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 AX25 Appropriate AX.25 support is enabled. 10 CLK Common clock infrastructure is enabled. 11 CMA Contiguous Memory Area support is enabled. 12 DRM Direct Rendering Management support is enabled. 13 DYNAMIC_DEBUG Build in debug messages and enable them at runtime 14 EARLY Parameter processed too early to be embedded in initrd. 15 EDD BIOS Enhanced Disk Drive Services (EDD) is enabled 16 EFI EFI Partitioning (GPT) is enabled 17 EVM Extended Verification Module 18 FB The frame buffer device is enabled. 19 FTRACE Function tracing enabled. 20 GCOV GCOV profiling is enabled. 21 HIBERNATION HIBERNATION is enabled. 22 HW Appropriate hardware is enabled. 23 HYPER_V HYPERV support is enabled. 24 IMA Integrity measurement architecture is enabled. 25 IP_PNP IP DHCP, BOOTP, or RARP is enabled. 26 IPV6 IPv6 support is enabled. 27 ISAPNP ISA PnP code is enabled. 28 ISDN Appropriate ISDN support 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 MDA MDA console support is enabled. 41 MIPS MIPS architecture is enabled. 42 MOUSE Appropriate mouse support is enabled. 43 MSI Message Signaled Interrupts (PCI). 44 MTD MTD (Memory Technology Device) support is enabled. 45 NET Appropriate network support is enabled. 46 NFS Appropriate NFS support is enabled. 47 NUMA NUMA support is enabled. 48 OF Devicetree is enabled. 49 PARISC The PA-RISC architecture is enabled. 50 PCI PCI bus support is enabled. 51 PCIE PCI Express support is enabled. 52 PCMCIA The PCMCIA subsystem 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_enforce_resources= [ACPI] 194 { strict | lax | no } 195 Check for resource conflicts between native drivers 196 and ACPI OperationRegions (SystemIO and SystemMemory 197 only). IO ports and memory declared in ACPI might be 198 used by the ACPI subsystem in arbitrary AML code and 199 can interfere with legacy drivers. 200 strict (default): access to resources claimed by ACPI 201 is denied; legacy drivers trying to access reserved 202 resources will fail to bind to device using them. 203 lax: access to resources claimed by ACPI is allowed; 204 legacy drivers trying to access reserved resources 205 will bind successfully but a warning message is logged. 206 no: ACPI OperationRegions are not marked as reserved, 207 no further checks are performed. 208 209 acpi_force_table_verification [HW,ACPI,EARLY] 210 Enable table checksum verification during early stage. 211 By default, this is disabled due to x86 early mapping 212 size limitation. 213 214 acpi_irq_balance [HW,ACPI] 215 ACPI will balance active IRQs 216 default in APIC mode 217 218 acpi_irq_nobalance [HW,ACPI] 219 ACPI will not move active IRQs (default) 220 default in PIC mode 221 222 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 223 Format: <irq>,<irq>... 224 225 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 226 use by PCI 227 Format: <irq>,<irq>... 228 229 acpi_mask_gpe= [HW,ACPI] 230 Due to the existence of _Lxx/_Exx, some GPEs triggered 231 by unsupported hardware/firmware features can result in 232 GPE floodings that cannot be automatically disabled by 233 the GPE dispatcher. 234 This facility can be used to prevent such uncontrolled 235 GPE floodings. 236 Format: <byte> or <bitmap-list> 237 238 acpi_no_auto_serialize [HW,ACPI] 239 Disable auto-serialization of AML methods 240 AML control methods that contain the opcodes to create 241 named objects will be marked as "Serialized" by the 242 auto-serialization feature. 243 This feature is enabled by default. 244 This option allows to turn off the feature. 245 246 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 247 kernels. 248 249 acpi_no_static_ssdt [HW,ACPI,EARLY] 250 Disable installation of static SSDTs at early boot time 251 By default, SSDTs contained in the RSDT/XSDT will be 252 installed automatically and they will appear under 253 /sys/firmware/acpi/tables. 254 This option turns off this feature. 255 Note that specifying this option does not affect 256 dynamic table installation which will install SSDT 257 tables to /sys/firmware/acpi/tables/dynamic. 258 259 acpi_no_watchdog [HW,ACPI,WDT] 260 Ignore the ACPI-based watchdog interface (WDAT) and let 261 a native driver control the watchdog device instead. 262 263 acpi_rsdp= [ACPI,EFI,KEXEC,EARLY] 264 Pass the RSDP address to the kernel, mostly used 265 on machines running EFI runtime service to boot the 266 second kernel for kdump. 267 268 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 269 Format: To spoof as Windows 98: ="Microsoft Windows" 270 271 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 272 of 2 which is mandated by ACPI 6) as the supported ACPI 273 specification revision (when using this switch, it may 274 be necessary to carry out a cold reboot _twice_ in a 275 row to make it take effect on the platform firmware). 276 277 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 278 acpi_osi="string1" # add string1 279 acpi_osi="!string2" # remove string2 280 acpi_osi=!* # remove all strings 281 acpi_osi=! # disable all built-in OS vendor 282 strings 283 acpi_osi=!! # enable all built-in OS vendor 284 strings 285 acpi_osi= # disable all strings 286 287 'acpi_osi=!' can be used in combination with single or 288 multiple 'acpi_osi="string1"' to support specific OS 289 vendor string(s). Note that such command can only 290 affect the default state of the OS vendor strings, thus 291 it cannot affect the default state of the feature group 292 strings and the current state of the OS vendor strings, 293 specifying it multiple times through kernel command line 294 is meaningless. This command is useful when one do not 295 care about the state of the feature group strings which 296 should be controlled by the OSPM. 297 Examples: 298 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 299 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 300 can make '_OSI("Windows 2000")' TRUE. 301 302 'acpi_osi=' cannot be used in combination with other 303 'acpi_osi=' command lines, the _OSI method will not 304 exist in the ACPI namespace. NOTE that such command can 305 only affect the _OSI support state, thus specifying it 306 multiple times through kernel command line is also 307 meaningless. 308 Examples: 309 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 310 FALSE. 311 312 'acpi_osi=!*' can be used in combination with single or 313 multiple 'acpi_osi="string1"' to support specific 314 string(s). Note that such command can affect the 315 current state of both the OS vendor strings and the 316 feature group strings, thus specifying it multiple times 317 through kernel command line is meaningful. But it may 318 still not able to affect the final state of a string if 319 there are quirks related to this string. This command 320 is useful when one want to control the state of the 321 feature group strings to debug BIOS issues related to 322 the OSPM features. 323 Examples: 324 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 325 '_OSI("Module Device")' FALSE. 326 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 327 '_OSI("Module Device")' TRUE. 328 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 329 equivalent to 330 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 331 and 332 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 333 they all will make '_OSI("Windows 2000")' TRUE. 334 335 acpi_pm_good [X86] 336 Override the pmtimer bug detection: force the kernel 337 to assume that this machine's pmtimer latches its value 338 and always returns good values. 339 340 acpi_sci= [HW,ACPI,EARLY] ACPI System Control Interrupt trigger mode 341 Format: { level | edge | high | low } 342 343 acpi_skip_timer_override [HW,ACPI,EARLY] 344 Recognize and ignore IRQ0/pin2 Interrupt Override. 345 For broken nForce2 BIOS resulting in XT-PIC timer. 346 347 acpi_sleep= [HW,ACPI] Sleep options 348 Format: { s3_bios, s3_mode, s3_beep, s4_hwsig, 349 s4_nohwsig, old_ordering, nonvs, 350 sci_force_enable, nobl } 351 See Documentation/power/video.rst for information on 352 s3_bios and s3_mode. 353 s3_beep is for debugging; it makes the PC's speaker beep 354 as soon as the kernel's real-mode entry point is called. 355 s4_hwsig causes the kernel to check the ACPI hardware 356 signature during resume from hibernation, and gracefully 357 refuse to resume if it has changed. This complies with 358 the ACPI specification but not with reality, since 359 Windows does not do this and many laptops do change it 360 on docking. So the default behaviour is to allow resume 361 and simply warn when the signature changes, unless the 362 s4_hwsig option is enabled. 363 s4_nohwsig prevents ACPI hardware signature from being 364 used (or even warned about) during resume. 365 old_ordering causes the ACPI 1.0 ordering of the _PTS 366 control method, with respect to putting devices into 367 low power states, to be enforced (the ACPI 2.0 ordering 368 of _PTS is used by default). 369 nonvs prevents the kernel from saving/restoring the 370 ACPI NVS memory during suspend/hibernation and resume. 371 sci_force_enable causes the kernel to set SCI_EN directly 372 on resume from S1/S3 (which is against the ACPI spec, 373 but some broken systems don't work without it). 374 nobl causes the internal blacklist of systems known to 375 behave incorrectly in some ways with respect to system 376 suspend and resume to be ignored (use wisely). 377 378 acpi_use_timer_override [HW,ACPI,EARLY] 379 Use timer override. For some broken Nvidia NF5 boards 380 that require a timer override, but don't have HPET 381 382 add_efi_memmap [EFI,X86,EARLY] Include EFI memory map in 383 kernel's map of available physical RAM. 384 385 agp= [AGP] 386 { off | try_unsupported } 387 off: disable AGP support 388 try_unsupported: try to drive unsupported chipsets 389 (may crash computer or cause data corruption) 390 391 ALSA [HW,ALSA] 392 See Documentation/sound/alsa-configuration.rst 393 394 alignment= [KNL,ARM] 395 Allow the default userspace alignment fault handler 396 behaviour to be specified. Bit 0 enables warnings, 397 bit 1 enables fixups, and bit 2 sends a segfault. 398 399 align_va_addr= [X86-64] 400 Align virtual addresses by clearing slice [14:12] when 401 allocating a VMA at process creation time. This option 402 gives you up to 3% performance improvement on AMD F15h 403 machines (where it is enabled by default) for a 404 CPU-intensive style benchmark, and it can vary highly in 405 a microbenchmark depending on workload and compiler. 406 407 32: only for 32-bit processes 408 64: only for 64-bit processes 409 on: enable for both 32- and 64-bit processes 410 off: disable for both 32- and 64-bit processes 411 412 alloc_snapshot [FTRACE] 413 Allocate the ftrace snapshot buffer on boot up when the 414 main buffer is allocated. This is handy if debugging 415 and you need to use tracing_snapshot() on boot up, and 416 do not want to use tracing_snapshot_alloc() as it needs 417 to be done where GFP_KERNEL allocations are allowed. 418 419 allow_mismatched_32bit_el0 [ARM64,EARLY] 420 Allow execve() of 32-bit applications and setting of the 421 PER_LINUX32 personality on systems where only a strict 422 subset of the CPUs support 32-bit EL0. When this 423 parameter is present, the set of CPUs supporting 32-bit 424 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0 425 and hot-unplug operations may be restricted. 426 427 See Documentation/arch/arm64/asymmetric-32bit.rst for more 428 information. 429 430 amd_iommu= [HW,X86-64] 431 Pass parameters to the AMD IOMMU driver in the system. 432 Possible values are: 433 fullflush - Deprecated, equivalent to iommu.strict=1 434 off - do not initialize any AMD IOMMU found in 435 the system 436 force_isolation - Force device isolation for all 437 devices. The IOMMU driver is not 438 allowed anymore to lift isolation 439 requirements as needed. This option 440 does not override iommu=pt 441 force_enable - Force enable the IOMMU on platforms known 442 to be buggy with IOMMU enabled. Use this 443 option with care. 444 pgtbl_v1 - Use v1 page table for DMA-API (Default). 445 pgtbl_v2 - Use v2 page table for DMA-API. 446 irtcachedis - Disable Interrupt Remapping Table (IRT) caching. 447 nohugepages - Limit page-sizes used for v1 page-tables 448 to 4 KiB. 449 v2_pgsizes_only - Limit page-sizes used for v1 page-tables 450 to 4KiB/2Mib/1GiB. 451 452 453 amd_iommu_dump= [HW,X86-64] 454 Enable AMD IOMMU driver option to dump the ACPI table 455 for AMD IOMMU. With this option enabled, AMD IOMMU 456 driver will print ACPI tables for AMD IOMMU during 457 IOMMU initialization. 458 459 amd_iommu_intr= [HW,X86-64] 460 Specifies one of the following AMD IOMMU interrupt 461 remapping modes: 462 legacy - Use legacy interrupt remapping mode. 463 vapic - Use virtual APIC mode, which allows IOMMU 464 to inject interrupts directly into guest. 465 This mode requires kvm-amd.avic=1. 466 (Default when IOMMU HW support is present.) 467 468 amd_pstate= [X86,EARLY] 469 disable 470 Do not enable amd_pstate as the default 471 scaling driver for the supported processors 472 passive 473 Use amd_pstate with passive mode as a scaling driver. 474 In this mode autonomous selection is disabled. 475 Driver requests a desired performance level and platform 476 tries to match the same performance level if it is 477 satisfied by guaranteed performance level. 478 active 479 Use amd_pstate_epp driver instance as the scaling driver, 480 driver provides a hint to the hardware if software wants 481 to bias toward performance (0x0) or energy efficiency (0xff) 482 to the CPPC firmware. then CPPC power algorithm will 483 calculate the runtime workload and adjust the realtime cores 484 frequency. 485 guided 486 Activate guided autonomous mode. Driver requests minimum and 487 maximum performance level and the platform autonomously 488 selects a performance level in this range and appropriate 489 to the current workload. 