vector registers. 1856 1857 This issue is mitigated by default in updated microcode. 1858 The mitigation may have a performance impact but can be 1859 disabled. On systems without the microcode mitigation 1860 disabling AVX serves as a mitigation. 1861 1862 force: Disable AVX to mitigate systems without 1863 microcode mitigation. No effect if the microcode 1864 mitigation is present. Known to cause crashes in 1865 userspace with buggy AVX enumeration. 1866 1867 off: Disable GDS mitigation. 1868 1869 gbpages [X86] Use GB pages for kernel direct mappings. 1870 1871 gcov_persist= [GCOV] When non-zero (default), profiling data for 1872 kernel modules is saved and remains accessible via 1873 debugfs, even when the module is unloaded/reloaded. 1874 When zero, profiling data is discarded and associated 1875 debugfs files are removed at module unload time. 1876 1877 goldfish [X86] Enable the goldfish android emulator platform. 1878 Don't use this when you are not running on the 1879 android emulator 1880 1881 gpio-mockup.gpio_mockup_ranges 1882 [HW] Sets the ranges of gpiochip of for this device. 1883 Format: <start1>,<end1>,<start2>,<end2>... 1884 gpio-mockup.gpio_mockup_named_lines 1885 [HW] Let the driver know GPIO lines should be named. 1886 1887 gpt [EFI] Forces disk with valid GPT signature but 1888 invalid Protective MBR to be treated as GPT. If the 1889 primary GPT is corrupted, it enables the backup/alternate 1890 GPT to be used instead. 1891 1892 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1893 the "Enable 0" bit of the configuration register. 1894 Format: 0 | 1 1895 Default: 0 1896 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1897 the "Enable 0" bit of the configuration register. 1898 Format: 0 | 1 1899 Default: 0 1900 grcan.select= [HW] Select which physical interface to use. 1901 Format: 0 | 1 1902 Default: 0 1903 grcan.txsize= [HW] Sets the size of the tx buffer. 1904 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1905 Default: 1024 1906 grcan.rxsize= [HW] Sets the size of the rx buffer. 1907 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1908 Default: 1024 1909 1910 hardened_usercopy= 1911 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 1912 hardening is enabled for this boot. Hardened 1913 usercopy checking is used to protect the kernel 1914 from reading or writing beyond known memory 1915 allocation boundaries as a proactive defense 1916 against bounds-checking flaws in the kernel's 1917 copy_to_user()/copy_from_user() interface. 1918 The default is determined by 1919 CONFIG_HARDENED_USERCOPY_DEFAULT_ON. 1920 on Perform hardened usercopy checks. 1921 off Disable hardened usercopy checks. 1922 1923 hardlockup_all_cpu_backtrace= 1924 [KNL] Should the hard-lockup detector generate 1925 backtraces on all cpus. 1926 Format: 0 | 1 1927 1928 hash_pointers= 1929 [KNL,EARLY] 1930 By default, when pointers are printed to the console 1931 or buffers via the %p format string, that pointer is 1932 "hashed", i.e. obscured by hashing the pointer value. 1933 This is a security feature that hides actual kernel 1934 addresses from unprivileged users, but it also makes 1935 debugging the kernel more difficult since unequal 1936 pointers can no longer be compared. The choices are: 1937 Format: { auto | always | never } 1938 Default: auto 1939 1940 auto - Hash pointers unless slab_debug is enabled. 1941 always - Always hash pointers (even if slab_debug is 1942 enabled). 1943 never - Never hash pointers. This option should only 1944 be specified when debugging the kernel. Do 1945 not use on production kernels. The boot 1946 param "no_hash_pointers" is an alias for 1947 this mode. 1948 1949 For controlling hashing dynamically at runtime, 1950 use the "kernel.kptr_restrict" sysctl instead. 1951 1952 hashdist= [KNL,NUMA] Large hashes allocated during boot 1953 are distributed across NUMA nodes. Defaults on 1954 for 64-bit NUMA, off otherwise. 1955 Format: 0 | 1 (for off | on) 1956 1957 hd= [EIDE] (E)IDE hard drive subsystem geometry 1958 Format: <cyl>,<head>,<sect> 1959 1960 hest_disable [ACPI] 1961 Disable Hardware Error Source Table (HEST) support; 1962 corresponding firmware-first mode error processing 1963 logic will be disabled. 1964 1965 hibernate= [HIBERNATION] 1966 noresume Don't check if there's a hibernation image 1967 present during boot. 1968 nocompress Don't compress/decompress hibernation images. 1969 no Disable hibernation and resume. 1970 protect_image Turn on image protection during restoration 1971 (that will set all pages holding image data 1972 during restoration read-only). 1973 1974 hibernate.compressor= [HIBERNATION] Compression algorithm to be 1975 used with hibernation. 1976 Format: { lzo | lz4 } 1977 Default: lzo 1978 1979 lzo: Select LZO compression algorithm to 1980 compress/decompress hibernation image. 1981 1982 lz4: Select LZ4 compression algorithm to 1983 compress/decompress hibernation image. 1984 1985 hibernate.pm_test_delay= 1986 [HIBERNATION] 1987 Sets the number of seconds to remain in a hibernation test 1988 mode before resuming the system (see 1989 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 1990 is set. Default value is 5. 1991 1992 hibernate_compression_threads= 1993 [HIBERNATION] 1994 Set the number of threads used for compressing or decompressing 1995 hibernation images. 1996 1997 Format: <integer> 1998 Default: 3 1999 Minimum: 1 2000 Example: hibernate_compression_threads=4 2001 2002 highmem=nn[KMG] [KNL,BOOT,EARLY] forces the highmem zone to have an exact 2003 size of <nn>. This works even on boxes that have no 2004 highmem otherwise. This also works to reduce highmem 2005 size on bigger boxes. 2006 2007 highres= [KNL] Enable/disable high resolution timer mode. 2008 Valid parameters: "on", "off" 2009 Default: "on" 2010 2011 hlt [BUGS=ARM,SH] 2012 2013 hostname= [KNL,EARLY] Set the hostname (aka UTS nodename). 2014 Format: <string> 2015 This allows setting the system's hostname during early 2016 startup. This sets the name returned by gethostname. 2017 Using this parameter to set the hostname makes it 2018 possible to ensure the hostname is correctly set before 2019 any userspace processes run, avoiding the possibility 2020 that a process may call gethostname before the hostname 2021 has been explicitly set, resulting in the calling 2022 process getting an incorrect result. The string must 2023 not exceed the maximum allowed hostname length (usually 2024 64 characters) and will be truncated otherwise. 2025 2026 hpet= [X86-32,HPET] option to control HPET usage 2027 Format: { enable (default) | disable | force | 2028 verbose } 2029 disable: disable HPET and use PIT instead 2030 force: allow force enabled of undocumented chips (ICH4, 2031 VIA, nVidia) 2032 verbose: show contents of HPET registers during setup 2033 2034 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 2035 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 2036 2037 hugepages= [HW,EARLY] Number of HugeTLB pages to allocate at boot. 2038 If this follows hugepagesz (below), it specifies 2039 the number of pages of hugepagesz to be allocated. 2040 If this is the first HugeTLB parameter on the command 2041 line, it specifies the number of pages to allocate for 2042 the default huge page size. If using node format, the 2043 number of pages to allocate per-node can be specified. 2044 See also Documentation/admin-guide/mm/hugetlbpage.rst. 2045 Format: <integer> or (node format) 2046 <node>:<integer>[,<node>:<integer>] 2047 2048 hugepagesz= 2049 [HW,EARLY] The size of the HugeTLB pages. This is 2050 used in conjunction with hugepages (above) to 2051 allocate huge pages of a specific size at boot. The 2052 pair hugepagesz=X hugepages=Y can be specified once 2053 for each supported huge page size. Huge page sizes 2054 are architecture dependent. See also 2055 Documentation/admin-guide/mm/hugetlbpage.rst. 2056 Format: size[KMG] 2057 2058 hugepage_alloc_threads= 2059 [HW] The number of threads that should be used to 2060 allocate hugepages during boot. This option can be 2061 used to improve system bootup time when allocating 2062 a large amount of huge pages. 2063 The default value is 25% of the available hardware threads. 2064 2065 Note that this parameter only applies to non-gigantic huge pages. 2066 2067 hugetlb_cma= [HW,CMA,EARLY] The size of a CMA area used for allocation 2068 of gigantic hugepages. Or using node format, the size 2069 of a CMA area per node can be specified. 2070 The size can be an absolute value (e.g., 2G) or a 2071 percentage of the total memory or node memory (e.g., 20%). 2072 Percentage-derived sizes are rounded down to a multiple of 2073 the architecture's gigantic hugepage size and may become 2074 zero. 2075 Format: nn[KMGTPE] or nn% or (node format) 2076 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]] or 2077 <node>:nn%[,<node>:nn%] 2078 2079 The size must be a multiple of the gigantic page size. 2080 When using node format, this applies to each per-node size. 2081 Missaligned values are dropped with a warning. 2082 2083 Reserve a CMA area of given size and allocate gigantic 2084 hugepages using the CMA allocator. If enabled, the 2085 boot-time allocation of gigantic hugepages is skipped. 2086 2087 hugetlb_cma_only= 2088 [HW,CMA,EARLY] When allocating new HugeTLB pages, only 2089 try to allocate from the CMA areas. 2090 2091 This option does nothing if hugetlb_cma= is not also 2092 specified. 2093 2094 hugetlb_free_vmemmap= 2095 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP 2096 enabled. 2097 Control if HugeTLB Vmemmap Optimization (HVO) is enabled. 2098 Allows heavy hugetlb users to free up some more 2099 memory (7 * PAGE_SIZE for each 2MB hugetlb page). 2100 Format: { on | off (default) } 2101 2102 on: enable HVO 2103 off: disable HVO 2104 2105 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y, 2106 the default is on. 2107 2108 Note that the vmemmap pages may be allocated from the added 2109 memory block itself when memory_hotplug.memmap_on_memory is 2110 enabled, those vmemmap pages cannot be optimized even if this 2111 feature is enabled. Other vmemmap pages not allocated from 2112 the added memory block itself do not be affected. 2113 2114 hung_task_panic= 2115 [KNL] Number of hung tasks to trigger kernel panic. 2116 Format: <int> 2117 2118 When set to a non-zero value, a kernel panic will be triggered if 2119 the number of detected hung tasks reaches this value. 2120 2121 0: don't panic 2122 1: panic immediately on first hung task 2123 N: panic after N hung tasks are detected in a single scan 2124 2125 The default value is controlled by the 2126 CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time option. The value 2127 selected by this boot parameter can be changed later by the 2128 kernel.hung_task_panic sysctl. 2129 2130 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 2131 terminal devices. Valid values: 0..8 2132 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 2133 If specified, z/VM IUCV HVC accepts connections 2134 from listed z/VM user IDs only. 2135 2136 hv_nopvspin [X86,HYPER_V,EARLY] 2137 Disables the paravirt spinlock optimizations 2138 which allow the hypervisor to 'idle' the guest 2139 on lock contention. 2140 2141 hw_protection= [HW] 2142 Format: reboot | shutdown 2143 2144 Hardware protection action taken on critical events like 2145 overtemperature or imminent voltage loss. 2146 2147 i2c_bus= [HW] Override the default board specific I2C bus speed 2148 or register an additional I2C bus that is not 2149 registered from board initialization code. 2150 Format: 2151 <bus_id>,<clkrate> 2152 2153 i2c_touchscreen_props= [HW,ACPI,X86] 2154 Set device-properties for ACPI-enumerated I2C-attached 2155 touchscreen, to e.g. fix coordinates of upside-down 2156 mounted touchscreens. If you need this option please 2157 submit a drivers/platform/x86/touchscreen_dmi.c patch 2158 adding a DMI quirk for this. 2159 2160 Format: 2161 <ACPI_HW_ID>:<prop_name>=<val>[:prop_name=val][:...] 2162 Where <val> is one of: 2163 Omit "=<val>" entirely Set a boolean device-property 2164 Unsigned number Set a u32 device-property 2165 Anything else Set a string device-property 2166 2167 Examples (split over multiple lines): 2168 i2c_touchscreen_props=GDIX1001:touchscreen-inverted-x: 2169 touchscreen-inverted-y 2170 2171 i2c_touchscreen_props=MSSL1680:touchscreen-size-x=1920: 2172 touchscreen-size-y=1080:touchscreen-inverted-y: 2173 firmware-name=gsl1680-vendor-model.fw:silead,home-button 2174 2175 i8042.debug [HW] Toggle i8042 debug mode 2176 i8042.unmask_kbd_data 2177 [HW] Enable printing of interrupt data from the KBD port 2178 (disabled by default, and as a pre-condition 2179 requires that i8042.debug=1 be enabled) 2180 i8042.direct [HW] Put keyboard port into non-translated mode 2181 i8042.dumbkbd [HW] Pretend that controller can only read data from 2182 keyboard and cannot control its state 2183 (Don't attempt to blink the leds) 2184 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 2185 i8042.nokbd [HW] Don't check/create keyboard port 2186 i8042.noloop [HW] Disable the AUX Loopback command while probing 2187 for the AUX port 2188 i8042.nomux [HW] Don't check presence of an active multiplexing 2189 controller 2190 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 2191 controllers 2192 i8042.notimeout [HW] Ignore timeout condition signalled by controller 2193 i8042.reset [HW] Reset the controller during init, cleanup and 2194 suspend-to-ram transitions, only during s2r 2195 transitions, or never reset 2196 Format: { 1 | Y | y | 0 | N | n } 2197 1, Y, y: always reset controller 2198 0, N, n: don't ever reset controller 2199 Default: only on s2r transitions on x86; most other 2200 architectures force reset to be always executed 2201 i8042.unlock [HW] Unlock (ignore) the keylock 2202 i8042.kbdreset [HW] Reset device connected to KBD port 2203 i8042.probe_defer 2204 [HW] Allow deferred probing upon i8042 probe errors 2205 2206 i810= [HW,DRM] 2207 2208 i915.invert_brightness= 2209 [DRM] Invert the sense of the variable that is used to 2210 set the brightness of the panel backlight. Normally a 2211 brightness value of 0 indicates backlight switched off, 2212 and the maximum of the brightness value sets the backlight 2213 to maximum brightness. If this parameter is set to 0 2214 (default) and the machine requires it, or this parameter 2215 is set to 1, a brightness value of 0 sets the backlight 2216 to maximum brightness, and the maximum of the brightness 2217 value switches the backlight off. 2218 -1 -- never invert brightness 2219 0 -- machine default 2220 1 -- force brightness inversion 2221 2222 ia32_emulation= [X86-64] 2223 Format: <bool> 2224 When true, allows loading 32-bit programs and executing 32-bit 2225 syscalls, essentially overriding IA32_EMULATION_DEFAULT_DISABLED at 2226 boot time. When false, unconditionally disables IA32 emulation. 2227 2228 2229 idle= [X86,EARLY] 2230 Format: idle=poll, idle=halt, idle=nomwait 2231 2232 idle=poll: Don't do power saving in the idle loop 2233 using HLT, but poll for rescheduling event. This will 2234 make the CPUs eat a lot more power, but may be useful 2235 to get slightly better performance in multiprocessor 2236 benchmarks. It also makes some profiling using 2237 performance counters more accurate. Please note that 2238 on systems with MONITOR/MWAIT support (like Intel 2239 EM64T CPUs) this option has no performance advantage 2240 over the normal idle loop. It may also interact badly 2241 with hyperthreading. 2242 2243 idle=halt: Halt is forced to be used for CPU idle. 2244 In such case C2/C3 won't be used again. 2245 2246 idle=nomwait: Disable mwait for CPU C-states 2247 2248 [ARM64,EARLY] 2249 Format: idle=wfi, idle=yield, idle=nop 2250 2251 idle=wfi: Use the WFI (Wait For Interrupt) hint 2252 instruction in the idle loop. This is the default and 2253 allows the CPU to enter a low-power state until an 2254 interrupt arrives. 2255 2256 idle=yield: Use the YIELD hint instruction instead of 2257 WFI. CPUs supporting simultaneous multi-threading (SMT), 2258 can continue executing another thread when the current 2259 thread reaches the idle loop. This will make the CPUs 2260 eat more power, but may be useful to get slightly better 2261 performance in some applications, since the CPUs will 2262 not enter a low-power state. 2263 2264 idle=nop: Do not execute any idle instruction in the 2265 idle loop. This is useful on platforms where WFI 2266 misbehaves, leading to system instability or loss of CPU 2267 state. This will make the CPUs eat more power, but may 2268 give slightly better performance in some applications, 2269 since the CPUs will not enter a low-power state. 2270 2271 idxd.sva= [HW] 2272 Format: <bool> 2273 Allow force disabling of Shared Virtual Memory (SVA) 2274 support for the idxd driver. By default it is set to 2275 true (1). 2276 2277 idxd.tc_override= [HW] 2278 Format: <bool> 2279 Allow override of default traffic class configuration 2280 for the device. By default it is set to false (0). 2281 2282 ieee754= [MIPS] Select IEEE Std 754 conformance mode 2283 Format: { strict | legacy | 2008 | relaxed | emulated } 2284 Default: strict 2285 2286 Choose which programs will be accepted for execution 2287 based on the IEEE 754 NaN encoding(s) supported by 2288 the FPU and the NaN encoding requested with the value 2289 of an ELF file header flag individually set by each 2290 binary. Hardware implementations are permitted to 2291 support either or both of the legacy and the 2008 NaN 2292 encoding mode. 2293 2294 Available settings are as follows: 2295 strict accept binaries that request a NaN encoding 2296 supported by the FPU 2297 legacy only accept legacy-NaN binaries, if supported 2298 by the FPU 2299 2008 only accept 2008-NaN binaries, if supported 2300 by the FPU 2301 relaxed accept any binaries regardless of whether 2302 supported by the FPU 2303 emulated accept any binaries but enable FPU emulator 2304 if binary mode is unsupported by the FPU. 2305 2306 The FPU emulator is always able to support both NaN 2307 encodings, so if no FPU hardware is present or it has 2308 been disabled with 'nofpu', then the settings of 2309 'legacy' and '2008' strap the emulator accordingly, 2310 'relaxed' straps the emulator for both legacy-NaN and 2311 2008-NaN, whereas 'strict' enables legacy-NaN only on 2312 legacy processors and both NaN encodings on MIPS32 or 2313 MIPS64 CPUs. 2314 2315 The setting for ABS.fmt/NEG.fmt instruction execution 2316 mode generally follows that for the NaN encoding, 2317 except where unsupported by hardware. 2318 2319 ignore_loglevel [KNL,EARLY] 2320 Ignore loglevel setting - this will print /all/ 2321 kernel messages to the console. Useful for debugging. 2322 We also add it as printk module parameter, so users 2323 could change it dynamically, usually by 2324 /sys/module/printk/parameters/ignore_loglevel. 2325 2326 ignore_rlimit_data 2327 Ignore RLIMIT_DATA setting for data mappings, 2328 print warning at first misuse. Can be changed via 2329 /sys/module/kernel/parameters/ignore_rlimit_data. 2330 2331 ihash_entries= [KNL] 2332 Set number of hash buckets for inode cache. 2333 2334 ima_appraise= [IMA] appraise integrity measurements 2335 Format: { "off" | "enforce" | "fix" | "log" } 2336 default: "enforce" 2337 2338 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 2339 The builtin appraise policy appraises all files 2340 owned by uid=0. 2341 2342 ima_canonical_fmt [IMA] 2343 Use the canonical format for the binary runtime 2344 measurements, instead of host native format. 2345 2346 ima_flush_htable [IMA] 2347 Flush the IMA hash table when deleting all the 2348 staged measurement records, to achieve maximum 2349 memory saving at the cost of having duplicate 2350 records across the staged measurement lists. 2351 2352 ima_hash= [IMA] 2353 Format: { md5 | sha1 | rmd160 | sha256 | sha384 2354 | sha512 | ... } 2355 default: "sha1" 2356 2357 The list of supported hash algorithms is defined 2358 in crypto/hash_info.h. 2359 2360 ima_policy= [IMA] 2361 The builtin policies to load during IMA setup. 2362 Format: "tcb | appraise_tcb | secure_boot | 2363 fail_securely | critical_data" 2364 2365 The "tcb" policy measures all programs exec'd, files 2366 mmap'd for exec, and all files opened with the read 2367 mode bit set by either the effective uid (euid=0) or 2368 uid=0. 2369 2370 The "appraise_tcb" policy appraises the integrity of 2371 all files owned by root. 2372 2373 The "secure_boot" policy appraises the integrity 2374 of files (eg. kexec kernel image, kernel modules, 2375 firmware, policy, etc) based on file signatures. 2376 2377 The "fail_securely" policy forces file signature 2378 verification failure also on privileged mounted 2379 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 2380 flag. 2381 2382 The "critical_data" policy measures kernel integrity 2383 critical data. 2384 2385 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 2386 Load a policy which meets the needs of the Trusted 2387 Computing Base. This means IMA will measure all 2388 programs exec'd, files mmap'd for exec, and all files 2389 opened for read by uid=0. 2390 2391 ima_template= [IMA] 2392 Select one of defined IMA measurements template formats. 2393 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" | 2394 "ima-sigv2" } 2395 Default: "ima-ng" 2396 2397 ima_template_fmt= 2398 [IMA] Define a custom template format. 2399 Format: { "field1|...|fieldN" } 2400 2401 ima= [IMA] Enable or disable IMA 2402 Format: { "off" | "on" } 2403 Default: "on" 2404 Note that disabling IMA is limited to kdump kernel. 2405 2406 indirect_target_selection= [X86,Intel] Mitigation control for Indirect 2407 Target Selection(ITS) bug in Intel CPUs. Updated 2408 microcode is also required for a fix in IBPB. 2409 2410 on: Enable mitigation (default). 2411 off: Disable mitigation. 2412 force: Force the ITS bug and deploy default 2413 mitigation. 2414 vmexit: Only deploy mitigation if CPU is affected by 2415 guest/host isolation part of ITS. 2416 stuff: Deploy RSB-fill mitigation when retpoline is 2417 also deployed. Otherwise, deploy the default 2418 mitigation. 2419 2420 For details see: 2421 Documentation/admin-guide/hw-vuln/indirect-target-selection.rst 2422 2423 init= [KNL] 2424 Format: <full_path> 2425 Run specified binary instead of /sbin/init as init 2426 process. 2427 2428 initcall_debug [KNL] Trace initcalls as they are executed. Useful 2429 for working out where the kernel is dying during 2430 startup. 2431 2432 initcall_blacklist= [KNL] Do not execute a comma-separated list of 2433 initcall functions. Useful for debugging built-in 2434 modules and initcalls. 2435 2436 initramfs_async= [KNL] 2437 Format: <bool> 2438 Default: 1 2439 This parameter controls whether the initramfs 2440 image is unpacked asynchronously, concurrently 2441 with devices being probed and 2442 initialized. This should normally just work, 2443 but as a debugging aid, one can get the 2444 historical behaviour of the initramfs 2445 unpacking being completed before device_ and 2446 late_ initcalls. 2447 2448 initrd= [BOOT,EARLY] Specify the location of the initial ramdisk 2449 2450 initrdmem= [KNL,EARLY] Specify a physical address and size from which to 2451 load the initrd. If an initrd is compiled in or 2452 specified in the bootparams, it takes priority over this 2453 setting. 2454 Format: ss[KMG],nn[KMG] 2455 Default is 0, 0 2456 2457 init_on_alloc= [MM,EARLY] Fill newly allocated pages and heap objects with 2458 zeroes. 2459 Format: 0 | 1 2460 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 2461 2462 init_on_free= [MM,EARLY] Fill freed pages and heap objects with zeroes. 2463 Format: 0 | 1 2464 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 2465 2466 init_pkru= [X86] Specify the default memory protection keys rights 2467 register contents for all processes. 0x55555554 by 2468 default (disallow access to all but pkey 0). Can 2469 override in debugfs after boot. 2470 2471 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 2472 Format: <irq> 2473 2474 int_pln_enable [X86] Enable power limit notification interrupt 2475 2476 integrity_audit=[IMA] 2477 Format: { "0" | "1" } 2478 0 -- basic integrity auditing messages. (Default) 2479 1 -- additional integrity auditing messages. 2480 2481 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 2482 on 2483 Enable intel iommu driver. 2484 off 2485 Disable intel iommu driver. 2486 igfx_off [Default Off] 2487 By default, gfx is mapped as normal device. If a gfx 2488 device has a dedicated DMAR unit, the DMAR unit is 2489 bypassed by not enabling DMAR with this option. In 2490 this case, gfx device will use physical address for 2491 DMA. 2492 strict [Default Off] 2493 Deprecated, equivalent to iommu.strict=1. 2494 sp_off [Default Off] 2495 By default, super page will be supported if Intel IOMMU 2496 has the capability. With this option, super page will 2497 not be supported. 2498 sm_on 2499 Enable the Intel IOMMU scalable mode if the hardware 2500 advertises that it has support for the scalable mode 2501 translation. 2502 sm_off 2503 Disallow use of the Intel IOMMU scalable mode. 2504 tboot_noforce [Default Off] 2505 Do not force the Intel IOMMU enabled under tboot. 2506 By default, tboot will force Intel IOMMU on, which 2507 could harm performance of some high-throughput 2508 devices like 40GBit network cards, even if identity 2509 mapping is enabled. 2510 Note that using this option lowers the security 2511 provided by tboot because it makes the system 2512 vulnerable to DMA attacks. 2513 2514 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 2515 0 disables intel_idle and fall back on acpi_idle. 2516 1 to 9 specify maximum depth of C-state. 2517 2518 intel_pstate= [X86,EARLY] 2519 disable 2520 Do not enable intel_pstate as the default 2521 scaling driver for the supported processors 2522 active 2523 Use intel_pstate driver to bypass the scaling 2524 governors layer of cpufreq and provides it own 2525 algorithms for p-state selection. There are two 2526 P-state selection algorithms provided by 2527 intel_pstate in the active mode: powersave and 2528 performance. The way they both operate depends 2529 on whether or not the hardware managed P-states 2530 (HWP) feature has been enabled in the processor 2531 and possibly on the processor model. 2532 passive 2533 Use intel_pstate as a scaling driver, but configure it 2534 to work with generic cpufreq governors (instead of 2535 enabling its internal governor). This mode cannot be 2536 used along with the hardware-managed P-states (HWP) 2537 feature. 2538 force 2539 Enable intel_pstate on systems that prohibit it by default 2540 in favor of acpi-cpufreq. Forcing the intel_pstate driver 2541 instead of acpi-cpufreq may disable platform features, such 2542 as thermal controls and power capping, that rely on ACPI 2543 P-States information being indicated to OSPM and therefore 2544 should be used with caution. This option does not work with 2545 processors that aren't supported by the intel_pstate driver 2546 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 2547 no_hwp 2548 Do not enable hardware P state control (HWP) 2549 if available. 2550 hwp_only 2551 Only load intel_pstate on systems which support 2552 hardware P state control (HWP) if available. 2553 support_acpi_ppc 2554 Enforce ACPI _PPC performance limits. If the Fixed ACPI 2555 Description Table, specifies preferred power management 2556 profile as "Enterprise Server" or "Performance Server", 2557 then this feature is turned on by default. 2558 per_cpu_perf_limits 2559 Allow per-logical-CPU P-State performance control limits using 2560 cpufreq sysfs interface 2561 no_cas 2562 Do not enable capacity-aware scheduling (CAS) on 2563 hybrid systems 2564 2565 intremap= [X86-64,Intel-IOMMU,EARLY] 2566 on enable Interrupt Remapping (default) 2567 off disable Interrupt Remapping 2568 nosid disable Source ID checking 2569 no_x2apic_optout 2570 BIOS x2APIC opt-out request will be ignored 2571 nopost disable Interrupt Posting 2572 posted_msi 2573 enable MSIs delivered as posted interrupts 2574 2575 iomem= Disable strict checking of access to MMIO memory 2576 strict regions from userspace. 2577 relaxed 2578 2579 iommu= [X86,EARLY] 2580 2581 off 2582 Don't initialize and use any kind of IOMMU. 2583 2584 force 2585 Force the use of the hardware IOMMU even when 2586 it is not actually needed (e.g. because < 3 GB 2587 memory). 2588 2589 noforce 2590 Don't force hardware IOMMU usage when it is not 2591 needed. (default). 2592 2593 merge 2594 Do scatter-gather (SG) merging. Implies "force" 2595 (experimental). 2596 2597 nomerge 2598 Don't do scatter-gather (SG) merging. 2599 2600 biomerge 2601 Do scatter-gather (SG) merging. Implies "force" 2602 (experimental). [same as "merge"] 2603 2604 panic 2605 Always panic when IOMMU overflows. 2606 2607 nopanic 2608 Don't panic on IOMMU overflows. 2609 2610 pt 2611 Use passththrough mode by default 2612 (Equivalent to iommu.passthrough=1) 2613 2614 nopt 2615 Use translated mode for DMA by default 2616 (Equivalent to iommu.passthrough=0) 2617 2618 soft 2619 Use software bounce buffering (SWIOTLB) (default for 2620 Intel machines). This can be used to prevent the usage 2621 of an available hardware IOMMU. 2622 2623 usedac 2624 Use the DAC on VIA PCI bridge 2625 (default: disable the VIA PCI bridge DAC) 2626 2627 AMD Gart HW IOMMU-specific options: (CONFIG_GART_IOMMU) 2628 2629 <size> 2630 Set the size of the remapping area in bytes. 2631 2632 allowed 2633 Overwrite iommu off workarounds for specific chipsets 2634 2635 force 2636 Overwrite iommu off workarounds for specific chipsets 2637 2638 fullflush 2639 Flush IOMMU on each allocation (default). 2640 2641 nofullflush 2642 Don't use IOMMU fullflush. 2643 2644 memaper[=<order>] 2645 Allocate an own aperture over RAM with size 2646 32MB<<order. (default: order=1, i.e. 64MB) 2647 2648 noaperture 2649 Ask the IOMMU not to touch the aperture for AGP. 2650 2651 noagp 2652 Don't initialize the AGP driver and use full aperture. 2653 2654 iommu= [PPC/POWERNV] 2655 2656 nobypass 2657 Disable IOMMU bypass, using IOMMU for PCI devices. 2658 2659 iommu.forcedac= [ARM64,X86,EARLY] Control IOVA allocation for PCI devices. 2660 Format: { "0" | "1" } 2661 0 - Try to allocate a 32-bit DMA address first, before 2662 falling back to the full range if needed. 2663 1 - Allocate directly from the full usable range, 2664 forcing Dual Address Cycle for PCI cards supporting 2665 greater than 32-bit addressing. 2666 2667 iommu.strict= [ARM64,X86,S390,EARLY] Configure TLB invalidation behaviour 2668 Format: { "0" | "1" } 2669 0 - Lazy mode. 2670 Request that DMA unmap operations use deferred 2671 invalidation of hardware TLBs, for increased 2672 throughput at the cost of reduced device isolation. 2673 Will fall back to strict mode if not supported by 2674 the relevant IOMMU driver. 2675 1 - Strict mode. 2676 DMA unmap operations invalidate IOMMU hardware TLBs 2677 synchronously. 2678 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}. 2679 Note: on x86, strict mode specified via one of the 2680 legacy driver-specific options takes precedence. 2681 2682 iommu.passthrough= 2683 [ARM64,X86,EARLY] Configure DMA to bypass the IOMMU by default. 2684 Format: { "0" | "1" } 2685 0 - Use IOMMU translation for DMA. 2686 1 - Bypass the IOMMU for DMA. 2687 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2688 2689 iommu.debug_pagealloc= 2690 [KNL,EARLY] When CONFIG_IOMMU_DEBUG_PAGEALLOC is set, this 2691 parameter enables the feature at boot time. By default, it 2692 is disabled and the system behaves the same way as a kernel 2693 built without CONFIG_IOMMU_DEBUG_PAGEALLOC. 2694 Format: { "0" | "1" } 2695 0 - Sanitizer disabled. 2696 1 - Sanitizer enabled, expect runtime overhead. 2697 2698 io7= [HW] IO7 for Marvel-based Alpha systems 2699 See comment before marvel_specify_io7 in 2700 arch/alpha/kernel/core_marvel.c. 2701 2702 io_delay= [X86,EARLY] I/O delay method 2703 0x80 2704 Standard port 0x80 based delay 2705 0xed 2706 Alternate port 0xed based delay (needed on some systems) 2707 udelay 2708 Simple two microseconds delay 2709 none 2710 No delay 2711 2712 ip= [IP_PNP] 2713 See Documentation/admin-guide/nfs/nfsroot.rst. 2714 2715 ipcmni_extend [KNL,EARLY] Extend the maximum number of unique System V 2716 IPC identifiers from 32,768 to 16,777,216. 2717 2718 ipe.enforce= [IPE] 2719 Format: <bool> 2720 Determine whether IPE starts in permissive (0) or 2721 enforce (1) mode. The default is enforce. 2722 2723 ipe.success_audit= 2724 [IPE] 2725 Format: <bool> 2726 Start IPE with success auditing enabled, emitting 2727 an audit event when a binary is allowed. The default 2728 is 0. 2729 2730 irqaffinity= [SMP] Set the default irq affinity mask 2731 The argument is a cpu list, as described above. 2732 2733 irqchip.gicv2_force_probe= 2734 [ARM,ARM64,EARLY] 2735 Format: <bool> 2736 Force the kernel to look for the second 4kB page 2737 of a GICv2 controller even if the memory range 2738 exposed by the device tree is too small. 2739 2740 irqchip.gicv3_nolpi= 2741 [ARM,ARM64,EARLY] 2742 Force the kernel to ignore the availability of 2743 LPIs (and by consequence ITSs). Intended for system 2744 that use the kernel as a bootloader, and thus want 2745 to let secondary kernels in charge of setting up 2746 LPIs. 2747 2748 irqchip.gicv3_pseudo_nmi= [ARM64,EARLY] 2749 Enables support for pseudo-NMIs in the kernel. This 2750 requires the kernel to be built with 2751 CONFIG_ARM64_PSEUDO_NMI. 2752 2753 irqchip.riscv_imsic_noipi 2754 [RISC-V,EARLY] 2755 Force the kernel to not use IMSIC software injected MSIs 2756 as IPIs. Intended for system where IMSIC is trap-n-emulated, 2757 and thus want to reduce MMIO traps when triggering IPIs 2758 to multiple harts. 2759 2760 irqfixup [HW] 2761 When an interrupt is not handled search all handlers 2762 for it. Intended to get systems with badly broken 2763 firmware running. 2764 2765 irqhandler.duration_warn_us= [KNL] 2766 Warn if an IRQ handler exceeds the specified duration 2767 threshold in microseconds. Useful for identifying 2768 long-running IRQs in the system. 2769 2770 irqpoll [HW] 2771 When an interrupt is not handled search all handlers 2772 for it. Also check all handlers each timer 2773 interrupt. Intended to get systems with badly broken 2774 firmware running. 2775 2776 isapnp= [ISAPNP] 2777 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2778 2779 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2780 Format: [flag-list,]<cpu-list> 2781 2782 Specify one or more CPUs to isolate from disturbances 2783 specified in the flag list (default: domain): 2784 2785 nohz 2786 Disable the tick when a single task runs as well as 2787 disabling other kernel noises like having RCU callbacks 2788 offloaded. This is equivalent to the nohz_full parameter. 2789 2790 A residual 1Hz tick is offloaded to workqueues, which you 2791 need to affine to housekeeping through the global 2792 workqueue's affinity configured via the 2793 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2794 by using the 'domain' flag described below. 2795 2796 NOTE: by default the global workqueue runs on all CPUs, 2797 so to protect individual CPUs the 'cpumask' file has to 2798 be configured manually after bootup. 2799 2800 domain 2801 Isolate from the general SMP balancing and scheduling 2802 algorithms. Note that performing domain isolation this way 2803 is irreversible: it's not possible to bring back a CPU to 2804 the domains once isolated through this boot time 2805 configuration. Use cpusets for a dynamic configuration 2806 which can be altered at runtime. For details see 2807 Documentation/admin-guide/cpu-isolation.rst. 2808 2809 You can move a process onto or off an "isolated" CPU via 2810 the CPU affinity syscalls or cpuset. 2811 <cpu number> begins at 0 and the maximum value is 2812 "number of CPUs in system - 1". 2813 2814 managed_irq 2815 2816 Isolate from being targeted by managed interrupts 2817 which have an interrupt mask containing isolated 2818 CPUs. The affinity of managed interrupts is 2819 handled by the kernel and cannot be changed via 2820 the /proc/irq/* interfaces. 2821 2822 This isolation is best effort and only effective 2823 if the automatically assigned interrupt mask of a 2824 device queue contains isolated and housekeeping 2825 CPUs. If housekeeping CPUs are online then such 2826 interrupts are directed to the housekeeping CPU 2827 so that IO submitted on the housekeeping CPU 2828 cannot disturb the isolated CPU. 2829 2830 If a queue's affinity mask contains only isolated 2831 CPUs then this parameter has no effect on the 2832 interrupt routing decision, though interrupts are 2833 only delivered when tasks running on those 2834 isolated CPUs submit IO. IO submitted on 2835 housekeeping CPUs has no influence on those 2836 queues. 2837 2838 The format of <cpu-list> is described above. 2839 2840 iucv= [HW,NET] 2841 2842 ivrs_ioapic [HW,X86-64] 2843 Provide an override to the IOAPIC-ID<->DEVICE-ID 2844 mapping provided in the IVRS ACPI table. 2845 By default, PCI segment is 0, and can be omitted. 2846 2847 For example, to map IOAPIC-ID decimal 10 to 2848 PCI segment 0x1 and PCI device 00:14.0, 2849 write the parameter as: 2850 ivrs_ioapic=10@0001:00:14.0 2851 2852 Deprecated formats: 2853 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0 2854 write the parameter as: 2855 ivrs_ioapic[10]=00:14.0 2856 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and 2857 PCI device 00:14.0 write the parameter as: 2858 ivrs_ioapic[10]=0001:00:14.0 2859 2860 ivrs_hpet [HW,X86-64] 2861 Provide an override to the HPET-ID<->DEVICE-ID 2862 mapping provided in the IVRS ACPI table. 2863 By default, PCI segment is 0, and can be omitted. 2864 2865 For example, to map HPET-ID decimal 10 to 2866 PCI segment 0x1 and PCI device 00:14.0, 2867 write the parameter as: 2868 ivrs_hpet=10@0001:00:14.0 2869 2870 Deprecated formats: 2871 * To map HPET-ID decimal 0 to PCI device 00:14.0 2872 write the parameter as: 2873 ivrs_hpet[0]=00:14.0 2874 * To map HPET-ID decimal 10 to PCI segment 0x1 and 2875 PCI device 00:14.0 write the parameter as: 2876 ivrs_ioapic[10]=0001:00:14.0 2877 2878 ivrs_acpihid [HW,X86-64] 2879 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2880 mapping provided in the IVRS ACPI table. 2881 By default, PCI segment is 0, and can be omitted. 2882 2883 For example, to map UART-HID:UID AMD0020:0 to 2884 PCI segment 0x1 and PCI device ID 00:14.5, 2885 write the parameter as: 2886 ivrs_acpihid=AMD0020:0@0001:00:14.5 2887 2888 Deprecated formats: 2889 * To map UART-HID:UID AMD0020:0 to PCI segment is 0, 2890 PCI device ID 00:14.5, write the parameter as: 2891 ivrs_acpihid[00:14.5]=AMD0020:0 2892 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and 2893 PCI device ID 00:14.5, write the parameter as: 2894 ivrs_acpihid[0001:00:14.5]=AMD0020:0 2895 2896 js= [HW,JOY] Analog joystick 2897 See Documentation/input/joydev/joystick.rst. 2898 2899 kasan_multi_shot 2900 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2901 report on every invalid memory access. Without this 2902 parameter KASAN will print report only for the first 2903 invalid access. 2904 2905 keep_bootcon [KNL,EARLY] 2906 Do not unregister boot console at start. This is only 2907 useful for debugging when something happens in the window 2908 between unregistering the boot console and initializing 2909 the real console. 2910 2911 keepinitrd [HW,ARM] See retain_initrd. 2912 2913 kernelcore= [KNL,X86,PPC,EARLY] 2914 Format: nn[KMGTPE] | nn% | "mirror" 2915 This parameter specifies the amount of memory usable by 2916 the kernel for non-movable allocations. The requested 2917 amount is spread evenly throughout all nodes in the 2918 system as ZONE_NORMAL. The remaining memory is used for 2919 movable memory in its own zone, ZONE_MOVABLE. In the 2920 event, a node is too small to have both ZONE_NORMAL and 2921 ZONE_MOVABLE, kernelcore memory will take priority and 2922 other nodes will have a larger ZONE_MOVABLE. 2923 2924 ZONE_MOVABLE is used for the allocation of pages that 2925 may be reclaimed or moved by the page migration 2926 subsystem. Note that allocations like PTEs-from-HighMem 2927 still use the HighMem zone if it exists, and the Normal 2928 zone if it does not. 2929 2930 It is possible to specify the exact amount of memory in 2931 the form of "nn[KMGTPE]", a percentage of total system 2932 memory in the form of "nn%", or "mirror". If "mirror" 2933 option is specified, mirrored (reliable) memory is used 2934 for non-movable allocations and remaining memory is used 2935 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2936 are exclusive, so you cannot specify multiple forms. 2937 2938 kfence.burst= [MM,KFENCE] The number of additional successive 2939 allocations to be attempted through KFENCE for each 2940 sample interval. 2941 Format: <unsigned integer> 2942 Default: 0 2943 2944 kfence.check_on_panic= 2945 [MM,KFENCE] Whether to check all KFENCE-managed objects' 2946 canaries on panic. 2947 Format: <bool> 2948 Default: false 2949 2950 kfence.deferrable= 2951 [MM,KFENCE] Whether to use a deferrable timer to trigger 2952 allocations. This avoids forcing CPU wake-ups if the 2953 system is idle, at the risk of a less predictable 2954 sample interval. 2955 Format: <bool> 2956 Default: CONFIG_KFENCE_DEFERRABLE 2957 2958 kfence.fault= [MM,KFENCE] Controls the behavior when a KFENCE 2959 error is detected. 2960 report - print the error report and continue (default). 2961 oops - print the error report and oops. 2962 panic - print the error report and panic. 2963 2964 kfence.sample_interval= 2965 [MM,KFENCE] KFENCE's sample interval in milliseconds. 2966 Format: <unsigned integer> 2967 0 - Disable KFENCE. 2968 >0 - Enabled KFENCE with given sample interval. 2969 Default: CONFIG_KFENCE_SAMPLE_INTERVAL 2970 2971 kfence.skip_covered_thresh= 2972 [MM,KFENCE] If pool utilization reaches this threshold 2973 (pool usage%), KFENCE limits currently covered 2974 allocations of the same source from further filling 2975 up the pool. 2976 Format: <unsigned integer> 2977 Default: 75 2978 2979 kgdbdbgp= [KGDB,HW,EARLY] kgdb over EHCI usb debug port. 2980 Format: <Controller#>[,poll interval] 2981 The controller # is the number of the ehci usb debug 2982 port as it is probed via PCI. The poll interval is 2983 optional and is the number seconds in between 2984 each poll cycle to the debug port in case you need 2985 the functionality for interrupting the kernel with 2986 gdb or control-c on the dbgp connection. When 2987 not using this parameter you use sysrq-g to break into 2988 the kernel debugger. 2989 2990 kgdboc= [KGDB,HW] kgdb over consoles. 2991 Requires a tty driver that supports console polling, 2992 or a supported polling keyboard driver (non-usb). 2993 Serial only format: <serial_device>[,baud] 2994 keyboard only format: kbd 2995 keyboard and serial format: kbd,<serial_device>[,baud] 2996 Optional Kernel mode setting: 2997 kms, kbd format: kms,kbd 2998 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2999 3000 kgdboc_earlycon= [KGDB,HW,EARLY] 3001 If the boot console provides the ability to read 3002 characters and can work in polling mode, you can use 3003 this parameter to tell kgdb to use it as a backend 3004 until the normal console is registered. Intended to 3005 be used together with the kgdboc parameter which 3006 specifies the normal console to transition to. 3007 3008 The name of the early console should be specified 3009 as the value of this parameter. Note that the name of 3010 the early console might be different than the tty 3011 name passed to kgdboc. It's OK to leave the value 3012 blank and the first boot console that implements 3013 read() will be picked. 3014 3015 kgdbwait [KGDB,EARLY] Stop kernel execution and enter the 3016 kernel debugger at the earliest opportunity. 3017 3018 kho= [KEXEC,EARLY] 3019 Format: { "0" | "1" | "off" | "on" | "y" | "n" } 3020 Enables or disables Kexec HandOver. 3021 "0" | "off" | "n" - kexec handover is disabled 3022 "1" | "on" | "y" - kexec handover is enabled 3023 3024 kho_scratch= [KEXEC,EARLY] 3025 Format: ll[KMG],mm[KMG],nn[KMG] | nn% 3026 Defines the size of the KHO scratch region. The KHO 3027 scratch regions are physically contiguous memory 3028 ranges that can only be used for non-kernel 3029 allocations. That way, even when memory is heavily 3030 fragmented with handed over memory, the kexeced 3031 kernel will always have enough contiguous ranges to 3032 bootstrap itself. 3033 3034 It is possible to specify the exact amount of 3035 memory in the form of "ll[KMG],mm[KMG],nn[KMG]" 3036 where the first parameter defines the size of a low 3037 memory scratch area, the second parameter defines 3038 the size of a global scratch area and the third 3039 parameter defines the size of additional per-node 3040 scratch areas. The form "nn%" defines scale factor 3041 (in percents) of memory that was used during boot. 3042 3043 kmac= [MIPS] Korina ethernet MAC address. 3044 Configure the RouterBoard 532 series on-chip 3045 Ethernet adapter MAC address. 3046 3047 kmemleak= [KNL,EARLY] Boot-time kmemleak enable/disable 3048 Valid arguments: on, off 3049 Default: on 3050 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 3051 the default is off. 3052 3053 kprobe_event=[probe-list] 3054 [FTRACE] Add kprobe events and enable at boot time. 3055 The probe-list is a semicolon delimited list of probe 3056 definitions. Each definition is same as kprobe_events 3057 interface, but the parameters are comma delimited. 3058 For example, to add a kprobe event on vfs_read with 3059 arg1 and arg2, add to the command line; 3060 3061 kprobe_event=p,vfs_read,$arg1,$arg2 3062 3063 See also Documentation/trace/kprobetrace.rst "Kernel 3064 Boot Parameter" section. 3065 3066 kpti= [ARM64,EARLY] Control page table isolation of 3067 user and kernel address spaces. 3068 Default: enabled on cores which need mitigation. 3069 0: force disabled 3070 1: force enabled 3071 3072 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires 3073 CONFIG_KUNIT to be set to be fully enabled. The 3074 default value can be overridden via 3075 KUNIT_DEFAULT_ENABLED. 3076 Default is 1 (enabled) 3077 3078 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 3079 Default is 0 (don't ignore, but inject #GP) 3080 3081 kvm.eager_page_split= 3082 [KVM,X86,RISCV] Controls whether or not KVM will try to 3083 proactively split all huge pages during dirty logging. 3084 Eager page splitting reduces interruptions to vCPU 3085 execution by eliminating the write-protection faults 3086 and MMU lock contention that would otherwise be 3087 required to split huge pages lazily. 3088 3089 VM workloads that rarely perform writes or that write 3090 only to a small region of VM memory may benefit from 3091 disabling eager page splitting to allow huge pages to 3092 still be used for reads. 3093 3094 The behavior of eager page splitting depends on whether 3095 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If 3096 disabled, all huge pages in a memslot will be eagerly 3097 split when dirty logging is enabled on that memslot. If 3098 enabled, eager page splitting will be performed during 3099 the KVM_CLEAR_DIRTY ioctl, and only for the pages being 3100 cleared. 3101 3102 On x86, eager page splitting is only supported when 3103 kvm.tdp_mmu=Y. 3104 3105 On RISCV, eager page splitting is supported by default. 3106 3107 Default is Y (on). 3108 3109 kvm.enable_pmu=[KVM,X86] 3110 If enabled, KVM will virtualize PMU functionality based 3111 on the virtual CPU model defined by userspace. This 3112 can be overridden on a per-VM basis via 3113 KVM_CAP_PMU_CAPABILITY. 3114 3115 If disabled, KVM will not virtualize PMU functionality, 3116 e.g. MSRs, PMCs, PMIs, etc., even if userspace defines 3117 a virtual CPU model that contains PMU assets. 3118 3119 Note, KVM's vPMU support implicitly requires running 3120 with an in-kernel local APIC, e.g. to deliver PMIs to 3121 the guest. Running without an in-kernel local APIC is 3122 not supported, though KVM will allow such a combination 3123 (with severely degraded functionality). 3124 3125 See also enable_mediated_pmu. 3126 3127 Default is Y (on). 3128 3129 kvm.enable_virt_at_load=[KVM,ARM64,LOONGARCH,MIPS,RISCV,X86] 3130 If enabled, KVM will enable virtualization in hardware 3131 when KVM is loaded, and disable virtualization when KVM 3132 is unloaded (if KVM is built as a module). 3133 3134 If disabled, KVM will dynamically enable and disable 3135 virtualization on-demand when creating and destroying 3136 VMs, i.e. on the 0=>1 and 1=>0 transitions of the 3137 number of VMs. 3138 3139 Enabling virtualization at module load avoids potential 3140 latency for creation of the 0=>1 VM, as KVM serializes 3141 virtualization enabling across all online CPUs. The 3142 "cost" of enabling virtualization when KVM is loaded, 3143 is that doing so may interfere with using out-of-tree 3144 hypervisors that want to "own" virtualization hardware. 3145 3146 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 3147 Default is false (don't support). 3148 3149 kvm.gmem_in_place_conversion= 3150 [KVM] Controls whether KVM enables in-place conversion 3151 support for guest_memfd and tracks the private/shared 3152 state of memory per guest_memfd instead of per VM. 3153 3154 If enabled, KVM enables the KVM_SET_MEMORY_ATTRIBUTES2 3155 ioctl on guest_memfd file descriptors and disables the 3156 legacy VM-scoped KVM_SET_MEMORY_ATTRIBUTES ioctl for 3157 private memory state tracking. Only the 3158 KVM_MEMORY_ATTRIBUTE_PRIVATE attribute moves to 3159 per-guest_memfd tracking; other attributes remain 3160 per-VM. 3161 3162 This parameter toggles KVM's in-place conversion 3163 capability support. Whether a VMM uses separate backends 3164 or out-of-place memory management is determined by 3165 userspace VMM design. 3166 3167 Note, this parameter is only available when 3168 CONFIG_KVM_PRIVATE_MEMATTR_VM=y. When 3169 CONFIG_KVM_PRIVATE_MEMATTR_VM is not set, in-place 3170 conversion is enabled (and cannot be disabled) if KVM 3171 is built with private memory support. 3172 3173 Default is N (off). 3174 3175 kvm.nx_huge_pages= 3176 [KVM] Controls the software workaround for the 3177 X86_BUG_ITLB_MULTIHIT bug. 3178 force : Always deploy workaround. 3179 off : Never deploy workaround. 3180 auto : Deploy workaround based on the presence of 3181 X86_BUG_ITLB_MULTIHIT. 3182 3183 Default is 'auto'. 3184 3185 If the software workaround is enabled for the host, 3186 guests do need not to enable it for nested guests. 3187 3188 kvm.nx_huge_pages_recovery_ratio= 3189 [KVM] Controls how many 4KiB pages are periodically zapped 3190 back to huge pages. 0 disables the recovery, otherwise if 3191 the value is N KVM will zap 1/Nth of the 4KiB pages every 3192 period (see below). The default is 60. 3193 3194 kvm.nx_huge_pages_recovery_period_ms= 3195 [KVM] Controls the time period at which KVM zaps 4KiB pages 3196 back to huge pages. If the value is a non-zero N, KVM will 3197 zap a portion (see ratio above) of the pages every N msecs. 3198 If the value is 0 (the default), KVM will pick a period based 3199 on the ratio, such that a page is zapped after 1 hour on average. 3200 3201 kvm-{amd,intel}.enable_mediated_pmu=[KVM,AMD,INTEL] 3202 If enabled, KVM will provide a mediated virtual PMU, 3203 instead of the default perf-based virtual PMU (if 3204 kvm.enable_pmu is true and PMU is enumerated via the 3205 virtual CPU model). 3206 3207 With a perf-based vPMU, KVM operates as a user of perf, 3208 i.e. emulates guest PMU counters using perf events. 3209 KVM-created perf events are managed by perf as regular 3210 (guest-only) events, e.g. are scheduled in/out, contend 3211 for hardware resources, etc. Using a perf-based vPMU 3212 allows guest and host usage of the PMU to co-exist, but 3213 incurs non-trivial overhead and can result in silently 3214 dropped guest events (due to resource contention). 3215 3216 With a mediated vPMU, hardware PMU state is context 3217 switched around the world switch to/from the guest. 3218 KVM mediates which events the guest can utilize, but 3219 gives the guest direct access to all other PMU assets 3220 when possible (KVM may intercept some accesses if the 3221 virtual CPU model provides a subset of hardware PMU 3222 functionality). Using a mediated vPMU significantly 3223 reduces PMU virtualization overhead and eliminates lost 3224 guest events, but is mutually exclusive with using perf 3225 to profile KVM guests and adds latency to most VM-Exits 3226 (to context switch PMU state). 3227 3228 Default is N (off). 3229 3230 kvm-amd.nested= [KVM,AMD] Control nested virtualization feature in 3231 KVM/SVM. Default is 1 (enabled). 3232 3233 kvm-amd.npt= [KVM,AMD] Control KVM's use of Nested Page Tables, 3234 a.k.a. Two-Dimensional Page Tables. Default is 1 3235 (enabled). Disable by KVM if hardware lacks support 3236 for NPT. 3237 3238 kvm-amd.ciphertext_hiding_asids= 3239 [KVM,AMD] Ciphertext hiding prevents disallowed accesses 3240 to SNP private memory from reading ciphertext. Instead, 3241 reads will see constant default values (0xff). 3242 3243 If ciphertext hiding is enabled, the joint SEV-ES and 3244 SEV-SNP ASID space is partitioned into separate SEV-ES 3245 and SEV-SNP ASID ranges, with the SEV-SNP range being 3246 [1..max_snp_asid] and the SEV-ES range being 3247 (max_snp_asid..min_sev_asid), where min_sev_asid is 3248 enumerated by CPUID.0x.8000_001F[EDX]. 3249 3250 A non-zero value enables SEV-SNP ciphertext hiding and 3251 adjusts the ASID ranges for SEV-ES and SEV-SNP guests. 3252 KVM caps the number of SEV-SNP ASIDs at the maximum 3253 possible value, e.g. specifying -1u will assign all 3254 joint SEV-ES and SEV-SNP ASIDs to SEV-SNP. Note, 3255 assigning all joint ASIDs to SEV-SNP, i.e. configuring 3256 max_snp_asid == min_sev_asid-1, will effectively make 3257 SEV-ES unusable. 3258 3259 kvm-arm.mode= 3260 [KVM,ARM,EARLY] Select one of KVM/arm64's modes of 3261 operation. 3262 3263 none: Forcefully disable KVM. 3264 3265 nvhe: Standard nVHE-based mode, without support for 3266 protected guests. 3267 3268 protected: Mode with support for guests whose state is 3269 kept private from the host, using VHE or 3270 nVHE depending on HW support. 3271 3272 nested: VHE-based mode with support for nested 3273 virtualization. Requires at least ARMv8.4 3274 hardware (with FEAT_NV2). 3275 3276 Defaults to VHE/nVHE based on hardware support. Setting 3277 mode to "protected" will disable kexec and hibernation 3278 for the host. To force nVHE on VHE hardware, add 3279 "arm64_sw.hvhe=0 id_aa64mmfr1.vh=0" to the 3280 command-line. 3281 "nested" and "protected" are experimental and should be 3282 used with extreme caution. 3283 3284 kvm-arm.vgic_v3_group0_trap= 3285 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-0 3286 system registers 3287 3288 kvm-arm.vgic_v3_group1_trap= 3289 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-1 3290 system registers 3291 3292 kvm-arm.vgic_v3_common_trap= 3293 [KVM,ARM,EARLY] Trap guest accesses to GICv3 common 3294 system registers 3295 3296 kvm-arm.vgic_v4_enable= 3297 [KVM,ARM,EARLY] Allow use of GICv4 for direct 3298 injection of LPIs. 3299 3300 kvm-arm.wfe_trap_policy= 3301 [KVM,ARM] Control when to set WFE instruction trap for 3302 KVM VMs. Traps are allowed but not guaranteed by the 3303 CPU architecture. 3304 3305 trap: set WFE instruction trap 3306 3307 notrap: clear WFE instruction trap 3308 3309 kvm-arm.wfi_trap_policy= 3310 [KVM,ARM] Control when to set WFI instruction trap for 3311 KVM VMs. Traps are allowed but not guaranteed by the 3312 CPU architecture. 3313 3314 trap: set WFI instruction trap 3315 3316 notrap: clear WFI instruction trap 3317 3318 kvm_cma_resv_ratio=n [PPC,EARLY] 3319 Reserves given percentage from system memory area for 3320 contiguous memory allocation for KVM hash pagetable 3321 allocation. 3322 By default it reserves 5% of total system memory. 3323 Format: <integer> 3324 Default: 5 3325 3326 kvm-intel.ept= [KVM,Intel] Control KVM's use of Extended Page Tables, 3327 a.k.a. Two-Dimensional Page Tables. Default is 1 3328 (enabled). Disable by KVM if hardware lacks support 3329 for EPT. 3330 3331 kvm-intel.emulate_invalid_guest_state= 3332 [KVM,Intel] Control whether to emulate invalid guest 3333 state. Ignored if kvm-intel.enable_unrestricted_guest=1, 3334 as guest state is never invalid for unrestricted 3335 guests. This param doesn't apply to nested guests (L2), 3336 as KVM never emulates invalid L2 guest state. 3337 Default is 1 (enabled). 3338 3339 kvm-intel.flexpriority= 3340 [KVM,Intel] Control KVM's use of FlexPriority feature 3341 (TPR shadow). Default is 1 (enabled). Disable by KVM if 3342 hardware lacks support for it. 3343 3344 kvm-intel.nested= 3345 [KVM,Intel] Control nested virtualization feature in 3346 KVM/VMX. Default is 1 (enabled). 3347 3348 kvm-intel.unrestricted_guest= 3349 [KVM,Intel] Control KVM's use of unrestricted guest 3350 feature (virtualized real and unpaged mode). Default 3351 is 1 (enabled). Disable by KVM if EPT is disabled or 3352 hardware lacks support for it. 3353 3354 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 3355 CVE-2018-3620. 3356 3357 Valid arguments: never, cond, always 3358 3359 always: L1D cache flush on every VMENTER. 3360 cond: Flush L1D on VMENTER only when the code between 3361 VMEXIT and VMENTER can leak host memory. 3362 never: Disables the mitigation 3363 3364 Default is cond (do L1 cache flush in specific instances) 3365 3366 kvm-intel.vpid= [KVM,Intel] Control KVM's use of Virtual Processor 3367 Identification feature (tagged TLBs). Default is 1 3368 (enabled). Disable by KVM if hardware lacks support 3369 for it. 3370 3371 l1d_flush= [X86,INTEL,EARLY] 3372 Control mitigation for L1D based snooping vulnerability. 3373 3374 Certain CPUs are vulnerable to an exploit against CPU 3375 internal buffers which can forward information to a 3376 disclosure gadget under certain conditions. 3377 3378 In vulnerable processors, the speculatively 3379 forwarded data can be used in a cache side channel 3380 attack, to access data to which the attacker does 3381 not have direct access. 3382 3383 This parameter controls the mitigation. The 3384 options are: 3385 3386 on - enable the interface for the mitigation 3387 3388 l1tf= [X86,EARLY] Control mitigation of the L1TF vulnerability on 3389 affected CPUs 3390 3391 The kernel PTE inversion protection is unconditionally 3392 enabled and cannot be disabled. 3393 3394 full 3395 Provides all available mitigations for the 3396 L1TF vulnerability. Disables SMT and 3397 enables all mitigations in the 3398 hypervisors, i.e. unconditional L1D flush. 3399 3400 SMT control and L1D flush control via the 3401 sysfs interface is still possible after 3402 boot. Hypervisors will issue a warning 3403 when the first VM is started in a 3404 potentially insecure configuration, 3405 i.e. SMT enabled or L1D flush disabled. 3406 3407 full,force 3408 Same as 'full', but disables SMT and L1D 3409 flush runtime control. Implies the 3410 'nosmt=force' command line option. 3411 (i.e. sysfs control of SMT is disabled.) 3412 3413 flush 3414 Leaves SMT enabled and enables the default 3415 hypervisor mitigation, i.e. conditional 3416 L1D flush. 3417 3418 SMT control and L1D flush control via the 3419 sysfs interface is still possible after 3420 boot. Hypervisors will issue a warning 3421 when the first VM is started in a 3422 potentially insecure configuration, 3423 i.e. SMT enabled or L1D flush disabled. 3424 3425 flush,nosmt 3426 3427 Disables SMT and enables the default 3428 hypervisor mitigation. 3429 3430 SMT control and L1D flush control via the 3431 sysfs interface is still possible after 3432 boot. Hypervisors will issue a warning 3433 when the first VM is started in a 3434 potentially insecure configuration, 3435 i.e. SMT enabled or L1D flush disabled. 3436 3437 flush,nowarn 3438 Same as 'flush', but hypervisors will not 3439 warn when a VM is started in a potentially 3440 insecure configuration. 3441 3442 off 3443 Disables hypervisor mitigations and doesn't 3444 emit any warnings. 3445 It also drops the swap size and available 3446 RAM limit restriction on both hypervisor and 3447 bare metal. 3448 3449 Default is 'flush'. 3450 3451 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 3452 3453 l2cr= [PPC] 3454 3455 l3cr= [PPC] 3456 3457 lapic [X86-32,APIC,EARLY] Enable the local APIC even if BIOS 3458 disabled it. 3459 3460 lapic= [X86,APIC] Do not use TSC deadline 3461 value for LAPIC timer one-shot implementation. Default 3462 back to the programmable timer unit in the LAPIC. 3463 Format: notscdeadline 3464 3465 lapic_timer_c2_ok [X86,APIC,EARLY] trust the local apic timer 3466 in C2 power state. 3467 3468 libata.dma= [LIBATA] DMA control 3469 libata.dma=0 Disable all PATA and SATA DMA 3470 libata.dma=1 PATA and SATA Disk DMA only 3471 libata.dma=2 ATAPI (CDROM) DMA only 3472 libata.dma=4 Compact Flash DMA only 3473 Combinations also work, so libata.dma=3 enables DMA 3474 for disks and CDROMs, but not CFs. 3475 3476 libata.ignore_hpa= [LIBATA] Ignore HPA limit 3477 libata.ignore_hpa=0 keep BIOS limits (default) 3478 libata.ignore_hpa=1 ignore limits, using full disk 3479 3480 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 3481 when set. 3482 Format: <int> 3483 3484 libata.force= [LIBATA] Force configurations. The format is a comma- 3485 separated list of "[ID:]VAL" where ID is PORT[.DEVICE]. 3486 PORT and DEVICE are decimal numbers matching port, link 3487 or device. Basically, it matches the ATA ID string 3488 printed on console by libata. If the whole ID part is 3489 omitted, the last PORT and DEVICE values are used. If 3490 ID hasn't been specified yet, the configuration applies 3491 to all ports, links and devices. 3492 3493 If only DEVICE is omitted, the parameter applies to 3494 the port and all links and devices behind it. DEVICE 3495 number of 0 either selects the first device or the 3496 first fan-out link behind PMP device. It does not 3497 select the host link. DEVICE number of 15 selects the 3498 host link and device attached to it. 3499 3500 The VAL specifies the configuration to force. As long 3501 as there is no ambiguity, shortcut notation is allowed. 3502 For example, both 1.5 and 1.5G would work for 1.5Gbps. 3503 The following configurations can be forced. 3504 3505 * Cable type: 40c, 80c, short40c, unk, ign or sata. 3506 Any ID with matching PORT is used. 3507 3508 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 3509 3510 * Transfer mode: pio[0-6], mwdma[0-4] and udma[0-7]. 3511 udma[/][16,25,33,44,66,100,133] notation is also 3512 allowed. 3513 3514 * nohrst, nosrst, norst: suppress hard, soft and both 3515 resets. 3516 3517 * rstonce: only attempt one reset during hot-unplug 3518 link recovery. 3519 3520 * [no]dbdelay: Enable or disable the extra 200ms delay 3521 before debouncing a link PHY and device presence 3522 detection. 3523 3524 * [no]ncq: Turn on or off NCQ. 3525 3526 * [no]ncqtrim: Enable or disable queued DSM TRIM. 3527 3528 * [no]ncqati: Enable or disable NCQ on ATI chipsets. 3529 3530 * [no]trim: Enable or disable (unqueued) TRIM. 3531 3532 * trim_zero: Indicate that TRIM command zeroes data. 3533 3534 * max_trim_128m: Set 128M maximum trim size limit. 3535 3536 * [no]dma: Turn on or off DMA transfers. 3537 3538 * atapi_dmadir: Enable ATAPI DMADIR bridge support. 3539 3540 * atapi_mod16_dma: Enable the use of ATAPI DMA for 3541 commands that are not a multiple of 16 bytes. 3542 3543 * [no]dmalog: Enable or disable the use of the 3544 READ LOG DMA EXT command to access logs. 3545 3546 * [no]iddevlog: Enable or disable access to the 3547 identify device data log. 3548 3549 * [no]logdir: Enable or disable access to the general 3550 purpose log directory. 3551 3552 * max_sec=<sectors>: Set the transfer size limit, in 3553 number of 512-byte sectors, to the value specified in 3554 <sectors>. The value specified in <sectors> has to be 3555 a non-zero positive integer. 3556 3557 * max_sec_128: Set transfer size limit to 128 sectors. 3558 3559 * max_sec_1024: Set transfer size limit to 1024 sectors. 3560 3561 * max_sec_lba48: Set transfer size limit to 65535 sectors. 3562 3563 * external: Mark port as external (hotplug-capable). 3564 3565 * [no]lpm: Enable or disable link power management. 3566 3567 * [no]setxfer: Indicate if transfer speed mode setting 3568 should be skipped. 3569 3570 * [no]fua: Disable or enable FUA (Force Unit Access) 3571 support for devices supporting this feature. 3572 3573 * dump_id: Dump IDENTIFY data. 3574 3575 * disable: Disable this device. 3576 3577 If there are multiple matching configurations changing 3578 the same attribute, the last one is used. 3579 3580 liveupdate= [KNL,EARLY] 3581 Format: <bool> 3582 Enable Live Update Orchestrator (LUO). 3583 Default: off. 3584 3585 lockd.nlm_grace_period=P [NFS] Assign grace period. 3586 Format: <integer> 3587 3588 lockd.nlm_tcpport=N [NFS] Assign TCP port. 3589 Format: <integer> 3590 3591 lockd.nlm_timeout=T [NFS] Assign timeout value. 3592 Format: <integer> 3593 3594 lockd.nlm_udpport=M [NFS] Assign UDP port. 3595 Format: <integer> 3596 3597 lockdown= [SECURITY,EARLY] 3598 { integrity | confidentiality } 3599 Enable the kernel lockdown feature. If set to 3600 integrity, kernel features that allow userland to 3601 modify the running kernel are disabled. If set to 3602 confidentiality, kernel features that allow userland 3603 to extract confidential information from the kernel 3604 are also disabled. 3605 3606 locktorture.acq_writer_lim= [KNL] 3607 Set the time limit in jiffies for a lock 3608 acquisition. Acquisitions exceeding this limit 3609 will result in a splat once they do complete. 3610 3611 locktorture.bind_readers= [KNL] 3612 Specify the list of CPUs to which the readers are 3613 to be bound. 