490 491 amd_prefcore= 492 [X86] 493 disable 494 Disable amd-pstate preferred core. 495 496 amijoy.map= [HW,JOY] Amiga joystick support 497 Map of devices attached to JOY0DAT and JOY1DAT 498 Format: <a>,<b> 499 See also Documentation/input/joydev/joystick.rst 500 501 analog.map= [HW,JOY] Analog joystick and gamepad support 502 Specifies type or capabilities of an analog joystick 503 connected to one of 16 gameports 504 Format: <type1>,<type2>,..<type16> 505 506 apc= [HW,SPARC] 507 Power management functions (SPARCstation-4/5 + deriv.) 508 Format: noidle 509 Disable APC CPU standby support. SPARCstation-Fox does 510 not play well with APC CPU idle - disable it if you have 511 APC and your system crashes randomly. 512 513 apic [APIC,X86-64] Use IO-APIC. Default. 514 515 apic= [APIC,X86,EARLY] Advanced Programmable Interrupt Controller 516 Change the output verbosity while booting 517 Format: { quiet (default) | verbose | debug } 518 Change the amount of debugging information output 519 when initialising the APIC and IO-APIC components. 520 521 apic_extnmi= [APIC,X86,EARLY] External NMI delivery setting 522 Format: { bsp (default) | all | none } 523 bsp: External NMI is delivered only to CPU 0 524 all: External NMIs are broadcast to all CPUs as a 525 backup of CPU 0 526 none: External NMI is masked for all CPUs. This is 527 useful so that a dump capture kernel won't be 528 shot down by NMI 529 530 apicpmtimer Do APIC timer calibration using the pmtimer. Implies 531 apicmaintimer. Useful when your PIT timer is totally 532 broken. 533 534 autoconf= [IPV6] 535 See Documentation/networking/ipv6.rst. 536 537 apm= [APM] Advanced Power Management 538 See header of arch/x86/kernel/apm_32.c. 539 540 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 541 Format: { "0" | "1" } 542 See security/apparmor/Kconfig help text 543 0 -- disable. 544 1 -- enable. 545 Default value is set via kernel config option. 546 547 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards 548 Format: <io>,<irq>,<nodeID> 549 550 arm64.no32bit_el0 [ARM64] Unconditionally disable the execution of 551 32 bit applications. 552 553 arm64.nobti [ARM64] Unconditionally disable Branch Target 554 Identification support 555 556 arm64.nogcs [ARM64] Unconditionally disable Guarded Control Stack 557 support 558 559 arm64.nomops [ARM64] Unconditionally disable Memory Copy and Memory 560 Set instructions support 561 562 arm64.nompam [ARM64] Unconditionally disable Memory Partitioning And 563 Monitoring support 564 565 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension 566 support 567 568 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 569 support 570 571 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix 572 Extension support 573 574 arm64.nosve [ARM64] Unconditionally disable Scalable Vector 575 Extension support 576 577 ataflop= [HW,M68k] 578 579 atarimouse= [HW,MOUSE] Atari Mouse 580 581 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 582 EzKey and similar keyboards 583 584 atkbd.reset= [HW] Reset keyboard during initialization 585 586 atkbd.set= [HW] Select keyboard code set 587 Format: <int> (2 = AT (default), 3 = PS/2) 588 589 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 590 keyboards 591 592 atkbd.softraw= [HW] Choose between synthetic and real raw mode 593 Format: <bool> (0 = real, 1 = synthetic (default)) 594 595 atkbd.softrepeat= [HW] 596 Use software keyboard repeat 597 598 audit= [KNL] Enable the audit sub-system 599 Format: { "0" | "1" | "off" | "on" } 600 0 | off - kernel audit is disabled and can not be 601 enabled until the next reboot 602 unset - kernel audit is initialized but disabled and 603 will be fully enabled by the userspace auditd. 604 1 | on - kernel audit is initialized and partially 605 enabled, storing at most audit_backlog_limit 606 messages in RAM until it is fully enabled by the 607 userspace auditd. 608 Default: unset 609 610 audit_backlog_limit= [KNL] Set the audit queue size limit. 611 Format: <int> (must be >=0) 612 Default: 64 613 614 bau= [X86_UV] Enable the BAU on SGI UV. The default 615 behavior is to disable the BAU (i.e. bau=0). 616 Format: { "0" | "1" } 617 0 - Disable the BAU. 618 1 - Enable the BAU. 619 unset - Disable the BAU. 620 621 baycom_epp= [HW,AX25] 622 Format: <io>,<mode> 623 624 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem 625 Format: <io>,<mode> 626 See header of drivers/net/hamradio/baycom_par.c. 627 628 baycom_ser_fdx= [HW,AX25] 629 BayCom Serial Port AX.25 Modem (Full Duplex Mode) 630 Format: <io>,<irq>,<mode>[,<baud>] 631 See header of drivers/net/hamradio/baycom_ser_fdx.c. 632 633 baycom_ser_hdx= [HW,AX25] 634 BayCom Serial Port AX.25 Modem (Half Duplex Mode) 635 Format: <io>,<irq>,<mode> 636 See header of drivers/net/hamradio/baycom_ser_hdx.c. 637 638 bdev_allow_write_mounted= 639 Format: <bool> 640 Control the ability to open a mounted block device 641 for writing, i.e., allow / disallow writes that bypass 642 the FS. This was implemented as a means to prevent 643 fuzzers from crashing the kernel by overwriting the 644 metadata underneath a mounted FS without its awareness. 645 This also prevents destructive formatting of mounted 646 filesystems by naive storage tooling that don't use 647 O_EXCL. Default is Y and can be changed through the 648 Kconfig option CONFIG_BLK_DEV_WRITE_MOUNTED. 649 650 bert_disable [ACPI] 651 Disable BERT OS support on buggy BIOSes. 652 653 bgrt_disable [ACPI,X86,EARLY] 654 Disable BGRT to avoid flickering OEM logo. 655 656 blkdevparts= Manual partition parsing of block device(s) for 657 embedded devices based on command line input. 658 See Documentation/block/cmdline-partition.rst 659 660 boot_delay= [KNL,EARLY] 661 Milliseconds to delay each printk during boot. 662 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled, 663 and you may also have to specify "lpj=". Boot_delay 664 values larger than 10 seconds (10000) are assumed 665 erroneous and ignored. 666 Format: integer 667 668 bootconfig [KNL,EARLY] 669 Extended command line options can be added to an initrd 670 and this will cause the kernel to look for it. 671 672 See Documentation/admin-guide/bootconfig.rst 673 674 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 675 bttv.radio= Most important insmod options are available as 676 kernel args too. 677 bttv.pll= See Documentation/admin-guide/media/bttv.rst 678 bttv.tuner= 679 680 bulk_remove=off [PPC] This parameter disables the use of the pSeries 681 firmware feature for flushing multiple hpte entries 682 at a time. 683 684 c101= [NET] Moxa C101 synchronous serial card 685 686 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 687 Sometimes CPU hardware bugs make them report the cache 688 size incorrectly. The kernel will attempt work arounds 689 to fix known problems, but for some CPUs it is not 690 possible to determine what the correct size should be. 691 This option provides an override for these situations. 692 693 carrier_timeout= 694 [NET] Specifies amount of time (in seconds) that 695 the kernel should wait for a network carrier. By default 696 it waits 120 seconds. 697 698 ca_keys= [KEYS] This parameter identifies a specific key(s) on 699 the system trusted keyring to be used for certificate 700 trust validation. 701 format: { id:<keyid> | builtin } 702 703 cca= [MIPS,EARLY] Override the kernel pages' cache coherency 704 algorithm. Accepted values range from 0 to 7 705 inclusive. See arch/mips/include/asm/pgtable-bits.h 706 for platform specific values (SB1, Loongson3 and 707 others). 708 709 ccw_timeout_log [S390] 710 See Documentation/arch/s390/common_io.rst for details. 711 712 cfi= [X86-64] Set Control Flow Integrity checking features 713 when CONFIG_FINEIBT is enabled. 714 Format: feature[,feature...] 715 Default: auto 716 717 auto: Use FineIBT if IBT available, otherwise kCFI. 718 Under FineIBT, enable "paranoid" mode when 719 FRED is not available. 720 off: Turn off CFI checking. 721 kcfi: Use kCFI (disable FineIBT). 722 fineibt: Use FineIBT (even if IBT not available). 723 norand: Do not re-randomize CFI hashes. 724 paranoid: Add caller hash checking under FineIBT. 725 bhi: Enable register poisoning to stop speculation 726 across FineIBT. (Disabled by default.) 727 warn: Do not enforce CFI checking: warn only. 728 debug: Report CFI initialization details. 729 730 cgroup_disable= [KNL] Disable a particular controller or optional feature 731 Format: {name of the controller(s) or feature(s) to disable} 732 The effects of cgroup_disable=foo are: 733 - foo isn't auto-mounted if you mount all cgroups in 734 a single hierarchy 735 - foo isn't visible as an individually mountable 736 subsystem 737 - if foo is an optional feature then the feature is 738 disabled and corresponding cgroup files are not 739 created 740 {Currently only "memory" controller deal with this and 741 cut the overhead, others just disable the usage. So 742 only cgroup_disable=memory is actually worthy} 743 Specifying "pressure" disables per-cgroup pressure 744 stall information accounting feature 745 746 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 747 Format: { { controller | "all" | "named" } 748 [,{ controller | "all" | "named" }...] } 749 Like cgroup_disable, but only applies to cgroup v1; 750 the blacklisted controllers remain available in cgroup2. 751 "all" blacklists all controllers and "named" disables 752 named mounts. Specifying both "all" and "named" disables 753 all v1 hierarchies. 754 755 cgroup_v1_proc= [KNL] Show also missing controllers in /proc/cgroups 756 Format: { "true" | "false" } 757 /proc/cgroups lists only v1 controllers by default. 758 This compatibility option enables listing also v2 759 controllers (whose v1 code is not compiled!), so that 760 semi-legacy software can check this file to decide 761 about usage of v2 (sic) controllers. 762 763 cgroup_favordynmods= [KNL] Enable or Disable favordynmods. 764 Format: { "true" | "false" } 765 Defaults to the value of CONFIG_CGROUP_FAVOR_DYNMODS. 766 767 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 768 Format: <string> 769 nosocket -- Disable socket memory accounting. 770 nokmem -- Disable kernel memory accounting. 771 nobpf -- Disable BPF memory accounting. 772 773 check_pages= [MM,EARLY] Enable sanity checking of pages after 774 allocations / before freeing. This adds checks to catch 775 double-frees, use-after-frees, and other sources of 776 page corruption by inspecting page internals (flags, 777 mapcount/refcount, memcg_data, etc.). 778 Format: { "0" | "1" } 779 Default: 0 (1 if CONFIG_DEBUG_VM is set) 780 781 checkreqprot= [SELINUX] Set initial checkreqprot flag value. 782 Format: { "0" | "1" } 783 See security/selinux/Kconfig help text. 784 0 -- check protection applied by kernel (includes 785 any implied execute protection). 786 1 -- check protection requested by application. 787 Default value is set via a kernel config option. 788 Value can be changed at runtime via 789 /sys/fs/selinux/checkreqprot. 790 Setting checkreqprot to 1 is deprecated. 791 792 cio_ignore= [S390] 793 See Documentation/arch/s390/common_io.rst for details. 794 795 clearcpuid=X[,X...] [X86] 796 Disable CPUID feature X for the kernel. See 797 arch/x86/include/asm/cpufeatures.h for the valid bit 798 numbers X. Note the Linux-specific bits are not necessarily 799 stable over kernel options, but the vendor-specific 800 ones should be. 801 X can also be a string as appearing in the flags: line 802 in /proc/cpuinfo which does not have the above 803 instability issue. However, not all features have names 804 in /proc/cpuinfo. 805 Note that using this option will taint your kernel. 806 Also note that user programs calling CPUID directly 807 or using the feature without checking anything 808 will still see it. This just prevents it from 809 being used by the kernel or shown in /proc/cpuinfo. 810 Also note the kernel might malfunction if you disable 811 some critical bits. 812 813 clk_ignore_unused 814 [CLK] 815 Prevents the clock framework from automatically gating 816 clocks that have not been explicitly enabled by a Linux 817 device driver but are enabled in hardware at reset or 818 by the bootloader/firmware. Note that this does not 819 force such clocks to be always-on nor does it reserve 820 those clocks in any way. This parameter is useful for 821 debug and development, but should not be needed on a 822 platform with proper driver support. For more 823 information, see Documentation/driver-api/clk.rst. 824 825 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 826 [Deprecated] 827 Forces specified clocksource (if available) to be used 828 when calculating gettimeofday(). If specified 829 clocksource is not available, it defaults to PIT. 830 Format: { pit | tsc | cyclone | pmtmr } 831 832 clocksource= Override the default clocksource 833 Format: <string> 834 Override the default clocksource and use the clocksource 835 with the name specified. 836 Some clocksource names to choose from, depending on 837 the platform: 838 [all] jiffies (this is the base, fallback clocksource) 839 [ACPI] acpi_pm 840 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 841 pxa_timer,timer3,32k_counter,timer0_1 842 [X86-32] pit,hpet,tsc; 843 scx200_hrt on Geode; cyclone on IBM x440 844 [MIPS] MIPS 845 [PARISC] cr16 846 [S390] tod 847 [SH] SuperH 848 [SPARC64] tick 849 [X86-64] hpet,tsc 850 851 clocksource.arm_arch_timer.evtstrm= 852 [ARM,ARM64,EARLY] 853 Format: <bool> 854 Enable/disable the eventstream feature of the ARM 855 architected timer so that code using WFE-based polling 856 loops can be debugged more effectively on production 857 systems. 