3614 3615 locktorture.bind_writers= [KNL] 3616 Specify the list of CPUs to which the writers are 3617 to be bound. 3618 3619 locktorture.call_rcu_chains= [KNL] 3620 Specify the number of self-propagating call_rcu() 3621 chains to set up. These are used to ensure that 3622 there is a high probability of an RCU grace period 3623 in progress at any given time. Defaults to 0, 3624 which disables these call_rcu() chains. 3625 3626 locktorture.long_hold= [KNL] 3627 Specify the duration in milliseconds for the 3628 occasional long-duration lock hold time. Defaults 3629 to 100 milliseconds. Select 0 to disable. 3630 3631 locktorture.nested_locks= [KNL] 3632 Specify the maximum lock nesting depth that 3633 locktorture is to exercise, up to a limit of 8 3634 (MAX_NESTED_LOCKS). Specify zero to disable. 3635 Note that this parameter is ineffective on types 3636 of locks that do not support nested acquisition. 3637 3638 locktorture.nreaders_stress= [KNL] 3639 Set the number of locking read-acquisition kthreads. 3640 Defaults to being automatically set based on the 3641 number of online CPUs. 3642 3643 locktorture.nwriters_stress= [KNL] 3644 Set the number of locking write-acquisition kthreads. 3645 3646 locktorture.onoff_holdoff= [KNL] 3647 Set time (s) after boot for CPU-hotplug testing. 3648 3649 locktorture.onoff_interval= [KNL] 3650 Set time (s) between CPU-hotplug operations, or 3651 zero to disable CPU-hotplug testing. 3652 3653 locktorture.rt_boost= [KNL] 3654 Do periodic testing of real-time lock priority 3655 boosting. Select 0 to disable, 1 to boost 3656 only rt_mutex, and 2 to boost unconditionally. 3657 Defaults to 2, which might seem to be an 3658 odd choice, but which should be harmless for 3659 non-real-time spinlocks, due to their disabling 3660 of preemption. Note that non-realtime mutexes 3661 disable boosting. 3662 3663 locktorture.rt_boost_factor= [KNL] 3664 Number that determines how often and for how 3665 long priority boosting is exercised. This is 3666 scaled down by the number of writers, so that the 3667 number of boosts per unit time remains roughly 3668 constant as the number of writers increases. 3669 On the other hand, the duration of each boost 3670 increases with the number of writers. 3671 3672 locktorture.shuffle_interval= [KNL] 3673 Set task-shuffle interval (jiffies). Shuffling 3674 tasks allows some CPUs to go into dyntick-idle 3675 mode during the locktorture test. 3676 3677 locktorture.shutdown_secs= [KNL] 3678 Set time (s) after boot system shutdown. This 3679 is useful for hands-off automated testing. 3680 3681 locktorture.stat_interval= [KNL] 3682 Time (s) between statistics printk()s. 3683 3684 locktorture.stutter= [KNL] 3685 Time (s) to stutter testing, for example, 3686 specifying five seconds causes the test to run for 3687 five seconds, wait for five seconds, and so on. 3688 This tests the locking primitive's ability to 3689 transition abruptly to and from idle. 3690 3691 locktorture.torture_type= [KNL] 3692 Specify the locking implementation to test. 3693 3694 locktorture.verbose= [KNL] 3695 Enable additional printk() statements. 3696 3697 locktorture.writer_fifo= [KNL] 3698 Run the write-side locktorture kthreads at 3699 sched_set_fifo() real-time priority. 3700 3701 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 3702 Format: <irq> 3703 3704 loglevel= [KNL,EARLY] 3705 All Kernel Messages with a loglevel smaller than the 3706 console loglevel will be printed to the console. It can 3707 also be changed with klogd or other programs. The 3708 loglevels are defined as follows: 3709 3710 0 (KERN_EMERG) system is unusable 3711 1 (KERN_ALERT) action must be taken immediately 3712 2 (KERN_CRIT) critical conditions 3713 3 (KERN_ERR) error conditions 3714 4 (KERN_WARNING) warning conditions 3715 5 (KERN_NOTICE) normal but significant condition 3716 6 (KERN_INFO) informational 3717 7 (KERN_DEBUG) debug-level messages 3718 3719 log_buf_len=n[KMG] [KNL,EARLY] 3720 Sets the size of the printk ring buffer, in bytes. 3721 n must be a power of two and greater than the 3722 minimal size. The minimal size is defined by 3723 LOG_BUF_SHIFT kernel config parameter. There 3724 is also CONFIG_LOG_CPU_MAX_BUF_SHIFT config 3725 parameter that allows to increase the default size 3726 depending on the number of CPUs. See init/Kconfig 3727 for more details. 3728 3729 logo.nologo [FB] Disables display of the built-in Linux logo. 3730 This may be used to provide more screen space for 3731 kernel log messages and is useful when debugging 3732 kernel boot problems. 3733 3734 lp=0 [LP] Specify parallel ports to use, e.g, 3735 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 3736 lp=reset first parallel port). 'lp=0' disables the 3737 lp=auto printer driver. 'lp=reset' (which can be 3738 specified in addition to the ports) causes 3739 attached printers to be reset. Using 3740 lp=port1,port2,... specifies the parallel ports 3741 to associate lp devices with, starting with 3742 lp0. A port specification may be 'none' to skip 3743 that lp device, or a parport name such as 3744 'parport0'. Specifying 'lp=auto' instead of a 3745 port specification list means that device IDs 3746 from each port should be examined, to see if 3747 an IEEE 1284-compliant printer is attached; if 3748 so, the driver will manage that printer. 3749 See also header of drivers/char/lp.c. 3750 3751 lpj=n [KNL] 3752 Sets loops_per_jiffy to given constant, thus avoiding 3753 time-consuming boot-time autodetection (up to 250 ms per 3754 CPU). 0 enables autodetection (default). To determine 3755 the correct value for your kernel, boot with normal 3756 autodetection and see what value is printed. Note that 3757 on SMP systems the preset will be applied to all CPUs, 3758 which is likely to cause problems if your CPUs need 3759 significantly divergent settings. An incorrect value 3760 will cause delays in the kernel to be wrong, leading to 3761 unpredictable I/O errors and other breakage. Although 3762 unlikely, in the extreme case this might damage your 3763 hardware. 3764 3765 lsm.debug [SECURITY] Enable LSM initialization debugging output. 3766 3767 lsm=lsm1,...,lsmN 3768 [SECURITY] Choose order of LSM initialization. This 3769 overrides CONFIG_LSM, and the "security=" parameter. 3770 3771 machtype= [Loongson] Share the same kernel image file between 3772 different yeeloong laptops. 3773 Example: machtype=lemote-yeeloong-2f-7inch 3774 3775 maxcpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 3776 will bring up during bootup. maxcpus=n : n >= 0 limits 3777 the kernel to bring up 'n' processors. Surely after 3778 bootup you can bring up the other plugged cpu by executing 3779 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 3780 only takes effect during system bootup. 3781 While n=0 is a special case, it is equivalent to "nosmp", 3782 which also disables the IO APIC. 3783 3784 max_loop= [LOOP] The number of loop block devices that get 3785 (loop.max_loop) unconditionally pre-created at init time. The default 3786 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 3787 of statically allocating a predefined number, loop 3788 devices can be requested on-demand with the 3789 /dev/loop-control interface. 3790 3791 mce= [X86-{32,64}] 3792 3793 Please see Documentation/arch/x86/x86_64/machinecheck.rst for sysfs runtime tunables. 3794 3795 off 3796 disable machine check 3797 3798 no_cmci 3799 disable CMCI(Corrected Machine Check Interrupt) that 3800 Intel processor supports. Usually this disablement is 3801 not recommended, but it might be handy if your 3802 hardware is misbehaving. 3803 3804 Note that you'll get more problems without CMCI than 3805 with due to the shared banks, i.e. you might get 3806 duplicated error logs. 3807 3808 dont_log_ce 3809 don't make logs for corrected errors. All events 3810 reported as corrected are silently cleared by OS. This 3811 option will be useful if you have no interest in any 3812 of corrected errors. 3813 3814 ignore_ce 3815 disable features for corrected errors, e.g. 3816 polling timer and CMCI. All events reported as 3817 corrected are not cleared by OS and remained in its 3818 error banks. 3819 3820 Usually this disablement is not recommended, however 3821 if there is an agent checking/clearing corrected 3822 errors (e.g. BIOS or hardware monitoring 3823 applications), conflicting with OS's error handling, 3824 and you cannot deactivate the agent, then this option 3825 will be a help. 3826 3827 no_lmce 3828 do not opt-in to Local MCE delivery. Use legacy method 3829 to broadcast MCEs. 3830 3831 bootlog 3832 enable logging of machine checks left over from 3833 booting. Disabled by default on AMD Fam10h and older 3834 because some BIOS leave bogus ones. 3835 3836 If your BIOS doesn't do that it's a good idea to 3837 enable though to make sure you log even machine check 3838 events that result in a reboot. On Intel systems it is 3839 enabled by default. 3840 3841 nobootlog 3842 disable boot machine check logging. 3843 3844 print_all 3845 print all machine check logs to the console. 3846 3847 monarchtimeout (number) 3848 sets the time in us to wait for other CPUs on machine 3849 checks. 0 to disable. 3850 3851 bios_cmci_threshold 3852 don't overwrite the bios-set CMCI threshold. This boot 3853 option prevents Linux from overwriting the CMCI 3854 threshold set by the bios. Without this option, Linux 3855 always sets the CMCI threshold to 1. Enabling this may 3856 make memory predictive failure analysis less effective 3857 if the bios sets thresholds for memory errors since we 3858 will not see details for all errors. 3859 3860 recovery 3861 force-enable recoverable machine check code paths 3862 3863 Everything else is in sysfs now. 3864 3865 3866 md= [HW] RAID subsystems devices and level 3867 See Documentation/admin-guide/md.rst. 3868 3869 mds= [X86,INTEL,EARLY] 3870 Control mitigation for the Micro-architectural Data 3871 Sampling (MDS) vulnerability. 3872 3873 Certain CPUs are vulnerable to an exploit against CPU 3874 internal buffers which can forward information to a 3875 disclosure gadget under certain conditions. 3876 3877 In vulnerable processors, the speculatively 3878 forwarded data can be used in a cache side channel 3879 attack, to access data to which the attacker does 3880 not have direct access. 3881 3882 This parameter controls the MDS mitigation. The 3883 options are: 3884 3885 full - Enable MDS mitigation on vulnerable CPUs 3886 full,nosmt - Enable MDS mitigation and disable 3887 SMT on vulnerable CPUs 3888 off - Unconditionally disable MDS mitigation 3889 3890 On TAA-affected machines, mds=off can be prevented by 3891 an active TAA mitigation as both vulnerabilities are 3892 mitigated with the same mechanism so in order to disable 3893 this mitigation, you need to specify tsx_async_abort=off 3894 too. 3895 3896 Not specifying this option is equivalent to 3897 mds=full. 3898 3899 For details see: Documentation/admin-guide/hw-vuln/mds.rst 3900 3901 mem=nn[KMG] [HEXAGON,EARLY] Set the memory size. 3902 Must be specified, otherwise memory size will be 0. 3903 3904 mem=nn[KMG] [KNL,BOOT,EARLY] Force usage of a specific amount 3905 of memory Amount of memory to be used in cases 3906 as follows: 3907 3908 1 for test; 3909 2 when the kernel is not able to see the whole system memory; 3910 3 memory that lies after 'mem=' boundary is excluded from 3911 the hypervisor, then assigned to KVM guests. 3912 4 to limit the memory available for kdump kernel. 3913 3914 [ARC,MICROBLAZE] - the limit applies only to low memory, 3915 high memory is not affected. 3916 3917 [ARM64] - only limits memory covered by the linear 3918 mapping. The NOMAP regions are not affected. 3919 3920 [X86] Work as limiting max address. Use together 3921 with memmap= to avoid physical address space collisions. 3922 Without memmap= PCI devices could be placed at addresses 3923 belonging to unused RAM. 3924 3925 Note that this only takes effects during boot time since 3926 in above case 3, memory may need be hot added after boot 3927 if system memory of hypervisor is not sufficient. 3928 3929 mem=nn[KMG]@ss[KMG] 3930 [ARM,MIPS,EARLY] - override the memory layout 3931 reported by firmware. 3932 Define a memory region of size nn[KMG] starting at 3933 ss[KMG]. 3934 Multiple different regions can be specified with 3935 multiple mem= parameters on the command line. 3936 3937 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 3938 memory. 3939 3940 memblock=debug [KNL,EARLY] Enable memblock debug messages. 3941 3942 memchunk=nn[KMG] 3943 [KNL,SH] Allow user to override the default size for 3944 per-device physically contiguous DMA buffers. 3945 3946 memhp_default_state=online/offline/online_kernel/online_movable 3947 [KNL] Set the initial state for the memory hotplug 3948 onlining policy. If not specified, the default value is 3949 set according to the 3950 CONFIG_MHP_DEFAULT_ONLINE_TYPE kernel config 3951 options. 3952 See Documentation/admin-guide/mm/memory-hotplug.rst. 3953 3954 memmap=exactmap [KNL,X86,EARLY] Enable setting of an exact 3955 E820 memory map, as specified by the user. 3956 Such memmap=exactmap lines can be constructed based on 3957 BIOS output or other requirements. See the memmap=nn@ss 3958 option description. 3959 3960 memmap=nn[KMG]@ss[KMG] 3961 [KNL, X86,MIPS,XTENSA,EARLY] Force usage of a specific region of memory. 3962 Region of memory to be used is from ss to ss+nn. 3963 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 3964 which limits max address to nn[KMG]. 3965 Multiple different regions can be specified, 3966 comma delimited. 3967 Example: 3968 memmap=100M@2G,100M#3G,1G!1024G 3969 3970 memmap=nn[KMG]#ss[KMG] 3971 [KNL,ACPI,EARLY] Mark specific memory as ACPI data. 3972 Region of memory to be marked is from ss to ss+nn. 3973 3974 memmap=nn[KMG]$ss[KMG] 3975 [KNL,ACPI,EARLY] Mark specific memory as reserved. 3976 Region of memory to be reserved is from ss to ss+nn. 3977 Example: Exclude memory from 0x18690000-0x1869ffff 3978 memmap=64K$0x18690000 3979 or 3980 memmap=0x10000$0x18690000 3981 Some bootloaders may need an escape character before '$', 3982 like Grub2, otherwise '$' and the following number 3983 will be eaten. 3984 3985 memmap=nn[KMG]!ss[KMG] 3986 [KNL,X86,EARLY] Mark specific memory as protected. 3987 Region of memory to be used, from ss to ss+nn. 3988 The memory region may be marked as e820 type 12 (0xc) 3989 and is NVDIMM or ADR memory. 3990 3991 memmap=<size>%<offset>-<oldtype>+<newtype> 3992 [KNL,ACPI,EARLY] Convert memory within the specified region 3993 from <oldtype> to <newtype>. If "-<oldtype>" is left 3994 out, the whole region will be marked as <newtype>, 3995 even if previously unavailable. If "+<newtype>" is left 3996 out, matching memory will be removed. Types are 3997 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 3998 3 = ACPI, 12 = PRAM. 3999 4000 memory_corruption_check=0/1 [X86,EARLY] 4001 Some BIOSes seem to corrupt the first 64k of 4002 memory when doing things like suspend/resume. 4003 Setting this option will scan the memory 4004 looking for corruption. Enabling this will 4005 both detect corruption and prevent the kernel 4006 from using the memory being corrupted. 4007 However, it's intended as a diagnostic tool; if 4008 repeatable BIOS-originated corruption always 4009 affects the same memory, you can use memmap= 4010 to prevent the kernel from using that memory. 4011 4012 memory_corruption_check_size=size [X86,EARLY] 4013 By default it checks for corruption in the low 4014 64k, making this memory unavailable for normal 4015 use. Use this parameter to scan for 4016 corruption in more or less memory. 4017 4018 memory_corruption_check_period=seconds [X86,EARLY] 4019 By default it checks for corruption every 60 4020 seconds. Use this parameter to check at some 4021 other rate. 0 disables periodic checking. 4022 4023 memory_hotplug.memmap_on_memory 4024 [KNL,X86,ARM] Boolean flag to enable this feature. 4025 Format: {on | off (default)} 4026 When enabled, runtime hotplugged memory will 4027 allocate its internal metadata (struct pages, 4028 those vmemmap pages cannot be optimized even 4029 if hugetlb_free_vmemmap is enabled) from the 4030 hotadded memory which will allow to hotadd a 4031 lot of memory without requiring additional 4032 memory to do so. 4033 This feature is disabled by default because it 4034 has some implication on large (e.g. GB) 4035 allocations in some configurations (e.g. small 4036 memory blocks). 4037 The state of the flag can be read in 4038 /sys/module/memory_hotplug/parameters/memmap_on_memory. 4039 Note that even when enabled, there are a few cases where 4040 the feature is not effective. 4041 4042 mempool_debug [MM] 4043 Enable mempool debugging. This enables element 4044 poison checking when freeing elements back to the 4045 pool. Useful for debugging mempool corruption. 4046 4047 memtest= [KNL,ALPHA,X86,ARM,M68K,PPC,RISCV,EARLY] Enable memtest 4048 Format: <integer> 4049 default : 0 <disable> 4050 Specifies the number of memtest passes to be 4051 performed. Each pass selects another test 4052 pattern from a given set of patterns. Memtest 4053 fills the memory with this pattern, validates 4054 memory contents and reserves bad memory 4055 regions that are detected. 4056 4057 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 4058 Valid arguments: on, off 4059 Default: off 4060 mem_encrypt=on: Activate SME 4061 mem_encrypt=off: Do not activate SME 4062 4063 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst 4064 for details on when memory encryption can be activated. 4065 4066 mem_sleep_default= [SUSPEND] Default system suspend mode: 4067 s2idle - Suspend-To-Idle 4068 shallow - Power-On Suspend or equivalent (if supported) 4069 deep - Suspend-To-RAM or equivalent (if supported) 4070 See Documentation/admin-guide/pm/sleep-states.rst. 4071 4072 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 4073 the BIOS has incorrectly applied a workaround. TinyBIOS 4074 version 0.98 is known to be affected, 0.99 fixes the 4075 problem by letting the user disable the workaround. 4076 4077 mga= [HW,DRM] 4078 4079 microcode= [X86] Control the behavior of the microcode loader. 4080 Available options, comma separated: 4081 4082 base_rev=X - with <X> with format: <u32> 4083 Set the base microcode revision of each thread when in 4084 debug mode. 4085 4086 dis_ucode_ldr: disable the microcode loader 4087 4088 force_minrev: 4089 Enable or disable the microcode minimal revision 4090 enforcement for the runtime microcode loader. 4091 4092 mitigations= 4093 [X86,PPC,S390,ARM64,EARLY] Control optional mitigations for 4094 CPU vulnerabilities. This is a set of curated, 4095 arch-independent options, each of which is an 4096 aggregation of existing arch-specific options. 4097 4098 Note, "mitigations" is supported if and only if the 4099 kernel was built with CPU_MITIGATIONS=y. 4100 4101 off 4102 Disable all optional CPU mitigations. This 4103 improves system performance, but it may also 4104 expose users to several CPU vulnerabilities. 4105 Equivalent to: if nokaslr then kpti=0 [ARM64] 4106 gather_data_sampling=off [X86] 4107 indirect_target_selection=off [X86] 4108 kvm.nx_huge_pages=off [X86] 4109 l1tf=off [X86] 4110 mds=off [X86] 4111 mmio_stale_data=off [X86] 4112 no_entry_flush [PPC] 4113 no_uaccess_flush [PPC] 4114 nobp=0 [S390] 4115 nopti [X86,PPC] 4116 nospectre_bhb [ARM64] 4117 nospectre_v1 [X86,PPC] 4118 nospectre_v2 [X86,PPC,S390,ARM64] 4119 reg_file_data_sampling=off [X86] 4120 retbleed=off [X86] 4121 spec_rstack_overflow=off [X86] 4122 spec_store_bypass_disable=off [X86,PPC] 4123 spectre_bhi=off [X86] 4124 spectre_v2_user=off [X86] 4125 srbds=off [X86,INTEL] 4126 ssbd=force-off [ARM64] 4127 tsa=off [X86,AMD] 4128 tsx_async_abort=off [X86] 4129 vmscape=off [X86] 4130 4131 Exceptions: 4132 This does not have any effect on 4133 kvm.nx_huge_pages when 4134 kvm.nx_huge_pages=force. 4135 4136 auto (default) 4137 Mitigate all CPU vulnerabilities, but leave SMT 4138 enabled, even if it's vulnerable. This is for 4139 users who don't want to be surprised by SMT 4140 getting disabled across kernel upgrades, or who 4141 have other ways of avoiding SMT-based attacks. 4142 Equivalent to: (default behavior) 4143 4144 auto,nosmt 4145 Mitigate all CPU vulnerabilities, disabling SMT 4146 if needed. This is for users who always want to 4147 be fully mitigated, even if it means losing SMT. 4148 Equivalent to: l1tf=flush,nosmt [X86] 4149 mds=full,nosmt [X86] 4150 tsx_async_abort=full,nosmt [X86] 4151 mmio_stale_data=full,nosmt [X86] 4152 retbleed=auto,nosmt [X86] 4153 4154 [X86] After one of the above options, additionally 4155 supports attack-vector based controls as documented in 4156 Documentation/admin-guide/hw-vuln/attack_vector_controls.rst 4157 4158 mminit_loglevel= 4159 [KNL,EARLY] When CONFIG_DEBUG_MEMORY_INIT is set, this 4160 parameter allows control of the logging verbosity for 4161 the additional memory initialisation checks. A value 4162 of 0 disables mminit logging and a level of 4 will 4163 log everything. Information is printed at KERN_DEBUG 4164 so loglevel=8 may also need to be specified. 4165 4166 mmio_stale_data= 4167 [X86,INTEL,EARLY] Control mitigation for the Processor 4168 MMIO Stale Data vulnerabilities. 4169 4170 Processor MMIO Stale Data is a class of 4171 vulnerabilities that may expose data after an MMIO 4172 operation. Exposed data could originate or end in 4173 the same CPU buffers as affected by MDS and TAA. 4174 Therefore, similar to MDS and TAA, the mitigation 4175 is to clear the affected CPU buffers. 4176 4177 This parameter controls the mitigation. The 4178 options are: 4179 4180 full - Enable mitigation on vulnerable CPUs 4181 4182 full,nosmt - Enable mitigation and disable SMT on 4183 vulnerable CPUs. 4184 4185 off - Unconditionally disable mitigation 4186 4187 On MDS or TAA affected machines, 4188 mmio_stale_data=off can be prevented by an active 4189 MDS or TAA mitigation as these vulnerabilities are 4190 mitigated with the same mechanism so in order to 4191 disable this mitigation, you need to specify 4192 mds=off and tsx_async_abort=off too. 4193 4194 Not specifying this option is equivalent to 4195 mmio_stale_data=full. 4196 4197 For details see: 4198 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst 4199 4200 <module>.async_probe[=<bool>] [KNL] 4201 If no <bool> value is specified or if the value 4202 specified is not a valid <bool>, enable asynchronous 4203 probe on this module. Otherwise, enable/disable 4204 asynchronous probe on this module as indicated by the 4205 <bool> value. See also: module.async_probe 4206 4207 module.async_probe=<bool> 4208 [KNL] When set to true, modules will use async probing 4209 by default. To enable/disable async probing for a 4210 specific module, use the module specific control that 4211 is documented under <module>.async_probe. When both 4212 module.async_probe and <module>.async_probe are 4213 specified, <module>.async_probe takes precedence for 4214 the specific module. 4215 4216 module.enable_dups_trace 4217 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set, 4218 this means that duplicate request_module() calls will 4219 trigger a WARN_ON() instead of a pr_warn(). Note that 4220 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s 4221 will always be issued and this option does nothing. 4222 module.sig_enforce 4223 [KNL] When CONFIG_MODULE_SIG is set, this means that 4224 modules without (valid) signatures will fail to load. 4225 Note that if CONFIG_MODULE_SIG_FORCE is set, that 4226 is always true, so this option does nothing. 4227 4228 module_blacklist= [KNL] Do not load a comma-separated list of 4229 modules. Useful for debugging problem modules. 4230 4231 mousedev.tap_time= 4232 [MOUSE] Maximum time between finger touching and 4233 leaving touchpad surface for touch to be considered 4234 a tap and be reported as a left button click (for 4235 touchpads working in absolute mode only). 4236 Format: <msecs> 4237 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 4238 reporting absolute coordinates, such as tablets 4239 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 4240 reporting absolute coordinates, such as tablets 4241 4242 movablecore= [KNL,X86,PPC,EARLY] 4243 Format: nn[KMGTPE] | nn% 4244 This parameter is the complement to kernelcore=, it 4245 specifies the amount of memory used for migratable 4246 allocations. If both kernelcore and movablecore is 4247 specified, then kernelcore will be at *least* the 4248 specified value but may be more. If movablecore on its 4249 own is specified, the administrator must be careful 4250 that the amount of memory usable for all allocations 4251 is not too small. 4252 4253 movable_node [KNL,EARLY] Boot-time switch to make hotplugable memory 4254 NUMA nodes to be movable. This means that the memory 4255 of such nodes will be usable only for movable 4256 allocations which rules out almost all kernel 4257 allocations. Use with caution! 4258 4259 MTD_Partition= [MTD] 4260 Format: <name>,<region-number>,<size>,<offset> 4261 4262 MTD_Region= [MTD] Format: 4263 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 4264 4265 mtdparts= [MTD] 4266 See drivers/mtd/parsers/cmdlinepart.c 4267 4268 mtouchusb.raw_coordinates= 4269 [HW] Make the MicroTouch USB driver use raw coordinates 4270 ('y', default) or cooked coordinates ('n') 4271 4272 mtrr=debug [X86,EARLY] 4273 Enable printing debug information related to MTRR 4274 registers at boot time. 4275 4276 mtrr_chunk_size=nn[KMG] [X86,EARLY] 4277 used for mtrr cleanup. It is largest continuous chunk 4278 that could hold holes aka. UC entries. 4279 4280 mtrr_gran_size=nn[KMG] [X86,EARLY] 4281 Used for mtrr cleanup. It is granularity of mtrr block. 4282 Default is 1. 4283 Large value could prevent small alignment from 4284 using up MTRRs. 4285 4286 mtrr_spare_reg_nr=n [X86,EARLY] 4287 Format: <integer> 4288 Range: 0,7 : spare reg number 4289 Default : 1 4290 Used for mtrr cleanup. It is spare mtrr entries number. 4291 Set to 2 or more if your graphical card needs more. 4292 4293 multitce=off [PPC] This parameter disables the use of the pSeries 4294 firmware feature for updating multiple TCE entries 4295 at a time. 4296 4297 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 4298 4299 netdev= [NET] NE2000 ISA network devices parameters 4300 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 4301 4302 netpoll.carrier_timeout= 4303 [NET] Specifies amount of time (in seconds) that 4304 netpoll should wait for a carrier. By default netpoll 4305 waits 4 seconds. 4306 4307 nf_conntrack.acct= 4308 [NETFILTER] Enable connection tracking flow accounting 4309 0 to disable accounting 4310 1 to enable accounting 4311 Default value is 0. 4312 4313 nfs.cache_getent= 4314 [NFS] sets the pathname to the program which is used 4315 to update the NFS client cache entries. 4316 4317 nfs.cache_getent_timeout= 4318 [NFS] sets the timeout after which an attempt to 4319 update a cache entry is deemed to have failed. 4320 4321 nfs.callback_nr_threads= 4322 [NFSv4] set the total number of threads that the 4323 NFS client will assign to service NFSv4 callback 4324 requests. 4325 4326 nfs.callback_tcpport= 4327 [NFS] set the TCP port on which the NFSv4 callback 4328 channel should listen. 4329 4330 nfs.delay_retrans= 4331 [NFS] specifies the number of times the NFSv4 client 4332 retries the request before returning an EAGAIN error, 4333 after a reply of NFS4ERR_DELAY from the server. 4334 Only applies if the softerr mount option is enabled, 4335 and the specified value is >= 0. 4336 4337 nfs.idmap_cache_timeout= 4338 [NFS] set the maximum lifetime for idmapper cache 4339 entries. 4340 4341 nfs.max_session_cb_slots= 4342 [NFSv4.1] Sets the maximum number of session 4343 slots the client will assign to the callback 4344 channel. This determines the maximum number of 4345 callbacks the client will process in parallel for 4346 a particular server. 4347 4348 nfs.max_session_slots= 4349 [NFSv4.1] Sets the maximum number of session slots 4350 the client will attempt to negotiate with the server. 4351 This limits the number of simultaneous RPC requests 4352 that the client can send to the NFSv4.1 server. 4353 Note that there is little point in setting this 4354 value higher than the max_tcp_slot_table_limit. 4355 4356 nfs.nfs4_disable_idmapping= 4357 [NFSv4] When set to the default of '1', this option 4358 ensures that both the RPC level authentication 4359 scheme and the NFS level operations agree to use 4360 numeric uids/gids if the mount is using the 4361 'sec=sys' security flavour. In effect it is 4362 disabling idmapping, which can make migration from 4363 legacy NFSv2/v3 systems to NFSv4 easier. 4364 Servers that do not support this mode of operation 4365 will be autodetected by the client, and it will fall 4366 back to using the idmapper. 4367 To turn off this behaviour, set the value to '0'. 4368 4369 nfs.nfs4_unique_id= 4370 [NFS4] Specify an additional fixed unique ident- 4371 ification string that NFSv4 clients can insert into 4372 their nfs_client_id4 string. This is typically a 4373 UUID that is generated at system install time. 4374 4375 nfs.recover_lost_locks= 4376 [NFSv4] Attempt to recover locks that were lost due 4377 to a lease timeout on the server. Please note that 4378 doing this risks data corruption, since there are 4379 no guarantees that the file will remain unchanged 4380 after the locks are lost. 4381 If you want to enable the kernel legacy behaviour of 4382 attempting to recover these locks, then set this 4383 parameter to '1'. 4384 The default parameter value of '0' causes the kernel 4385 not to attempt recovery of lost locks. 4386 4387 nfs.send_implementation_id= 4388 [NFSv4.1] Send client implementation identification 4389 information in exchange_id requests. 4390 If zero, no implementation identification information 4391 will be sent. 4392 The default is to send the implementation identification 4393 information. 4394 4395 nfs4.layoutstats_timer= 4396 [NFSv4.2] Change the rate at which the kernel sends 4397 layoutstats to the pNFS metadata server. 4398 4399 Setting this to value to 0 causes the kernel to use 4400 whatever value is the default set by the layout 4401 driver. A non-zero value sets the minimum interval 4402 in seconds between layoutstats transmissions. 4403 4404 nfsd.inter_copy_offload_enable= 4405 [NFSv4.2] When set to 1, the server will support 4406 server-to-server copies for which this server is 4407 the destination of the copy. 4408 4409 nfsd.nfs4_disable_idmapping= 4410 [NFSv4] When set to the default of '1', the NFSv4 4411 server will return only numeric uids and gids to 4412 clients using auth_sys, and will accept numeric uids 4413 and gids from such clients. This is intended to ease 4414 migration from NFSv2/v3. 4415 4416 nfsd.nfsd4_ssc_umount_timeout= 4417 [NFSv4.2] When used as the destination of a 4418 server-to-server copy, knfsd temporarily mounts 4419 the source server. It caches the mount in case 4420 it will be needed again, and discards it if not 4421 used for the number of milliseconds specified by 4422 this parameter. 4423 4424 nfsaddrs= [NFS] Deprecated. Use ip= instead. 4425 See Documentation/admin-guide/nfs/nfsroot.rst. 4426 4427 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 4428 See Documentation/admin-guide/nfs/nfsroot.rst. 4429 4430 nfsrootdebug [NFS] enable nfsroot debugging messages. 4431 See Documentation/admin-guide/nfs/nfsroot.rst. 4432 4433 nmi_backtrace.backtrace_idle [KNL] 4434 Dump stacks even of idle CPUs in response to an 4435 NMI stack-backtrace request. 4436 4437 nmi_debug= [KNL,SH] Specify one or more actions to take 4438 when a NMI is triggered. 4439 Format: [state][,regs][,debounce][,die] 4440 4441 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 4442 Format: [panic,][nopanic,][rNNN,][num] 4443 Valid num: 0 or 1 4444 0 - turn hardlockup detector in nmi_watchdog off 4445 1 - turn hardlockup detector in nmi_watchdog on 4446 rNNN - configure the watchdog with raw perf event 0xNNN 4447 4448 When panic is specified, panic when an NMI watchdog 4449 timeout occurs (or 'nopanic' to not panic on an NMI 4450 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 4451 To disable both hard and soft lockup detectors, 4452 please see 'nowatchdog'. 4453 This is useful when you use a panic=... timeout and 4454 need the box quickly up again. 4455 4456 These settings can be accessed at runtime via 4457 the nmi_watchdog and hardlockup_panic sysctls. 4458 4459 no4lvl [RISCV,EARLY] Disable 4-level and 5-level paging modes. 4460 Forces kernel to use 3-level paging instead. 4461 4462 no5lvl [X86-64,RISCV,EARLY] Disable 5-level paging mode. Forces 4463 kernel to use 4-level paging instead. 4464 4465 noalign [KNL,ARM] 4466 4467 noapic [SMP,APIC,EARLY] Tells the kernel to not make use of any 4468 IOAPICs that may be present in the system. 4469 4470 noapictimer [APIC,X86] Don't set up the APIC timer 4471 4472 noautogroup Disable scheduler automatic task group creation. 4473 4474 nocache [ARM,EARLY] 4475 4476 no_console_suspend 4477 [HW] Never suspend the console 4478 Disable suspending of consoles during suspend and 4479 hibernate operations. Once disabled, debugging 4480 messages can reach various consoles while the rest 4481 of the system is being put to sleep (ie, while 4482 debugging driver suspend/resume hooks). This may 4483 not work reliably with all consoles, but is known 4484 to work with serial and VGA consoles. 4485 To facilitate more flexible debugging, we also add 4486 console_suspend, a printk module parameter to control 4487 it. Users could use console_suspend (usually 4488 /sys/module/printk/parameters/console_suspend) to 4489 turn on/off it dynamically. 4490 4491 no_debug_objects 4492 [KNL,EARLY] Disable object debugging 4493 4494 nodsp [SH] Disable hardware DSP at boot time. 4495 4496 noefi [EFI,EARLY] Disable EFI runtime services support. 4497 4498 no_entry_flush [PPC,EARLY] Don't flush the L1-D cache when entering the kernel. 4499 4500 noexec32 [X86-64] 4501 This affects only 32-bit executables. 4502 noexec32=on: enable non-executable mappings (default) 4503 read doesn't imply executable mappings 4504 noexec32=off: disable non-executable mappings 4505 read implies executable mappings 4506 4507 no_file_caps Tells the kernel not to honor file capabilities. The 4508 only way then for a file to be executed with privilege 4509 is to be setuid root or executed by root. 4510 4511 nofpu [MIPS,SH] Disable hardware FPU at boot time. 4512 4513 nofsgsbase [X86] Disables FSGSBASE instructions. 4514 4515 nofxsr [BUGS=X86-32] Disables x86 floating point extended 4516 register save and restore. The kernel will only save 4517 legacy floating-point registers on task switch. 4518 4519 nogbpages [X86] Do not use GB pages for kernel direct mappings. 4520 4521 no_hash_pointers 4522 [KNL,EARLY] 4523 Alias for "hash_pointers=never". 4524 4525 nohibernate [HIBERNATION] Disable hibernation and resume. 4526 4527 nohlt [ARM,ARM64,MICROBLAZE,MIPS,PPC,RISCV,SH] Forces the kernel to 4528 busy wait in do_idle() and not use the arch_cpu_idle() 4529 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP 4530 to be effective. This is useful on platforms where the 4531 sleep(SH) or wfi(ARM,ARM64) instructions do not work 4532 correctly or when doing power measurements to evaluate 4533 the impact of the sleep instructions. This is also 4534 useful when using JTAG debugger. 4535 4536 nohpet [X86] Don't use the HPET timer. 4537 4538 nohugeiomap [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge I/O mappings. 4539 4540 nohugevmalloc [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge vmalloc mappings. 4541 4542 nohz= [KNL] Boottime enable/disable dynamic ticks 4543 Valid arguments: on, off 4544 Default: on 4545 4546 nohz_full= [KNL,BOOT,SMP,ISOL] 4547 The argument is a cpu list, as described above. 4548 In kernels built with CONFIG_NO_HZ_FULL=y, set 4549 the specified list of CPUs whose tick will be stopped 4550 whenever possible. The boot CPU will be forced outside 4551 the range to maintain the timekeeping. Any CPUs 4552 in this list will have their RCU callbacks offloaded, 4553 just as if they had also been called out in the 4554 rcu_nocbs= boot parameter. 4555 4556 Note that this argument takes precedence over 4557 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 4558 4559 noinitrd [Deprecated,RAM] Tells the kernel not to load any configured 4560 initial RAM disk. Currently this parameter applies to 4561 initrd only, not to initramfs. But it applies to both 4562 in EFI mode. 4563 4564 nointremap [X86-64,Intel-IOMMU,EARLY] Do not enable interrupt 4565 remapping. 4566 [Deprecated - use intremap=off] 4567 4568 noinvpcid [X86,EARLY] Disable the INVPCID cpu feature. 4569 4570 noiotrap [SH] Disables trapped I/O port accesses. 4571 4572 noirqdebug [X86-32] Disables the code which attempts to detect and 4573 disable unhandled interrupt sources. 4574 4575 noisapnp [ISAPNP] Disables ISA PnP code. 4576 4577 nokaslr [KNL,EARLY] 4578 When CONFIG_RANDOMIZE_BASE is set, this disables 4579 kernel and module base offset ASLR (Address Space 4580 Layout Randomization). 4581 4582 no-kvmapf [X86,KVM,EARLY] Disable paravirtualized asynchronous page 4583 fault handling. 4584 4585 no-kvmclock [X86,KVM,EARLY] Disable paravirtualized KVM clock driver 4586 4587 nolapic [X86-32,APIC,EARLY] Do not enable or use the local APIC. 4588 4589 nolapic_timer [X86-32,APIC,EARLY] Do not use the local APIC timer. 4590 4591 nomce [X86-32] Disable Machine Check Exception 4592 4593 nomodeset Disable kernel modesetting. Most systems' firmware 4594 sets up a display mode and provides framebuffer memory 4595 for output. With nomodeset, DRM and fbdev drivers will 4596 not load if they could possibly displace the pre- 4597 initialized output. Only the system framebuffer will 4598 be available for use. The respective drivers will not 4599 perform display-mode changes or accelerated rendering. 4600 4601 Useful as error fallback, or for testing and debugging. 4602 4603 nomodule Disable module load 4604 4605 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 4606 shutdown the other cpus. Instead use the REBOOT_VECTOR 4607 irq. 4608 4609 nopat [X86,EARLY] Disable PAT (page attribute table extension of 4610 pagetables) support. 4611 4612 nopcid [X86-64,EARLY] Disable the PCID cpu feature. 4613 4614 nopku [X86] Disable Memory Protection Keys CPU feature found 4615 in some Intel CPUs. 4616 4617 nopti [X86-64,EARLY] 4618 Equivalent to pti=off 4619 4620 nopv= [X86,XEN,KVM,HYPER_V,VMWARE,EARLY] 4621 Disables the PV optimizations forcing the guest to run 4622 as generic guest with no PV drivers. Currently support 4623 XEN HVM, KVM, HYPER_V and VMWARE guest. 4624 4625 nopvspin [X86,XEN,KVM,EARLY] 4626 Disables the qspinlock slow path using PV optimizations 4627 which allow the hypervisor to 'idle' the guest on lock 4628 contention. 4629 4630 norandmaps Don't use address space randomization. Equivalent to 4631 echo 0 > /proc/sys/kernel/randomize_va_space 4632 4633 noresume [SWSUSP] Disables resume and restores original swap 4634 space. 4635 4636 no-scroll [VGA] Disables scrollback. 4637 This is required for the Braillex ib80-piezo Braille 4638 reader made by F.H. Papenmeier (Germany). 4639 4640 nosgx [X86-64,SGX,EARLY] Disables Intel SGX kernel support. 4641 4642 nosmap [PPC,EARLY] 4643 Disable SMAP (Supervisor Mode Access Prevention) 4644 even if it is supported by processor. 4645 4646 nosmep [PPC64s,EARLY] 4647 Disable SMEP (Supervisor Mode Execution Prevention) 4648 even if it is supported by processor. 4649 4650 nosmp [SMP,EARLY] Tells an SMP kernel to act as a UP kernel, 4651 and disable the IO APIC. legacy for "maxcpus=0". 4652 4653 nosmt [KNL,MIPS,PPC,EARLY] Disable symmetric multithreading (SMT). 4654 Equivalent to smt=1. 4655 4656 [KNL,LOONGARCH,X86,ARM64,PPC,S390] Disable symmetric multithreading (SMT). 4657 nosmt=force: Force disable SMT, cannot be undone 4658 via the sysfs control file. 4659 4660 nosoftlockup [KNL] Disable the soft-lockup detector. 4661 4662 nospec_store_bypass_disable 4663 [HW,EARLY] Disable all mitigations for the Speculative 4664 Store Bypass vulnerability 4665 4666 nospectre_bhb [ARM64,EARLY] Disable all mitigations for Spectre-BHB (branch 4667 history injection) vulnerability. System may allow data leaks 4668 with this option. 4669 4670 nospectre_v1 [X86,PPC,EARLY] Disable mitigations for Spectre Variant 1 4671 (bounds check bypass). With this option data leaks are 4672 possible in the system. 4673 4674 nospectre_v2 [X86,PPC_E500,ARM64,EARLY] Disable all mitigations 4675 for the Spectre variant 2 (indirect branch 4676 prediction) vulnerability. System may allow data 4677 leaks with this option. 4678 4679 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES,RISCV,LOONGARCH,EARLY] 4680 Disable paravirtualized steal time accounting. steal time 4681 is computed, but won't influence scheduler behaviour 4682 4683 nosync [HW,M68K] Disables sync negotiation for all devices. 4684 4685 no_timer_check [X86,APIC] Disables the code which tests for broken 4686 timer IRQ sources, i.e., the IO-APIC timer. This can 4687 work around problems with incorrect timer 4688 initialization on some boards. 4689 4690 no_uaccess_flush 4691 [PPC,EARLY] Don't flush the L1-D cache after accessing user data. 4692 4693 novmcoredd [KNL,KDUMP] 4694 Disable device dump. Device dump allows drivers to 4695 append dump data to vmcore so you can collect driver 4696 specified debug info. Drivers can append the data 4697 without any limit and this data is stored in memory, 4698 so this may cause significant memory stress. Disabling 4699 device dump can help save memory but the driver debug 4700 data will be no longer available. This parameter 4701 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 4702 is set. 4703 4704 no-vmw-sched-clock 4705 [X86,PV_OPS,EARLY] Disable paravirtualized VMware 4706 scheduler clock and use the default one. 4707 4708 nowatchdog [KNL] Disable both lockup detectors, i.e. 4709 soft-lockup and NMI watchdog (hard-lockup). 4710 4711 nowb [ARM,EARLY] 4712 4713 nox2apic [X86-64,APIC,EARLY] Do not enable x2APIC mode. 4714 4715 NOTE: this parameter will be ignored on systems with the 4716 LEGACY_XAPIC_DISABLED bit set in the 4717 IA32_XAPIC_DISABLE_STATUS MSR. 4718 4719 noxsave [BUGS=X86] Disables x86 extended register state save 4720 and restore using xsave. The kernel will fallback to 4721 enabling legacy floating-point and sse state. 4722 4723 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 4724 register states. The kernel will fall back to use 4725 xsave to save the states. By using this parameter, 4726 performance of saving the states is degraded because 4727 xsave doesn't support modified optimization while 4728 xsaveopt supports it on xsaveopt enabled systems. 4729 4730 noxsaves [X86] Disables xsaves and xrstors used in saving and 4731 restoring x86 extended register state in compacted 4732 form of xsave area. The kernel will fall back to use 4733 xsaveopt and xrstor to save and restore the states 4734 in standard form of xsave area. By using this 4735 parameter, xsave area per process might occupy more 4736 memory on xsaves enabled systems. 4737 4738 nr_cpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 4739 could support. nr_cpus=n : n >= 1 limits the kernel to 4740 support 'n' processors. It could be larger than the 4741 number of already plugged CPU during bootup, later in 4742 runtime you can physically add extra cpu until it reaches 4743 n. So during boot up some boot time memory for per-cpu 4744 variables need be pre-allocated for later physical cpu 4745 hot plugging. 4746 4747 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 4748 4749 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86, EARLY] 4750 Disable NUMA, Only set up a single NUMA node 4751 spanning all memory. 4752 4753 numa=fake=<size>[MG] 4754 [KNL, ARM64, RISCV, X86, EARLY] 4755 If given as a memory unit, fills all system RAM with 4756 nodes of size interleaved over physical nodes. 4757 4758 numa=fake=<N> 4759 [KNL, ARM64, RISCV, X86, EARLY] 4760 If given as an integer, fills all system RAM with N 4761 fake nodes interleaved over physical nodes. 4762 4763 numa=fake=<N>U 4764 [KNL, ARM64, RISCV, X86, EARLY] 4765 If given as an integer followed by 'U', it will 4766 divide each physical node into N emulated nodes. 4767 4768 numa=noacpi [X86] Don't parse the SRAT table for NUMA setup 4769 4770 numa=nohmat [X86] Don't parse the HMAT table for NUMA setup, or 4771 soft-reserved memory partitioning. 4772 4773 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic 4774 NUMA balancing. 4775 Allowed values are enable and disable 4776 4777 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 4778 'node', 'default' can be specified 4779 This can be set from sysctl after boot. 4780 See Documentation/admin-guide/sysctl/vm.rst for details. 4781 4782 nvme.quirks= [NVME] A list of quirk entries to augment the built-in 4783 nvme quirk list. List entries are separated by a 4784 '-' character. 4785 Each entry has the form VendorID:ProductID:quirk_names. 4786 The IDs are 4-digits hex numbers and quirk_names is a 4787 list of quirk names separated by commas. A quirk name 4788 can be prefixed by '^', meaning that the specified 4789 quirk must be disabled. 4790 4791 Example: 4792 nvme.quirks=7710:2267:bogus_nid,^identify_cns-9900:7711:broken_msi 4793 4794 ohci1394_dma=early [HW,EARLY] enable debugging via the ohci1394 driver. 4795 See Documentation/core-api/debugging-via-ohci1394.rst for more 4796 info. 4797 4798 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 4799 Rather than timing out after 20 ms if an EC 4800 command is not properly ACKed, override the length 4801 of the timeout. We have interrupts disabled while 4802 waiting for the ACK, so if this is set too high 4803 interrupts *may* be lost! 4804 4805 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 4806 4807 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 4808 4809 boundary - index of last SLC block on Flex-OneNAND. 4810 The remaining blocks are configured as MLC blocks. 4811 lock - Configure if Flex-OneNAND boundary should be locked. 4812 Once locked, the boundary cannot be changed. 4813 1 indicates lock status, 0 indicates unlock status. 4814 4815 oops=panic [KNL,EARLY] 4816 Always panic on oopses. Default is to just kill the 4817 process, but there is a small probability of 4818 deadlocking the machine. 4819 This will also cause panics on machine check exceptions. 4820 Useful together with panic=30 to trigger a reboot. 4821 4822 page_alloc.shuffle= 4823 [KNL] Boolean flag to control whether the page allocator 4824 should randomize its free lists. This parameter can be 4825 used to enable/disable page randomization. The state of 4826 the flag can be read from sysfs at: 4827 /sys/module/page_alloc/parameters/shuffle. 4828 This parameter is only available if CONFIG_SHUFFLE_PAGE_ALLOCATOR=y. 4829 4830 page_owner= [KNL,EARLY] Boot-time page_owner enabling option. 4831 Storage of the information about who allocated 4832 each page is disabled in default. With this switch, 4833 we can turn it on. 4834 on: enable the feature 4835 4836 page_poison= [KNL,EARLY] Boot-time parameter changing the state of 4837 poisoning on the buddy allocator, available with 4838 CONFIG_PAGE_POISONING=y. 4839 off: turn off poisoning (default) 4840 on: turn on poisoning 4841 4842 page_reporting.page_reporting_order= 4843 [KNL] Minimal page reporting order 4844 Format: <integer> 4845 Adjust the minimal page reporting order. The page 4846 reporting is disabled when it exceeds MAX_PAGE_ORDER. 4847 4848 page_reporting.page_reporting_delay_ms= 4849 [KNL] Free page reporting delay in milliseconds 4850 Format: <unsigned integer> 4851 Adjust the delay in milliseconds between free page 4852 reporting intervals. Default is 2000 (2 seconds). 4853 4854 panic= [KNL] Kernel behaviour on panic: delay <timeout> 4855 timeout > 0: seconds before rebooting 4856 timeout = 0: wait forever 4857 timeout < 0: reboot immediately 4858 Format: <timeout> 4859 4860 panic_on_taint= [KNL,EARLY] 4861 Bitmask for conditionally calling panic() in add_taint() 4862 Format: <hex>[,nousertaint] 4863 Hexadecimal bitmask representing the set of TAINT flags 4864 that will cause the kernel to panic when add_taint() is 4865 called with any of the flags in this set. 4866 The optional switch "nousertaint" can be utilized to 4867 prevent userspace forced crashes by writing to sysctl 4868 /proc/sys/kernel/tainted any flagset matching with the 4869 bitmask set on panic_on_taint. 4870 See Documentation/admin-guide/tainted-kernels.rst for 4871 extra details on the taint flags that users can pick 4872 to compose the bitmask to assign to panic_on_taint. 4873 4874 panic_on_warn=1 panic() instead of WARN(). Useful to cause kdump 4875 on a WARN(). 4876 4877 panic_force_cpu= 4878 [KNL,SMP] Force panic handling to execute on a specific CPU. 4879 Format: <cpu number> 4880 Some platforms require panic handling to occur on a 4881 specific CPU for the crash kernel to function correctly. 4882 This can be due to firmware limitations, interrupt routing 4883 constraints, or platform-specific requirements where only 4884 a particular CPU can safely enter the crash kernel. 4885 When set, panic() will redirect execution to the specified 4886 CPU before proceeding with the normal panic and kexec flow. 4887 If the target CPU is offline or unavailable, panic proceeds 4888 on the current CPU. 4889 This option should only be used for systems with the above 4890 constraints as it might cause the panic operation to be less reliable. 4891 4892 panic_print= Bitmask for printing system info when panic happens. 4893 User can chose combination of the following bits: 4894 4895 bit 0: print all tasks info 4896 bit 1: print system memory info 4897 bit 2: print timer info 4898 bit 3: print locks info if CONFIG_LOCKDEP is on 4899 bit 4: print ftrace buffer 4900 bit 5: replay all kernel messages on consoles at the end of panic 4901 bit 6: print all CPUs backtrace (if available in the arch) 4902 bit 7: print only tasks in uninterruptible (blocked) state 4903 4904 *Be aware* that this option may print a _lot_ of lines. 4905 4906 There is an increased risk of losing older messages in 4907 the log. Maybe worth to setup a bigger log buffer with 4908 "log_buf_len" along with this. 4909 4910 Also consider using "printk.console_no_auto_verbose=Y" 4911 if using this along with pstore, to avoid extra delays 4912 due to increased console verbosity during panic. 4913 4914 panic_sys_info= A comma separated list of extra information to be dumped 4915 on panic. 4916 Format: val[,val...] 4917 Where @val can be any of the following: 4918 4919 tasks: print all tasks info 4920 mem: print system memory info 4921 timers: print timers info 4922 locks: print locks info if CONFIG_LOCKDEP is on 4923 ftrace: print ftrace buffer 4924 all_bt: print all CPUs backtrace (if available in the arch) 4925 blocked_tasks: print only tasks in uninterruptible (blocked) state 4926 4927 This is a human readable alternative to the 'panic_print' 4928 option above. Please be aware of the caveats using this, 4929 check the 'panic_print' parameter for more details. 4930 4931 panic_console_replay 4932 When panic happens, replay all kernel messages on 4933 consoles at the end of panic. 4934 4935 parkbd.port= [HW] Parallel port number the keyboard adapter is 4936 connected to, default is 0. 4937 Format: <parport#> 4938 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 4939 0 for XT, 1 for AT (default is AT). 4940 Format: <mode> 4941 4942 parport= [HW,PPT] Specify parallel ports. 0 disables. 4943 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 4944 Use 'auto' to force the driver to use any 4945 IRQ/DMA settings detected (the default is to 4946 ignore detected IRQ/DMA settings because of 4947 possible conflicts). You can specify the base 4948 address, IRQ, and DMA settings; IRQ and DMA 4949 should be numbers, or 'auto' (for using detected 4950 settings on that particular port), or 'nofifo' 4951 (to avoid using a FIFO even if it is detected). 4952 Parallel ports are assigned in the order they 4953 are specified on the command line, starting 4954 with parport0. 4955 4956 parport_init_mode= [HW,PPT] 4957 Configure VIA parallel port to operate in 4958 a specific mode. This is necessary on Pegasos 4959 computer where firmware has no options for setting 4960 up parallel port mode and sets it to spp. 4961 Currently this function knows 686a and 8231 chips. 4962 Format: [spp|ps2|epp|ecp|ecpepp] 4963 4964 pata_legacy.all= [HW,LIBATA] 4965 Format: <int> 4966 Set to non-zero to probe primary and secondary ISA 4967 port ranges on PCI systems where no PCI PATA device 4968 has been found at either range. Disabled by default. 4969 4970 pata_legacy.autospeed= [HW,LIBATA] 4971 Format: <int> 4972 Set to non-zero if a chip is present that snoops speed 4973 changes. Disabled by default. 4974 4975 pata_legacy.iordy_mask= [HW,LIBATA] 4976 Format: <int> 4977 IORDY enable mask. Set individual bits to allow IORDY 4978 for the respective channel. Bit 0 is for the first 4979 legacy channel handled by this driver, bit 1 is for 4980 the second channel, and so on. The sequence will often 4981 correspond to the primary legacy channel, the secondary 4982 legacy channel, and so on, but the handling of a PCI 4983 bus and the use of other driver options may interfere 4984 with the sequence. By default IORDY is allowed across 4985 all channels. 4986 4987 pata_legacy.pio_mask= [HW,LIBATA] 4988 Format: <int> 4989 PIO mode mask for autospeed devices. Set individual 4990 bits to allow the use of the respective PIO modes. 4991 Bit 0 is for mode 0, bit 1 is for mode 1, and so on. 4992 All modes allowed by default. 4993 4994 pata_legacy.probe_all= [HW,LIBATA] 4995 Format: <int> 4996 Set to non-zero to probe tertiary and further ISA 4997 port ranges on PCI systems. Disabled by default. 4998 4999 pata_legacy.probe_mask= [HW,LIBATA] 5000 Format: <int> 5001 Probe mask for legacy ISA PATA ports. Depending on 5002 platform configuration and the use of other driver 5003 options up to 6 legacy ports are supported: 0x1f0, 5004 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing 5005 of individual ports can be disabled by setting the 5006 corresponding bits in the mask to 1. Bit 0 is for 5007 the first port in the list above (0x1f0), and so on. 5008 By default all supported ports are probed. 5009 5010 pata_platform.pio_mask= [HW,LIBATA] 5011 Format: <int> 5012 Supported PIO mode mask. Set individual bits to allow 5013 the use of the respective PIO modes. Bit 0 is for 5014 mode 0, bit 1 is for mode 1, and so on. Mode 0 only 5015 allowed by default. 5016 5017 pause_on_oops=<int> 5018 Halt all CPUs after the first oops has been printed for 5019 the specified number of seconds. This is to be used if 5020 your oopses keep scrolling off the screen. 5021 5022 pci=option[,option...] [PCI,EARLY] various PCI subsystem options. 5023 5024 Some options herein operate on a specific device 5025 or a set of devices (<pci_dev>). These are 5026 specified in one of the following formats: 5027 5028 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 5029 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 5030 5031 Note: the first format specifies a PCI 5032 bus/device/function address which may change 5033 if new hardware is inserted, if motherboard 5034 firmware changes, or due to changes caused 5035 by other kernel parameters. If the 5036 domain is left unspecified, it is 5037 taken to be zero. Optionally, a path 5038 to a device through multiple device/function 5039 addresses can be specified after the base 5040 address (this is more robust against 5041 renumbering issues). The second format 5042 selects devices using IDs from the 5043 configuration space which may match multiple 5044 devices in the system. 5045 5046 earlydump dump PCI config space before the kernel 5047 changes anything 5048 off [X86] don't probe for the PCI bus 5049 bios [X86-32] force use of PCI BIOS, don't access 5050 the hardware directly. Use this if your machine 5051 has a non-standard PCI host bridge. 5052 nobios [X86-32] disallow use of PCI BIOS, only direct 5053 hardware access methods are allowed. Use this 5054 if you experience crashes upon bootup and you 5055 suspect they are caused by the BIOS. 5056 conf1 [X86] Force use of PCI Configuration Access 5057 Mechanism 1 (config address in IO port 0xCF8, 5058 data in IO port 0xCFC, both 32-bit). 5059 conf2 [X86] Force use of PCI Configuration Access 5060 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 5061 the function, IO port 0xCFA, also 8-bit, sets 5062 bus number. The config space is then accessed 5063 through ports 0xC000-0xCFFF). 5064 See http://wiki.osdev.org/PCI for more info 5065 on the configuration access mechanisms. 5066 noaer [PCIE] If the PCIEAER kernel config parameter is 5067 enabled, this kernel boot option can be used to 5068 disable the use of PCIE advanced error reporting. 5069 nodomains [PCI] Disable support for multiple PCI 5070 root domains (aka PCI segments, in ACPI-speak). 5071 nommconf [X86] Disable use of MMCONFIG for PCI 5072 Configuration 5073 check_enable_amd_mmconf [X86] check for and enable 5074 properly configured MMIO access to PCI 5075 config space on AMD family 10h CPU 5076 nomsi [MSI] If the PCI_MSI kernel config parameter is 5077 enabled, this kernel boot option can be used to 5078 disable the use of MSI interrupts system-wide. 5079 noioapicquirk [APIC] Disable all boot interrupt quirks. 5080 Safety option to keep boot IRQs enabled. This 5081 should never be necessary. 5082 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 5083 primary IO-APIC for bridges that cannot disable 5084 boot IRQs. This fixes a source of spurious IRQs 5085 when the system masks IRQs. 5086 noioapicreroute [APIC] Disable workaround that uses the 5087 boot IRQ equivalent of an IRQ that connects to 5088 a chipset where boot IRQs cannot be disabled. 5089 The opposite of ioapicreroute. 5090 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 5091 routing table. These calls are known to be buggy 5092 on several machines and they hang the machine 5093 when used, but on other computers it's the only 5094 way to get the interrupt routing table. Try 5095 this option if the kernel is unable to allocate 5096 IRQs or discover secondary PCI buses on your 5097 motherboard. 5098 rom [X86] Assign address space to expansion ROMs. 5099 Use with caution as certain devices share 5100 address decoders between ROMs and other 5101 resources. 5102 norom [X86] Do not assign address space to 5103 expansion ROMs that do not already have 5104 BIOS assigned address ranges. 5105 nobar [X86] Do not assign address space to the 5106 BARs that weren't assigned by the BIOS. 5107 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 5108 assigned automatically to PCI devices. You can 5109 make the kernel exclude IRQs of your ISA cards 5110 this way. 5111 pirqaddr=0xAAAAA [X86] Specify the physical address 5112 of the PIRQ table (normally generated 5113 by the BIOS) if it is outside the 5114 F0000h-100000h range. 5115 lastbus=N [X86] Scan all buses thru bus #N. Can be 5116 useful if the kernel is unable to find your 5117 secondary buses and you want to tell it 5118 explicitly which ones they are. 5119 assign-busses [X86] Always assign all PCI bus 5120 numbers ourselves, overriding 5121 whatever the firmware may have done. 5122 usepirqmask [X86] Honor the possible IRQ mask stored 5123 in the BIOS $PIR table. This is needed on 5124 some systems with broken BIOSes, notably 5125 some HP Pavilion N5400 and Omnibook XE3 5126 notebooks. This will have no effect if ACPI 5127 IRQ routing is enabled. 5128 noacpi [X86] Do not use ACPI for IRQ routing 5129 or for PCI scanning. 5130 use_crs [X86] Use PCI host bridge window information 5131 from ACPI. On BIOSes from 2008 or later, this 5132 is enabled by default. If you need to use this, 5133 please report a bug. 5134 nocrs [X86] Ignore PCI host bridge windows from ACPI. 5135 If you need to use this, please report a bug. 5136 use_e820 [X86] Use E820 reservations to exclude parts of 5137 PCI host bridge windows. This is a workaround 5138 for BIOS defects in host bridge _CRS methods. 5139 If you need to use this, please report a bug to 5140 <linux-pci@vger.kernel.org>. 5141 no_e820 [X86] Ignore E820 reservations for PCI host 5142 bridge windows. This is the default on modern 5143 hardware. If you need to use this, please report 5144 a bug to <linux-pci@vger.kernel.org>. 5145 routeirq Do IRQ routing for all PCI devices. 5146 This is normally done in pci_enable_device(), 5147 so this option is a temporary workaround 5148 for broken drivers that don't call it. 5149 skip_isa_align [X86] do not align io start addr, so can 5150 handle more pci cards 5151 noearly [X86] Don't do any early type 1 scanning. 5152 This might help on some broken boards which 5153 machine check when some devices' config space 5154 is read. But various workarounds are disabled 5155 and some IOMMU drivers will not work. 5156 bfsort Sort PCI devices into breadth-first order. 5157 This sorting is done to get a device 5158 order compatible with older (<= 2.4) kernels. 5159 nobfsort Don't sort PCI devices into breadth-first order. 5160 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 5161 tuning and use the BIOS-configured MPS defaults. 5162 pcie_bus_safe Set every device's MPS to the largest value 5163 supported by all devices below the root complex. 5164 pcie_bus_perf Set device MPS to the largest allowable MPS 5165 based on its parent bus. Also set MRRS (Max 5166 Read Request Size) to the largest supported 5167 value (no larger than the MPS that the device 5168 or bus can support) for best performance. 5169 pcie_bus_peer2peer Set every device's MPS to 128B, which 5170 every device is guaranteed to support. This 5171 configuration allows peer-to-peer DMA between 5172 any pair of devices, possibly at the cost of 5173 reduced performance. This also guarantees 5174 that hot-added devices will work. 5175 cbiosize=nn[KMG] The fixed amount of bus space which is 5176 reserved for the CardBus bridge's IO window. 5177 The default value is 256 bytes. 5178 cbmemsize=nn[KMG] The fixed amount of bus space which is 5179 reserved for the CardBus bridge's memory 5180 window. The default value is 64 megabytes. 5181 resource_alignment= 5182 Format: 5183 [<order of align>@]<pci_dev>[; ...] 5184 Specifies alignment and device to reassign 5185 aligned memory resources. How to 5186 specify the device is described above. 5187 If <order of align> is not specified, 5188 PAGE_SIZE is used as alignment. 5189 A PCI-PCI bridge can be specified if resource 5190 windows need to be expanded. 5191 To specify the alignment for several 5192 instances of a device, the PCI vendor, 5193 device, subvendor, and subdevice may be 5194 specified, e.g., 12@pci:8086:9c22:103c:198f 5195 for 4096-byte alignment. 5196 ecrc= Enable/disable PCIe ECRC (transaction layer 5197 end-to-end CRC checking). Only effective if 5198 OS has native AER control (either granted by 5199 ACPI _OSC or forced via "pcie_ports=native") 5200 bios: Use BIOS/firmware settings. This is the 5201 the default. 5202 off: Turn ECRC off 5203 on: Turn ECRC on. 5204 hpiosize=nn[KMG] The fixed amount of bus space which is 5205 reserved for hotplug bridge's IO window. 5206 Default size is 256 bytes. 5207 hpmmiosize=nn[KMG] The fixed amount of bus space which is 5208 reserved for hotplug bridge's MMIO window. 5209 Default size is 2 megabytes. 5210 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 5211 reserved for hotplug bridge's MMIO_PREF window. 5212 Default size is 2 megabytes. 5213 hpmemsize=nn[KMG] The fixed amount of bus space which is 5214 reserved for hotplug bridge's MMIO and 5215 MMIO_PREF window. 5216 Default size is 2 megabytes. 5217 hpbussize=nn The minimum amount of additional bus numbers 5218 reserved for buses below a hotplug bridge. 5219 Default is 1. 5220 realloc= Enable/disable reallocating PCI bridge resources 5221 if allocations done by BIOS are too small to 5222 accommodate resources required by all child 5223 devices. 5224 off: Turn realloc off 5225 on: Turn realloc on 5226 realloc same as realloc=on 5227 noari do not use PCIe ARI. 5228 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 5229 do not use PCIe ATS (and IOMMU device IOTLB). 5230 pcie_scan_all Scan all possible PCIe devices. Otherwise we 5231 only look for one device below a PCIe downstream 5232 port. 5233 big_root_window Try to add a big 64bit memory window to the PCIe 5234 root complex on AMD CPUs. Some GFX hardware 5235 can resize a BAR to allow access to all VRAM. 5236 Adding the window is slightly risky (it may 5237 conflict with unreported devices), so this 5238 taints the kernel. 5239 disable_acs_redir=<pci_dev>[; ...] 5240 Specify one or more PCI devices (in the format 5241 specified above) separated by semicolons. 5242 Each device specified will have the PCI ACS 5243 redirect capabilities forced off which will 5244 allow P2P traffic between devices through 5245 bridges without forcing it upstream. Note: 5246 this removes isolation between devices and 5247 may put more devices in an IOMMU group. 5248 config_acs= 5249 Format: 5250 <ACS flags>@<pci_dev>[; ...] 5251 Specify one or more PCI devices (in the format 5252 specified above) optionally prepended with flags 5253 and separated by semicolons. The respective 5254 capabilities will be enabled, disabled or 5255 unchanged based on what is specified in 5256 flags. 5257 5258 ACS Flags is defined as follows: 5259 bit-0 : ACS Source Validation 5260 bit-1 : ACS Translation Blocking 5261 bit-2 : ACS P2P Request Redirect 5262 bit-3 : ACS P2P Completion Redirect 5263 bit-4 : ACS Upstream Forwarding 5264 bit-5 : ACS P2P Egress Control 5265 bit-6 : ACS Direct Translated P2P 5266 Each bit can be marked as: 5267 '0' – force disabled 5268 '1' – force enabled 5269 'x' – unchanged 5270 For example, 5271 pci=config_acs=10x@pci:0:0 5272 would configure all devices that support 5273 ACS to enable P2P Request Redirect, disable 5274 Translation Blocking, and leave Source 5275 Validation unchanged from whatever power-up 5276 or firmware set it to. 5277 5278 Note: this may remove isolation between devices 5279 and may put more devices in an IOMMU group. 