858 859 clocksource.verify_n_cpus= [KNL] 860 Limit the number of CPUs checked for clocksources 861 marked with CLOCK_SOURCE_VERIFY_PERCPU that 862 are marked unstable due to excessive skew. 863 A negative value says to check all CPUs, while 864 zero says not to check any. Values larger than 865 nr_cpu_ids are silently truncated to nr_cpu_ids. 866 The actual CPUs are chosen randomly, with 867 no replacement if the same CPU is chosen twice. 868 869 clocksource-wdtest.holdoff= [KNL] 870 Set the time in seconds that the clocksource 871 watchdog test waits before commencing its tests. 872 Defaults to zero when built as a module and to 873 10 seconds when built into the kernel. 874 875 cma=nn[MG]@[start[MG][-end[MG]]] 876 [KNL,CMA,EARLY] 877 Sets the size of kernel global memory area for 878 contiguous memory allocations and optionally the 879 placement constraint by the physical address range of 880 memory allocations. A value of 0 disables CMA 881 altogether. For more information, see 882 kernel/dma/contiguous.c 883 884 cma_pernuma=nn[MG] 885 [KNL,CMA,EARLY] 886 Sets the size of kernel per-numa memory area for 887 contiguous memory allocations. A value of 0 disables 888 per-numa CMA altogether. And If this option is not 889 specified, the default value is 0. 890 With per-numa CMA enabled, DMA users on node nid will 891 first try to allocate buffer from the pernuma area 892 which is located in node nid, if the allocation fails, 893 they will fallback to the global default memory area. 894 895 numa_cma=<node>:nn[MG][,<node>:nn[MG]] 896 [KNL,CMA,EARLY] 897 Sets the size of kernel numa memory area for 898 contiguous memory allocations. It will reserve CMA 899 area for the specified node. 900 901 With numa CMA enabled, DMA users on node nid will 902 first try to allocate buffer from the numa area 903 which is located in node nid, if the allocation fails, 904 they will fallback to the global default memory area. 905 906 cmo_free_hint= [PPC] Format: { yes | no } 907 Specify whether pages are marked as being inactive 908 when they are freed. This is used in CMO environments 909 to determine OS memory pressure for page stealing by 910 a hypervisor. 911 Default: yes 912 913 coherent_pool=nn[KMG] [ARM,KNL,EARLY] 914 Sets the size of memory pool for coherent, atomic dma 915 allocations, by default set to 256K. 916 917 com20020= [HW,NET] ARCnet - COM20020 chipset 918 Format: 919 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 920 921 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 922 Format: <io>[,<irq>] 923 924 com90xx= [HW,NET] 925 ARCnet - COM90xx chipset (memory-mapped buffers) 926 Format: <io>[,<irq>[,<memstart>]] 927 928 condev= [HW,S390] console device 929 conmode= 930 931 con3215_drop= [S390,EARLY] 3215 console drop mode. 932 Format: y|n|Y|N|1|0 933 When set to true, drop data on the 3215 console when 934 the console buffer is full. In this case the 935 operator using a 3270 terminal emulator (for example 936 x3270) does not have to enter the clear key for the 937 console output to advance and the kernel to continue. 938 This leads to a much faster boot time when a 3270 939 terminal emulator is active. If no 3270 terminal 940 emulator is used, this parameter has no effect. 941 942 console= [KNL] Output console device and options. 943 944 tty<n> Use the virtual console device <n>. 945 946 ttyS<n>[,options] 947 ttyUSB0[,options] 948 Use the specified serial port. The options are of 949 the form "bbbbpnf", where "bbbb" is the baud rate, 950 "p" is parity ("n", "o", or "e"), "n" is number of 951 bits, and "f" is flow control ("r" for RTS or 952 omit it). Default is "9600n8". 953 954 See Documentation/admin-guide/serial-console.rst for more 955 information. See 956 Documentation/networking/netconsole.rst for an 957 alternative. 958 959 <DEVNAME>:<n>.<n>[,options] 960 Use the specified serial port on the serial core bus. 961 The addressing uses DEVNAME of the physical serial port 962 device, followed by the serial core controller instance, 963 and the serial port instance. The options are the same 964 as documented for the ttyS addressing above. 965 966 The mapping of the serial ports to the tty instances 967 can be viewed with: 968 969 $ ls -d /sys/bus/serial-base/devices/*:*.*/tty/* 970 /sys/bus/serial-base/devices/00:04:0.0/tty/ttyS0 971 972 In the above example, the console can be addressed with 973 console=00:04:0.0. Note that a console addressed this 974 way will only get added when the related device driver 975 is ready. The use of an earlycon parameter in addition to 976 the console may be desired for console output early on. 977 978 uart[8250],io,<addr>[,options] 979 uart[8250],mmio,<addr>[,options] 980 uart[8250],mmio16,<addr>[,options] 981 uart[8250],mmio32,<addr>[,options] 982 uart[8250],0x<addr>[,options] 983 Start an early, polled-mode console on the 8250/16550 984 UART at the specified I/O port or MMIO address, 985 switching to the matching ttyS device later. 986 MMIO inter-register address stride is either 8-bit 987 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 988 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 989 to be equivalent to 'mmio'. 'options' are specified in 990 the same format described for ttyS above; if unspecified, 991 the h/w is not re-initialized. 992 993 hvc<n> Use the hypervisor console device <n>. This is for 994 both Xen and PowerPC hypervisors. 995 996 { null | "" } 997 Use to disable console output, i.e., to have kernel 998 console messages discarded. 999 This must be the only console= parameter used on the 1000 kernel command line. 1001 1002 If the device connected to the port is not a TTY but a braille 1003 device, prepend "brl," before the device type, for instance 1004 console=brl,ttyS0 1005 For now, only VisioBraille is supported. 1006 1007 console_msg_format= 1008 [KNL] Change console messages format 1009 default 1010 By default we print messages on consoles in 1011 "[time stamp] text\n" format (time stamp may not be 1012 printed, depending on CONFIG_PRINTK_TIME or 1013 `printk_time' param). 1014 syslog 1015 Switch to syslog format: "<%u>[time stamp] text\n" 1016 IOW, each message will have a facility and loglevel 1017 prefix. The format is similar to one used by syslog() 1018 syscall, or to executing "dmesg -S --raw" or to reading 1019 from /proc/kmsg. 1020 1021 consoleblank= [KNL] The console blank (screen saver) timeout in 1022 seconds. A value of 0 disables the blank timer. 1023 Defaults to 0. 1024 1025 coredump_filter= 1026 [KNL] Change the default value for 1027 /proc/<pid>/coredump_filter. 1028 See also Documentation/filesystems/proc.rst. 1029 1030 coresight_cpu_debug.enable 1031 [ARM,ARM64] 1032 Format: <bool> 1033 Enable/disable the CPU sampling based debugging. 1034 0: default value, disable debugging 1035 1: enable debugging at boot time 1036 1037 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 1038 Format: 1039 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 1040 1041 cpuidle.off=1 [CPU_IDLE] 1042 disable the cpuidle sub-system 1043 1044 cpuidle.governor= 1045 [CPU_IDLE] Name of the cpuidle governor to use. 1046 1047 cpufreq.off=1 [CPU_FREQ] 1048 disable the cpufreq sub-system 1049 1050 cpufreq.default_governor= 1051 [CPU_FREQ] Name of the default cpufreq governor or 1052 policy to use. This governor must be registered in the 1053 kernel before the cpufreq driver probes. 1054 1055 cpu_init_udelay=N 1056 [X86,EARLY] Delay for N microsec between assert and de-assert 1057 of APIC INIT to start processors. This delay occurs 1058 on every CPU online, such as boot, and resume from suspend. 1059 Default: 10000 1060 1061 cpuhp.parallel= 1062 [SMP] Enable/disable parallel bringup of secondary CPUs 1063 Format: <bool> 1064 Default is enabled if CONFIG_HOTPLUG_PARALLEL=y. Otherwise 1065 the parameter has no effect. 1066 1067 crash_kexec_post_notifiers 1068 Only jump to kdump kernel after running the panic 1069 notifiers and dumping kmsg. This option increases 1070 the risks of a kdump failure, since some panic 1071 notifiers can make the crashed kernel more unstable. 1072 In configurations where kdump may not be reliable, 1073 running the panic notifiers could allow collecting 1074 more data on dmesg, like stack traces from other CPUS 1075 or extra data dumped by panic_print. Note that some 1076 configurations enable this option unconditionally, 1077 like Hyper-V, PowerPC (fadump) and AMD SEV-SNP. 1078 1079 crashkernel=size[KMG][@offset[KMG]] 1080 [KNL,EARLY] Using kexec, Linux can switch to a 'crash kernel' 1081 upon panic. This parameter reserves the physical 1082 memory region [offset, offset + size] for that kernel 1083 image. If '@offset' is omitted, then a suitable offset 1084 is selected automatically. 1085 [KNL, X86-64, ARM64, RISCV, LoongArch] Select a region 1086 under 4G first, and fall back to reserve region above 1087 4G when '@offset' hasn't been specified. 1088 See Documentation/admin-guide/kdump/kdump.rst for further details. 1089 1090 crashkernel=range1:size1[,range2:size2,...][@offset] 1091 [KNL] Same as above, but depends on the memory 1092 in the running system. The syntax of range is 1093 start-[end] where start and end are both 1094 a memory unit (amount[KMG]). See also 1095 Documentation/admin-guide/kdump/kdump.rst for an example. 1096 1097 crashkernel=size[KMG],high 1098 [KNL, X86-64, ARM64, RISCV, LoongArch] range could be 1099 above 4G. 1100 Allow kernel to allocate physical memory region from top, 1101 so could be above 4G if system have more than 4G ram 1102 installed. Otherwise memory region will be allocated 1103 below 4G, if available. 1104 It will be ignored if crashkernel=X is specified. 1105 crashkernel=size[KMG],low 1106 [KNL, X86-64, ARM64, RISCV, LoongArch] range under 4G. 1107 When crashkernel=X,high is passed, kernel could allocate 1108 physical memory region above 4G, that cause second kernel 1109 crash on system that require some amount of low memory, 1110 e.g. swiotlb requires at least 64M+32K low memory, also 1111 enough extra low memory is needed to make sure DMA buffers 1112 for 32-bit devices won't run out. Kernel would try to allocate 1113 default size of memory below 4G automatically. The default 1114 size is platform dependent. 1115 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB) 1116 --> arm64: 128MiB 1117 --> riscv: 128MiB 1118 --> loongarch: 128MiB 1119 This one lets the user specify own low range under 4G 1120 for second kernel instead. 1121 0: to disable low allocation. 1122 It will be ignored when crashkernel=X,high is not used 1123 or memory reserved is below 4G. 1124 crashkernel=size[KMG],cma 1125 [KNL, X86, ppc] Reserve additional crash kernel memory from 1126 CMA. This reservation is usable by the first system's 1127 userspace memory and kernel movable allocations (memory 1128 balloon, zswap). Pages allocated from this memory range 1129 will not be included in the vmcore so this should not 1130 be used if dumping of userspace memory is intended and 1131 it has to be expected that some movable kernel pages 1132 may be missing from the dump. 1133 1134 A standard crashkernel reservation, as described above, 1135 is still needed to hold the crash kernel and initrd. 1136 1137 This option increases the risk of a kdump failure: DMA 1138 transfers configured by the first kernel may end up 1139 corrupting the second kernel's memory. 1140 1141 This reservation method is intended for systems that 1142 can't afford to sacrifice enough memory for standard 1143 crashkernel reservation and where less reliable and 1144 possibly incomplete kdump is preferable to no kdump at 1145 all. 1146 1147 cryptomgr.notests 1148 [KNL] Disable crypto self-tests 1149 1150 cs89x0_dma= [HW,NET] 1151 Format: <dma> 1152 1153 cs89x0_media= [HW,NET] 1154 Format: { rj45 | aui | bnc } 1155 1156 csdlock_debug= [KNL] Enable or disable debug add-ons of cross-CPU 1157 function call handling. When switched on, 1158 additional debug data is printed to the console 1159 in case a hanging CPU is detected, and that 1160 CPU is pinged again in order to try to resolve 1161 the hang situation. The default value of this 1162 option depends on the CSD_LOCK_WAIT_DEBUG_DEFAULT 1163 Kconfig option. 1164 1165 dasd= [HW,NET] 1166 See header of drivers/s390/block/dasd_devmap.c. 1167 1168 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 1169 (one device per port) 1170 Format: <port#>,<type> 1171 See also Documentation/input/devices/joystick-parport.rst 1172 1173 debug [KNL,EARLY] Enable kernel debugging (events log level). 1174 1175 debug_boot_weak_hash 1176 [KNL,EARLY] Enable printing [hashed] pointers early in the 1177 boot sequence. If enabled, we use a weak hash instead 1178 of siphash to hash pointers. Use this option if you are 1179 seeing instances of '(___ptrval___)') and need to see a 1180 value (hashed pointer) instead. Cryptographically 1181 insecure, please do not use on production kernels. 1182 1183 debug_locks_verbose= 1184 [KNL] verbose locking self-tests 1185 Format: <int> 1186 Print debugging info while doing the locking API 1187 self-tests. 1188 Bitmask for the various LOCKTYPE_ tests. Defaults to 0 1189 (no extra messages), setting it to -1 (all bits set) 1190 will print _a_lot_ more information - normally only 1191 useful to lockdep developers. 1192 1193 debug_objects [KNL,EARLY] Enable object debugging 1194 1195 debug_guardpage_minorder= 1196 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this 1197 parameter allows control of the order of pages that will 1198 be intentionally kept free (and hence protected) by the 1199 buddy allocator. Bigger value increase the probability 1200 of catching random memory corruption, but reduce the 1201 amount of memory for normal system use. The maximum 1202 possible value is MAX_PAGE_ORDER/2. Setting this 1203 parameter to 1 or 2 should be enough to identify most 1204 random memory corruption problems caused by bugs in 1205 kernel or driver code when a CPU writes to (or reads 1206 from) a random memory location. Note that there exists 1207 a class of memory corruptions problems caused by buggy 1208 H/W or F/W or by drivers badly programming DMA 1209 (basically when memory is written at bus level and the 1210 CPU MMU is bypassed) which are not detectable by 1211 CONFIG_DEBUG_PAGEALLOC, hence this option will not 1212 help tracking down these problems. 