5280 force_floating [S390] Force usage of floating interrupts. 5281 nomio [S390] Do not use MIO instructions. 5282 norid [S390] ignore the RID field and force use of 5283 one PCI domain per PCI function 5284 notph [PCIE] If the PCIE_TPH kernel config parameter 5285 is enabled, this kernel boot option can be used 5286 to disable PCIe TLP Processing Hints support 5287 system-wide. 5288 5289 pcie_aspm= [PCIE] Forcibly enable or ignore PCIe Active State Power 5290 Management. 5291 off Don't touch ASPM configuration at all. Leave any 5292 configuration done by firmware unchanged. 5293 force Enable ASPM even on devices that claim not to support it. 5294 WARNING: Forcing ASPM on may cause system lockups. 5295 5296 pcie_ports= [PCIE] PCIe port services handling: 5297 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 5298 even if the platform doesn't give the OS permission to 5299 use them. This may cause conflicts if the platform 5300 also tries to use these services. 5301 dpc-native Use native PCIe service for DPC only. May 5302 cause conflicts if firmware uses AER or DPC. 5303 compat Disable native PCIe services (PME, AER, DPC, PCIe 5304 hotplug). 5305 5306 pcie_port_pm= [PCIE] PCIe port power management handling: 5307 off Disable power management of all PCIe ports 5308 force Forcibly enable power management of all PCIe ports 5309 5310 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 5311 nomsi Do not use MSI for native PCIe PME signaling (this makes 5312 all PCIe root ports use INTx for all services). 5313 5314 pd_ignore_unused 5315 [PM] 5316 Keep all power-domains already enabled by bootloader on, 5317 even if no driver has claimed them. This is useful 5318 for debug and development, but should not be 5319 needed on a platform with proper driver support. 5320 5321 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 5322 boot time. 5323 Format: { 0 | 1 } 5324 See arch/parisc/kernel/pdc_chassis.c 5325 5326 percpu_alloc= [MM,EARLY] 5327 Select which percpu first chunk allocator to use. 5328 Currently supported values are "embed" and "page". 5329 Archs may support subset or none of the selections. 5330 See comments in mm/percpu.c for details on each 5331 allocator. This parameter is primarily for debugging 5332 and performance comparison. 5333 5334 pirq= [SMP,APIC] Manual mp-table setup 5335 See Documentation/arch/x86/i386/IO-APIC.rst. 5336 5337 plip= [PPT,NET] Parallel port network link 5338 Format: { parport<nr> | timid | 0 } 5339 See also Documentation/admin-guide/parport.rst. 5340 5341 pmtmr= [X86] Manual setup of pmtmr I/O Port. 5342 Override pmtimer IOPort with a hex value. 5343 e.g. pmtmr=0x508 5344 5345 pmu_override= [PPC] Override the PMU. 5346 This option takes over the PMU facility, so it is no 5347 longer usable by perf. Setting this option starts the 5348 PMU counters by setting MMCR0 to 0 (the FC bit is 5349 cleared). If a number is given, then MMCR1 is set to 5350 that number, otherwise (e.g., 'pmu_override=on'), MMCR1 5351 remains 0. 5352 5353 pm_async= [PM] 5354 Format: off 5355 This parameter sets the initial value of the 5356 /sys/power/pm_async sysfs knob at boot time. 5357 If set to "off", disables asynchronous suspend and 5358 resume of devices during system-wide power transitions. 5359 This can be useful on platforms where device 5360 dependencies are not well-defined, or for debugging 5361 power management issues. Asynchronous operations are 5362 enabled by default. 5363 5364 5365 pm_debug_messages [SUSPEND,KNL] 5366 Enable suspend/resume debug messages during boot up. 5367 5368 pm.dpm_watchdog_enabled= 5369 [PM] Enable or disable the DPM watchdog. Requires 5370 CONFIG_PM_SLEEP and CONFIG_DPM_WATCHDOG enabled. 5371 Format: <bool> 5372 Default value is set by CONFIG_DPM_WATCHDOG_ENABLED. 5373 5374 pnp.debug=1 [PNP] 5375 Enable PNP debug messages (depends on the 5376 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 5377 via /sys/module/pnp/parameters/debug. We always show 5378 current resource usage; turning this on also shows 5379 possible settings and some assignment information. 5380 5381 pnpacpi= [ACPI] 5382 { off } 5383 5384 pnpbios= [ISAPNP] 5385 { on | off | curr | res | no-curr | no-res } 5386 5387 pnp_reserve_irq= 5388 [ISAPNP] Exclude IRQs for the autoconfiguration 5389 5390 pnp_reserve_dma= 5391 [ISAPNP] Exclude DMAs for the autoconfiguration 5392 5393 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 5394 Ranges are in pairs (I/O port base and size). 5395 5396 pnp_reserve_mem= 5397 [ISAPNP] Exclude memory regions for the 5398 autoconfiguration. 5399 Ranges are in pairs (memory base and size). 5400 5401 ports= [IP_VS_FTP] IPVS ftp helper module 5402 Default is 21. 5403 Up to 8 (IP_VS_APP_MAX_PORTS) ports 5404 may be specified. 5405 Format: <port>,<port>.... 5406 5407 possible_cpus= [SMP,S390,X86] 5408 Format: <unsigned int> 5409 Set the number of possible CPUs, overriding the 5410 regular discovery mechanisms (such as ACPI/FW, etc). 5411 5412 powersave=off [PPC] This option disables power saving features. 5413 It specifically disables cpuidle and sets the 5414 platform machine description specific power_save 5415 function to NULL. On Idle the CPU just reduces 5416 execution priority. 5417 5418 ppc_strict_facility_enable 5419 [PPC,ENABLE] This option catches any kernel floating point, 5420 Altivec, VSX and SPE outside of regions specifically 5421 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 5422 There is some performance impact when enabling this. 5423 5424 ppc_tm= [PPC,EARLY] 5425 Format: {"off"} 5426 Disable Hardware Transactional Memory 5427 5428 preempt= [KNL] 5429 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC 5430 none - Limited to cond_resched() calls 5431 voluntary - Limited to cond_resched() and might_sleep() calls 5432 full - Any section that isn't explicitly preempt disabled 5433 can be preempted anytime. Tasks will also yield 5434 contended spinlocks (if the critical section isn't 5435 explicitly preempt disabled beyond the lock itself). 5436 lazy - Scheduler controlled. Similar to full but instead 5437 of preempting the task immediately, the task gets 5438 one HZ tick time to yield itself before the 5439 preemption will be forced. One such preemption 5440 point is when the task returns to user space. 5441 5442 print-fatal-signals= 5443 [KNL] debug: print fatal signals 5444 5445 If enabled, warn about various signal handling 5446 related application anomalies: too many signals, 5447 too many POSIX.1 timers, fatal signals causing a 5448 coredump - etc. 5449 5450 If you hit the warning due to signal overflow, 5451 you might want to try "ulimit -i unlimited". 5452 5453 default: off. 5454 5455 printk.always_kmsg_dump= 5456 Trigger kmsg_dump for cases other than kernel oops or 5457 panics 5458 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5459 default: disabled 5460 5461 printk.console_no_auto_verbose= 5462 Disable console loglevel raise on oops, panic 5463 or lockdep-detected issues (only if lock debug is on). 5464 With an exception to setups with low baudrate on 5465 serial console, keeping this 0 is a good choice 5466 in order to provide more debug information. 5467 Format: <bool> 5468 default: 0 (auto_verbose is enabled) 5469 5470 printk.debug_non_panic_cpus= 5471 Allows storing messages from non-panic CPUs into 5472 the printk log buffer during panic(). They are 5473 flushed to consoles by the panic-CPU on 5474 a best-effort basis. 5475 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5476 Default: disabled 5477 5478 printk.devkmsg={on,off,ratelimit} 5479 Control writing to /dev/kmsg. 5480 on - unlimited logging to /dev/kmsg from userspace 5481 off - logging to /dev/kmsg disabled 5482 ratelimit - ratelimit the logging 5483 Default: ratelimit 5484 5485 printk.time= Show timing data prefixed to each printk message line 5486 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5487 5488 proc_mem.force_override= [KNL] 5489 Format: {always | ptrace | never} 5490 Traditionally /proc/pid/mem allows memory permissions to be 5491 overridden without restrictions. This option may be set to 5492 restrict that. Can be one of: 5493 - 'always': traditional behavior always allows mem overrides. 5494 - 'ptrace': only allow mem overrides for active ptracers. 5495 - 'never': never allow mem overrides. 5496 If not specified, default is the CONFIG_PROC_MEM_* choice. 5497 5498 processor.max_cstate= [HW,ACPI] 5499 Limit processor to maximum C-state 5500 max_cstate=9 overrides any DMI blacklist limit. 5501 5502 processor.nocst [HW,ACPI] 5503 Ignore the _CST method to determine C-states, 5504 instead using the legacy FADT method 5505 5506 profile= [KNL] Enable kernel profiling via /proc/profile 5507 Format: [<profiletype>,]<number> 5508 Param: <profiletype>: "schedule" or "kvm" 5509 [defaults to kernel profiling] 5510 Param: "schedule" - profile schedule points. 5511 Param: "kvm" - profile VM exits. 5512 Param: <number> - step/bucket size as a power of 2 for 5513 statistical time based profiling. 5514 5515 prot_virt= [S390] enable hosting protected virtual machines 5516 isolated from the hypervisor (if hardware supports 5517 that). If enabled, the default kernel base address 5518 might be overridden even when Kernel Address Space 5519 Layout Randomization is disabled. 5520 Format: <bool> 5521 5522 psi= [KNL] Enable or disable pressure stall information 5523 tracking. 5524 Format: <bool> 5525 5526 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 5527 probe for; one of (bare|imps|exps|lifebook|any). 5528 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 5529 per second. 5530 psmouse.resetafter= [HW,MOUSE] 5531 Try to reset the device after so many bad packets 5532 (0 = never). 5533 psmouse.resolution= 5534 [HW,MOUSE] Set desired mouse resolution, in dpi. 5535 psmouse.smartscroll= 5536 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 5537 0 = disabled, 1 = enabled (default). 5538 5539 pstore.backend= Specify the name of the pstore backend to use 5540 5541 pti= [X86-64] Control Page Table Isolation of user and 5542 kernel address spaces. Disabling this feature 5543 removes hardening, but improves performance of 5544 system calls and interrupts. 5545 5546 on - unconditionally enable 5547 off - unconditionally disable 5548 auto - kernel detects whether your CPU model is 5549 vulnerable to issues that PTI mitigates 5550 5551 Not specifying this option is equivalent to pti=auto. 5552 5553 pty.legacy_count= 5554 [KNL] Number of legacy pty's. Overwrites compiled-in 5555 default number. 5556 5557 quiet [KNL,EARLY] Disable most log messages 5558 5559 r128= [HW,DRM] 5560 5561 radix_hcall_invalidate=on [PPC/PSERIES] 5562 Disable RADIX GTSE feature and use hcall for TLB 5563 invalidate. 5564 5565 raid= [HW,RAID] 5566 See Documentation/admin-guide/md.rst. 5567 5568 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 5569 See Documentation/admin-guide/blockdev/ramdisk.rst. 5570 5571 ramdisk_start= [Deprecated,RAM] RAM disk image start address 5572 5573 random.trust_cpu=off 5574 [KNL,EARLY] Disable trusting the use of the CPU's 5575 random number generator (if available) to 5576 initialize the kernel's RNG. 5577 5578 random.trust_bootloader=off 5579 [KNL,EARLY] Disable trusting the use of the a seed 5580 passed by the bootloader (if available) to 5581 initialize the kernel's RNG. 5582 5583 randomize_kstack_offset= 5584 [KNL,EARLY] Enable or disable kernel stack offset 5585 randomization, which provides roughly 5 bits of 5586 entropy, frustrating memory corruption attacks 5587 that depend on stack address determinism or 5588 cross-syscall address exposures. This is only 5589 available on architectures that have defined 5590 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET. 5591 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5592 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT. 5593 5594 ras=option[,option,...] [KNL] RAS-specific options 5595 5596 cec_disable [X86] 5597 Disable the Correctable Errors Collector, 5598 see CONFIG_RAS_CEC help text. 5599 5600 rcu_nocbs[=cpu-list] 5601 [KNL] The optional argument is a cpu list, 5602 as described above. 5603 5604 In kernels built with CONFIG_RCU_NOCB_CPU=y, 5605 enable the no-callback CPU mode, which prevents 5606 such CPUs' callbacks from being invoked in 5607 softirq context. Invocation of such CPUs' RCU 5608 callbacks will instead be offloaded to "rcuox/N" 5609 kthreads created for that purpose, where "x" is 5610 "p" for RCU-preempt, "s" for RCU-sched, and "g" 5611 for the kthreads that mediate grace periods; and 5612 "N" is the CPU number. This reduces OS jitter on 5613 the offloaded CPUs, which can be useful for HPC 5614 and real-time workloads. It can also improve 5615 energy efficiency for asymmetric multiprocessors. 5616 5617 If a cpulist is passed as an argument, the specified 5618 list of CPUs is set to no-callback mode from boot. 5619 5620 Otherwise, if the '=' sign and the cpulist 5621 arguments are omitted, no CPU will be set to 5622 no-callback mode from boot but the mode may be 5623 toggled at runtime via cpusets. 5624 5625 Note that this argument takes precedence over 5626 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 5627 5628 rcu_nocb_poll [KNL] 5629 Rather than requiring that offloaded CPUs 5630 (specified by rcu_nocbs= above) explicitly 5631 awaken the corresponding "rcuoN" kthreads, 5632 make these kthreads poll for callbacks. 5633 This improves the real-time response for the 5634 offloaded CPUs by relieving them of the need to 5635 wake up the corresponding kthread, but degrades 5636 energy efficiency by requiring that the kthreads 5637 periodically wake up to do the polling. 5638 5639 rcutree.blimit= [KNL] 5640 Set maximum number of finished RCU callbacks to 5641 process in one batch. 5642 5643 rcutree.csd_lock_suppress_rcu_stall= [KNL] 5644 Do only a one-line RCU CPU stall warning when 5645 there is an ongoing too-long CSD-lock wait. 5646 5647 rcutree.do_rcu_barrier= [KNL] 5648 Wait for deferred kfree_rcu() frees and ordinary 5649 call_rcu() callbacks queued before this request to 5650 complete. This does not prevent new work from being 5651 queued concurrently. Requests are throttled so that 5652 userspace tests can safely hammer on the sysfs 5653 variable if they so choose. 5654 If triggered before the RCU grace-period machinery 5655 is fully active, this will error out with EAGAIN. 5656 5657 rcutree.dump_tree= [KNL] 5658 Dump the structure of the rcu_node combining tree 5659 out at early boot. This is used for diagnostic 5660 purposes, to verify correct tree setup. 5661 5662 rcutree.gp_cleanup_delay= [KNL] 5663 Set the number of jiffies to delay each step of 5664 RCU grace-period cleanup. 5665 5666 rcutree.gp_init_delay= [KNL] 5667 Set the number of jiffies to delay each step of 5668 RCU grace-period initialization. 5669 5670 rcutree.gp_preinit_delay= [KNL] 5671 Set the number of jiffies to delay each step of 5672 RCU grace-period pre-initialization, that is, 5673 the propagation of recent CPU-hotplug changes up 5674 the rcu_node combining tree. 5675 5676 rcutree.jiffies_till_first_fqs= [KNL] 5677 Set delay from grace-period initialization to 5678 first attempt to force quiescent states. 5679 Units are jiffies, minimum value is zero, 5680 and maximum value is HZ. 5681 5682 rcutree.jiffies_till_next_fqs= [KNL] 5683 Set delay between subsequent attempts to force 5684 quiescent states. Units are jiffies, minimum 5685 value is one, and maximum value is HZ. 5686 5687 rcutree.jiffies_till_sched_qs= [KNL] 5688 Set required age in jiffies for a 5689 given grace period before RCU starts 5690 soliciting quiescent-state help from 5691 rcu_note_context_switch() and cond_resched(). 5692 If not specified, the kernel will calculate 5693 a value based on the most recent settings 5694 of rcutree.jiffies_till_first_fqs 5695 and rcutree.jiffies_till_next_fqs. 5696 This calculated value may be viewed in 5697 rcutree.jiffies_to_sched_qs. Any attempt to set 5698 rcutree.jiffies_to_sched_qs will be cheerfully 5699 overwritten. 5700 5701 rcutree.kthread_prio= [KNL,BOOT] 5702 Set the SCHED_FIFO priority of the RCU per-CPU 5703 kthreads (rcuc/N). This value is also used for 5704 the priority of the RCU boost threads (rcub/N) 5705 and for the RCU grace-period kthreads (rcu_bh, 5706 rcu_preempt, and rcu_sched). If RCU_BOOST is 5707 set, valid values are 1-99 and the default is 1 5708 (the least-favored priority). Otherwise, when 5709 RCU_BOOST is not set, valid values are 0-99 and 5710 the default is zero (non-realtime operation). 5711 When RCU_NOCB_CPU is set, also adjust the 5712 priority of NOCB callback kthreads. 5713 5714 rcutree.nocb_nobypass_lim_per_jiffy= [KNL] 5715 On callback-offloaded (rcu_nocbs) CPUs, 5716 RCU reduces the lock contention that would 5717 otherwise be caused by callback floods through 5718 use of the ->nocb_bypass list. However, in the 5719 common non-flooded case, RCU queues directly to 5720 the main ->cblist in order to avoid the extra 5721 overhead of the ->nocb_bypass list and its lock. 5722 But if there are too many callbacks queued during 5723 a single jiffy, RCU pre-queues the callbacks into 5724 the ->nocb_bypass queue. The definition of "too 5725 many" is supplied by this kernel boot parameter. 5726 5727 rcutree.nohz_full_patience_delay= [KNL] 5728 On callback-offloaded (rcu_nocbs) CPUs, avoid 5729 disturbing RCU unless the grace period has 5730 reached the specified age in milliseconds. 5731 Defaults to zero. Large values will be capped 5732 at five seconds. All values will be rounded down 5733 to the nearest value representable by jiffies. 5734 5735 rcutree.qhimark= [KNL] 5736 Set threshold of queued RCU callbacks beyond which 5737 batch limiting is disabled. 5738 5739 rcutree.qlowmark= [KNL] 5740 Set threshold of queued RCU callbacks below which 5741 batch limiting is re-enabled. 5742 5743 rcutree.qovld= [KNL] 5744 Set threshold of queued RCU callbacks beyond which 5745 RCU's force-quiescent-state scan will aggressively 5746 enlist help from cond_resched() and sched IPIs to 5747 help CPUs more quickly reach quiescent states. 5748 Set to less than zero to make this be set based 5749 on rcutree.qhimark at boot time and to zero to 5750 disable more aggressive help enlistment. 5751 5752 rcutree.rcu_delay_page_cache_fill_msec= [KNL] 5753 Set the page-cache refill delay (in milliseconds) 5754 in response to low-memory conditions. The range 5755 of permitted values is in the range 0:100000. 5756 5757 rcutree.rcu_divisor= [KNL] 5758 Set the shift-right count to use to compute 5759 the callback-invocation batch limit bl from 5760 the number of callbacks queued on this CPU. 5761 The result will be bounded below by the value of 5762 the rcutree.blimit kernel parameter. Every bl 5763 callbacks, the softirq handler will exit in 5764 order to allow the CPU to do other work. 5765 5766 Please note that this callback-invocation batch 5767 limit applies only to non-offloaded callback 5768 invocation. Offloaded callbacks are instead 5769 invoked in the context of an rcuoc kthread, which 5770 scheduler will preempt as it does any other task. 5771 5772 rcutree.rcu_fanout_exact= [KNL] 5773 Disable autobalancing of the rcu_node combining 5774 tree. This is used by rcutorture, and might 5775 possibly be useful for architectures having high 5776 cache-to-cache transfer latencies. 5777 5778 rcutree.rcu_fanout_leaf= [KNL] 5779 Change the number of CPUs assigned to each 5780 leaf rcu_node structure. Useful for very 5781 large systems, which will choose the value 64, 5782 and for NUMA systems with large remote-access 5783 latencies, which will choose a value aligned 5784 with the appropriate hardware boundaries. 5785 5786 rcutree.rcu_min_cached_objs= [KNL] 5787 Minimum number of objects which are cached and 5788 maintained per one CPU. Object size is equal 5789 to PAGE_SIZE. The cache allows to reduce the 5790 pressure to page allocator, also it makes the 5791 whole algorithm to behave better in low memory 5792 condition. 5793 5794 rcutree.rcu_nocb_gp_stride= [KNL] 5795 Set the number of NOCB callback kthreads in 5796 each group, which defaults to the square root 5797 of the number of CPUs. Larger numbers reduce 5798 the wakeup overhead on the global grace-period 5799 kthread, but increases that same overhead on 5800 each group's NOCB grace-period kthread. 5801 5802 rcutree.rcu_kick_kthreads= [KNL] 5803 Cause the grace-period kthread to get an extra 5804 wake_up() if it sleeps three times longer than 5805 it should at force-quiescent-state time. 5806 This wake_up() will be accompanied by a 5807 WARN_ONCE() splat and an ftrace_dump(). 5808 5809 rcutree.rcu_resched_ns= [KNL] 5810 Limit the time spend invoking a batch of RCU 5811 callbacks to the specified number of nanoseconds. 5812 By default, this limit is checked only once 5813 every 32 callbacks in order to limit the pain 5814 inflicted by local_clock() overhead. 5815 5816 rcutree.rcu_unlock_delay= [KNL] 5817 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 5818 this specifies an rcu_read_unlock()-time delay 5819 in microseconds. This defaults to zero. 5820 Larger delays increase the probability of 5821 catching RCU pointer leaks, that is, buggy use 5822 of RCU-protected pointers after the relevant 5823 rcu_read_unlock() has completed. 5824 5825 rcutree.sysrq_rcu= [KNL] 5826 Commandeer a sysrq key to dump out Tree RCU's 5827 rcu_node tree with an eye towards determining 5828 why a new grace period has not yet started. 5829 5830 rcutree.use_softirq= [KNL] 5831 If set to zero, move all RCU_SOFTIRQ processing to 5832 per-CPU rcuc kthreads. Defaults to a non-zero 5833 value, meaning that RCU_SOFTIRQ is used by default. 5834 Specify rcutree.use_softirq=0 to use rcuc kthreads. 5835 5836 But note that CONFIG_PREEMPT_RT=y kernels disable 5837 this kernel boot parameter, forcibly setting it 5838 to zero. 5839 5840 rcutree.enable_rcu_lazy= [KNL] 5841 To save power, batch RCU callbacks and flush after 5842 delay, memory pressure or callback list growing too 5843 big. 5844 5845 rcutree.rcu_normal_wake_from_gp= [KNL] 5846 Reduces a latency of synchronize_rcu() call. This approach 5847 maintains its own track of synchronize_rcu() callers, so it 5848 does not interact with regular callbacks because it does not 5849 use a call_rcu[_hurry]() path. Please note, this is for a 5850 normal grace period. 5851 5852 How to disable it: 5853 5854 echo 0 > /sys/module/rcutree/parameters/rcu_normal_wake_from_gp 5855 or pass a boot parameter "rcutree.rcu_normal_wake_from_gp=0" 5856 5857 Default is 1 if it is not explicitly disabled by the boot parameter 5858 passing 0. 5859 5860 rcuscale.gp_async= [KNL] 5861 Measure performance of asynchronous 5862 grace-period primitives such as call_rcu(). 5863 5864 rcuscale.gp_async_max= [KNL] 5865 Specify the maximum number of outstanding 5866 callbacks per writer thread. When a writer 5867 thread exceeds this limit, it invokes the 5868 corresponding flavor of rcu_barrier() to allow 5869 previously posted callbacks to drain. 5870 5871 rcuscale.gp_exp= [KNL] 5872 Measure performance of expedited synchronous 5873 grace-period primitives. 5874 5875 rcuscale.holdoff= [KNL] 5876 Set test-start holdoff period. The purpose of 5877 this parameter is to delay the start of the 5878 test until boot completes in order to avoid 5879 interference. 5880 5881 rcuscale.kfree_by_call_rcu= [KNL] 5882 In kernels built with CONFIG_RCU_LAZY=y, test 5883 call_rcu() instead of kfree_rcu(). 5884 5885 rcuscale.kfree_mult= [KNL] 5886 Instead of allocating an object of size kfree_obj, 5887 allocate one of kfree_mult * sizeof(kfree_obj). 5888 Defaults to 1. 5889 5890 rcuscale.kfree_rcu_test= [KNL] 5891 Set to measure performance of kfree_rcu() flooding. 5892 5893 rcuscale.kfree_rcu_test_double= [KNL] 5894 Test the double-argument variant of kfree_rcu(). 5895 If this parameter has the same value as 5896 rcuscale.kfree_rcu_test_single, both the single- 5897 and double-argument variants are tested. 5898 5899 rcuscale.kfree_rcu_test_single= [KNL] 5900 Test the single-argument variant of kfree_rcu(). 5901 If this parameter has the same value as 5902 rcuscale.kfree_rcu_test_double, both the single- 5903 and double-argument variants are tested. 5904 5905 rcuscale.kfree_nthreads= [KNL] 5906 The number of threads running loops of kfree_rcu(). 5907 5908 rcuscale.kfree_alloc_num= [KNL] 5909 Number of allocations and frees done in an iteration. 5910 5911 rcuscale.kfree_loops= [KNL] 5912 Number of loops doing rcuscale.kfree_alloc_num number 5913 of allocations and frees. 5914 5915 rcuscale.minruntime= [KNL] 5916 Set the minimum test run time in seconds. This 5917 does not affect the data-collection interval, 5918 but instead allows better measurement of things 5919 like CPU consumption. 5920 5921 rcuscale.nreaders= [KNL] 5922 Set number of RCU readers. The value -1 selects 5923 N, where N is the number of CPUs. A value 5924 "n" less than -1 selects N-n+1, where N is again 5925 the number of CPUs. For example, -2 selects N 5926 (the number of CPUs), -3 selects N+1, and so on. 5927 A value of "n" less than or equal to -N selects 5928 a single reader. 5929 5930 rcuscale.nwriters= [KNL] 5931 Set number of RCU writers. The values operate 5932 the same as for rcuscale.nreaders. 5933 N, where N is the number of CPUs 5934 5935 rcuscale.scale_type= [KNL] 5936 Specify the RCU implementation to test. 5937 5938 rcuscale.shutdown= [KNL] 5939 Shut the system down after performance tests 5940 complete. This is useful for hands-off automated 5941 testing. 5942 5943 rcuscale.verbose= [KNL] 5944 Enable additional printk() statements. 5945 5946 rcuscale.writer_holdoff= [KNL] 5947 Write-side holdoff between grace periods, 5948 in microseconds. The default of zero says 5949 no holdoff. 5950 5951 rcuscale.writer_holdoff_jiffies= [KNL] 5952 Additional write-side holdoff between grace 5953 periods, but in jiffies. The default of zero 5954 says no holdoff. 5955 5956 rcutorture.fqs_duration= [KNL] 5957 Set duration of force_quiescent_state bursts 5958 in microseconds. 5959 5960 rcutorture.fqs_holdoff= [KNL] 5961 Set holdoff time within force_quiescent_state bursts 5962 in microseconds. 5963 5964 rcutorture.fqs_stutter= [KNL] 5965 Set wait time between force_quiescent_state bursts 5966 in seconds. 5967 5968 rcutorture.fwd_progress= [KNL] 5969 Specifies the number of kthreads to be used 5970 for RCU grace-period forward-progress testing 5971 for the types of RCU supporting this notion. 5972 Defaults to 1 kthread, values less than zero or 5973 greater than the number of CPUs cause the number 5974 of CPUs to be used. 5975 5976 rcutorture.fwd_progress_div= [KNL] 5977 Specify the fraction of a CPU-stall-warning 5978 period to do tight-loop forward-progress testing. 5979 5980 rcutorture.fwd_progress_holdoff= [KNL] 5981 Number of seconds to wait between successive 5982 forward-progress tests. 5983 5984 rcutorture.fwd_progress_need_resched= [KNL] 5985 Enclose cond_resched() calls within checks for 5986 need_resched() during tight-loop forward-progress 5987 testing. 5988 5989 rcutorture.gp_cond= [KNL] 5990 Use conditional/asynchronous update-side 5991 normal-grace-period primitives, if available. 5992 5993 rcutorture.gp_cond_exp= [KNL] 5994 Use conditional/asynchronous update-side 5995 expedited-grace-period primitives, if available. 5996 5997 rcutorture.gp_cond_full= [KNL] 5998 Use conditional/asynchronous update-side 5999 normal-grace-period primitives that also take 6000 concurrent expedited grace periods into account, 6001 if available. 6002 6003 rcutorture.gp_cond_exp_full= [KNL] 6004 Use conditional/asynchronous update-side 6005 expedited-grace-period primitives that also take 6006 concurrent normal grace periods into account, 6007 if available. 6008 6009 rcutorture.gp_cond_wi= [KNL] 6010 Nominal wait interval for normal conditional 6011 grace periods (specified by rcutorture's 6012 gp_cond and gp_cond_full module parameters), 6013 in microseconds. The actual wait interval will 6014 be randomly selected to nanosecond granularity up 6015 to this wait interval. Defaults to 16 jiffies, 6016 for example, 16,000 microseconds on a system 6017 with HZ=1000. 6018 6019 rcutorture.gp_cond_wi_exp= [KNL] 6020 Nominal wait interval for expedited conditional 6021 grace periods (specified by rcutorture's 6022 gp_cond_exp and gp_cond_exp_full module 6023 parameters), in microseconds. The actual wait 6024 interval will be randomly selected to nanosecond 6025 granularity up to this wait interval. Defaults to 6026 128 microseconds. 6027 6028 rcutorture.gp_exp= [KNL] 6029 Use expedited update-side primitives, if available. 6030 6031 rcutorture.gp_normal= [KNL] 6032 Use normal (non-expedited) asynchronous 6033 update-side primitives, if available. 6034 6035 rcutorture.gp_poll= [KNL] 6036 Use polled update-side normal-grace-period 6037 primitives, if available. 6038 6039 rcutorture.gp_poll_exp= [KNL] 6040 Use polled update-side expedited-grace-period 6041 primitives, if available. 6042 6043 rcutorture.gp_poll_full= [KNL] 6044 Use polled update-side normal-grace-period 6045 primitives that also take concurrent expedited 6046 grace periods into account, if available. 6047 6048 rcutorture.gp_poll_exp_full= [KNL] 6049 Use polled update-side expedited-grace-period 6050 primitives that also take concurrent normal 6051 grace periods into account, if available. 6052 6053 rcutorture.gp_poll_wi= [KNL] 6054 Nominal wait interval for normal conditional 6055 grace periods (specified by rcutorture's 6056 gp_poll and gp_poll_full module parameters), 6057 in microseconds. The actual wait interval will 6058 be randomly selected to nanosecond granularity up 6059 to this wait interval. Defaults to 16 jiffies, 6060 for example, 16,000 microseconds on a system 6061 with HZ=1000. 6062 6063 rcutorture.gp_poll_wi_exp= [KNL] 6064 Nominal wait interval for expedited conditional 6065 grace periods (specified by rcutorture's 6066 gp_poll_exp and gp_poll_exp_full module 6067 parameters), in microseconds. The actual wait 6068 interval will be randomly selected to nanosecond 6069 granularity up to this wait interval. Defaults to 6070 128 microseconds. 6071 6072 rcutorture.gp_sync= [KNL] 6073 Use normal (non-expedited) synchronous 6074 update-side primitives, if available. If all 6075 of rcutorture.gp_cond=, rcutorture.gp_exp=, 6076 rcutorture.gp_normal=, and rcutorture.gp_sync= 6077 are zero, rcutorture acts as if is interpreted 6078 they are all non-zero. 6079 6080 rcutorture.gpwrap_lag= [KNL] 6081 Enable grace-period wrap lag testing. Setting 6082 to false prevents the gpwrap lag test from 6083 running. Default is true. 6084 6085 rcutorture.gpwrap_lag_gps= [KNL] 6086 Set the value for grace-period wrap lag during 6087 active lag testing periods. This controls how many 6088 grace periods differences we tolerate between 6089 rdp and rnp's gp_seq before setting overflow flag. 6090 The default is always set to 8. 6091 6092 rcutorture.gpwrap_lag_cycle_mins= [KNL] 6093 Set the total cycle duration for gpwrap lag 6094 testing in minutes. This is the total time for 6095 one complete cycle of active and inactive 6096 testing periods. Default is 30 minutes. 6097 6098 rcutorture.gpwrap_lag_active_mins= [KNL] 6099 Set the duration for which gpwrap lag is active 6100 within each cycle, in minutes. During this time, 6101 the grace-period wrap lag will be set to the 6102 value specified by gpwrap_lag_gps. Default is 6103 5 minutes. 6104 6105 rcutorture.irqreader= [KNL] 6106 Run RCU readers from irq handlers, or, more 6107 accurately, from a timer handler. Not all RCU 6108 flavors take kindly to this sort of thing. 6109 6110 rcutorture.leakpointer= [KNL] 6111 Leak an RCU-protected pointer out of the reader. 6112 This can of course result in splats, and is 6113 intended to test the ability of things like 6114 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 6115 such leaks. 6116 6117 rcutorture.n_barrier_cbs= [KNL] 6118 Set callbacks/threads for rcu_barrier() testing. 6119 6120 rcutorture.nfakewriters= [KNL] 6121 Set number of concurrent RCU writers. These just 6122 stress RCU, they don't participate in the actual 6123 test, hence the "fake". 6124 6125 rcutorture.nmi_calls= [KNL] 6126 Enable issuing RCU callbacks from an NMI, on the 6127 RCU flavors that support it, to exercise the 6128 any-context callback path. Requires 6129 CONFIG_PERF_EVENTS and a hardware PMU; without 6130 both, nothing is issued. Defaults to enabled. 6131 6132 rcutorture.nocbs_nthreads= [KNL] 6133 Set number of RCU callback-offload togglers. 6134 Zero (the default) disables toggling. 