1213 1214 debug_pagealloc= 1215 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 1216 enables the feature at boot time. By default, it is 1217 disabled and the system will work mostly the same as a 1218 kernel built without CONFIG_DEBUG_PAGEALLOC. 1219 Note: to get most of debug_pagealloc error reports, it's 1220 useful to also enable the page_owner functionality. 1221 on: enable the feature 1222 1223 debugfs= [KNL,EARLY] This parameter enables what is exposed to 1224 userspace and debugfs internal clients. 1225 Format: { on, off } 1226 on: All functions are enabled. 1227 off: Filesystem is not registered and clients 1228 get a -EPERM as result when trying to register files 1229 or directories within debugfs. 1230 This is equivalent of the runtime functionality if 1231 debugfs was not enabled in the kernel at all. 1232 Default value is set in build-time with a kernel configuration. 1233 1234 debugpat [X86] Enable PAT debugging 1235 1236 default_hugepagesz= 1237 [HW] The size of the default HugeTLB page. This is 1238 the size represented by the legacy /proc/ hugepages 1239 APIs. In addition, this is the default hugetlb size 1240 used for shmget(), mmap() and mounting hugetlbfs 1241 filesystems. If not specified, defaults to the 1242 architecture's default huge page size. Huge page 1243 sizes are architecture dependent. See also 1244 Documentation/admin-guide/mm/hugetlbpage.rst. 1245 Format: size[KMG] 1246 1247 deferred_probe_timeout= 1248 [KNL] Debugging option to set a timeout in seconds for 1249 deferred probe to give up waiting on dependencies to 1250 probe. Only specific dependencies (subsystems or 1251 drivers) that have opted in will be ignored. A timeout 1252 of 0 will timeout at the end of initcalls. If the time 1253 out hasn't expired, it'll be restarted by each 1254 successful driver registration. This option will also 1255 dump out devices still on the deferred probe list after 1256 retrying. 1257 1258 delayacct [KNL] Enable per-task delay accounting 1259 1260 dell_smm_hwmon.ignore_dmi= 1261 [HW] Continue probing hardware even if DMI data 1262 indicates that the driver is running on unsupported 1263 hardware. 1264 1265 dell_smm_hwmon.force= 1266 [HW] Activate driver even if SMM BIOS signature does 1267 not match list of supported models and enable otherwise 1268 blacklisted features. 1269 1270 dell_smm_hwmon.power_status= 1271 [HW] Report power status in /proc/i8k 1272 (disabled by default). 1273 1274 dell_smm_hwmon.restricted= 1275 [HW] Allow controlling fans only if SYS_ADMIN 1276 capability is set. 1277 1278 dell_smm_hwmon.fan_mult= 1279 [HW] Factor to multiply fan speed with. 1280 1281 dell_smm_hwmon.fan_max= 1282 [HW] Maximum configurable fan speed. 1283 1284 dfltcc= [HW,S390] 1285 Format: { on | off | def_only | inf_only | always } 1286 on: s390 zlib hardware support for compression on 1287 level 1 and decompression (default) 1288 off: No s390 zlib hardware support 1289 def_only: s390 zlib hardware support for deflate 1290 only (compression on level 1) 1291 inf_only: s390 zlib hardware support for inflate 1292 only (decompression) 1293 always: Same as 'on' but ignores the selected compression 1294 level always using hardware support (used for debugging) 1295 1296 dhash_entries= [KNL] 1297 Set number of hash buckets for dentry cache. 1298 1299 disable_1tb_segments [PPC,EARLY] 1300 Disables the use of 1TB hash page table segments. This 1301 causes the kernel to fall back to 256MB segments which 1302 can be useful when debugging issues that require an SLB 1303 miss to occur. 1304 1305 disable= [IPV6] 1306 See Documentation/networking/ipv6.rst. 1307 1308 disable_radix [PPC,EARLY] 1309 Disable RADIX MMU mode on POWER9 1310 1311 disable_tlbie [PPC] 1312 Disable TLBIE instruction. Currently does not work 1313 with KVM, with HASH MMU, or with coherent accelerators. 1314 1315 disable_ddw [PPC/PSERIES,EARLY] 1316 Disable Dynamic DMA Window support. Use this 1317 to workaround buggy firmware. 1318 1319 disable_ipv6= [IPV6] 1320 See Documentation/networking/ipv6.rst. 1321 1322 disable_mtrr_cleanup [X86,EARLY] 1323 The kernel tries to adjust MTRR layout from continuous 1324 to discrete, to make X server driver able to add WB 1325 entry later. This parameter disables that. 1326 1327 disable_mtrr_trim [X86, Intel and AMD only,EARLY] 1328 By default the kernel will trim any uncacheable 1329 memory out of your available memory pool based on 1330 MTRR settings. This parameter disables that behavior, 1331 possibly causing your machine to run very slowly. 1332 1333 disable_timer_pin_1 [X86,EARLY] 1334 Disable PIN 1 of APIC timer 1335 Can be useful to work around chipset bugs. 1336 1337 dis_ucode_ldr [X86] Disable the microcode loader. 1338 1339 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 1340 this option disables the debugging code at boot. 1341 1342 dma_debug_entries=<number> 1343 This option allows to tune the number of preallocated 1344 entries for DMA-API debugging code. One entry is 1345 required per DMA-API allocation. Use this if the 1346 DMA-API debugging code disables itself because the 1347 architectural default is too low. 1348 1349 dma_debug_driver=<driver_name> 1350 With this option the DMA-API debugging driver 1351 filter feature can be enabled at boot time. Just 1352 pass the driver to filter for as the parameter. 1353 The filter can be disabled or changed to another 1354 driver later using sysfs. 1355 1356 reg_file_data_sampling= 1357 [X86] Controls mitigation for Register File Data 1358 Sampling (RFDS) vulnerability. RFDS is a CPU 1359 vulnerability which may allow userspace to infer 1360 kernel data values previously stored in floating point 1361 registers, vector registers, or integer registers. 1362 RFDS only affects Intel Atom processors. 1363 1364 on: Turns ON the mitigation. 1365 off: Turns OFF the mitigation. 1366 1367 This parameter overrides the compile time default set 1368 by CONFIG_MITIGATION_RFDS. Mitigation cannot be 1369 disabled when other VERW based mitigations (like MDS) 1370 are enabled. In order to disable RFDS mitigation all 1371 VERW based mitigations need to be disabled. 1372 1373 For details see: 1374 Documentation/admin-guide/hw-vuln/reg-file-data-sampling.rst 1375 1376 dm_verity.keyring_unsealed= 1377 [KNL] When set to 1, leave the dm-verity keyring 1378 unsealed after initialization so userspace can 1379 provision keys. Once the keyring is restricted 1380 it becomes active and is searched during signature 1381 verification. 1382 1383 driver_async_probe= [KNL] 1384 List of driver names to be probed asynchronously. * 1385 matches with all driver names. If * is specified, the 1386 rest of the listed driver names are those that will NOT 1387 match the *. 1388 Format: <driver_name1>,<driver_name2>... 1389 1390 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1391 Broken monitors, graphic adapters, KVMs and EDIDless 1392 panels may send no or incorrect EDID data sets. 1393 This parameter allows to specify an EDID data sets 1394 in the /lib/firmware directory that are used instead. 1395 An EDID data set will only be used for a particular 1396 connector, if its name and a colon are prepended to 1397 the EDID name. Each connector may use a unique EDID 1398 data set by separating the files with a comma. An EDID 1399 data set with no connector name will be used for 1400 any connectors not explicitly specified. 1401 1402 dscc4.setup= [NET] 1403 1404 dt_cpu_ftrs= [PPC,EARLY] 1405 Format: {"off" | "known"} 1406 Control how the dt_cpu_ftrs device-tree binding is 1407 used for CPU feature discovery and setup (if it 1408 exists). 1409 off: Do not use it, fall back to legacy cpu table. 1410 known: Do not pass through unknown features to guests 1411 or userspace, only those that the kernel is aware of. 1412 1413 dump_apple_properties [X86] 1414 Dump name and content of EFI device properties on 1415 x86 Macs. Useful for driver authors to determine 1416 what data is available or for reverse-engineering. 1417 1418 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1419 <module>.dyndbg[="val"] 1420 Enable debug messages at boot time. See 1421 Documentation/admin-guide/dynamic-debug-howto.rst 1422 for details. 1423 1424 early_ioremap_debug [KNL,EARLY] 1425 Enable debug messages in early_ioremap support. This 1426 is useful for tracking down temporary early mappings 1427 which are not unmapped. 1428 1429 earlycon= [KNL,EARLY] Output early console device and options. 1430 1431 When used with no options, the early console is 1432 determined by stdout-path property in device tree's 1433 chosen node or the ACPI SPCR table if supported by 1434 the platform. 1435 1436 cdns,<addr>[,options] 1437 Start an early, polled-mode console on a Cadence 1438 (xuartps) serial port at the specified address. Only 1439 supported option is baud rate. If baud rate is not 1440 specified, the serial port must already be setup and 1441 configured. 1442 1443 uart[8250],io,<addr>[,options[,uartclk]] 1444 uart[8250],mmio,<addr>[,options[,uartclk]] 1445 uart[8250],mmio32,<addr>[,options[,uartclk]] 1446 uart[8250],mmio32be,<addr>[,options[,uartclk]] 1447 uart[8250],0x<addr>[,options] 1448 Start an early, polled-mode console on the 8250/16550 1449 UART at the specified I/O port or MMIO address. 1450 MMIO inter-register address stride is either 8-bit 1451 (mmio) or 32-bit (mmio32 or mmio32be). 1452 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1453 to be equivalent to 'mmio'. 'options' are specified 1454 in the same format described for "console=ttyS<n>"; if 1455 unspecified, the h/w is not initialized. 'uartclk' is 1456 the uart clock frequency; if unspecified, it is set 1457 to 'BASE_BAUD' * 16. 1458 1459 pl011,<addr> 1460 pl011,mmio32,<addr> 1461 Start an early, polled-mode console on a pl011 serial 1462 port at the specified address. The pl011 serial port 1463 must already be setup and configured. Options are not 1464 yet supported. If 'mmio32' is specified, then only 1465 the driver will use only 32-bit accessors to read/write 1466 the device registers. 1467 1468 liteuart,<addr> 1469 Start an early console on a litex serial port at the 1470 specified address. The serial port must already be 1471 setup and configured. Options are not yet supported. 1472 1473 meson,<addr> 1474 Start an early, polled-mode console on a meson serial 1475 port at the specified address. The serial port must 1476 already be setup and configured. Options are not yet 1477 supported. 1478 1479 msm_serial,<addr> 1480 Start an early, polled-mode console on an msm serial 1481 port at the specified address. The serial port 1482 must already be setup and configured. Options are not 1483 yet supported. 1484 1485 msm_serial_dm,<addr> 1486 Start an early, polled-mode console on an msm serial 1487 dm port at the specified address. The serial port 1488 must already be setup and configured. Options are not 1489 yet supported. 1490 1491 owl,<addr> 1492 Start an early, polled-mode console on a serial port 1493 of an Actions Semi SoC, such as S500 or S900, at the 1494 specified address. The serial port must already be 1495 setup and configured. Options are not yet supported. 1496 1497 rda,<addr> 1498 Start an early, polled-mode console on a serial port 1499 of an RDA Micro SoC, such as RDA8810PL, at the 1500 specified address. The serial port must already be 1501 setup and configured. Options are not yet supported. 1502 1503 sbi 1504 Use RISC-V SBI (Supervisor Binary Interface) for early 1505 console. 1506 1507 smh Use ARM semihosting calls for early console. 1508 1509 s3c2410,<addr> 1510 s3c2412,<addr> 1511 s3c2440,<addr> 1512 s3c6400,<addr> 1513 s5pv210,<addr> 1514 exynos4210,<addr> 1515 Use early console provided by serial driver available 1516 on Samsung SoCs, requires selecting proper type and 1517 a correct base address of the selected UART port. The 1518 serial port must already be setup and configured. 1519 Options are not yet supported. 1520 1521 lantiq,<addr> 1522 Start an early, polled-mode console on a lantiq serial 1523 (lqasc) port at the specified address. The serial port 1524 must already be setup and configured. Options are not 1525 yet supported. 1526 1527 lpuart,<addr> 1528 lpuart32,<addr> 1529 Use early console provided by Freescale LP UART driver 1530 found on Freescale Vybrid and QorIQ LS1021A processors. 1531 A valid base address must be provided, and the serial 1532 port must already be setup and configured. 1533 1534 ec_imx21,<addr> 1535 ec_imx6q,<addr> 1536 Start an early, polled-mode, output-only console on the 1537 Freescale i.MX UART at the specified address. The UART 1538 must already be setup and configured. 1539 1540 ar3700_uart,<addr> 1541 Start an early, polled-mode console on the 1542 Armada 3700 serial port at the specified 1543 address. The serial port must already be setup 1544 and configured. Options are not yet supported. 1545 1546 qcom_geni,<addr> 1547 Start an early, polled-mode console on a Qualcomm 1548 Generic Interface (GENI) based serial port at the 1549 specified address. The serial port must already be 1550 setup and configured. Options are not yet supported. 1551 1552 efifb,[options] 1553 Start an early, unaccelerated console on the EFI 1554 memory mapped framebuffer (if available). On cache 1555 coherent non-x86 systems that use system memory for 1556 the framebuffer, pass the 'ram' option so that it is 1557 mapped with the correct attributes. 1558 1559 linflex,<addr> 1560 Use early console provided by Freescale LINFlexD UART 1561 serial driver for NXP S32V234 SoCs. A valid base 1562 address must be provided, and the serial port must 1563 already be setup and configured. 1564 1565 earlyprintk= [X86,SH,ARM,M68k,S390,UM,EARLY] 1566 earlyprintk=vga 1567 earlyprintk=sclp 1568 earlyprintk=xen 1569 earlyprintk=serial[,ttySn[,baudrate]] 1570 earlyprintk=serial[,0x...[,baudrate]] 1571 earlyprintk=ttySn[,baudrate] 1572 earlyprintk=dbgp[debugController#] 1573 earlyprintk=mmio32,membase[,{nocfg|baudrate}] 1574 earlyprintk=pciserial[,force],bus:device.function[,{nocfg|baudrate}] 1575 earlyprintk=xdbc[xhciController#] 1576 earlyprintk=bios 1577 1578 earlyprintk is useful when the kernel crashes before 1579 the normal console is initialized. It is not enabled by 1580 default because it has some cosmetic problems. 1581 1582 Use "nocfg" to skip UART configuration, assume 1583 BIOS/firmware has configured UART correctly. 1584 1585 Append ",keep" to not disable it when the real console 1586 takes over. 1587 1588 Only one of vga, serial, or usb debug port can 1589 be used at a time. 1590 1591 Currently only ttyS0 and ttyS1 may be specified by 1592 name. Other I/O ports may be explicitly specified 1593 on some architectures (x86 and arm at least) by 1594 replacing ttySn with an I/O port address, like this: 1595 earlyprintk=serial,0x1008,115200 1596 You can find the port for a given device in 1597 /proc/tty/driver/serial: 1598 2: uart:ST16650V2 port:00001008 irq:18 ... 1599 1600 Interaction with the standard serial driver is not 1601 very good. 1602 1603 The VGA output is eventually overwritten by 1604 the real console. 1605 1606 The xen option can only be used in Xen domains. 1607 1608 The sclp output can only be used on s390. 1609 1610 The bios output can only be used on SuperH. 1611 1612 The optional "force" to "pciserial" enables use of a 1613 PCI device even when its classcode is not of the 1614 UART class. 1615 1616 edac_report= [HW,EDAC] Control how to report EDAC event 1617 Format: {"on" | "off" | "force"} 1618 on: enable EDAC to report H/W event. May be overridden 1619 by other higher priority error reporting module. 1620 off: disable H/W event reporting through EDAC. 1621 force: enforce the use of EDAC to report H/W event. 1622 default: on. 1623 1624 edd= [EDD] 1625 Format: {"off" | "on" | "skip[mbr]"} 1626 1627 efi= [EFI,EARLY] 1628 Format: { "debug", "disable_early_pci_dma", 1629 "nochunk", "noruntime", "nosoftreserve", 1630 "novamap", "no_disable_early_pci_dma" } 1631 debug: enable misc debug output. 1632 disable_early_pci_dma: disable the busmaster bit on all 1633 PCI bridges while in the EFI boot stub. 1634 nochunk: disable reading files in "chunks" in the EFI 1635 boot stub, as chunking can cause problems with some 1636 firmware implementations. 1637 noruntime : disable EFI runtime services support 1638 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1639 attribute may cause the kernel to reserve the 1640 memory range for a memory mapping driver to 1641 claim. Specify efi=nosoftreserve to disable this 1642 reservation and treat the memory by its base type 1643 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1644 novamap: do not call SetVirtualAddressMap(). 1645 no_disable_early_pci_dma: Leave the busmaster bit set 1646 on all PCI bridges while in the EFI boot stub 1647 1648 efi_no_storage_paranoia [EFI,X86,EARLY] 1649 Using this parameter you can use more than 50% of 1650 your efi variable storage. Use this parameter only if 1651 you are really sure that your UEFI does sane gc and 1652 fulfills the spec otherwise your board may brick. 1653 1654 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1655 that is to be dynamically loaded by Linux. If there are 1656 multiple variables with the same name but with different 1657 vendor GUIDs, all of them will be loaded. See 1658 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1659 1660 1661 eisa_irq_edge= [PARISC,HW] 1662 See header of drivers/parisc/eisa.c. 1663 1664 ekgdboc= [X86,KGDB,EARLY] Allow early kernel console debugging 1665 Format: ekgdboc=kbd 1666 1667 This is designed to be used in conjunction with 1668 the boot argument: earlyprintk=vga 1669 1670 This parameter works in place of the kgdboc parameter 1671 but can only be used if the backing tty is available 1672 very early in the boot process. For early debugging 1673 via a serial port see kgdboc_earlycon instead. 1674 1675 elanfreq= [X86-32] 1676 See comment before function elanfreq_setup() in 1677 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1678 1679 elfcorehdr=[size[KMG]@]offset[KMG] [PPC,SH,X86,S390,EARLY] 1680 Specifies physical address of start of kernel core 1681 image elf header and optionally the size. Generally 1682 kexec loader will pass this option to capture kernel. 1683 See Documentation/admin-guide/kdump/kdump.rst for details. 1684 1685 enable_mtrr_cleanup [X86,EARLY] 1686 The kernel tries to adjust MTRR layout from continuous 1687 to discrete, to make X server driver able to add WB 1688 entry later. This parameter enables that. 1689 1690 enable_timer_pin_1 [X86] 1691 Enable PIN 1 of APIC timer 1692 Can be useful to work around chipset bugs 1693 (in particular on some ATI chipsets). 1694 The kernel tries to set a reasonable default. 1695 1696 enforcing= [SELINUX] Set initial enforcing status. 1697 Format: {"0" | "1"} 1698 See security/selinux/Kconfig help text. 1699 0 -- permissive (log only, no denials). 1700 1 -- enforcing (deny and log). 1701 Default value is 0. 1702 Value can be changed at runtime via 1703 /sys/fs/selinux/enforce. 1704 1705 erst_disable [ACPI] 1706 Disable Error Record Serialization Table (ERST) 1707 support. 1708 1709 ether= [HW,NET] Ethernet cards parameters 1710 This option is obsoleted by the "netdev=" option, which 1711 has equivalent usage. See its documentation for details. 1712 1713 evm= [EVM] 1714 Format: { "fix" } 1715 Permit 'security.evm' to be updated regardless of 1716 current integrity status. 1717 1718 early_page_ext [KNL,EARLY] Enforces page_ext initialization to earlier 1719 stages so cover more early boot allocations. 1720 Please note that as side effect some optimizations 1721 might be disabled to achieve that (e.g. parallelized 1722 memory initialization is disabled) so the boot process 1723 might take longer, especially on systems with a lot of 1724 memory. Available with CONFIG_PAGE_EXTENSION=y. 1725 1726 failslab= 1727 fail_usercopy= 1728 fail_page_alloc= 1729 fail_skb_realloc= 1730 fail_make_request=[KNL] 1731 General fault injection mechanism. 1732 Format: <interval>,<probability>,<space>,<times> 1733 See also Documentation/fault-injection/. 1734 1735 fb_tunnels= [NET] 1736 Format: { initns | none } 1737 See Documentation/admin-guide/sysctl/net.rst for 1738 fb_tunnels_only_for_init_ns 1739 1740 floppy= [HW] 1741 See Documentation/admin-guide/blockdev/floppy.rst. 1742 1743 forcepae [X86-32] 1744 Forcefully enable Physical Address Extension (PAE). 1745 Many Pentium M systems disable PAE but may have a 1746 functionally usable PAE implementation. 1747 Warning: use of this parameter will taint the kernel 1748 and may cause unknown problems. 1749 1750 fred= [X86-64] 1751 Enable/disable Flexible Return and Event Delivery. 1752 Format: { on | off } 1753 on: enable FRED when it's present. 1754 off: disable FRED, the default setting. 1755 1756 ftrace=[tracer] 1757 [FTRACE] will set and start the specified tracer 1758 as early as possible in order to facilitate early 1759 boot debugging. 1760 1761 ftrace_boot_snapshot 1762 [FTRACE] On boot up, a snapshot will be taken of the 1763 ftrace ring buffer that can be read at: 1764 /sys/kernel/tracing/snapshot. 1765 This is useful if you need tracing information from kernel 1766 boot up that is likely to be overridden by user space 1767 start up functionality. 1768 1769 Optionally, the snapshot can also be defined for a tracing 1770 instance that was created by the trace_instance= command 1771 line parameter. 1772 1773 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo 1774 1775 The above will cause the "foo" tracing instance to trigger 1776 a snapshot at the end of boot up. 1777 1778 ftrace_dump_on_oops[=2(orig_cpu) | =<instance>][,<instance> | 1779 ,<instance>=2(orig_cpu)] 1780 [FTRACE] will dump the trace buffers on oops. 1781 If no parameter is passed, ftrace will dump global 1782 buffers of all CPUs, if you pass 2 or orig_cpu, it 1783 will dump only the buffer of the CPU that triggered 1784 the oops, or the specific instance will be dumped if 1785 its name is passed. Multiple instance dump is also 1786 supported, and instances are separated by commas. Each 1787 instance supports only dump on CPU that triggered the 1788 oops by passing 2 or orig_cpu to it. 1789 1790 ftrace_dump_on_oops=foo=orig_cpu 1791 1792 The above will dump only the buffer of "foo" instance 1793 on CPU that triggered the oops. 1794 1795 ftrace_dump_on_oops,foo,bar=orig_cpu 1796 1797 The above will dump global buffer on all CPUs, the 1798 buffer of "foo" instance on all CPUs and the buffer 1799 of "bar" instance on CPU that triggered the oops. 1800 1801 ftrace_filter=[function-list] 1802 [FTRACE] Limit the functions traced by the function 1803 tracer at boot up. function-list is a comma-separated 1804 list of functions. This list can be changed at run 1805 time by the set_ftrace_filter file in the debugfs 1806 tracing directory. 1807 1808 ftrace_notrace=[function-list] 1809 [FTRACE] Do not trace the functions specified in 1810 function-list. This list can be changed at run time 1811 by the set_ftrace_notrace file in the debugfs 1812 tracing directory. 1813 1814 ftrace_graph_filter=[function-list] 1815 [FTRACE] Limit the top level callers functions traced 1816 by the function graph tracer at boot up. 1817 function-list is a comma-separated list of functions 1818 that can be changed at run time by the 1819 set_graph_function file in the debugfs tracing directory. 1820 1821 ftrace_graph_notrace=[function-list] 1822 [FTRACE] Do not trace from the functions specified in 1823 function-list. This list is a comma-separated list of 1824 functions that can be changed at run time by the 1825 set_graph_notrace file in the debugfs tracing directory. 1826 1827 ftrace_graph_max_depth=<uint> 1828 [FTRACE] Used with the function graph tracer. This is 1829 the max depth it will trace into a function. This value 1830 can be changed at run time by the max_graph_depth file 1831 in the tracefs tracing directory. default: 0 (no limit) 1832 1833 fw_devlink= [KNL,EARLY] Create device links between consumer and supplier 1834 devices by scanning the firmware to infer the 1835 consumer/supplier relationships. This feature is 1836 especially useful when drivers are loaded as modules as 1837 it ensures proper ordering of tasks like device probing 1838 (suppliers first, then consumers), supplier boot state 1839 clean up (only after all consumers have probed), 1840 suspend/resume & runtime PM (consumers first, then 1841 suppliers). 1842 Format: { off | permissive | on | rpm } 1843 off -- Don't create device links from firmware info. 1844 permissive -- Create device links from firmware info 1845 but use it only for ordering boot state clean 1846 up (sync_state() calls). 1847 on -- Create device links from firmware info and use it 1848 to enforce probe and suspend/resume ordering. 1849 rpm -- Like "on", but also use to order runtime PM. 1850 1851 fw_devlink.strict=<bool> 1852 [KNL,EARLY] Treat all inferred dependencies as mandatory 1853 dependencies. This only applies for fw_devlink=on|rpm. 1854 Format: <bool> 1855 1856 fw_devlink.sync_state = 1857 [KNL,EARLY] When all devices that could probe have finished 1858 probing, this parameter controls what to do with 1859 devices that haven't yet received their sync_state() 1860 calls. 1861 Format: { strict | timeout } 1862 strict -- Default. Continue waiting on consumers to 1863 probe successfully. 1864 timeout -- Give up waiting on consumers and call 1865 sync_state() on any devices that haven't yet 1866 received their sync_state() calls after 1867 deferred_probe_timeout has expired or by 1868 late_initcall() if !CONFIG_MODULES. 1869 1870 gamecon.map[2|3]= 1871 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1872 support via parallel port (up to 5 devices per port) 1873 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1874 See also Documentation/input/devices/joystick-parport.rst 1875 1876 gamma= [HW,DRM] 1877 1878 gart_fix_e820= [X86-64,EARLY] disable the fix e820 for K8 GART 1879 Format: off | on 1880 default: on 1881 1882 gather_data_sampling= 1883 [X86,INTEL,EARLY] Control the Gather Data Sampling (GDS) 1884 mitigation. 1885 1886 Gather Data Sampling is a hardware vulnerability which 1887 allows unprivileged speculative access to data which was 1888 previously stored in vector registers. 1889 1890 This issue is mitigated by default in updated microcode. 1891 The mitigation may have a performance impact but can be 1892 disabled. On systems without the microcode mitigation 1893 disabling AVX serves as a mitigation. 1894 1895 force: Disable AVX to mitigate systems without 1896 microcode mitigation. No effect if the microcode 1897 mitigation is present. Known to cause crashes in 1898 userspace with buggy AVX enumeration. 1899 1900 off: Disable GDS mitigation. 1901 1902 gbpages [X86] Use GB pages for kernel direct mappings. 1903 1904 gcov_persist= [GCOV] When non-zero (default), profiling data for 1905 kernel modules is saved and remains accessible via 1906 debugfs, even when the module is unloaded/reloaded. 1907 When zero, profiling data is discarded and associated 1908 debugfs files are removed at module unload time. 1909 1910 goldfish [X86] Enable the goldfish android emulator platform. 1911 Don't use this when you are not running on the 1912 android emulator 1913 1914 gpio-mockup.gpio_mockup_ranges 1915 [HW] Sets the ranges of gpiochip of for this device. 1916 Format: <start1>,<end1>,<start2>,<end2>... 