6135 6136 rcutorture.nocbs_toggle= [KNL] 6137 Set the delay in milliseconds between successive 6138 callback-offload toggling attempts. 6139 6140 rcutorture.nreaders= [KNL] 6141 Set number of RCU readers. The value -1 selects 6142 N-1, where N is the number of CPUs. A value 6143 "n" less than -1 selects N-n-2, where N is again 6144 the number of CPUs. For example, -2 selects N 6145 (the number of CPUs), -3 selects N+1, and so on. 6146 6147 rcutorture.nwriters= [KNL] 6148 Set number of RCU writers, which must be either 6149 zero or one. For additional writers, use instead 6150 the rcutorture.nfakewriters parameter. 6151 6152 rcutorture.object_debug= [KNL] 6153 Enable debug-object double-call_rcu() testing. 6154 6155 rcutorture.onoff_holdoff= [KNL] 6156 Set time (s) after boot for CPU-hotplug testing. 6157 6158 rcutorture.onoff_interval= [KNL] 6159 Set time (jiffies) between CPU-hotplug operations, 6160 or zero to disable CPU-hotplug testing. 6161 6162 rcutorture.preempt_duration= [KNL] 6163 Set duration (in milliseconds) of preemptions 6164 by a high-priority FIFO real-time task. Set to 6165 zero (the default) to disable. The CPUs to 6166 preempt are selected randomly from the set that 6167 are online at a given point in time. Races with 6168 CPUs going offline are ignored, with that attempt 6169 at preemption skipped. 6170 6171 rcutorture.preempt_interval= [KNL] 6172 Set interval (in milliseconds, defaulting to one 6173 second) between preemptions by a high-priority 6174 FIFO real-time task. This delay is mediated 6175 by an hrtimer and is further fuzzed to avoid 6176 inadvertent synchronizations. 6177 6178 rcutorture.read_exit_burst= [KNL] 6179 The number of times in a given read-then-exit 6180 episode that a set of read-then-exit kthreads 6181 is spawned. 6182 6183 rcutorture.read_exit_delay= [KNL] 6184 The delay, in seconds, between successive 6185 read-then-exit testing episodes. 6186 6187 rcutorture.reader_flavor= [KNL] 6188 A bit mask indicating which readers to use. 6189 If there is more than one bit set, the readers 6190 are entered from low-order bit up, and are 6191 exited in the opposite order. For SRCU, the 6192 0x1 bit is normal readers, 0x2 NMI-safe readers, 6193 and 0x4 light-weight readers. 6194 6195 rcutorture.shuffle_interval= [KNL] 6196 Set task-shuffle interval (s). Shuffling tasks 6197 allows some CPUs to go into dyntick-idle mode 6198 during the rcutorture test. 6199 6200 rcutorture.shutdown_secs= [KNL] 6201 Set time (s) after boot system shutdown. This 6202 is useful for hands-off automated testing. 6203 6204 rcutorture.stall_cpu= [KNL] 6205 Duration of CPU stall (s) to test RCU CPU stall 6206 warnings, zero to disable. 6207 6208 rcutorture.stall_cpu_block= [KNL] 6209 Sleep while stalling if set. This will result 6210 in warnings from preemptible RCU in addition to 6211 any other stall-related activity. Note that 6212 in kernels built with CONFIG_PREEMPTION=n and 6213 CONFIG_PREEMPT_COUNT=y, this parameter will 6214 cause the CPU to pass through a quiescent state. 6215 Given CONFIG_PREEMPTION=n, this will suppress 6216 RCU CPU stall warnings, but will instead result 6217 in scheduling-while-atomic splats. 6218 6219 Use of this module parameter results in splats. 6220 6221 6222 rcutorture.stall_cpu_holdoff= [KNL] 6223 Time to wait (s) after boot before inducing stall. 6224 6225 rcutorture.stall_cpu_irqsoff= [KNL] 6226 Disable interrupts while stalling if set, but only 6227 on the first stall in the set. 6228 6229 rcutorture.stall_cpu_repeat= [KNL] 6230 Number of times to repeat the stall sequence, 6231 so that rcutorture.stall_cpu_repeat=3 will result 6232 in four stall sequences. 6233 6234 rcutorture.stall_gp_kthread= [KNL] 6235 Duration (s) of forced sleep within RCU 6236 grace-period kthread to test RCU CPU stall 6237 warnings, zero to disable. If both stall_cpu 6238 and stall_gp_kthread are specified, the 6239 kthread is starved first, then the CPU. 6240 6241 rcutorture.stall_only= [KNL] 6242 Shut off all rcutorture kthreads other than the 6243 RCU CPU stall-warning test kthreads. The purpose 6244 of this is to test production applictions' 6245 reactions to CPU stalls, and with minimal 6246 additional overhead. Or you can omit the 6247 stall-warning tests as well and get a heavy 6248 no-op, your choice! 6249 6250 rcutorture.stat_interval= [KNL] 6251 Time (s) between statistics printk()s. 6252 6253 rcutorture.stutter= [KNL] 6254 Time (s) to stutter testing, for example, specifying 6255 five seconds causes the test to run for five seconds, 6256 wait for five seconds, and so on. This tests RCU's 6257 ability to transition abruptly to and from idle. 6258 6259 rcutorture.test_boost= [KNL] 6260 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 6261 "Maybe" means test if the RCU implementation 6262 under test support RCU priority boosting. 6263 6264 rcutorture.test_boost_duration= [KNL] 6265 Duration (s) of each individual boost test. 6266 6267 rcutorture.test_boost_holdoff= [KNL] 6268 Holdoff time (s) from start of test to the start 6269 of RCU priority-boost testing. Defaults to zero, 6270 that is, no holdoff. 6271 6272 rcutorture.test_boost_interval= [KNL] 6273 Interval (s) between each boost test. 6274 6275 rcutorture.test_no_idle_hz= [KNL] 6276 Test RCU's dyntick-idle handling. See also the 6277 rcutorture.shuffle_interval parameter. 6278 6279 rcutorture.torture_type= [KNL] 6280 Specify the RCU implementation to test. 6281 6282 rcutorture.verbose= [KNL] 6283 Enable additional printk() statements. 6284 6285 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 6286 Dump ftrace buffer after reporting RCU CPU 6287 stall warning. 6288 6289 rcupdate.rcu_cpu_stall_notifiers= [KNL] 6290 Provide RCU CPU stall notifiers, but see the 6291 warnings in the RCU_CPU_STALL_NOTIFIER Kconfig 6292 option's help text. TL;DR: You almost certainly 6293 do not want rcupdate.rcu_cpu_stall_notifiers. 6294 6295 rcupdate.rcu_cpu_stall_suppress= [KNL] 6296 Suppress RCU CPU stall warning messages. 6297 6298 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 6299 Suppress RCU CPU stall warning messages and 6300 rcutorture writer stall warnings that occur 6301 during early boot, that is, during the time 6302 before the init task is spawned. 6303 6304 rcupdate.rcu_cpu_stall_timeout= [KNL] 6305 Set timeout for RCU CPU stall warning messages. 6306 The value is in seconds and the maximum allowed 6307 value is 300 seconds. 6308 6309 rcupdate.rcu_exp_cpu_stall_timeout= [KNL] 6310 Set timeout for expedited RCU CPU stall warning 6311 messages. The value is in milliseconds 6312 and the maximum allowed value is 21000 6313 milliseconds. Please note that this value is 6314 adjusted to an arch timer tick resolution. 6315 Setting this to zero causes the value from 6316 rcupdate.rcu_cpu_stall_timeout to be used (after 6317 conversion from seconds to milliseconds). 6318 6319 rcupdate.rcu_cpu_stall_cputime= [KNL] 6320 Provide statistics on the cputime and count of 6321 interrupts and tasks during the sampling period. For 6322 multiple continuous RCU stalls, all sampling periods 6323 begin at half of the first RCU stall timeout. 6324 6325 rcupdate.rcu_exp_stall_task_details= [KNL] 6326 Print stack dumps of any tasks blocking the 6327 current expedited RCU grace period during an 6328 expedited RCU CPU stall warning. 6329 6330 rcupdate.rcu_expedited= [KNL] 6331 Use expedited grace-period primitives, for 6332 example, synchronize_rcu_expedited() instead 6333 of synchronize_rcu(). This reduces latency, 6334 but can increase CPU utilization, degrade 6335 real-time latency, and degrade energy efficiency. 6336 No effect on CONFIG_TINY_RCU kernels. 6337 6338 rcupdate.rcu_normal= [KNL] 6339 Use only normal grace-period primitives, 6340 for example, synchronize_rcu() instead of 6341 synchronize_rcu_expedited(). This improves 6342 real-time latency, CPU utilization, and 6343 energy efficiency, but can expose users to 6344 increased grace-period latency. This parameter 6345 overrides rcupdate.rcu_expedited. No effect on 6346 CONFIG_TINY_RCU kernels. 6347 6348 rcupdate.rcu_normal_after_boot= [KNL] 6349 Once boot has completed (that is, after 6350 rcu_end_inkernel_boot() has been invoked), use 6351 only normal grace-period primitives. No effect 6352 on CONFIG_TINY_RCU kernels. 6353 6354 But note that CONFIG_PREEMPT_RT=y kernels enables 6355 this kernel boot parameter, forcibly setting 6356 it to the value one, that is, converting any 6357 post-boot attempt at an expedited RCU grace 6358 period to instead use normal non-expedited 6359 grace-period processing. 6360 6361 rcupdate.rcu_task_collapse_debug= [KNL] 6362 Enable debugging prints that record when RCU Tasks 6363 and RCU Tasks Trace expand to per-CPU callback 6364 queuing and collapse back to CPU-0 queuing. 6365 This is default-disabled due to the fact that 6366 some workloads can make it quite noisy. 6367 6368 rcupdate.rcu_task_collapse_lim= [KNL] 6369 Set the maximum number of callbacks present 6370 at the beginning of a grace period that allows 6371 the RCU Tasks flavors to collapse back to using 6372 a single callback queue. This switching only 6373 occurs when rcupdate.rcu_task_enqueue_lim is 6374 set to the default value of -1. 6375 6376 rcupdate.rcu_task_contend_lim= [KNL] 6377 Set the minimum number of callback-queuing-time 6378 lock-contention events per jiffy required to 6379 cause the RCU Tasks flavors to switch to per-CPU 6380 callback queuing. This switching only occurs 6381 when rcupdate.rcu_task_enqueue_lim is set to 6382 the default value of -1. 6383 6384 rcupdate.rcu_task_enqueue_lim= [KNL] 6385 Set the number of callback queues to use for the 6386 RCU Tasks family of RCU flavors. The default 6387 of -1 allows this to be automatically (and 6388 dynamically) adjusted. This parameter is intended 6389 for use in testing. 6390 6391 rcupdate.rcu_task_lazy_lim= [KNL] 6392 Number of callbacks on a given CPU that will 6393 cancel laziness on that CPU. Use -1 to disable 6394 cancellation of laziness, but be advised that 6395 doing so increases the danger of OOM due to 6396 callback flooding. 6397 6398 rcupdate.rcu_task_stall_info= [KNL] 6399 Set initial timeout in jiffies for RCU task stall 6400 informational messages, which give some indication 6401 of the problem for those not patient enough to 6402 wait for ten minutes. Informational messages are 6403 only printed prior to the stall-warning message 6404 for a given grace period. Disable with a value 6405 less than or equal to zero. Defaults to ten 6406 seconds. A change in value does not take effect 6407 until the beginning of the next grace period. 6408 6409 rcupdate.rcu_task_stall_info_mult= [KNL] 6410 Multiplier for time interval between successive 6411 RCU task stall informational messages for a given 6412 RCU tasks grace period. This value is clamped 6413 to one through ten, inclusive. It defaults to 6414 the value three, so that the first informational 6415 message is printed 10 seconds into the grace 6416 period, the second at 40 seconds, the third at 6417 160 seconds, and then the stall warning at 600 6418 seconds would prevent a fourth at 640 seconds. 6419 6420 rcupdate.rcu_task_stall_timeout= [KNL] 6421 Set timeout in jiffies for RCU task stall 6422 warning messages. Disable with a value less 6423 than or equal to zero. Defaults to ten minutes. 6424 A change in value does not take effect until 6425 the beginning of the next grace period. 6426 6427 rcupdate.rcu_tasks_lazy_ms= [KNL] 6428 Set timeout in milliseconds RCU Tasks asynchronous 6429 callback batching for call_rcu_tasks(). 6430 A negative value will take the default. A value 6431 of zero will disable batching. Batching is 6432 always disabled for synchronize_rcu_tasks(). 6433 6434 rcupdate.rcu_self_test= [KNL] 6435 Run the RCU early boot self tests 6436 6437 rdinit= [KNL] 6438 Format: <full_path> 6439 Run specified binary instead of /init from the ramdisk, 6440 used for early userspace startup. See initrd. 6441 6442 rdrand= [X86,EARLY] 6443 force - Override the decision by the kernel to hide the 6444 advertisement of RDRAND support (this affects 6445 certain AMD processors because of buggy BIOS 6446 support, specifically around the suspend/resume 6447 path). 6448 6449 rdt= [HW,X86,RDT] 6450 Turn on/off individual RDT features. List is: 6451 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 6452 mba, smba, bmec, abmc, sdciae, energy[:guid], 6453 perf[:guid]. 6454 E.g. to turn on cmt and turn off mba use: 6455 rdt=cmt,!mba 6456 To turn off all energy telemetry monitoring and ensure that 6457 perf telemetry monitoring associated with guid 0x12345 6458 is enabled use: 6459 rdt=!energy,perf:0x12345 6460 6461 reboot= [KNL] 6462 Format (x86 or x86_64): 6463 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \ 6464 [[,]s[mp]####] \ 6465 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 6466 [[,]f[orce]] 6467 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 6468 (prefix with 'panic_' to set mode for panic 6469 reboot only), 6470 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 6471 reboot_force is either force or not specified, 6472 reboot_cpu is s[mp]#### with #### being the processor 6473 to be used for rebooting. 6474 6475 acpi 6476 Use the ACPI RESET_REG in the FADT. If ACPI is not 6477 configured or the ACPI reset does not work, the reboot 6478 path attempts the reset using the keyboard controller. 6479 6480 bios 6481 Use the CPU reboot vector for warm reset 6482 6483 cold 6484 Set the cold reboot flag 6485 6486 default 6487 There are some built-in platform specific "quirks" 6488 - you may see: "reboot: <name> series board detected. 6489 Selecting <type> for reboots." In the case where you 6490 think the quirk is in error (e.g. you have newer BIOS, 6491 or newer board) using this option will ignore the 6492 built-in quirk table, and use the generic default 6493 reboot actions. 6494 6495 efi 6496 Use efi reset_system runtime service. If EFI is not 6497 configured or the EFI reset does not work, the reboot 6498 path attempts the reset using the keyboard controller. 6499 6500 force 6501 Don't stop other CPUs on reboot. This can make reboot 6502 more reliable in some cases. 6503 6504 kbd 6505 Use the keyboard controller. cold reset (default) 6506 6507 pci 6508 Use a write to the PCI config space register 0xcf9 to 6509 trigger reboot. 6510 6511 triple 6512 Force a triple fault (init) 6513 6514 warm 6515 Don't set the cold reboot flag 6516 6517 Using warm reset will be much faster especially on big 6518 memory systems because the BIOS will not go through 6519 the memory check. Disadvantage is that not all 6520 hardware will be completely reinitialized on reboot so 6521 there may be boot problems on some systems. 6522 6523 6524 refscale.holdoff= [KNL] 6525 Set test-start holdoff period. The purpose of 6526 this parameter is to delay the start of the 6527 test until boot completes in order to avoid 6528 interference. 6529 6530 refscale.lookup_instances= [KNL] 6531 Number of data elements to use for the forms of 6532 SLAB_TYPESAFE_BY_RCU testing. A negative number 6533 is negated and multiplied by nr_cpu_ids, while 6534 zero specifies nr_cpu_ids. 6535 6536 refscale.loops= [KNL] 6537 Set the number of loops over the synchronization 6538 primitive under test. Increasing this number 6539 reduces noise due to loop start/end overhead, 6540 but the default has already reduced the per-pass 6541 noise to a handful of picoseconds on ca. 2020 6542 x86 laptops. 6543 6544 refscale.nreaders= [KNL] 6545 Set number of readers. The default value of -1 6546 selects N, where N is roughly 75% of the number 6547 of CPUs. A value of zero is an interesting choice. 6548 6549 refscale.nruns= [KNL] 6550 Set number of runs, each of which is dumped onto 6551 the console log. 6552 6553 refscale.readdelay= [KNL] 6554 Set the read-side critical-section duration, 6555 measured in microseconds. 6556 6557 refscale.scale_type= [KNL] 6558 Specify the read-protection implementation to test. 6559 6560 refscale.shutdown= [KNL] 6561 Shut down the system at the end of the performance 6562 test. This defaults to 1 (shut it down) when 6563 refscale is built into the kernel and to 0 (leave 6564 it running) when refscale is built as a module. 6565 6566 refscale.verbose= [KNL] 6567 Enable additional printk() statements. 6568 6569 refscale.verbose_batched= [KNL] 6570 Batch the additional printk() statements. If zero 6571 (the default) or negative, print everything. Otherwise, 6572 print every Nth verbose statement, where N is the value 6573 specified. 6574 6575 regulator_ignore_unused 6576 [REGULATOR] 6577 Prevents regulator framework from disabling regulators 6578 that are unused, due no driver claiming them. This may 6579 be useful for debug and development, but should not be 6580 needed on a platform with proper driver support. 6581 6582 relax_domain_level= 6583 [KNL, SMP] Set scheduler's default relax_domain_level. 6584 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 6585 6586 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 6587 Format: <base1>,<size1>[,<base2>,<size2>,...] 6588 Reserve I/O ports or memory so the kernel won't use 6589 them. If <base> is less than 0x10000, the region 6590 is assumed to be I/O ports; otherwise it is memory. 6591 6592 reserve_mem= [MM] 6593 Format: nn[KMG]:<align>:<label> 6594 Reserve physical memory and label it with a name that 6595 other subsystems can use to access it. This is typically 6596 used for systems that do not wipe the RAM, and this command 6597 line will try to reserve the same physical memory on 6598 soft reboots. Note, it is not guaranteed to be the same 6599 location. For example, if anything about the system changes 6600 or if booting a different kernel. It can also fail if KASLR 6601 places the kernel at the location of where the RAM reservation 6602 was from a previous boot, the new reservation will be at a 6603 different location. 6604 Any subsystem using this feature must add a way to verify 6605 that the contents of the physical memory is from a previous 6606 boot, as there may be cases where the memory will not be 6607 located at the same location. 6608 6609 The format is size:align:label for example, to request 6610 12 megabytes of 4096 alignment for ramoops: 6611 6612 reserve_mem=12M:4096:oops ramoops.mem_name=oops 6613 6614 reservetop= [X86-32,EARLY] 6615 Format: nn[KMG] 6616 Reserves a hole at the top of the kernel virtual 6617 address space. 6618 6619 reset_devices [KNL] Force drivers to reset the underlying device 6620 during initialization. 6621 6622 resume= [SWSUSP] 6623 Specify the partition device for software suspend 6624 Format: 6625 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 6626 6627 resume_offset= [SWSUSP] 6628 Specify the offset from the beginning of the partition 6629 given by "resume=" at which the swap header is located, 6630 in <PAGE_SIZE> units (needed only for swap files). 6631 See Documentation/power/swsusp-and-swap-files.rst 6632 6633 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 6634 read the resume files 6635 6636 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 6637 Useful for devices that are detected asynchronously 6638 (e.g. USB and MMC devices). 6639 6640 retain_initrd [RAM] Keep initrd memory after extraction. After boot, it will 6641 be accessible via /sys/firmware/initrd. 6642 6643 retbleed= [X86] Control mitigation of RETBleed (Arbitrary 6644 Speculative Code Execution with Return Instructions) 6645 vulnerability. 6646 6647 AMD-based UNRET and IBPB mitigations alone do not stop 6648 sibling threads from influencing the predictions of other 6649 sibling threads. For that reason, STIBP is used on pro- 6650 cessors that support it, and mitigate SMT on processors 6651 that don't. 6652 6653 off - no mitigation 6654 auto - automatically select a mitigation 6655 auto,nosmt - automatically select a mitigation, 6656 disabling SMT if necessary for 6657 the full mitigation (only on Zen1 6658 and older without STIBP). 6659 ibpb - On AMD, mitigate short speculation 6660 windows on basic block boundaries too. 6661 Safe, highest perf impact. It also 6662 enables STIBP if present. Not suitable 6663 on Intel. 6664 ibpb,nosmt - Like "ibpb" above but will disable SMT 6665 when STIBP is not available. This is 6666 the alternative for systems which do not 6667 have STIBP. 6668 unret - Force enable untrained return thunks, 6669 only effective on AMD f15h-f17h based 6670 systems. 6671 unret,nosmt - Like unret, but will disable SMT when STIBP 6672 is not available. This is the alternative for 6673 systems which do not have STIBP. 6674 6675 Selecting 'auto' will choose a mitigation method at run 6676 time according to the CPU. 6677 6678 Not specifying this option is equivalent to retbleed=auto. 6679 6680 rfkill.default_state= 6681 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 6682 etc. communication is blocked by default. 6683 1 Unblocked. 6684 6685 rfkill.master_switch_mode= 6686 0 The "airplane mode" button does nothing. 6687 1 The "airplane mode" button toggles between everything 6688 blocked and the previous configuration. 6689 2 The "airplane mode" button toggles between everything 6690 blocked and everything unblocked. 6691 6692 ring3mwait=disable 6693 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 6694 CPUs. 6695 6696 riscv_isa_fallback [RISCV,EARLY] 6697 When CONFIG_RISCV_ISA_FALLBACK is not enabled, permit 6698 falling back to detecting extension support by parsing 6699 "riscv,isa" property on devicetree systems when the 6700 replacement properties are not found. See the Kconfig 6701 entry for RISCV_ISA_FALLBACK. 6702 6703 riscv_nousercfi= 6704 all Disable user CFI ABI to userspace even if cpu extension 6705 are available. 6706 bcfi Disable user backward CFI ABI to userspace even if 6707 the shadow stack extension is available. 6708 fcfi Disable user forward CFI ABI to userspace even if the 6709 landing pad extension is available. 6710 6711 ro [KNL] Mount root device read-only on boot 6712 6713 rodata= [KNL,EARLY] 6714 on Mark read-only kernel memory as read-only (default). 6715 off Leave read-only kernel memory writable for debugging. 6716 noalias Mark read-only kernel memory as read-only but retain 6717 writable aliases in the direct map for regions outside 6718 of the kernel image. [arm64] 6719 6720 rockchip.usb_uart 6721 [EARLY] 6722 Enable the uart passthrough on the designated usb port 6723 on Rockchip SoCs. When active, the signals of the 6724 debug-uart get routed to the D+ and D- pins of the usb 6725 port and the regular usb controller gets disabled. 6726 6727 root= [KNL] Root filesystem 6728 Usually this is a block device specifier of some kind, 6729 see the early_lookup_bdev comment in 6730 block/early-lookup.c for details. 6731 Alternatively this can be "ram" for the legacy initial 6732 ramdisk, "nfs" and "cifs" for root on a network file 6733 system, or "mtd" and "ubi" for mounting from raw flash. 6734 6735 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 6736 mount the root filesystem 6737 6738 rootflags= [KNL] Set root filesystem mount option string 6739 6740 rseq_slice_ext= [KNL] RSEQ based time slice extension 6741 Format: boolean 6742 Control enablement of RSEQ based time slice extension. 6743 Default is 'on'. 6744 6745 initramfs_options= [KNL] 6746 Specify mount options for the initramfs mount. 6747 6748 rootfstype= [KNL] Set root filesystem type 6749 6750 rootwait [KNL] Wait (indefinitely) for root device to show up. 6751 Useful for devices that are detected asynchronously 6752 (e.g. USB and MMC devices). 6753 6754 rootwait= [KNL] Maximum time (in seconds) to wait for root device 6755 to show up before attempting to mount the root 6756 filesystem. 6757 6758 rproc_mem=nn[KMG][@address] 6759 [KNL,ARM,CMA] Remoteproc physical memory block. 6760 Memory area to be used by remote processor image, 6761 managed by CMA. 6762 6763 rseq_debug= [KNL] Enable or disable restartable sequence 6764 debug mode. Defaults to CONFIG_RSEQ_DEBUG_DEFAULT_ENABLE. 6765 Format: <bool> 6766 6767 rt_group_sched= [KNL] Enable or disable SCHED_RR/FIFO group scheduling 6768 when CONFIG_RT_GROUP_SCHED=y. Defaults to 6769 !CONFIG_RT_GROUP_SCHED_DEFAULT_DISABLED. 6770 Format: <bool> 6771 6772 rw [KNL] Mount root device read-write on boot 6773 6774 S [KNL] Run init in single mode 6775 6776 s390_iommu= [HW,S390] 6777 Set s390 IOTLB flushing mode 6778 strict 6779 With strict flushing every unmap operation will result 6780 in an IOTLB flush. Default is lazy flushing before 6781 reuse, which is faster. Deprecated, equivalent to 6782 iommu.strict=1. 6783 6784 s390_iommu_aperture= [KNL,S390] 6785 Specifies the size of the per device DMA address space 6786 accessible through the DMA and IOMMU APIs as a decimal 6787 factor of the size of main memory. 6788 The default is 1 meaning that one can concurrently use 6789 as many DMA addresses as physical memory is installed, 6790 if supported by hardware, and thus map all of memory 6791 once. With a value of 2 one can map all of memory twice 6792 and so on. As a special case a factor of 0 imposes no 6793 restrictions other than those given by hardware at the 6794 cost of significant additional memory use for tables. 6795 6796 6797 sched_proxy_exec= [KNL] 6798 Enables or disables "proxy execution" style 6799 solution to mutex-based priority inversion. 6800 Format: <bool> 6801 6802 sched_smt_asym_packing= [KNL,SMP] 6803 Format: on 6804 Force asymmetric packing at the SMT scheduling domain. 6805 Idle CPU selection prefers siblings with a higher 6806 arch_asym_cpu_priority(). The default implementation 6807 prefers lower-numbered logical CPUs, but architecture 6808 overrides remain authoritative. Equal priorities do 6809 not establish a sibling preference. For example, x86 6810 with CONFIG_SCHED_MC_PRIO normally assigns the same 6811 core priority to both SMT siblings, so this option 6812 does not favor either sibling. 6813 6814 sched_verbose [KNL,EARLY] Enables verbose scheduler debug messages. 6815 6816 schedstats= [KNL,X86] Enable or disable scheduled statistics. 6817 Allowed values are enable and disable. This feature 6818 incurs a small amount of overhead in the scheduler 6819 but is useful for debugging and performance tuning. 6820 6821 sched_thermal_decay_shift= 6822 [Deprecated] 6823 [KNL, SMP] Set a decay shift for scheduler thermal 6824 pressure signal. Thermal pressure signal follows the 6825 default decay period of other scheduler pelt 6826 signals(usually 32 ms but configurable). Setting 6827 sched_thermal_decay_shift will left shift the decay 6828 period for the thermal pressure signal by the shift 6829 value. 6830 i.e. with the default pelt decay period of 32 ms 6831 sched_thermal_decay_shift thermal pressure decay pr 6832 1 64 ms 6833 2 128 ms 6834 and so on. 6835 Format: integer between 0 and 10 6836 Default is 0. 6837 6838 scftorture.holdoff= [KNL] 6839 Number of seconds to hold off before starting 6840 test. Defaults to zero for module insertion and 6841 to 10 seconds for built-in smp_call_function() 6842 tests. 6843 6844 scftorture.longwait= [KNL] 6845 Request ridiculously long waits randomly selected 6846 up to the chosen limit in seconds. Zero (the 6847 default) disables this feature. Please note 6848 that requesting even small non-zero numbers of 6849 seconds can result in RCU CPU stall warnings, 6850 softlockup complaints, and so on. 6851 6852 scftorture.nthreads= [KNL] 6853 Number of kthreads to spawn to invoke the 6854 smp_call_function() family of functions. 6855 The default of -1 specifies a number of kthreads 6856 equal to the number of CPUs. 6857 6858 scftorture.onoff_holdoff= [KNL] 6859 Number seconds to wait after the start of the 6860 test before initiating CPU-hotplug operations. 6861 6862 scftorture.onoff_interval= [KNL] 6863 Number seconds to wait between successive 6864 CPU-hotplug operations. Specifying zero (which 6865 is the default) disables CPU-hotplug operations. 6866 6867 scftorture.shutdown_secs= [KNL] 6868 The number of seconds following the start of the 6869 test after which to shut down the system. The 6870 default of zero avoids shutting down the system. 6871 Non-zero values are useful for automated tests. 6872 6873 scftorture.stat_interval= [KNL] 6874 The number of seconds between outputting the 6875 current test statistics to the console. A value 6876 of zero disables statistics output. 6877 6878 scftorture.stutter_cpus= [KNL] 6879 The number of jiffies to wait between each change 6880 to the set of CPUs under test. 6881 6882 scftorture.use_cpus_read_lock= [KNL] 6883 Use use_cpus_read_lock() instead of the default 6884 preempt_disable() to disable CPU hotplug 6885 while invoking one of the smp_call_function*() 6886 functions. 6887 6888 scftorture.verbose= [KNL] 6889 Enable additional printk() statements. 6890 6891 scftorture.weight_single= [KNL] 6892 The probability weighting to use for the 6893 smp_call_function_single() function with a zero 6894 "wait" parameter. A value of -1 selects the 6895 default if all other weights are -1. However, 6896 if at least one weight has some other value, a 6897 value of -1 will instead select a weight of zero. 6898 6899 scftorture.weight_single_wait= [KNL] 6900 The probability weighting to use for the 6901 smp_call_function_single() function with a 6902 non-zero "wait" parameter. See weight_single. 6903 6904 scftorture.weight_many= [KNL] 6905 The probability weighting to use for the 6906 smp_call_function_many() function with a zero 6907 "wait" parameter. See weight_single. 6908 Note well that setting a high probability for 6909 this weighting can place serious IPI load 6910 on the system. 6911 6912 scftorture.weight_many_wait= [KNL] 6913 The probability weighting to use for the 6914 smp_call_function_many() function with a 6915 non-zero "wait" parameter. See weight_single 6916 and weight_many. 6917 6918 scftorture.weight_all= [KNL] 6919 The probability weighting to use for the 6920 smp_call_function_all() function with a zero 6921 "wait" parameter. See weight_single and 6922 weight_many. 6923 6924 scftorture.weight_all_wait= [KNL] 6925 The probability weighting to use for the 6926 smp_call_function_all() function with a 6927 non-zero "wait" parameter. See weight_single 6928 and weight_many. 6929 6930 sdw_mclk_divider=[SDW] 6931 Specify the MCLK divider for Intel SoundWire buses in 6932 case the BIOS does not provide the clock rate properly. 6933 6934 skew_tick= [KNL,EARLY] Offset the periodic timer tick per cpu to mitigate 6935 xtime_lock contention on larger systems, and/or RCU lock 6936 contention on all systems with CONFIG_MAXSMP set. 6937 Format: { "0" | "1" } 6938 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1") 6939 1 -- enable. 6940 Note: increases power consumption, thus should only be 6941 enabled if running jitter sensitive (HPC/RT) workloads. 6942 6943 security= [SECURITY] Choose a legacy "major" security module to 6944 enable at boot. This has been deprecated by the 6945 "lsm=" parameter. 6946 6947 selinux= [SELINUX] Disable or enable SELinux at boot time. 6948 Format: { "0" | "1" } 6949 See security/selinux/Kconfig help text. 6950 0 -- disable. 6951 1 -- enable. 6952 Default value is 1. 6953 6954 serialnumber [BUGS=X86-32] 6955 6956 sev=option[,option...] [X86-64] 6957 6958 debug 6959 Enable debug messages. 6960 6961 nosnp 6962 Do not enable SEV-SNP (applies to host/hypervisor 6963 only). Setting 'nosnp' avoids the RMP check overhead 6964 in memory accesses when users do not want to run 6965 SEV-SNP guests. 6966 6967 shapers= [NET] 6968 Maximal number of shapers. 6969 6970 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 6971 Limit apic dumping. The parameter defines the maximal 6972 number of local apics being dumped. Also it is possible 6973 to set it to "all" by meaning -- no limit here. 6974 Format: { 1 (default) | 2 | ... | all }. 6975 The parameter valid if only apic=debug or 6976 apic=verbose is specified. 6977 Example: apic=debug show_lapic=all 6978 6979 slab_debug[=options[,slabs][;[options[,slabs]]...]] [MM] 6980 Enabling slab_debug allows one to determine the 6981 culprit if slab objects become corrupted. Enabling 6982 slab_debug can create guard zones around objects and 6983 may poison objects when not in use. Also tracks the 6984 last alloc / free. For more information see 6985 Documentation/admin-guide/mm/slab.rst. 6986 (slub_debug legacy name also accepted for now) 6987 6988 Using this option implies the "no_hash_pointers" 6989 option which can be undone by adding the 6990 "hash_pointers=always" option. 6991 6992 slab_max_order= [MM] 6993 Determines the maximum allowed order for slabs. 6994 A high setting may cause OOMs due to memory 6995 fragmentation. For more information see 6996 Documentation/admin-guide/mm/slab.rst. 6997 (slub_max_order legacy name also accepted for now) 6998 6999 slab_merge [MM] 7000 Enable merging of slabs with similar size when the 7001 kernel is built without CONFIG_SLAB_MERGE_DEFAULT. 7002 (slub_merge legacy name also accepted for now) 7003 7004 slab_min_objects= [MM] 7005 The minimum number of objects per slab. SLUB will 7006 increase the slab order up to slab_max_order to 7007 generate a sufficiently large slab able to contain 7008 the number of objects indicated. The higher the number 7009 of objects the smaller the overhead of tracking slabs 7010 and the less frequently locks need to be acquired. 7011 For more information see 7012 Documentation/admin-guide/mm/slab.rst. 7013 (slub_min_objects legacy name also accepted for now) 7014 7015 slab_min_order= [MM] 7016 Determines the minimum page order for slabs. Must be 7017 lower or equal to slab_max_order. For more information see 7018 Documentation/admin-guide/mm/slab.rst. 7019 (slub_min_order legacy name also accepted for now) 7020 7021 slab_nomerge [MM] 7022 Disable merging of slabs with similar size. May be 7023 necessary if there is some reason to distinguish 7024 allocs to different slabs, especially in hardened 7025 environments where the risk of heap overflows and 7026 layout control by attackers can usually be 7027 frustrated by disabling merging. This will reduce 7028 most of the exposure of a heap attack to a single 7029 cache (risks via metadata attacks are mostly 7030 unchanged). Debug options disable merging on their 7031 own. 7032 For more information see 7033 Documentation/admin-guide/mm/slab.rst. 7034 (slub_nomerge legacy name also accepted for now) 7035 7036 slab_strict_numa [MM] 7037 Support memory policies on a per object level 7038 in the slab allocator. The default is for memory 7039 policies to be applied at the folio level when 7040 a new folio is needed or a partial folio is 7041 retrieved from the lists. Increases overhead 7042 in the slab fastpaths but gains more accurate 7043 NUMA kernel object placement which helps with slow 7044 interconnects in NUMA systems. 7045 7046 slram= [HW,MTD] 7047 7048 smart2= [HW] 7049 Format: <io1>[,<io2>[,...,<io8>]] 7050 7051 smp.csd_lock_timeout= [KNL] 7052 Specify the period of time in milliseconds 7053 that smp_call_function() and friends will wait 7054 for a CPU to release the CSD lock. This is 7055 useful when diagnosing bugs involving CPUs 7056 disabling interrupts for extended periods 7057 of time. Defaults to 5,000 milliseconds, and 7058 setting a value of zero disables this feature. 7059 This feature may be more efficiently disabled 7060 using the csdlock_debug- kernel parameter. 7061 7062 smp.panic_on_ipistall= [KNL] 7063 If a csd_lock_timeout extends for more than 7064 the specified number of milliseconds, panic the 7065 system. By default, let CSD-lock acquisition 7066 take as long as they take. Specifying 300,000 7067 for this value provides a 5-minute timeout. 7068 7069 smt= [KNL,MIPS,S390,EARLY] Set the maximum number of threads 7070 (logical CPUs) to use per physical CPU on systems 7071 capable of symmetric multithreading (SMT). Will 7072 be capped to the actual hardware limit. 7073 Format: <integer> 7074 Default: -1 (no limit) 7075 7076 softlockup_panic= 7077 [KNL] Should the soft-lockup detector generate panics. 7078 Format: <int> 7079 7080 A value of non-zero instructs the soft-lockup detector 7081 to panic the machine when a soft-lockup duration exceeds 7082 N thresholds. It is also controlled by the kernel.softlockup_panic 7083 sysctl and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 7084 respective build-time switch to that functionality. 7085 7086 softlockup_all_cpu_backtrace= 7087 [KNL] Should the soft-lockup detector generate 7088 backtraces on all cpus. 7089 Format: 0 | 1 7090 7091 sonypi.*= [HW] Sony Programmable I/O Control Device driver 7092 See Documentation/admin-guide/laptops/sonypi.rst 7093 7094 spectre_bhi= [X86] Control mitigation of Branch History Injection 7095 (BHI) vulnerability. This setting affects the 7096 deployment of the HW BHI control and the SW BHB 7097 clearing sequence. 7098 7099 on - (default) Enable the HW or SW mitigation as 7100 needed. This protects the kernel from 7101 both syscalls and VMs. 7102 vmexit - On systems which don't have the HW mitigation 7103 available, enable the SW mitigation on vmexit 7104 ONLY. On such systems, the host kernel is 7105 protected from VM-originated BHI attacks, but 7106 may still be vulnerable to syscall attacks. 7107 off - Disable the mitigation. 7108 7109 spectre_v2= [X86,EARLY] Control mitigation of Spectre variant 2 7110 (indirect branch speculation) vulnerability. 7111 The default operation protects the kernel from 7112 user space attacks. 7113 7114 on - unconditionally enable, implies 7115 spectre_v2_user=on 7116 off - unconditionally disable, implies 7117 spectre_v2_user=off 7118 auto - kernel detects whether your CPU model is 7119 vulnerable 7120 7121 Selecting 'on' will, and 'auto' may, choose a 7122 mitigation method at run time according to the 7123 CPU, the available microcode, the setting of the 7124 CONFIG_MITIGATION_RETPOLINE configuration option, 7125 and the compiler with which the kernel was built. 7126 7127 Selecting 'on' will also enable the mitigation 7128 against user space to user space task attacks. 7129 Selecting specific mitigation does not force enable 7130 user mitigations. 7131 7132 Selecting 'off' will disable both the kernel and 7133 the user space protections. 7134 7135 Specific mitigations can also be selected manually: 7136 7137 retpoline - replace indirect branches 7138 retpoline,generic - Retpolines 7139 retpoline,lfence - LFENCE; indirect branch 7140 retpoline,amd - alias for retpoline,lfence 7141 eibrs - Enhanced/Auto IBRS 7142 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines 7143 eibrs,lfence - Enhanced/Auto IBRS + LFENCE 7144 ibrs - use IBRS to protect kernel 7145 7146 Not specifying this option is equivalent to 7147 spectre_v2=auto. 7148 7149 spectre_v2_user= 7150 [X86] Control mitigation of Spectre variant 2 7151 (indirect branch speculation) vulnerability between 7152 user space tasks 7153 7154 on - Unconditionally enable mitigations. Is 7155 enforced by spectre_v2=on 7156 7157 off - Unconditionally disable mitigations. Is 7158 enforced by spectre_v2=off 7159 7160 prctl - Indirect branch speculation is enabled, 7161 but mitigation can be enabled via prctl 7162 per thread. The mitigation control state 7163 is inherited on fork. 7164 7165 prctl,ibpb 7166 - Like "prctl" above, but only STIBP is 7167 controlled per thread. IBPB is issued 7168 always when switching between different user 7169 space processes. 7170 7171 seccomp 7172 - Same as "prctl" above, but all seccomp 7173 threads will enable the mitigation unless 7174 they explicitly opt out. 7175 7176 seccomp,ibpb 7177 - Like "seccomp" above, but only STIBP is 7178 controlled per thread. IBPB is issued 7179 always when switching between different 7180 user space processes. 7181 7182 auto - Kernel selects the mitigation depending on 7183 the available CPU features and vulnerability. 7184 7185 Default mitigation: "prctl" 7186 7187 Not specifying this option is equivalent to 7188 spectre_v2_user=auto. 7189 7190 spec_rstack_overflow= 7191 [X86,EARLY] Control RAS overflow mitigation on AMD Zen CPUs 7192 7193 off - Disable mitigation 7194 microcode - Enable microcode mitigation only 7195 safe-ret - Enable sw-only safe RET mitigation (default) 7196 ibpb - Enable mitigation by issuing IBPB on 7197 kernel entry 7198 ibpb-vmexit - Issue IBPB only on VMEXIT 7199 (cloud-specific mitigation) 7200 7201 spec_store_bypass_disable= 7202 [HW,EARLY] Control Speculative Store Bypass (SSB) Disable mitigation 7203 (Speculative Store Bypass vulnerability) 7204 7205 Certain CPUs are vulnerable to an exploit against a 7206 a common industry wide performance optimization known 7207 as "Speculative Store Bypass" in which recent stores 7208 to the same memory location may not be observed by 7209 later loads during speculative execution. The idea 7210 is that such stores are unlikely and that they can 7211 be detected prior to instruction retirement at the 7212 end of a particular speculation execution window. 7213 7214 In vulnerable processors, the speculatively forwarded 7215 store can be used in a cache side channel attack, for 7216 example to read memory to which the attacker does not 7217 directly have access (e.g. inside sandboxed code). 7218 7219 This parameter controls whether the Speculative Store 7220 Bypass optimization is used. 7221 7222 On x86 the options are: 7223 7224 on - Unconditionally disable Speculative Store Bypass 7225 off - Unconditionally enable Speculative Store Bypass 7226 auto - Kernel detects whether the CPU model contains an 7227 implementation of Speculative Store Bypass and 7228 picks the most appropriate mitigation. If the 7229 CPU is not vulnerable, "off" is selected. If the 7230 CPU is vulnerable the default mitigation is 7231 architecture and Kconfig dependent. See below. 7232 prctl - Control Speculative Store Bypass per thread 7233 via prctl. Speculative Store Bypass is enabled 7234 for a process by default. The state of the control 7235 is inherited on fork. 7236 seccomp - Same as "prctl" above, but all seccomp threads 7237 will disable SSB unless they explicitly opt out. 7238 7239 Default mitigations: 7240 X86: "prctl" 7241 7242 On powerpc the options are: 7243 7244 on,auto - On Power8 and Power9 insert a store-forwarding 7245 barrier on kernel entry and exit. On Power7 7246 perform a software flush on kernel entry and 7247 exit. 7248 off - No action. 7249 7250 Not specifying this option is equivalent to 7251 spec_store_bypass_disable=auto. 7252 7253 split_llc= 7254 [X86,EARLY] Split the LLC N-ways 7255 7256 When set, the LLC is split this many ways by matching 7257 'core_id % n'. This is setup before SMP bringup and 7258 used during SMP bringup before it knows the full 7259 topology. If your core count doesn't nicely divide by 7260 the number given, you get to keep the pieces. 7261 7262 This is mostly a debug feature to emulate multiple LLCs 7263 on hardware that only have a single LLC. 7264 7265 split_lock_detect= 7266 [X86] Enable split lock detection or bus lock detection 7267 7268 When enabled (and if hardware support is present), atomic 7269 instructions that access data across cache line 7270 boundaries will result in an alignment check exception 7271 for split lock detection or a debug exception for 7272 bus lock detection. 7273 7274 off - not enabled 7275 7276 warn - the kernel will emit rate-limited warnings 7277 about applications triggering the #AC 7278 exception or the #DB exception. This mode is 7279 the default on CPUs that support split lock 7280 detection or bus lock detection. Default 7281 behavior is by #AC if both features are 7282 enabled in hardware. 7283 7284 fatal - the kernel will send SIGBUS to applications 7285 that trigger the #AC exception or the #DB 7286 exception. Default behavior is by #AC if 7287 both features are enabled in hardware. 7288 7289 ratelimit:N - 7290 Set system wide rate limit to N bus locks 7291 per second for bus lock detection. 7292 0 < N <= 1000. 7293 7294 N/A for split lock detection. 7295 7296 7297 If an #AC exception is hit in the kernel or in 7298 firmware (i.e. not while executing in user mode) 7299 the kernel will oops in either "warn" or "fatal" 7300 mode. 7301 7302 #DB exception for bus lock is triggered only when 7303 CPL > 0. 7304 7305 srbds= [X86,INTEL,EARLY] 7306 Control the Special Register Buffer Data Sampling 7307 (SRBDS) mitigation. 7308 7309 Certain CPUs are vulnerable to an MDS-like 7310 exploit which can leak bits from the random 7311 number generator. 7312 7313 By default, this issue is mitigated by 7314 microcode. However, the microcode fix can cause 7315 the RDRAND and RDSEED instructions to become 7316 much slower. Among other effects, this will 7317 result in reduced throughput from /dev/urandom. 7318 7319 The microcode mitigation can be disabled with 7320 the following option: 7321 7322 off: Disable mitigation and remove 7323 performance impact to RDRAND and RDSEED 7324 7325 srcutree.big_cpu_lim [KNL] 7326 Specifies the number of CPUs constituting a 7327 large system, such that srcu_struct structures 7328 should immediately allocate an srcu_node array. 7329 This kernel-boot parameter defaults to 128, 7330 but takes effect only when the low-order four 7331 bits of srcutree.convert_to_big is equal to 3 7332 (decide at boot). 7333 7334 srcutree.convert_to_big [KNL] 7335 Specifies under what conditions an SRCU tree 7336 srcu_struct structure will be converted to big 7337 form, that is, with an rcu_node tree: 7338 7339 0: Never. 7340 1: At init_srcu_struct() time. 7341 2: When rcutorture decides to. 7342 3: Decide at boot time (default). 7343 0x1X: Above plus if high contention. 7344 7345 Either way, the srcu_node tree will be sized based 7346 on the actual runtime number of CPUs (nr_cpu_ids) 7347 instead of the compile-time CONFIG_NR_CPUS. 7348 7349 srcutree.counter_wrap_check [KNL] 7350 Specifies how frequently to check for 7351 grace-period sequence counter wrap for the 7352 srcu_data structure's ->srcu_gp_seq_needed field. 7353 The greater the number of bits set in this kernel 7354 parameter, the less frequently counter wrap will 7355 be checked for. Note that the bottom two bits 7356 are ignored. 7357 7358 srcutree.exp_holdoff [KNL] 7359 Specifies how many nanoseconds must elapse 7360 since the end of the last SRCU grace period for 7361 a given srcu_struct until the next normal SRCU 7362 grace period will be considered for automatic 7363 expediting. Set to zero to disable automatic 7364 expediting. 7365 7366 srcutree.srcu_max_nodelay [KNL] 7367 Specifies the number of no-delay instances 7368 per jiffy for which the SRCU grace period 7369 worker thread will be rescheduled with zero 7370 delay. Beyond this limit, worker thread will 7371 be rescheduled with a sleep delay of one jiffy. 7372 7373 srcutree.srcu_max_nodelay_phase [KNL] 7374 Specifies the per-grace-period phase, number of 7375 non-sleeping polls of readers. Beyond this limit, 7376 grace period worker thread will be rescheduled 7377 with a sleep delay of one jiffy, between each 7378 rescan of the readers, for a grace period phase. 7379 7380 srcutree.srcu_retry_check_delay [KNL] 7381 Specifies number of microseconds of non-sleeping 7382 delay between each non-sleeping poll of readers. 7383 7384 srcutree.small_contention_lim [KNL] 7385 Specifies the number of update-side contention 7386 events per jiffy will be tolerated before 7387 initiating a conversion of an srcu_struct 7388 structure to big form. Note that the value of 7389 srcutree.convert_to_big must have the 0x10 bit 7390 set for contention-based conversions to occur. 7391 7392 ssbd= [ARM64,HW,EARLY] 7393 Speculative Store Bypass Disable control 7394 7395 On CPUs that are vulnerable to the Speculative 7396 Store Bypass vulnerability and offer a 7397 firmware based mitigation, this parameter 7398 indicates how the mitigation should be used: 7399 7400 force-on: Unconditionally enable mitigation for 7401 for both kernel and userspace 7402 force-off: Unconditionally disable mitigation for 7403 for both kernel and userspace 7404 kernel: Always enable mitigation in the 7405 kernel, and offer a prctl interface 7406 to allow userspace to register its 7407 interest in being mitigated too. 7408 7409 stack_guard_gap= [MM] 7410 override the default stack gap protection. The value 7411 is in page units and it defines how many pages prior 7412 to (for stacks growing down) resp. after (for stacks 7413 growing up) the main stack are reserved for no other 7414 mapping. Default value is 256 pages. 7415 7416 stack_depot_disable= [KNL,EARLY] 7417 Setting this to true through kernel command line will 7418 disable the stack depot thereby saving the static memory 7419 consumed by the stack hash table. By default this is set 7420 to false. 7421 7422 stack_depot_max_pools= [KNL,EARLY] 7423 Specify the maximum number of pools to use for storing 7424 stack traces. Pools are allocated on-demand up to this 7425 limit. Default value is 8191 pools. 7426 7427 stacktrace [FTRACE] 7428 Enable the stack tracer on boot up. 7429 7430 stacktrace_filter=[function-list] 7431 [FTRACE] Limit the functions that the stack tracer 7432 will trace at boot up. function-list is a comma-separated 7433 list of functions. This list can be changed at run 7434 time by the stack_trace_filter file in the debugfs 7435 tracing directory. Note, this enables stack tracing 7436 and the stacktrace above is not needed. 7437 7438 sti= [PARISC,HW] 7439 Format: <num> 7440 Set the STI (builtin display/keyboard on the HP-PARISC 7441 machines) console (graphic card) which should be used 7442 as the initial boot-console. 7443 See also comment in drivers/video/sticore.c. 7444 7445 sti_font= [HW] 7446 See comment in drivers/video/sticore.c. 7447 7448 stifb= [HW] 7449 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 7450 7451 strict_sas_size= 7452 [X86] 7453 Format: <bool> 7454 Enable or disable strict sigaltstack size checks 7455 against the required signal frame size which 7456 depends on the supported FPU features. This can 7457 be used to filter out binaries which have 7458 not yet been made aware of AT_MINSIGSTKSZ. 7459 7460 stress_hpt [PPC,EARLY] 7461 Limits the number of kernel HPT entries in the hash 7462 page table to increase the rate of hash page table 7463 faults on kernel addresses. 7464 7465 stress_slb [PPC,EARLY] 7466 Limits the number of kernel SLB entries, and flushes 7467 them frequently to increase the rate of SLB faults 7468 on kernel addresses. 7469 7470 no_slb_preload [PPC,EARLY] 7471 Disables slb preloading for userspace. 7472 7473 sunrpc.min_resvport= 7474 sunrpc.max_resvport= 7475 [NFS,SUNRPC] 7476 SunRPC servers often require that client requests 7477 originate from a privileged port (i.e. a port in the 7478 range 0 < portnr < 1024). 7479 An administrator who wishes to reserve some of these 7480 ports for other uses may adjust the range that the 7481 kernel's sunrpc client considers to be privileged 7482 using these two parameters to set the minimum and 7483 maximum port values. 7484 7485 sunrpc.svc_rpc_per_connection_limit= 7486 [NFS,SUNRPC] 7487 Limit the number of requests that the server will 7488 process in parallel from a single connection. 7489 The default value is 0 (no limit). 7490 7491 sunrpc.pool_mode= 7492 [NFS] 7493 Deprecated. The NFS server now always uses one 7494 service thread pool per NUMA node (equivalent to a 7495 single global pool on non-NUMA machines). All of 7496 the previously accepted values (auto, global, 7497 percpu, pernode) are still accepted for backward 7498 compatibility but are ignored: the mode is always 7499 pernode, and reads always return "pernode". 7500 7501 sunrpc.tcp_slot_table_entries= 7502 sunrpc.udp_slot_table_entries= 7503 [NFS,SUNRPC] 7504 Sets the upper limit on the number of simultaneous 7505 RPC calls that can be sent from the client to a 7506 server. Increasing these values may allow you to 7507 improve throughput, but will also increase the 7508 amount of memory reserved for use by the client. 7509 7510 suspend.pm_test_delay= 7511 [SUSPEND] 7512 Sets the number of seconds to remain in a suspend test 7513 mode before resuming the system (see 7514 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 7515 is set. Default value is 5. 7516 7517 svm= [PPC] 7518 Format: { on | off | y | n | 1 | 0 } 7519 This parameter controls use of the Protected 7520 Execution Facility on pSeries. 7521 7522 swiotlb= [ARM,PPC,MIPS,X86,S390,EARLY] 7523 Format: { <int> [,<int>] | force | noforce | track_hiwater} 7524 <int> -- Number of I/O TLB slabs 7525 <int> -- Second integer after comma. Number of swiotlb 7526 areas with their own lock. Will be rounded up 7527 to a power of 2. 7528 force -- force using of bounce buffers even if they 7529 wouldn't be automatically used by the kernel 7530 noforce -- Never use bounce buffers (for debugging) 7531 track_hiwater -- Track high watermark of swiotlb buffers. 7532 Only available when CONFIG_DEBUG_FS is set. 7533 7534 switches= [HW,M68k,EARLY] 7535 7536 sysctl.*= [KNL] 7537 Set a sysctl parameter, right before loading the init 7538 process, as if the value was written to the respective 7539 /proc/sys/... file. Both '.' and '/' are recognized as 7540 separators. Unrecognized parameters and invalid values 7541 are reported in the kernel log. Sysctls registered 7542 later by a loaded module cannot be set this way. 7543 Example: sysctl.vm.swappiness=40 7544 7545 sysrq_always_enabled 7546 [KNL] 7547 Ignore sysrq setting - this boot parameter will 7548 neutralize any effect of /proc/sys/kernel/sysrq. 7549 Useful for debugging. 7550 7551 tcpmhash_entries= [KNL,NET] 7552 Set the number of tcp_metrics_hash slots. 7553 Default value is 8192 or 16384 depending on total 7554 ram pages. This is used to specify the TCP metrics 7555 cache size. See Documentation/networking/ip-sysctl.rst 7556 "tcp_no_metrics_save" section for more details. 7557 7558 tdfx= [HW,DRM] 7559 7560 test_suspend= [SUSPEND] 7561 Format: { "mem" | "standby" | "freeze" }[,N] 7562 Specify "mem" (for Suspend-to-RAM) or "standby" (for 7563 standby suspend) or "freeze" (for suspend type freeze) 7564 as the system sleep state during system startup with 7565 the optional capability to repeat N number of times. 7566 The system is woken from this state using a 7567 wakeup-capable RTC alarm. 7568 7569 thash_entries= [KNL,NET] 7570 Set number of hash buckets for TCP connection 7571 7572 thermal.act= [HW,ACPI] 7573 -1: disable all active trip points in all thermal zones 7574 <degrees C>: override all lowest active trip points 7575 7576 thermal.crt= [HW,ACPI] 7577 -1: disable all critical trip points in all thermal zones 7578 <degrees C>: override all critical trip points 7579 7580 thermal.off= [HW,ACPI] 7581 1: disable ACPI thermal control 7582 7583 thermal.psv= [HW,ACPI] 7584 -1: disable all passive trip points 7585 <degrees C>: override all passive trip points to this 7586 value 7587 7588 thermal.tzp= [HW,ACPI] 7589 Specify global default ACPI thermal zone polling rate 7590 <deci-seconds>: poll all this frequency 7591 0: no polling (default) 7592 7593 thp_anon= [KNL] 7594 Format: <size>[KMG],<size>[KMG]:<state>;<size>[KMG]-<size>[KMG]:<state> 7595 state is one of "always", "madvise", "never" or "inherit". 7596 Control the default behavior of the system with respect 7597 to anonymous transparent hugepages. 7598 Can be used multiple times for multiple anon THP sizes. 7599 See Documentation/admin-guide/mm/transhuge.rst for more 7600 details. 7601 7602 threadirqs [KNL,EARLY] 7603 Force threading of all interrupt handlers except those 7604 marked explicitly IRQF_NO_THREAD. 7605 7606 thp_shmem= [KNL] 7607 Format: <size>[KMG],<size>[KMG]:<policy>;<size>[KMG]-<size>[KMG]:<policy> 7608 Control the default policy of each hugepage size for the 7609 internal shmem mount. <policy> is one of policies available 7610 for the shmem mount ("always", "inherit", "never", "within_size", 7611 and "advise"). 7612 It can be used multiple times for multiple shmem THP sizes. 7613 See Documentation/admin-guide/mm/transhuge.rst for more 7614 details. 7615 7616 tlbi= [X86-64] 7617 Format: {ipi} 7618 ipi: switch to IPI-based TLB flushing 7619 7620 topology= [S390,EARLY] 7621 Format: {off | on} 7622 Specify if the kernel should make use of the cpu 7623 topology information if the hardware supports this. 7624 The scheduler will make use of this information and 7625 e.g. base its process migration decisions on it. 7626 Default is on. 7627 7628 torture.disable_onoff_at_boot= [KNL] 7629 Prevent the CPU-hotplug component of torturing 7630 until after init has spawned. 7631 7632 torture.ftrace_dump_at_shutdown= [KNL] 7633 Dump the ftrace buffer at torture-test shutdown, 7634 even if there were no errors. This can be a 7635 very costly operation when many torture tests 7636 are running concurrently, especially on systems 7637 with rotating-rust storage. 7638 7639 torture.verbose_sleep_frequency= [KNL] 7640 Specifies how many verbose printk()s should be 7641 emitted between each sleep. The default of zero 7642 disables verbose-printk() sleeping. 7643 7644 torture.verbose_sleep_duration= [KNL] 7645 Duration of each verbose-printk() sleep in jiffies. 7646 7647 tpm.disable_pcr_integrity= [HW,TPM] 7648 Do not protect PCR registers from unintended physical 7649 access, or interposers in the bus by the means of 7650 having an integrity protected session wrapped around 7651 TPM2_PCR_Extend command. Consider this in a situation 7652 where TPM is heavily utilized by IMA, thus protection 7653 causing a major performance hit, and the space where 7654 machines are deployed is by other means guarded. 7655 7656 tpm_crb_ffa.busy_timeout_ms= [ARM64,TPM] 7657 Maximum time in milliseconds to retry sending a message 7658 to the TPM service before giving up. This parameter controls 7659 how long the system will continue retrying when the TPM 7660 service is busy. 7661 Format: <unsigned int> 7662 Default: 2000 (2 seconds) 7663 7664 tpm_suspend_pcr=[HW,TPM] 7665 Format: integer pcr id 7666 Specify that at suspend time, the tpm driver 7667 should extend the specified pcr with zeros, 7668 as a workaround for some chips which fail to 7669 flush the last written pcr on TPM_SaveState. 7670 This will guarantee that all the other pcrs 7671 are saved. 7672 7673 tpm_tis.interrupts= [HW,TPM] 7674 Enable interrupts for the MMIO based physical layer 7675 for the FIFO interface. By default it is set to false 7676 (0). For more information about TPM hardware interfaces 7677 defined by Trusted Computing Group (TCG) see 7678 https://trustedcomputinggroup.org/resource/pc-client-platform-tpm-profile-ptp-specification/ 7679 7680 tp_printk [FTRACE] 7681 Have the tracepoints sent to printk as well as the 7682 tracing ring buffer. This is useful for early boot up 7683 where the system hangs or reboots and does not give the 7684 option for reading the tracing buffer or performing a 7685 ftrace_dump_on_oops. 7686 7687 To turn off having tracepoints sent to printk, 7688 echo 0 > /proc/sys/kernel/tracepoint_printk 7689 Note, echoing 1 into this file without the 7690 tp_printk kernel cmdline option has no effect. 7691 7692 The tp_printk_stop_on_boot (see below) can also be used 7693 to stop the printing of events to console at 7694 late_initcall_sync. 7695 7696 ** CAUTION ** 7697 7698 Having tracepoints sent to printk() and activating high 7699 frequency tracepoints such as irq or sched, can cause 7700 the system to live lock. 7701 7702 tp_printk_stop_on_boot [FTRACE] 7703 When tp_printk (above) is set, it can cause a lot of noise 7704 on the console. It may be useful to only include the 7705 printing of events during boot up, as user space may 7706 make the system inoperable. 7707 7708 This command line option will stop the printing of events 7709 to console at the late_initcall_sync() time frame. 7710 7711 trace_buf_size=nn[KMG] 7712 [FTRACE] will set tracing buffer size on each cpu. 7713 7714 trace_clock= [FTRACE] Set the clock used for tracing events 7715 at boot up. 7716 local - Use the per CPU time stamp counter 7717 (converted into nanoseconds). Fast, but 7718 depending on the architecture, may not be 7719 in sync between CPUs. 7720 global - Event time stamps are synchronized across 7721 CPUs. May be slower than the local clock, 7722 but better for some race conditions. 7723 counter - Simple counting of events (1, 2, ..) 7724 note, some counts may be skipped due to the 7725 infrastructure grabbing the clock more than 7726 once per event. 7727 uptime - Use jiffies as the time stamp. 7728 perf - Use the same clock that perf uses. 7729 mono - Use ktime_get_mono_fast_ns() for time stamps. 7730 mono_raw - Use ktime_get_raw_fast_ns() for time 7731 stamps. 7732 boot - Use ktime_get_boot_fast_ns() for time stamps. 7733 Architectures may add more clocks. See 7734 Documentation/trace/ftrace.rst for more details. 7735 7736 trace_event=[event-list] 7737 [FTRACE] Set and start specified trace events in order 7738 to facilitate early boot debugging. The event-list is a 7739 comma-separated list of trace events to enable. See 7740 also Documentation/trace/events.rst 7741 7742 To enable modules, use :mod: keyword: 7743 7744 trace_event=:mod:<module> 7745 7746 The value before :mod: will only enable specific events 7747 that are part of the module. See the above mentioned 7748 document for more information. 7749 7750 trace_instance=[instance-info] 7751 [FTRACE] Create a ring buffer instance early in boot up. 7752 This will be listed in: 7753 7754 /sys/kernel/tracing/instances 7755 7756 Events can be enabled at the time the instance is created 7757 via: 7758 7759 trace_instance=<name>,<system1>:<event1>,<system2>:<event2> 7760 7761 Note, the "<system*>:" portion is optional if the event is 7762 unique. 7763 7764 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall 7765 7766 will enable the "sched_switch" event (note, the "sched:" is optional, and 7767 the same thing would happen if it was left off). The irq_handler_entry 7768 event, and all events under the "initcall" system. 7769 7770 Flags can be added to the instance to modify its behavior when it is 7771 created. The flags are separated by '^'. 7772 7773 The available flags are: 7774 7775 traceoff - Have the tracing instance tracing disabled after it is created. 7776 traceprintk - Have trace_printk() write into this trace instance 7777 (note, "printk" and "trace_printk" can also be used) 7778 7779 trace_instance=foo^traceoff^traceprintk,sched,irq 7780 7781 The flags must come before the defined events. 7782 7783 If memory has been reserved (see memmap for x86), the instance 7784 can use that memory: 7785 7786 memmap=12M$0x284500000 trace_instance=boot_map@0x284500000:12M 7787 7788 The above will create a "boot_map" instance that uses the physical 7789 memory at 0x284500000 that is 12Megs. The per CPU buffers of that 7790 instance will be split up accordingly. 7791 7792 Alternatively, the memory can be reserved by the reserve_mem option: 7793 7794 reserve_mem=12M:4096:trace trace_instance=boot_map@trace 7795 7796 This will reserve 12 megabytes at boot up with a 4096 byte alignment 7797 and place the ring buffer in this memory. Note that due to KASLR, the 7798 memory may not be the same location each time, which will not preserve 7799 the buffer content. 7800 7801 Also note that the layout of the ring buffer data may change between 7802 kernel versions where the validator will fail and reset the ring buffer 7803 if the layout is not the same as the previous kernel. 7804 7805 If the ring buffer is used for persistent bootups and has events enabled, 7806 it is recommend to disable tracing so that events from a previous boot do not 7807 mix with events of the current boot (unless you are debugging a random crash 7808 at boot up). 7809 7810 reserve_mem=12M:4096:trace trace_instance=boot_map^traceoff^traceprintk@trace,sched,irq 7811 7812 Note, saving the trace buffer across reboots does require that the system 7813 is set up to not wipe memory. For instance, CONFIG_RESET_ATTACK_MITIGATION 7814 can force a memory reset on boot which will clear any trace that was stored. 7815 This is just one of many ways that can clear memory. Make sure your system 7816 keeps the content of memory across reboots before relying on this option. 7817 7818 NB: Both the mapped address and size must be page aligned for the architecture. 7819 7820 See also Documentation/trace/debugging.rst 7821 7822 7823 trace_options=[option-list] 7824 [FTRACE] Enable or disable tracer options at boot. 7825 The option-list is a comma delimited list of options 7826 that can be enabled or disabled just as if you were 7827 to echo the option name into 7828 7829 /sys/kernel/tracing/trace_options 7830 7831 For example, to enable stacktrace option (to dump the 7832 stack trace of each event), add to the command line: 7833 7834 trace_options=stacktrace 7835 7836 See also Documentation/trace/ftrace.rst "trace options" 7837 section. 7838 7839 trace_trigger=[trigger-list] 7840 [FTRACE] Add an event trigger on specific events. 7841 Set a trigger on top of a specific event, with an optional 7842 filter. 7843 7844 The format is "trace_trigger=<event>.<trigger>[ if <filter>],..." 7845 Where more than one trigger may be specified that are comma delimited. 7846 7847 For example: 7848 7849 trace_trigger="sched_switch.stacktrace if prev_state == 2" 7850 7851 The above will enable the "stacktrace" trigger on the "sched_switch" 7852 event but only trigger it if the "prev_state" of the "sched_switch" 7853 event is "2" (TASK_UNINTERRUPTIBLE). 7854 7855 See also "Event triggers" in Documentation/trace/events.rst 7856 7857 7858 traceoff_after_boot 7859 [FTRACE] Sometimes tracing is used to debug issues 7860 during the boot process. Since the trace buffer has a 7861 limited amount of storage, it may be prudent to 7862 disable tracing after the boot is finished, otherwise 7863 the critical information may be overwritten. With this 7864 option, the main tracing buffer will be turned off at 7865 the end of the boot process. 