1917 gpio-mockup.gpio_mockup_named_lines 1918 [HW] Let the driver know GPIO lines should be named. 1919 1920 gpt [EFI] Forces disk with valid GPT signature but 1921 invalid Protective MBR to be treated as GPT. If the 1922 primary GPT is corrupted, it enables the backup/alternate 1923 GPT to be used instead. 1924 1925 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1926 the "Enable 0" bit of the configuration register. 1927 Format: 0 | 1 1928 Default: 0 1929 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1930 the "Enable 0" bit of the configuration register. 1931 Format: 0 | 1 1932 Default: 0 1933 grcan.select= [HW] Select which physical interface to use. 1934 Format: 0 | 1 1935 Default: 0 1936 grcan.txsize= [HW] Sets the size of the tx buffer. 1937 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1938 Default: 1024 1939 grcan.rxsize= [HW] Sets the size of the rx buffer. 1940 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1941 Default: 1024 1942 1943 hardened_usercopy= 1944 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 1945 hardening is enabled for this boot. Hardened 1946 usercopy checking is used to protect the kernel 1947 from reading or writing beyond known memory 1948 allocation boundaries as a proactive defense 1949 against bounds-checking flaws in the kernel's 1950 copy_to_user()/copy_from_user() interface. 1951 The default is determined by 1952 CONFIG_HARDENED_USERCOPY_DEFAULT_ON. 1953 on Perform hardened usercopy checks. 1954 off Disable hardened usercopy checks. 1955 1956 hardlockup_all_cpu_backtrace= 1957 [KNL] Should the hard-lockup detector generate 1958 backtraces on all cpus. 1959 Format: 0 | 1 1960 1961 hash_pointers= 1962 [KNL,EARLY] 1963 By default, when pointers are printed to the console 1964 or buffers via the %p format string, that pointer is 1965 "hashed", i.e. obscured by hashing the pointer value. 1966 This is a security feature that hides actual kernel 1967 addresses from unprivileged users, but it also makes 1968 debugging the kernel more difficult since unequal 1969 pointers can no longer be compared. The choices are: 1970 Format: { auto | always | never } 1971 Default: auto 1972 1973 auto - Hash pointers unless slab_debug is enabled. 1974 always - Always hash pointers (even if slab_debug is 1975 enabled). 1976 never - Never hash pointers. This option should only 1977 be specified when debugging the kernel. Do 1978 not use on production kernels. The boot 1979 param "no_hash_pointers" is an alias for 1980 this mode. 1981 1982 For controlling hashing dynamically at runtime, 1983 use the "kernel.kptr_restrict" sysctl instead. 1984 1985 hashdist= [KNL,NUMA] Large hashes allocated during boot 1986 are distributed across NUMA nodes. Defaults on 1987 for 64-bit NUMA, off otherwise. 1988 Format: 0 | 1 (for off | on) 1989 1990 hd= [EIDE] (E)IDE hard drive subsystem geometry 1991 Format: <cyl>,<head>,<sect> 1992 1993 hest_disable [ACPI] 1994 Disable Hardware Error Source Table (HEST) support; 1995 corresponding firmware-first mode error processing 1996 logic will be disabled. 1997 1998 hibernate= [HIBERNATION] 1999 noresume Don't check if there's a hibernation image 2000 present during boot. 2001 nocompress Don't compress/decompress hibernation images. 2002 no Disable hibernation and resume. 2003 protect_image Turn on image protection during restoration 2004 (that will set all pages holding image data 2005 during restoration read-only). 2006 2007 hibernate.compressor= [HIBERNATION] Compression algorithm to be 2008 used with hibernation. 2009 Format: { lzo | lz4 } 2010 Default: lzo 2011 2012 lzo: Select LZO compression algorithm to 2013 compress/decompress hibernation image. 2014 2015 lz4: Select LZ4 compression algorithm to 2016 compress/decompress hibernation image. 2017 2018 hibernate.pm_test_delay= 2019 [HIBERNATION] 2020 Sets the number of seconds to remain in a hibernation test 2021 mode before resuming the system (see 2022 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 2023 is set. Default value is 5. 2024 2025 hibernate_compression_threads= 2026 [HIBERNATION] 2027 Set the number of threads used for compressing or decompressing 2028 hibernation images. 2029 2030 Format: <integer> 2031 Default: 3 2032 Minimum: 1 2033 Example: hibernate_compression_threads=4 2034 2035 highmem=nn[KMG] [KNL,BOOT,EARLY] forces the highmem zone to have an exact 2036 size of <nn>. This works even on boxes that have no 2037 highmem otherwise. This also works to reduce highmem 2038 size on bigger boxes. 2039 2040 highres= [KNL] Enable/disable high resolution timer mode. 2041 Valid parameters: "on", "off" 2042 Default: "on" 2043 2044 hlt [BUGS=ARM,SH] 2045 2046 hostname= [KNL,EARLY] Set the hostname (aka UTS nodename). 2047 Format: <string> 2048 This allows setting the system's hostname during early 2049 startup. This sets the name returned by gethostname. 2050 Using this parameter to set the hostname makes it 2051 possible to ensure the hostname is correctly set before 2052 any userspace processes run, avoiding the possibility 2053 that a process may call gethostname before the hostname 2054 has been explicitly set, resulting in the calling 2055 process getting an incorrect result. The string must 2056 not exceed the maximum allowed hostname length (usually 2057 64 characters) and will be truncated otherwise. 2058 2059 hpet= [X86-32,HPET] option to control HPET usage 2060 Format: { enable (default) | disable | force | 2061 verbose } 2062 disable: disable HPET and use PIT instead 2063 force: allow force enabled of undocumented chips (ICH4, 2064 VIA, nVidia) 2065 verbose: show contents of HPET registers during setup 2066 2067 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 2068 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 2069 2070 hugepages= [HW,EARLY] Number of HugeTLB pages to allocate at boot. 2071 If this follows hugepagesz (below), it specifies 2072 the number of pages of hugepagesz to be allocated. 2073 If this is the first HugeTLB parameter on the command 2074 line, it specifies the number of pages to allocate for 2075 the default huge page size. If using node format, the 2076 number of pages to allocate per-node can be specified. 2077 See also Documentation/admin-guide/mm/hugetlbpage.rst. 2078 Format: <integer> or (node format) 2079 <node>:<integer>[,<node>:<integer>] 2080 2081 hugepagesz= 2082 [HW,EARLY] The size of the HugeTLB pages. This is 2083 used in conjunction with hugepages (above) to 2084 allocate huge pages of a specific size at boot. The 2085 pair hugepagesz=X hugepages=Y can be specified once 2086 for each supported huge page size. Huge page sizes 2087 are architecture dependent. See also 2088 Documentation/admin-guide/mm/hugetlbpage.rst. 2089 Format: size[KMG] 2090 2091 hugepage_alloc_threads= 2092 [HW] The number of threads that should be used to 2093 allocate hugepages during boot. This option can be 2094 used to improve system bootup time when allocating 2095 a large amount of huge pages. 2096 The default value is 25% of the available hardware threads. 2097 2098 Note that this parameter only applies to non-gigantic huge pages. 2099 2100 hugetlb_cma= [HW,CMA,EARLY] The size of a CMA area used for allocation 2101 of gigantic hugepages. Or using node format, the size 2102 of a CMA area per node can be specified. 2103 Format: nn[KMGTPE] or (node format) 2104 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]] 2105 2106 Reserve a CMA area of given size and allocate gigantic 2107 hugepages using the CMA allocator. If enabled, the 2108 boot-time allocation of gigantic hugepages is skipped. 2109 2110 hugetlb_cma_only= 2111 [HW,CMA,EARLY] When allocating new HugeTLB pages, only 2112 try to allocate from the CMA areas. 2113 2114 This option does nothing if hugetlb_cma= is not also 2115 specified. 2116 2117 hugetlb_free_vmemmap= 2118 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP 2119 enabled. 2120 Control if HugeTLB Vmemmap Optimization (HVO) is enabled. 2121 Allows heavy hugetlb users to free up some more 2122 memory (7 * PAGE_SIZE for each 2MB hugetlb page). 2123 Format: { on | off (default) } 2124 2125 on: enable HVO 2126 off: disable HVO 2127 2128 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y, 2129 the default is on. 2130 2131 Note that the vmemmap pages may be allocated from the added 2132 memory block itself when memory_hotplug.memmap_on_memory is 2133 enabled, those vmemmap pages cannot be optimized even if this 2134 feature is enabled. Other vmemmap pages not allocated from 2135 the added memory block itself do not be affected. 2136 2137 hung_task_panic= 2138 [KNL] Number of hung tasks to trigger kernel panic. 2139 Format: <int> 2140 2141 When set to a non-zero value, a kernel panic will be triggered if 2142 the number of detected hung tasks reaches this value. 2143 2144 0: don't panic 2145 1: panic immediately on first hung task 2146 N: panic after N hung tasks are detected in a single scan 2147 2148 The default value is controlled by the 2149 CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time option. The value 2150 selected by this boot parameter can be changed later by the 2151 kernel.hung_task_panic sysctl. 2152 2153 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 2154 terminal devices. Valid values: 0..8 2155 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 2156 If specified, z/VM IUCV HVC accepts connections 2157 from listed z/VM user IDs only. 2158 2159 hv_nopvspin [X86,HYPER_V,EARLY] 2160 Disables the paravirt spinlock optimizations 2161 which allow the hypervisor to 'idle' the guest 2162 on lock contention. 2163 2164 hw_protection= [HW] 2165 Format: reboot | shutdown 2166 2167 Hardware protection action taken on critical events like 2168 overtemperature or imminent voltage loss. 2169 2170 i2c_bus= [HW] Override the default board specific I2C bus speed 2171 or register an additional I2C bus that is not 2172 registered from board initialization code. 2173 Format: 2174 <bus_id>,<clkrate> 2175 2176 i2c_touchscreen_props= [HW,ACPI,X86] 2177 Set device-properties for ACPI-enumerated I2C-attached 2178 touchscreen, to e.g. fix coordinates of upside-down 2179 mounted touchscreens. If you need this option please 2180 submit a drivers/platform/x86/touchscreen_dmi.c patch 2181 adding a DMI quirk for this. 2182 2183 Format: 2184 <ACPI_HW_ID>:<prop_name>=<val>[:prop_name=val][:...] 2185 Where <val> is one of: 2186 Omit "=<val>" entirely Set a boolean device-property 2187 Unsigned number Set a u32 device-property 2188 Anything else Set a string device-property 2189 2190 Examples (split over multiple lines): 2191 i2c_touchscreen_props=GDIX1001:touchscreen-inverted-x: 2192 touchscreen-inverted-y 2193 2194 i2c_touchscreen_props=MSSL1680:touchscreen-size-x=1920: 2195 touchscreen-size-y=1080:touchscreen-inverted-y: 2196 firmware-name=gsl1680-vendor-model.fw:silead,home-button 2197 2198 i8042.debug [HW] Toggle i8042 debug mode 2199 i8042.unmask_kbd_data 2200 [HW] Enable printing of interrupt data from the KBD port 2201 (disabled by default, and as a pre-condition 2202 requires that i8042.debug=1 be enabled) 2203 i8042.direct [HW] Put keyboard port into non-translated mode 2204 i8042.dumbkbd [HW] Pretend that controller can only read data from 2205 keyboard and cannot control its state 2206 (Don't attempt to blink the leds) 2207 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 2208 i8042.nokbd [HW] Don't check/create keyboard port 2209 i8042.noloop [HW] Disable the AUX Loopback command while probing 2210 for the AUX port 2211 i8042.nomux [HW] Don't check presence of an active multiplexing 2212 controller 2213 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 2214 controllers 2215 i8042.notimeout [HW] Ignore timeout condition signalled by controller 2216 i8042.reset [HW] Reset the controller during init, cleanup and 2217 suspend-to-ram transitions, only during s2r 2218 transitions, or never reset 2219 Format: { 1 | Y | y | 0 | N | n } 2220 1, Y, y: always reset controller 2221 0, N, n: don't ever reset controller 2222 Default: only on s2r transitions on x86; most other 2223 architectures force reset to be always executed 2224 i8042.unlock [HW] Unlock (ignore) the keylock 2225 i8042.kbdreset [HW] Reset device connected to KBD port 2226 i8042.probe_defer 2227 [HW] Allow deferred probing upon i8042 probe errors 2228 2229 i810= [HW,DRM] 2230 2231 i915.invert_brightness= 2232 [DRM] Invert the sense of the variable that is used to 2233 set the brightness of the panel backlight. Normally a 2234 brightness value of 0 indicates backlight switched off, 2235 and the maximum of the brightness value sets the backlight 2236 to maximum brightness. If this parameter is set to 0 2237 (default) and the machine requires it, or this parameter 2238 is set to 1, a brightness value of 0 sets the backlight 2239 to maximum brightness, and the maximum of the brightness 2240 value switches the backlight off. 2241 -1 -- never invert brightness 2242 0 -- machine default 2243 1 -- force brightness inversion 2244 2245 ia32_emulation= [X86-64] 2246 Format: <bool> 2247 When true, allows loading 32-bit programs and executing 32-bit 2248 syscalls, essentially overriding IA32_EMULATION_DEFAULT_DISABLED at 2249 boot time. When false, unconditionally disables IA32 emulation. 2250 2251 icn= [HW,ISDN] 2252 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 2253 2254 2255 idle= [X86,EARLY] 2256 Format: idle=poll, idle=halt, idle=nomwait 2257 2258 idle=poll: Don't do power saving in the idle loop 2259 using HLT, but poll for rescheduling event. This will 2260 make the CPUs eat a lot more power, but may be useful 2261 to get slightly better performance in multiprocessor 2262 benchmarks. It also makes some profiling using 2263 performance counters more accurate. Please note that 2264 on systems with MONITOR/MWAIT support (like Intel 2265 EM64T CPUs) this option has no performance advantage 2266 over the normal idle loop. It may also interact badly 2267 with hyperthreading. 2268 2269 idle=halt: Halt is forced to be used for CPU idle. 2270 In such case C2/C3 won't be used again. 2271 2272 idle=nomwait: Disable mwait for CPU C-states 2273 2274 idxd.sva= [HW] 2275 Format: <bool> 2276 Allow force disabling of Shared Virtual Memory (SVA) 2277 support for the idxd driver. By default it is set to 2278 true (1). 2279 2280 idxd.tc_override= [HW] 2281 Format: <bool> 2282 Allow override of default traffic class configuration 2283 for the device. By default it is set to false (0). 2284 2285 ieee754= [MIPS] Select IEEE Std 754 conformance mode 2286 Format: { strict | legacy | 2008 | relaxed | emulated } 2287 Default: strict 2288 2289 Choose which programs will be accepted for execution 2290 based on the IEEE 754 NaN encoding(s) supported by 2291 the FPU and the NaN encoding requested with the value 2292 of an ELF file header flag individually set by each 2293 binary. Hardware implementations are permitted to 2294 support either or both of the legacy and the 2008 NaN 2295 encoding mode. 2296 2297 Available settings are as follows: 2298 strict accept binaries that request a NaN encoding 2299 supported by the FPU 2300 legacy only accept legacy-NaN binaries, if supported 2301 by the FPU 2302 2008 only accept 2008-NaN binaries, if supported 2303 by the FPU 2304 relaxed accept any binaries regardless of whether 2305 supported by the FPU 2306 emulated accept any binaries but enable FPU emulator 2307 if binary mode is unsupported by the FPU. 2308 2309 The FPU emulator is always able to support both NaN 2310 encodings, so if no FPU hardware is present or it has 2311 been disabled with 'nofpu', then the settings of 2312 'legacy' and '2008' strap the emulator accordingly, 2313 'relaxed' straps the emulator for both legacy-NaN and 2314 2008-NaN, whereas 'strict' enables legacy-NaN only on 2315 legacy processors and both NaN encodings on MIPS32 or 2316 MIPS64 CPUs. 2317 2318 The setting for ABS.fmt/NEG.fmt instruction execution 2319 mode generally follows that for the NaN encoding, 2320 except where unsupported by hardware. 2321 2322 ignore_loglevel [KNL,EARLY] 2323 Ignore loglevel setting - this will print /all/ 2324 kernel messages to the console. Useful for debugging. 2325 We also add it as printk module parameter, so users 2326 could change it dynamically, usually by 2327 /sys/module/printk/parameters/ignore_loglevel. 2328 2329 ignore_rlimit_data 2330 Ignore RLIMIT_DATA setting for data mappings, 2331 print warning at first misuse. Can be changed via 2332 /sys/module/kernel/parameters/ignore_rlimit_data. 2333 2334 ihash_entries= [KNL] 2335 Set number of hash buckets for inode cache. 2336 2337 ima_appraise= [IMA] appraise integrity measurements 2338 Format: { "off" | "enforce" | "fix" | "log" } 2339 default: "enforce" 2340 2341 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 2342 The builtin appraise policy appraises all files 2343 owned by uid=0. 2344 2345 ima_canonical_fmt [IMA] 2346 Use the canonical format for the binary runtime 2347 measurements, instead of host native format. 2348 2349 ima_hash= [IMA] 2350 Format: { md5 | sha1 | rmd160 | sha256 | sha384 2351 | sha512 | ... } 2352 default: "sha1" 2353 2354 The list of supported hash algorithms is defined 2355 in crypto/hash_info.h. 2356 2357 ima_policy= [IMA] 2358 The builtin policies to load during IMA setup. 2359 Format: "tcb | appraise_tcb | secure_boot | 2360 fail_securely | critical_data" 2361 2362 The "tcb" policy measures all programs exec'd, files 2363 mmap'd for exec, and all files opened with the read 2364 mode bit set by either the effective uid (euid=0) or 2365 uid=0. 2366 2367 The "appraise_tcb" policy appraises the integrity of 2368 all files owned by root. 2369 2370 The "secure_boot" policy appraises the integrity 2371 of files (eg. kexec kernel image, kernel modules, 2372 firmware, policy, etc) based on file signatures. 2373 2374 The "fail_securely" policy forces file signature 2375 verification failure also on privileged mounted 2376 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 2377 flag. 2378 2379 The "critical_data" policy measures kernel integrity 2380 critical data. 2381 2382 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 2383 Load a policy which meets the needs of the Trusted 2384 Computing Base. This means IMA will measure all 2385 programs exec'd, files mmap'd for exec, and all files 2386 opened for read by uid=0. 2387 2388 ima_template= [IMA] 2389 Select one of defined IMA measurements template formats. 2390 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" | 2391 "ima-sigv2" } 2392 Default: "ima-ng" 2393 2394 ima_template_fmt= 2395 [IMA] Define a custom template format. 2396 Format: { "field1|...|fieldN" } 2397 2398 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 2399 Format: <min_file_size> 2400 Set the minimal file size for using asynchronous hash. 2401 If left unspecified, ahash usage is disabled. 2402 2403 ahash performance varies for different data sizes on 2404 different crypto accelerators. This option can be used 2405 to achieve the best performance for a particular HW. 2406 2407 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size 2408 Format: <bufsize> 2409 Set hashing buffer size. Default: 4k. 2410 2411 ahash performance varies for different chunk sizes on 2412 different crypto accelerators. This option can be used 2413 to achieve best performance for particular HW. 2414 2415 ima= [IMA] Enable or disable IMA 2416 Format: { "off" | "on" } 2417 Default: "on" 2418 Note that disabling IMA is limited to kdump kernel. 2419 2420 indirect_target_selection= [X86,Intel] Mitigation control for Indirect 2421 Target Selection(ITS) bug in Intel CPUs. Updated 2422 microcode is also required for a fix in IBPB. 2423 2424 on: Enable mitigation (default). 2425 off: Disable mitigation. 2426 force: Force the ITS bug and deploy default 2427 mitigation. 2428 vmexit: Only deploy mitigation if CPU is affected by 2429 guest/host isolation part of ITS. 2430 stuff: Deploy RSB-fill mitigation when retpoline is 2431 also deployed. Otherwise, deploy the default 2432 mitigation. 2433 2434 For details see: 2435 Documentation/admin-guide/hw-vuln/indirect-target-selection.rst 2436 2437 init= [KNL] 2438 Format: <full_path> 2439 Run specified binary instead of /sbin/init as init 2440 process. 2441 2442 initcall_debug [KNL] Trace initcalls as they are executed. Useful 2443 for working out where the kernel is dying during 2444 startup. 2445 2446 initcall_blacklist= [KNL] Do not execute a comma-separated list of 2447 initcall functions. Useful for debugging built-in 2448 modules and initcalls. 2449 2450 initramfs_async= [KNL] 2451 Format: <bool> 2452 Default: 1 2453 This parameter controls whether the initramfs 2454 image is unpacked asynchronously, concurrently 2455 with devices being probed and 2456 initialized. This should normally just work, 2457 but as a debugging aid, one can get the 2458 historical behaviour of the initramfs 2459 unpacking being completed before device_ and 2460 late_ initcalls. 2461 2462 initrd= [BOOT,EARLY] Specify the location of the initial ramdisk 2463 2464 initrdmem= [KNL,EARLY] Specify a physical address and size from which to 2465 load the initrd. If an initrd is compiled in or 2466 specified in the bootparams, it takes priority over this 2467 setting. 2468 Format: ss[KMG],nn[KMG] 2469 Default is 0, 0 2470 2471 init_on_alloc= [MM,EARLY] Fill newly allocated pages and heap objects with 2472 zeroes. 2473 Format: 0 | 1 2474 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 2475 2476 init_on_free= [MM,EARLY] Fill freed pages and heap objects with zeroes. 2477 Format: 0 | 1 2478 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 2479 2480 init_pkru= [X86] Specify the default memory protection keys rights 2481 register contents for all processes. 0x55555554 by 2482 default (disallow access to all but pkey 0). Can 2483 override in debugfs after boot. 2484 2485 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 2486 Format: <irq> 2487 2488 int_pln_enable [X86] Enable power limit notification interrupt 2489 2490 integrity_audit=[IMA] 2491 Format: { "0" | "1" } 2492 0 -- basic integrity auditing messages. (Default) 2493 1 -- additional integrity auditing messages. 2494 2495 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 2496 on 2497 Enable intel iommu driver. 2498 off 2499 Disable intel iommu driver. 2500 igfx_off [Default Off] 2501 By default, gfx is mapped as normal device. If a gfx 2502 device has a dedicated DMAR unit, the DMAR unit is 2503 bypassed by not enabling DMAR with this option. In 2504 this case, gfx device will use physical address for 2505 DMA. 2506 strict [Default Off] 2507 Deprecated, equivalent to iommu.strict=1. 2508 sp_off [Default Off] 2509 By default, super page will be supported if Intel IOMMU 2510 has the capability. With this option, super page will 2511 not be supported. 2512 sm_on 2513 Enable the Intel IOMMU scalable mode if the hardware 2514 advertises that it has support for the scalable mode 2515 translation. 2516 sm_off 2517 Disallow use of the Intel IOMMU scalable mode. 2518 tboot_noforce [Default Off] 2519 Do not force the Intel IOMMU enabled under tboot. 2520 By default, tboot will force Intel IOMMU on, which 2521 could harm performance of some high-throughput 2522 devices like 40GBit network cards, even if identity 2523 mapping is enabled. 2524 Note that using this option lowers the security 2525 provided by tboot because it makes the system 2526 vulnerable to DMA attacks. 2527 2528 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 2529 0 disables intel_idle and fall back on acpi_idle. 2530 1 to 9 specify maximum depth of C-state. 2531 2532 intel_pstate= [X86,EARLY] 2533 disable 2534 Do not enable intel_pstate as the default 2535 scaling driver for the supported processors 2536 active 2537 Use intel_pstate driver to bypass the scaling 2538 governors layer of cpufreq and provides it own 2539 algorithms for p-state selection. There are two 2540 P-state selection algorithms provided by 2541 intel_pstate in the active mode: powersave and 2542 performance. The way they both operate depends 2543 on whether or not the hardware managed P-states 2544 (HWP) feature has been enabled in the processor 2545 and possibly on the processor model. 2546 passive 2547 Use intel_pstate as a scaling driver, but configure it 2548 to work with generic cpufreq governors (instead of 2549 enabling its internal governor). This mode cannot be 2550 used along with the hardware-managed P-states (HWP) 2551 feature. 2552 force 2553 Enable intel_pstate on systems that prohibit it by default 2554 in favor of acpi-cpufreq. Forcing the intel_pstate driver 2555 instead of acpi-cpufreq may disable platform features, such 2556 as thermal controls and power capping, that rely on ACPI 2557 P-States information being indicated to OSPM and therefore 2558 should be used with caution. This option does not work with 2559 processors that aren't supported by the intel_pstate driver 2560 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 2561 no_hwp 2562 Do not enable hardware P state control (HWP) 2563 if available. 2564 hwp_only 2565 Only load intel_pstate on systems which support 2566 hardware P state control (HWP) if available. 2567 support_acpi_ppc 2568 Enforce ACPI _PPC performance limits. If the Fixed ACPI 2569 Description Table, specifies preferred power management 2570 profile as "Enterprise Server" or "Performance Server", 2571 then this feature is turned on by default. 2572 per_cpu_perf_limits 2573 Allow per-logical-CPU P-State performance control limits using 2574 cpufreq sysfs interface 2575 no_cas 2576 Do not enable capacity-aware scheduling (CAS) on 2577 hybrid systems 2578 2579 intremap= [X86-64,Intel-IOMMU,EARLY] 2580 on enable Interrupt Remapping (default) 2581 off disable Interrupt Remapping 2582 nosid disable Source ID checking 2583 no_x2apic_optout 2584 BIOS x2APIC opt-out request will be ignored 2585 nopost disable Interrupt Posting 2586 posted_msi 2587 enable MSIs delivered as posted interrupts 2588 2589 iomem= Disable strict checking of access to MMIO memory 2590 strict regions from userspace. 2591 relaxed 2592 2593 iommu= [X86,EARLY] 2594 2595 off 2596 Don't initialize and use any kind of IOMMU. 2597 2598 force 2599 Force the use of the hardware IOMMU even when 2600 it is not actually needed (e.g. because < 3 GB 2601 memory). 2602 2603 noforce 2604 Don't force hardware IOMMU usage when it is not 2605 needed. (default). 2606 2607 biomerge 2608 panic 2609 nopanic 2610 merge 2611 nomerge 2612 2613 soft 2614 Use software bounce buffering (SWIOTLB) (default for 2615 Intel machines). This can be used to prevent the usage 2616 of an available hardware IOMMU. 2617 2618 [X86] 2619 pt 2620 [X86] 2621 nopt 2622 [PPC/POWERNV] 2623 nobypass 2624 Disable IOMMU bypass, using IOMMU for PCI devices. 2625 2626 [X86] 2627 AMD Gart HW IOMMU-specific options: 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 fullflush 2636 Flush IOMMU on each allocation (default). 2637 2638 nofullflush 2639 Don't use IOMMU fullflush. 2640 2641 memaper[=<order>] 2642 Allocate an own aperture over RAM with size 2643 32MB<<order. (default: order=1, i.e. 64MB) 2644 2645 merge 2646 Do scatter-gather (SG) merging. Implies "force" 2647 (experimental). 2648 2649 nomerge 2650 Don't do scatter-gather (SG) merging. 2651 2652 noaperture 2653 Ask the IOMMU not to touch the aperture for AGP. 2654 2655 noagp 2656 Don't initialize the AGP driver and use full aperture. 2657 2658 panic 2659 Always panic when IOMMU overflows. 2660 2661 iommu.forcedac= [ARM64,X86,EARLY] Control IOVA allocation for PCI devices. 2662 Format: { "0" | "1" } 2663 0 - Try to allocate a 32-bit DMA address first, before 2664 falling back to the full range if needed. 