7866 7867 traceoff_on_warning 7868 [FTRACE] enable this option to disable tracing when a 7869 warning is hit. This turns off "tracing_on". Tracing can 7870 be enabled again by echoing '1' into the "tracing_on" 7871 file located in /sys/kernel/tracing/ 7872 7873 This option is useful, as it disables the trace before 7874 the WARNING dump is called, which prevents the trace to 7875 be filled with content caused by the warning output. 7876 7877 This option can also be set at run time via the sysctl 7878 option: kernel/traceoff_on_warning 7879 7880 transparent_hugepage= 7881 [KNL] 7882 Format: [always|madvise|never] 7883 Can be used to control the default behavior of the system 7884 with respect to transparent hugepages. 7885 See Documentation/admin-guide/mm/transhuge.rst 7886 for more details. 7887 7888 transparent_hugepage_shmem= [KNL] 7889 Format: [always|within_size|advise|never|deny|force] 7890 Can be used to control the hugepage allocation policy for 7891 the internal shmem mount. 7892 See Documentation/admin-guide/mm/transhuge.rst 7893 for more details. 7894 7895 transparent_hugepage_tmpfs= [KNL] 7896 Format: [always|within_size|advise|never] 7897 Can be used to control the default hugepage allocation policy 7898 for the tmpfs mount. 7899 See Documentation/admin-guide/mm/transhuge.rst 7900 for more details. 7901 7902 trusted.source= [KEYS] 7903 Format: <string> 7904 This parameter identifies the trust source as a backend 7905 for trusted keys implementation. Supported trust 7906 sources: 7907 - "tpm" 7908 - "tee" 7909 - "caam" 7910 - "dcp" 7911 - "pkwm" 7912 If not specified then it defaults to iterating through 7913 the trust source list starting with TPM and assigns the 7914 first trust source as a backend which is initialized 7915 successfully during iteration. 7916 7917 trusted.debug= [KEYS] 7918 Format: <bool> 7919 Enable trusted keys debug traces at runtime when 7920 CONFIG_TRUSTED_KEYS_DEBUG=y. 7921 7922 To make the traces visible after enabling the option, 7923 use trusted.dyndbg='+p' as needed. By convention, 7924 the subsystem uses pr_debug() for these traces. 7925 7926 SAFETY: The traces can leak sensitive data, so be 7927 cautious before enabling this. They remain inactive 7928 unless this parameter is set this option to a true 7929 value. 7930 7931 Default: false 7932 7933 trusted.rng= [KEYS] 7934 Format: <string> 7935 The RNG used to generate key material for trusted keys. 7936 Can be one of: 7937 - "kernel" 7938 - the same value as trusted.source: "tpm" or "tee" 7939 - "default" 7940 If not specified, "default" is used. In this case, 7941 the RNG's choice is left to each individual trust source. 7942 7943 trusted.dcp_use_otp_key 7944 This is intended to be used in combination with 7945 trusted.source=dcp and will select the DCP OTP key 7946 instead of the DCP UNIQUE key blob encryption. 7947 7948 trusted.dcp_skip_zk_test 7949 This is intended to be used in combination with 7950 trusted.source=dcp and will disable the check if the 7951 blob key is all zeros. This is helpful for situations where 7952 having this key zero'ed is acceptable. E.g. in testing 7953 scenarios. 7954 7955 tsa= [X86] Control mitigation for Transient Scheduler 7956 Attacks on AMD CPUs. Search the following in your 7957 favourite search engine for more details: 7958 7959 "Technical guidance for mitigating transient scheduler 7960 attacks". 7961 7962 off - disable the mitigation 7963 on - enable the mitigation (default) 7964 user - mitigate only user/kernel transitions 7965 vm - mitigate only guest/host transitions 7966 7967 7968 tsc= Disable clocksource stability checks for TSC. 7969 Format: <string> 7970 [x86] reliable: mark tsc clocksource as reliable, this 7971 disables clocksource verification at runtime, as well 7972 as the stability checks done at bootup. Used to enable 7973 high-resolution timer mode on older hardware, and in 7974 virtualized environment. 7975 [x86] noirqtime: Do not use TSC to do irq accounting. 7976 Used to run time disable IRQ_TIME_ACCOUNTING on any 7977 platforms where RDTSC is slow and this accounting 7978 can add overhead. 7979 [x86] unstable: mark the TSC clocksource as unstable, this 7980 marks the TSC unconditionally unstable at bootup and 7981 avoids any further wobbles once the TSC watchdog notices. 7982 [x86] nowatchdog: disable clocksource watchdog. Used 7983 in situations with strict latency requirements (where 7984 interruptions from clocksource watchdog are not 7985 acceptable). 7986 [x86] recalibrate: force recalibration against a HW timer 7987 (HPET or PM timer) on systems whose TSC frequency was 7988 obtained from HW or FW using either an MSR or CPUID(0x15). 7989 Warn if the difference is more than 500 ppm. 7990 [x86] watchdog: Enforce the clocksource watchdog on TSC 7991 7992 tsc_early_khz= [X86,EARLY] Skip early TSC calibration and use the given 7993 value instead. Useful when the early TSC frequency discovery 7994 procedure is not reliable, such as on overclocked systems 7995 with CPUID.16h support and partial CPUID.15h support. 7996 Format: <unsigned int> 7997 7998 tsx= [X86] Control Transactional Synchronization 7999 Extensions (TSX) feature in Intel processors that 8000 support TSX control. 8001 8002 This parameter controls the TSX feature. The options are: 8003 8004 on - Enable TSX on the system. Although there are 8005 mitigations for all known security vulnerabilities, 8006 TSX has been known to be an accelerator for 8007 several previous speculation-related CVEs, and 8008 so there may be unknown security risks associated 8009 with leaving it enabled. 8010 8011 off - Disable TSX on the system. (Note that this 8012 option takes effect only on newer CPUs which are 8013 not vulnerable to MDS, i.e., have 8014 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 8015 the new IA32_TSX_CTRL MSR through a microcode 8016 update. This new MSR allows for the reliable 8017 deactivation of the TSX functionality.) 8018 8019 auto - Disable TSX if X86_BUG_TAA is present, 8020 otherwise enable TSX on the system. 8021 8022 Not specifying this option is equivalent to tsx=off. 8023 8024 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 8025 for more details. 8026 8027 tsx_async_abort= [X86,INTEL,EARLY] Control mitigation for the TSX Async 8028 Abort (TAA) vulnerability. 8029 8030 Similar to Micro-architectural Data Sampling (MDS) 8031 certain CPUs that support Transactional 8032 Synchronization Extensions (TSX) are vulnerable to an 8033 exploit against CPU internal buffers which can forward 8034 information to a disclosure gadget under certain 8035 conditions. 8036 8037 In vulnerable processors, the speculatively forwarded 8038 data can be used in a cache side channel attack, to 8039 access data to which the attacker does not have direct 8040 access. 8041 8042 This parameter controls the TAA mitigation. The 8043 options are: 8044 8045 full - Enable TAA mitigation on vulnerable CPUs 8046 if TSX is enabled. 8047 8048 full,nosmt - Enable TAA mitigation and disable SMT on 8049 vulnerable CPUs. If TSX is disabled, SMT 8050 is not disabled because CPU is not 8051 vulnerable to cross-thread TAA attacks. 8052 off - Unconditionally disable TAA mitigation 8053 8054 On MDS-affected machines, tsx_async_abort=off can be 8055 prevented by an active MDS mitigation as both vulnerabilities 8056 are mitigated with the same mechanism so in order to disable 8057 this mitigation, you need to specify mds=off too. 8058 8059 Not specifying this option is equivalent to 8060 tsx_async_abort=full. On CPUs which are MDS affected 8061 and deploy MDS mitigation, TAA mitigation is not 8062 required and doesn't provide any additional 8063 mitigation. 8064 8065 For details see: 8066 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 8067 8068 turbografx.map[2|3]= [HW,JOY] 8069 TurboGraFX parallel port interface 8070 Format: 8071 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 8072 See also Documentation/input/devices/joystick-parport.rst 8073 8074 udbg-immortal [PPC] When debugging early kernel crashes that 8075 happen after console_init() and before a proper 8076 console driver takes over, this boot options might 8077 help "seeing" what's going on. 8078 8079 uhash_entries= [KNL,NET] 8080 Set number of hash buckets for UDP/UDP-Lite connections 8081 8082 uhci-hcd.ignore_oc= 8083 [USB] Ignore overcurrent events (default N). 8084 Some badly-designed motherboards generate lots of 8085 bogus events, for ports that aren't wired to 8086 anything. Set this parameter to avoid log spamming. 8087 Note that genuine overcurrent events won't be 8088 reported either. 8089 8090 unaligned_scalar_speed= 8091 [RISCV] 8092 Format: {slow | fast | unsupported} 8093 Allow skipping scalar unaligned access speed tests. This 8094 is useful for testing alternative code paths and to skip 8095 the tests in environments where they run too slowly. All 8096 CPUs must have the same scalar unaligned access speed. 8097 8098 unaligned_vector_speed= 8099 [RISCV] 8100 Format: {slow | fast | unsupported} 8101 Allow skipping vector unaligned access speed tests. This 8102 is useful for testing alternative code paths and to skip 8103 the tests in environments where they run too slowly. All 8104 CPUs must have the same vector unaligned access speed. 8105 8106 unknown_nmi_panic 8107 [X86] Cause panic on unknown NMI. 8108 8109 unwind_debug [X86-64,EARLY] 8110 Enable unwinder debug output. This can be 8111 useful for debugging certain unwinder error 8112 conditions, including corrupt stacks and 8113 bad/missing unwinder metadata. 8114 8115 usbcore.authorized_default= 8116 [USB] Default USB device authorization: 8117 (default -1 = authorized (same as 1), 8118 0 = not authorized, 1 = authorized, 2 = authorized 8119 if device connected to internal port) 8120 8121 usbcore.autosuspend= 8122 [USB] The autosuspend time delay (in seconds) used 8123 for newly-detected USB devices (default 2). This 8124 is the time required before an idle device will be 8125 autosuspended. Devices for which the delay is set 8126 to a negative value won't be autosuspended at all. 8127 8128 usbcore.usbfs_snoop= 8129 [USB] Set to log all usbfs traffic (default 0 = off). 8130 8131 usbcore.usbfs_snoop_max= 8132 [USB] Maximum number of bytes to snoop in each URB 8133 (default = 65536). 8134 8135 usbcore.blinkenlights= 8136 [USB] Set to cycle leds on hubs (default 0 = off). 8137 8138 usbcore.old_scheme_first= 8139 [USB] Start with the old device initialization 8140 scheme (default 0 = off). 8141 8142 usbcore.usbfs_memory_mb= 8143 [USB] Memory limit (in MB) for buffers allocated by 8144 usbfs (default = 16, 0 = max = 2047). 8145 8146 usbcore.use_both_schemes= 8147 [USB] Try the other device initialization scheme 8148 if the first one fails (default 1 = enabled). 8149 8150 usbcore.initial_descriptor_timeout= 8151 [USB] Specifies timeout for the initial 64-byte 8152 USB_REQ_GET_DESCRIPTOR request in milliseconds 8153 (default 5000 = 5.0 seconds). 8154 8155 usbcore.nousb [USB] Disable the USB subsystem 8156 8157 usbcore.quirks= 8158 [USB] A list of quirk entries to augment the built-in 8159 usb core quirk list. List entries are separated by 8160 commas. Each entry has the form 8161 VendorID:ProductID:Flags. The IDs are 4-digit hex 8162 numbers and Flags is a set of letters. Each letter 8163 will change the built-in quirk; setting it if it is 8164 clear and clearing it if it is set. The letters have 8165 the following meanings: 8166 a = USB_QUIRK_STRING_FETCH_255 (string 8167 descriptors must not be fetched using 8168 a 255-byte read); 8169 b = USB_QUIRK_RESET_RESUME (device can't resume 8170 correctly so reset it instead); 8171 c = USB_QUIRK_NO_SET_INTF (device can't handle 8172 Set-Interface requests); 8173 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 8174 handle its Configuration or Interface 8175 strings); 8176 e = USB_QUIRK_RESET (device can't be reset 8177 (e.g morph devices), don't use reset); 8178 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 8179 more interface descriptions than the 8180 bNumInterfaces count, and can't handle 8181 talking to these interfaces); 8182 g = USB_QUIRK_DELAY_INIT (device needs a pause 8183 during initialization, after we read 8184 the device descriptor); 8185 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 8186 high speed and super speed interrupt 8187 endpoints, the USB 2.0 and USB 3.0 spec 8188 require the interval in microframes (1 8189 microframe = 125 microseconds) to be 8190 calculated as interval = 2 ^ 8191 (bInterval-1). 8192 Devices with this quirk report their 8193 bInterval as the result of this 8194 calculation instead of the exponent 8195 variable used in the calculation); 8196 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 8197 handle device_qualifier descriptor 8198 requests); 8199 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 8200 generates spurious wakeup, ignore 8201 remote wakeup capability); 8202 k = USB_QUIRK_NO_LPM (device can't handle Link 8203 Power Management); 8204 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 8205 (Device reports its bInterval as linear 8206 frames instead of the USB 2.0 8207 calculation); 8208 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 8209 to be disconnected before suspend to 8210 prevent spurious wakeup); 8211 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 8212 pause after every control message); 8213 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 8214 delay after resetting its port); 8215 p = USB_QUIRK_SHORT_SET_ADDRESS_REQ_TIMEOUT 8216 (Reduce timeout of the SET_ADDRESS 8217 request from 5000 ms to 500 ms); 8218 q = USB_QUIRK_FORCE_ONE_CONFIG (Device 8219 claims zero configurations, 8220 forcing to 1); 8221 r = USB_QUIRK_WINDOWS_CONFIG_REQ_SIZE (Device 8222 fails during initialization when asked for 8223 9-bytes configuration descriptor request. 8224 Ask for 255-bytes request instead to mirror 8225 Windows' behavior); 8226 Example: quirks=0781:5580:bk,0a5c:5834:gij 8227 8228 usbhid.mousepoll= 8229 [USBHID] The interval which mice are to be polled at. 8230 8231 usbhid.jspoll= 8232 [USBHID] The interval which joysticks are to be polled at. 8233 8234 usbhid.kbpoll= 8235 [USBHID] The interval which keyboards are to be polled at. 8236 8237 usb-storage.delay_use= 8238 [UMS] The delay in seconds before a new device is 8239 scanned for Logical Units (default 1). 8240 Optionally the delay in milliseconds if the value has 8241 suffix with "ms". 8242 Example: delay_use=2567ms 8243 8244 usb-storage.quirks= 8245 [UMS] A list of quirks entries to supplement or 8246 override the built-in unusual_devs list. List 8247 entries are separated by commas. Each entry has 8248 the form VID:PID:Flags where VID and PID are Vendor 8249 and Product ID values (4-digit hex numbers) and 8250 Flags is a set of characters, each corresponding 8251 to a common usb-storage quirk flag as follows: 8252 a = SANE_SENSE (collect more than 18 bytes 8253 of sense data, not on uas); 8254 b = BAD_SENSE (don't collect more than 18 8255 bytes of sense data, not on uas); 8256 c = FIX_CAPACITY (decrease the reported 8257 device capacity by one sector); 8258 d = NO_READ_DISC_INFO (don't use 8259 READ_DISC_INFO command, not on uas); 8260 e = NO_READ_CAPACITY_16 (don't use 8261 READ_CAPACITY_16 command); 8262 f = NO_REPORT_OPCODES (don't use report opcodes 8263 command, uas only); 8264 g = MAX_SECTORS_240 (don't transfer more than 8265 240 sectors at a time, uas only); 8266 h = CAPACITY_HEURISTICS (decrease the 8267 reported device capacity by one 8268 sector if the number is odd); 8269 i = IGNORE_DEVICE (don't bind to this 8270 device); 8271 j = NO_REPORT_LUNS (don't use report luns 8272 command, uas only); 8273 k = NO_SAME (do not use WRITE_SAME, uas only) 8274 l = NOT_LOCKABLE (don't try to lock and 8275 unlock ejectable media, not on uas); 8276 m = MAX_SECTORS_64 (don't transfer more 8277 than 64 sectors = 32 KB at a time, 8278 not on uas); 8279 n = INITIAL_READ10 (force a retry of the 8280 initial READ(10) command, not on uas); 8281 o = CAPACITY_OK (accept the capacity 8282 reported by the device, not on uas); 8283 p = WRITE_CACHE (the device cache is ON 8284 by default, not on uas); 8285 r = IGNORE_RESIDUE (the device reports 8286 bogus residue values, not on uas); 8287 s = SINGLE_LUN (the device has only one 8288 Logical Unit); 8289 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 8290 commands, uas only); 8291 u = IGNORE_UAS (don't bind to the uas driver); 8292 w = NO_WP_DETECT (don't test whether the 8293 medium is write-protected). 8294 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 8295 even if the device claims no cache, 8296 not on uas) 8297 Example: quirks=0419:aaf5:rl,0421:0433:rc 8298 8299 user_debug= [KNL,ARM] 8300 Format: <int> 8301 See arch/arm/Kconfig.debug help text. 8302 1 - undefined instruction events 8303 2 - system calls 8304 4 - invalid data aborts 8305 8 - SIGSEGV faults 8306 16 - SIGBUS faults 8307 Example: user_debug=31 8308 8309 vdso= [X86,SH,SPARC] 8310 On X86_32, this is an alias for vdso32=. Otherwise: 8311 8312 vdso=1: enable VDSO (the default) 8313 vdso=0: disable VDSO mapping 8314 8315 vdso32= [X86] Control the 32-bit vDSO 8316 vdso32=1: enable 32-bit VDSO 8317 vdso32=0 or vdso32=2: disable 32-bit VDSO 8318 8319 See the help text for CONFIG_COMPAT_VDSO for more 8320 details. If CONFIG_COMPAT_VDSO is set, the default is 8321 vdso32=0; otherwise, the default is vdso32=1. 8322 8323 For compatibility with older kernels, vdso32=2 is an 8324 alias for vdso32=0. 8325 8326 Try vdso32=0 if you encounter an error that says: 8327 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 8328 8329 video= [FB,EARLY] Frame buffer configuration 8330 See Documentation/fb/modedb.rst. 8331 8332 video.brightness_switch_enabled= [ACPI] 8333 Format: [0|1] 8334 If set to 1, on receiving an ACPI notify event 8335 generated by hotkey, video driver will adjust brightness 8336 level and then send out the event to user space through 8337 the allocated input device. If set to 0, video driver 8338 will only send out the event without touching backlight 8339 brightness level. 8340 default: 1 8341 8342 virtio_mmio.device= 8343 [VMMIO] Memory mapped virtio (platform) device. 8344 8345 <size>@<baseaddr>:<irq>[:<id>] 8346 where: 8347 <size> := size (can use standard suffixes 8348 like K, M and G) 8349 <baseaddr> := physical base address 8350 <irq> := interrupt number (as passed to 8351 request_irq()) 8352 <id> := (optional) platform device id 8353 example: 8354 virtio_mmio.device=1K@0x100b0000:48:7 8355 8356 Can be used multiple times for multiple devices. 8357 8358 vga= [BOOT,X86-32] Select a particular video mode 8359 See Documentation/arch/x86/boot.rst and 8360 Documentation/admin-guide/svga.rst. 8361 Use vga=ask for menu. 8362 This is actually a boot loader parameter; the value is 8363 passed to the kernel using a special protocol. 8364 8365 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 8366 May slow down system boot speed, especially when 8367 enabled on systems with a large amount of memory. 8368 All options are enabled by default, and this 8369 interface is meant to allow for selectively 8370 enabling or disabling specific virtual memory 8371 debugging features. 8372 8373 Available options are: 8374 P Enable page structure init time poisoning 8375 - Disable all of the above options 8376 8377 vmalloc=nn[KMG] [KNL,BOOT,EARLY] Forces the vmalloc area to have an 8378 exact size of <nn>. This can be used to increase 8379 the minimum size (128MB on x86, arm32 platforms). 8380 It can also be used to decrease the size and leave more room 8381 for directly mapped kernel RAM. Note that this parameter does 8382 not exist on many other platforms (including arm64, alpha, 8383 loongarch, arc, csky, hexagon, microblaze, mips, nios2, openrisc, 8384 parisc, m64k, powerpc, riscv, sh, um, xtensa, s390, sparc). 8385 8386 vmcp_cma=nn[MG] [KNL,S390,EARLY] 8387 Sets the memory size reserved for contiguous memory 8388 allocations for the vmcp device driver. 8389 8390 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 8391 Format: <command> 8392 8393 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 8394 Format: <command> 8395 8396 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 8397 Format: <command> 8398 8399 vmscape= [X86] Controls mitigation for VMscape attacks. 8400 VMscape attacks can leak information from a userspace 8401 hypervisor to a guest via speculative side-channels. 8402 8403 off - disable the mitigation 8404 ibpb - use Indirect Branch Prediction Barrier 8405 (IBPB) mitigation (default) 8406 force - force vulnerability detection even on 8407 unaffected processors 8408 8409 vsyscall= [X86-64,EARLY] 8410 Controls the behavior of vsyscalls (i.e. calls to 8411 fixed addresses of 0xffffffffff600x00 from legacy 8412 code). Most statically-linked binaries and older 8413 versions of glibc use these calls. Because these 8414 functions are at fixed addresses, they make nice 8415 targets for exploits that can control RIP. 8416 8417 emulate Vsyscalls turn into traps and are emulated 8418 reasonably safely. The vsyscall page is 8419 readable. This disables the Linear 8420 Address Space Separation (LASS) security 8421 feature and makes the system less secure. 8422 8423 xonly [default] Vsyscalls turn into traps and are 8424 emulated reasonably safely. The vsyscall 8425 page is not readable. 8426 8427 none Vsyscalls don't work at all. This makes 8428 them quite hard to use for exploits but 8429 might break your system. 8430 8431 vt.color= [VT] Default text color. 8432 Format: 0xYX, X = foreground, Y = background. 8433 Default: 0x07 = light gray on black. 8434 8435 vt.cur_default= [VT] Default cursor shape. 8436 Format: 0xCCBBAA, where AA, BB, and CC are the same as 8437 the parameters of the <Esc>[?A;B;Cc escape sequence; 8438 see vga-softcursor.rst. Default: 2 = underline. 8439 8440 vt.default_blu= [VT] 8441 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 8442 Change the default blue palette of the console. 8443 This is a 16-member array composed of values 8444 ranging from 0-255. 8445 8446 vt.default_grn= [VT] 8447 Format: <green0>,<green1>,<green2>,...,<green15> 8448 Change the default green palette of the console. 8449 This is a 16-member array composed of values 8450 ranging from 0-255. 8451 8452 vt.default_red= [VT] 8453 Format: <red0>,<red1>,<red2>,...,<red15> 8454 Change the default red palette of the console. 8455 This is a 16-member array composed of values 8456 ranging from 0-255. 8457 8458 vt.default_utf8= 8459 [VT] 8460 Format=<0|1> 8461 Set system-wide default UTF-8 mode for all tty's. 8462 Default is 1, i.e. UTF-8 mode is enabled for all 8463 newly opened terminals. 8464 8465 vt.global_cursor_default= 8466 [VT] 8467 Format=<-1|0|1> 8468 Set system-wide default for whether a cursor 8469 is shown on new VTs. Default is -1, 8470 i.e. cursors will be created by default unless 8471 overridden by individual drivers. 0 will hide 8472 cursors, 1 will display them. 8473 8474 vt.italic= [VT] Default color for italic text; 0-15. 8475 Default: 2 = green. 8476 8477 vt.underline= [VT] Default color for underlined text; 0-15. 8478 Default: 3 = cyan. 8479 8480 watchdog timers [HW,WDT] For information on watchdog timers, 8481 see Documentation/watchdog/watchdog-parameters.rst 8482 or other driver-specific files in the 8483 Documentation/watchdog/ directory. 8484 8485 watchdog_thresh= 8486 [KNL] 8487 Set the hard lockup detector stall duration 8488 threshold in seconds. The soft lockup detector 8489 threshold is set to twice the value. A value of 0 8490 disables both lockup detectors. Default is 10 8491 seconds. 8492 8493 workqueue.unbound_cpus= 8494 [KNL,SMP] Specify to constrain one or some CPUs 8495 to use in unbound workqueues. 8496 Format: <cpu-list> 8497 By default, all online CPUs are available for 8498 unbound workqueues. 8499 8500 workqueue.watchdog_thresh= 8501 If CONFIG_WQ_WATCHDOG is configured, workqueue can 8502 warn stall conditions and dump internal state to 8503 help debugging. 0 disables workqueue stall 8504 detection; otherwise, it's the stall threshold 8505 duration in seconds. The default value is 30 and 8506 it can be updated at runtime by writing to the 8507 corresponding sysfs file. 8508 8509 workqueue.panic_on_stall=<uint> 8510 Panic when workqueue stall is detected by 8511 CONFIG_WQ_WATCHDOG. It sets the number times of the 8512 stall to trigger panic. 8513 8514 The default is set by CONFIG_BOOTPARAM_WQ_STALL_PANIC, 8515 which is 0 (disabled) if not configured. 8516 8517 workqueue.panic_on_stall_time=<uint> 8518 Panic when a workqueue stall has been continuous for 8519 the specified number of seconds. Unlike panic_on_stall 8520 which counts accumulated stall events, this triggers 8521 based on the duration of a single continuous stall. 8522 8523 The default is 0, which disables the time-based panic. 8524 8525 workqueue.cpu_intensive_thresh_us= 8526 Per-cpu work items which run for longer than this 8527 threshold are automatically considered CPU intensive 8528 and excluded from concurrency management to prevent 8529 them from noticeably delaying other per-cpu work 8530 items. Default is 10000 (10ms). 8531 8532 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8533 will report the work functions which violate this 8534 threshold repeatedly. They are likely good 8535 candidates for using WQ_UNBOUND workqueues instead. 8536 8537 workqueue.cpu_intensive_warning_thresh=<uint> 8538 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8539 will report the work functions which violate the 8540 intensive_threshold_us repeatedly. In order to prevent 8541 spurious warnings, start printing only after a work 8542 function has violated this threshold number of times. 8543 8544 The default is 4 times. 0 disables the warning. 8545 8546 workqueue.power_efficient 8547 Per-cpu workqueues are generally preferred because 8548 they show better performance thanks to cache 8549 locality; unfortunately, per-cpu workqueues tend to 8550 be more power hungry than unbound workqueues. 8551 8552 Enabling this makes the per-cpu workqueues which 8553 were observed to contribute significantly to power 8554 consumption unbound, leading to measurably lower 8555 power usage at the cost of small performance 8556 overhead. 8557 8558 The default value of this parameter is determined by 8559 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 8560 8561 workqueue.default_affinity_scope= 8562 Select the default affinity scope to use for unbound 8563 workqueues. Can be one of "cpu", "smt", "cache", 8564 "cache_shard", "numa" and "system". Default is 8565 "cache_shard". For more 8566 information, see the Affinity Scopes section in 8567 Documentation/core-api/workqueue.rst. 8568 8569 This can be changed after boot by writing to the 8570 matching /sys/module/workqueue/parameters file. All 8571 workqueues with the "default" affinity scope will be 8572 updated accordingly. 8573 8574 workqueue.debug_force_rr_cpu 8575 Workqueue used to implicitly guarantee that work 8576 items queued without explicit CPU specified are put 8577 on the local CPU. This guarantee is no longer true 8578 and while local CPU is still preferred work items 8579 may be put on foreign CPUs. This debug option 8580 forces round-robin CPU selection to flush out 8581 usages which depend on the now broken guarantee. 8582 When enabled, memory and cache locality will be 8583 impacted. 8584 8585 writecombine= [LOONGARCH,EARLY] Control the MAT (Memory Access 8586 Type) of ioremap_wc(). 8587 8588 on - Enable writecombine, use WUC for ioremap_wc() 8589 off - Disable writecombine, use SUC for ioremap_wc() 8590 8591 x2apic_phys [X86-64,APIC,EARLY] Use x2apic physical mode instead of 8592 default x2apic cluster mode on platforms 8593 supporting x2apic. 8594 8595 xen_512gb_limit [KNL,X86-64,XEN] 8596 Restricts the kernel running paravirtualized under Xen 8597 to use only up to 512 GB of RAM. The reason to do so is 8598 crash analysis tools and Xen tools for doing domain 8599 save/restore/migration must be enabled to handle larger 8600 domains. 8601 8602 xen_console_io [XEN,EARLY] 8603 Boolean option to enable/disable the usage of the Xen 8604 console_io hypercalls to read and write to the console. 8605 Mostly useful for debugging and development. 8606 8607 xen_emul_unplug= [HW,X86,XEN,EARLY] 8608 Unplug Xen emulated devices 8609 Format: [unplug0,][unplug1] 8610 ide-disks -- unplug primary master IDE devices 8611 aux-ide-disks -- unplug non-primary-master IDE devices 8612 nics -- unplug network devices 8613 all -- unplug all emulated devices (NICs and IDE disks) 8614 unnecessary -- unplugging emulated devices is 8615 unnecessary even if the host did not respond to 8616 the unplug protocol 8617 never -- do not unplug even if version check succeeds 8618 8619 xen_legacy_crash [X86,XEN,EARLY] 8620 Crash from Xen panic notifier, without executing late 8621 panic() code such as dumping handler. 8622 8623 xen_mc_debug [X86,XEN,EARLY] 8624 Enable multicall debugging when running as a Xen PV guest. 8625 Enabling this feature will reduce performance a little 8626 bit, so it should only be enabled for obtaining extended 8627 debug data in case of multicall errors. 8628 8629 xen_msr_safe= [X86,XEN,EARLY] 8630 Format: <bool> 8631 Select whether to always use non-faulting (safe) MSR 8632 access functions when running as Xen PV guest. The 8633 default value is controlled by CONFIG_XEN_PV_MSR_SAFE. 8634 8635 xen_nopv [X86] 8636 Disables the PV optimizations forcing the HVM guest to 8637 run as generic HVM guest with no PV drivers. 8638 This option is obsoleted by the "nopv" option, which 8639 has equivalent effect for XEN platform. 8640 8641 xen_no_vector_callback 8642 [KNL,X86,XEN,EARLY] Disable the vector callback for Xen 8643 event channel interrupts. 8644 8645 xen_scrub_pages= [XEN] 8646 Boolean option to control scrubbing pages before giving them back 8647 to Xen, for use by other domains. Can be also changed at runtime 8648 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 8649 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 8650 8651 xen_timer_slop= [X86-64,XEN,EARLY] 8652 Set the timer slop (in nanoseconds) for the virtual Xen 8653 timers (default is 100000). This adjusts the minimum 8654 delta of virtualized Xen timers, where lower values 8655 improve timer resolution at the expense of processing 8656 more timer interrupts. 8657 8658 xen.balloon_boot_timeout= [XEN] 8659 The time (in seconds) to wait before giving up to boot 8660 in case initial ballooning fails to free enough memory. 8661 Applies only when running as HVM or PVH guest and 8662 started with less memory configured than allowed at 8663 max. Default is 180. 8664 8665 xen.event_eoi_delay= [XEN] 8666 How long to delay EOI handling in case of event 8667 storms (jiffies). Default is 10. 8668 8669 xen.event_loop_timeout= [XEN] 8670 After which time (jiffies) the event handling loop 8671 should start to delay EOI handling. Default is 2. 8672 8673 xen.fifo_events= [XEN] 8674 Boolean parameter to disable using fifo event handling 8675 even if available. Normally fifo event handling is 8676 preferred over the 2-level event handling, as it is 8677 fairer and the number of possible event channels is 8678 much higher. Default is on (use fifo events). 8679 8680 xirc2ps_cs= [NET,PCMCIA] 8681 Format: 8682 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 8683 8684 xive= [PPC] 8685 By default on POWER9 and above, the kernel will 8686 natively use the XIVE interrupt controller. This option 8687 allows the fallback firmware mode to be used: 8688 8689 off Fallback to firmware control of XIVE interrupt 8690 controller on both pseries and powernv 8691 platforms. Only useful on POWER9 and above. 8692 8693 xive.store-eoi=off [PPC] 8694 By default on POWER10 and above, the kernel will use 8695 stores for EOI handling when the XIVE interrupt mode 8696 is active. This option allows the XIVE driver to use 8697 loads instead, as on POWER9. 8698 8699 xhci-hcd.quirks [USB,KNL] 8700 A hex value specifying bitmask with supplemental xhci 8701 host controller quirks. Meaning of each bit can be 8702 consulted in header drivers/usb/host/xhci.h. 8703 8704 xmon [PPC,EARLY] 8705 Format: { early | on | rw | ro | off } 8706 Controls if xmon debugger is enabled. Default is off. 8707 Passing only "xmon" is equivalent to "xmon=early". 8708 early Call xmon as early as possible on boot; xmon 8709 debugger is called from setup_arch(). 8710 on xmon debugger hooks will be installed so xmon 8711 is only called on a kernel crash. Default mode, 8712 i.e. either "ro" or "rw" mode, is controlled 8713 with CONFIG_XMON_DEFAULT_RO_MODE. 8714 rw xmon debugger hooks will be installed so xmon 8715 is called only on a kernel crash, mode is write, 8716 meaning SPR registers, memory and, other data 8717 can be written using xmon commands. 8718 ro same as "rw" option above but SPR registers, 8719 memory, and other data can't be written using 8720 xmon commands. 8721 off xmon is disabled. 8722