2665 1 - Allocate directly from the full usable range, 2666 forcing Dual Address Cycle for PCI cards supporting 2667 greater than 32-bit addressing. 2668 2669 iommu.strict= [ARM64,X86,S390,EARLY] Configure TLB invalidation behaviour 2670 Format: { "0" | "1" } 2671 0 - Lazy mode. 2672 Request that DMA unmap operations use deferred 2673 invalidation of hardware TLBs, for increased 2674 throughput at the cost of reduced device isolation. 2675 Will fall back to strict mode if not supported by 2676 the relevant IOMMU driver. 2677 1 - Strict mode. 2678 DMA unmap operations invalidate IOMMU hardware TLBs 2679 synchronously. 2680 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}. 2681 Note: on x86, strict mode specified via one of the 2682 legacy driver-specific options takes precedence. 2683 2684 iommu.passthrough= 2685 [ARM64,X86,EARLY] Configure DMA to bypass the IOMMU by default. 2686 Format: { "0" | "1" } 2687 0 - Use IOMMU translation for DMA. 2688 1 - Bypass the IOMMU for DMA. 2689 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2690 2691 iommu.debug_pagealloc= 2692 [KNL,EARLY] When CONFIG_IOMMU_DEBUG_PAGEALLOC is set, this 2693 parameter enables the feature at boot time. By default, it 2694 is disabled and the system behaves the same way as a kernel 2695 built without CONFIG_IOMMU_DEBUG_PAGEALLOC. 2696 Format: { "0" | "1" } 2697 0 - Sanitizer disabled. 2698 1 - Sanitizer enabled, expect runtime overhead. 2699 2700 io7= [HW] IO7 for Marvel-based Alpha systems 2701 See comment before marvel_specify_io7 in 2702 arch/alpha/kernel/core_marvel.c. 2703 2704 io_delay= [X86,EARLY] I/O delay method 2705 0x80 2706 Standard port 0x80 based delay 2707 0xed 2708 Alternate port 0xed based delay (needed on some systems) 2709 udelay 2710 Simple two microseconds delay 2711 none 2712 No delay 2713 2714 ip= [IP_PNP] 2715 See Documentation/admin-guide/nfs/nfsroot.rst. 2716 2717 ipcmni_extend [KNL,EARLY] Extend the maximum number of unique System V 2718 IPC identifiers from 32,768 to 16,777,216. 2719 2720 ipe.enforce= [IPE] 2721 Format: <bool> 2722 Determine whether IPE starts in permissive (0) or 2723 enforce (1) mode. The default is enforce. 2724 2725 ipe.success_audit= 2726 [IPE] 2727 Format: <bool> 2728 Start IPE with success auditing enabled, emitting 2729 an audit event when a binary is allowed. The default 2730 is 0. 2731 2732 irqaffinity= [SMP] Set the default irq affinity mask 2733 The argument is a cpu list, as described above. 2734 2735 irqchip.gicv2_force_probe= 2736 [ARM,ARM64,EARLY] 2737 Format: <bool> 2738 Force the kernel to look for the second 4kB page 2739 of a GICv2 controller even if the memory range 2740 exposed by the device tree is too small. 2741 2742 irqchip.gicv3_nolpi= 2743 [ARM,ARM64,EARLY] 2744 Force the kernel to ignore the availability of 2745 LPIs (and by consequence ITSs). Intended for system 2746 that use the kernel as a bootloader, and thus want 2747 to let secondary kernels in charge of setting up 2748 LPIs. 2749 2750 irqchip.gicv3_pseudo_nmi= [ARM64,EARLY] 2751 Enables support for pseudo-NMIs in the kernel. This 2752 requires the kernel to be built with 2753 CONFIG_ARM64_PSEUDO_NMI. 2754 2755 irqchip.riscv_imsic_noipi 2756 [RISC-V,EARLY] 2757 Force the kernel to not use IMSIC software injected MSIs 2758 as IPIs. Intended for system where IMSIC is trap-n-emulated, 2759 and thus want to reduce MMIO traps when triggering IPIs 2760 to multiple harts. 2761 2762 irqfixup [HW] 2763 When an interrupt is not handled search all handlers 2764 for it. Intended to get systems with badly broken 2765 firmware running. 2766 2767 irqhandler.duration_warn_us= [KNL] 2768 Warn if an IRQ handler exceeds the specified duration 2769 threshold in microseconds. Useful for identifying 2770 long-running IRQs in the system. 2771 2772 irqpoll [HW] 2773 When an interrupt is not handled search all handlers 2774 for it. Also check all handlers each timer 2775 interrupt. Intended to get systems with badly broken 2776 firmware running. 2777 2778 isapnp= [ISAPNP] 2779 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2780 2781 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2782 [Deprecated - use cpusets instead] 2783 Format: [flag-list,]<cpu-list> 2784 2785 Specify one or more CPUs to isolate from disturbances 2786 specified in the flag list (default: domain): 2787 2788 nohz 2789 Disable the tick when a single task runs as well as 2790 disabling other kernel noises like having RCU callbacks 2791 offloaded. This is equivalent to the nohz_full parameter. 2792 2793 A residual 1Hz tick is offloaded to workqueues, which you 2794 need to affine to housekeeping through the global 2795 workqueue's affinity configured via the 2796 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2797 by using the 'domain' flag described below. 2798 2799 NOTE: by default the global workqueue runs on all CPUs, 2800 so to protect individual CPUs the 'cpumask' file has to 2801 be configured manually after bootup. 2802 2803 domain 2804 Isolate from the general SMP balancing and scheduling 2805 algorithms. Note that performing domain isolation this way 2806 is irreversible: it's not possible to bring back a CPU to 2807 the domains once isolated through isolcpus. It's strongly 2808 advised to use cpusets instead to disable scheduler load 2809 balancing through the "cpuset.sched_load_balance" file. 2810 It offers a much more flexible interface where CPUs can 2811 move in and out of an isolated set anytime. 2812 2813 You can move a process onto or off an "isolated" CPU via 2814 the CPU affinity syscalls or cpuset. 2815 <cpu number> begins at 0 and the maximum value is 2816 "number of CPUs in system - 1". 2817 2818 managed_irq 2819 2820 Isolate from being targeted by managed interrupts 2821 which have an interrupt mask containing isolated 2822 CPUs. The affinity of managed interrupts is 2823 handled by the kernel and cannot be changed via 2824 the /proc/irq/* interfaces. 2825 2826 This isolation is best effort and only effective 2827 if the automatically assigned interrupt mask of a 2828 device queue contains isolated and housekeeping 2829 CPUs. If housekeeping CPUs are online then such 2830 interrupts are directed to the housekeeping CPU 2831 so that IO submitted on the housekeeping CPU 2832 cannot disturb the isolated CPU. 2833 2834 If a queue's affinity mask contains only isolated 2835 CPUs then this parameter has no effect on the 2836 interrupt routing decision, though interrupts are 2837 only delivered when tasks running on those 2838 isolated CPUs submit IO. IO submitted on 2839 housekeeping CPUs has no influence on those 2840 queues. 2841 2842 The format of <cpu-list> is described above. 2843 2844 iucv= [HW,NET] 2845 2846 ivrs_ioapic [HW,X86-64] 2847 Provide an override to the IOAPIC-ID<->DEVICE-ID 2848 mapping provided in the IVRS ACPI table. 2849 By default, PCI segment is 0, and can be omitted. 2850 2851 For example, to map IOAPIC-ID decimal 10 to 2852 PCI segment 0x1 and PCI device 00:14.0, 2853 write the parameter as: 2854 ivrs_ioapic=10@0001:00:14.0 2855 2856 Deprecated formats: 2857 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0 2858 write the parameter as: 2859 ivrs_ioapic[10]=00:14.0 2860 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and 2861 PCI device 00:14.0 write the parameter as: 2862 ivrs_ioapic[10]=0001:00:14.0 2863 2864 ivrs_hpet [HW,X86-64] 2865 Provide an override to the HPET-ID<->DEVICE-ID 2866 mapping provided in the IVRS ACPI table. 2867 By default, PCI segment is 0, and can be omitted. 2868 2869 For example, to map HPET-ID decimal 10 to 2870 PCI segment 0x1 and PCI device 00:14.0, 2871 write the parameter as: 2872 ivrs_hpet=10@0001:00:14.0 2873 2874 Deprecated formats: 2875 * To map HPET-ID decimal 0 to PCI device 00:14.0 2876 write the parameter as: 2877 ivrs_hpet[0]=00:14.0 2878 * To map HPET-ID decimal 10 to PCI segment 0x1 and 2879 PCI device 00:14.0 write the parameter as: 2880 ivrs_ioapic[10]=0001:00:14.0 2881 2882 ivrs_acpihid [HW,X86-64] 2883 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2884 mapping provided in the IVRS ACPI table. 2885 By default, PCI segment is 0, and can be omitted. 2886 2887 For example, to map UART-HID:UID AMD0020:0 to 2888 PCI segment 0x1 and PCI device ID 00:14.5, 2889 write the parameter as: 2890 ivrs_acpihid=AMD0020:0@0001:00:14.5 2891 2892 Deprecated formats: 2893 * To map UART-HID:UID AMD0020:0 to PCI segment is 0, 2894 PCI device ID 00:14.5, write the parameter as: 2895 ivrs_acpihid[00:14.5]=AMD0020:0 2896 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and 2897 PCI device ID 00:14.5, write the parameter as: 2898 ivrs_acpihid[0001:00:14.5]=AMD0020:0 2899 2900 js= [HW,JOY] Analog joystick 2901 See Documentation/input/joydev/joystick.rst. 2902 2903 kasan_multi_shot 2904 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2905 report on every invalid memory access. Without this 2906 parameter KASAN will print report only for the first 2907 invalid access. 2908 2909 keep_bootcon [KNL,EARLY] 2910 Do not unregister boot console at start. This is only 2911 useful for debugging when something happens in the window 2912 between unregistering the boot console and initializing 2913 the real console. 2914 2915 keepinitrd [HW,ARM] See retain_initrd. 2916 2917 kernelcore= [KNL,X86,PPC,EARLY] 2918 Format: nn[KMGTPE] | nn% | "mirror" 2919 This parameter specifies the amount of memory usable by 2920 the kernel for non-movable allocations. The requested 2921 amount is spread evenly throughout all nodes in the 2922 system as ZONE_NORMAL. The remaining memory is used for 2923 movable memory in its own zone, ZONE_MOVABLE. In the 2924 event, a node is too small to have both ZONE_NORMAL and 2925 ZONE_MOVABLE, kernelcore memory will take priority and 2926 other nodes will have a larger ZONE_MOVABLE. 2927 2928 ZONE_MOVABLE is used for the allocation of pages that 2929 may be reclaimed or moved by the page migration 2930 subsystem. Note that allocations like PTEs-from-HighMem 2931 still use the HighMem zone if it exists, and the Normal 2932 zone if it does not. 2933 2934 It is possible to specify the exact amount of memory in 2935 the form of "nn[KMGTPE]", a percentage of total system 2936 memory in the form of "nn%", or "mirror". If "mirror" 2937 option is specified, mirrored (reliable) memory is used 2938 for non-movable allocations and remaining memory is used 2939 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2940 are exclusive, so you cannot specify multiple forms. 2941 2942 kfence.burst= [MM,KFENCE] The number of additional successive 2943 allocations to be attempted through KFENCE for each 2944 sample interval. 2945 Format: <unsigned integer> 2946 Default: 0 2947 2948 kfence.check_on_panic= 2949 [MM,KFENCE] Whether to check all KFENCE-managed objects' 2950 canaries on panic. 2951 Format: <bool> 2952 Default: false 2953 2954 kfence.deferrable= 2955 [MM,KFENCE] Whether to use a deferrable timer to trigger 2956 allocations. This avoids forcing CPU wake-ups if the 2957 system is idle, at the risk of a less predictable 2958 sample interval. 2959 Format: <bool> 2960 Default: CONFIG_KFENCE_DEFERRABLE 2961 2962 kfence.fault= [MM,KFENCE] Controls the behavior when a KFENCE 2963 error is detected. 2964 report - print the error report and continue (default). 2965 oops - print the error report and oops. 2966 panic - print the error report and panic. 2967 2968 kfence.sample_interval= 2969 [MM,KFENCE] KFENCE's sample interval in milliseconds. 2970 Format: <unsigned integer> 2971 0 - Disable KFENCE. 2972 >0 - Enabled KFENCE with given sample interval. 2973 Default: CONFIG_KFENCE_SAMPLE_INTERVAL 2974 2975 kfence.skip_covered_thresh= 2976 [MM,KFENCE] If pool utilization reaches this threshold 2977 (pool usage%), KFENCE limits currently covered 2978 allocations of the same source from further filling 2979 up the pool. 2980 Format: <unsigned integer> 2981 Default: 75 2982 2983 kgdbdbgp= [KGDB,HW,EARLY] kgdb over EHCI usb debug port. 2984 Format: <Controller#>[,poll interval] 2985 The controller # is the number of the ehci usb debug 2986 port as it is probed via PCI. The poll interval is 2987 optional and is the number seconds in between 2988 each poll cycle to the debug port in case you need 2989 the functionality for interrupting the kernel with 2990 gdb or control-c on the dbgp connection. When 2991 not using this parameter you use sysrq-g to break into 2992 the kernel debugger. 2993 2994 kgdboc= [KGDB,HW] kgdb over consoles. 2995 Requires a tty driver that supports console polling, 2996 or a supported polling keyboard driver (non-usb). 2997 Serial only format: <serial_device>[,baud] 2998 keyboard only format: kbd 2999 keyboard and serial format: kbd,<serial_device>[,baud] 3000 Optional Kernel mode setting: 3001 kms, kbd format: kms,kbd 3002 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 3003 3004 kgdboc_earlycon= [KGDB,HW,EARLY] 3005 If the boot console provides the ability to read 3006 characters and can work in polling mode, you can use 3007 this parameter to tell kgdb to use it as a backend 3008 until the normal console is registered. Intended to 3009 be used together with the kgdboc parameter which 3010 specifies the normal console to transition to. 3011 3012 The name of the early console should be specified 3013 as the value of this parameter. Note that the name of 3014 the early console might be different than the tty 3015 name passed to kgdboc. It's OK to leave the value 3016 blank and the first boot console that implements 3017 read() will be picked. 3018 3019 kgdbwait [KGDB,EARLY] Stop kernel execution and enter the 3020 kernel debugger at the earliest opportunity. 3021 3022 kho= [KEXEC,EARLY] 3023 Format: { "0" | "1" | "off" | "on" | "y" | "n" } 3024 Enables or disables Kexec HandOver. 3025 "0" | "off" | "n" - kexec handover is disabled 3026 "1" | "on" | "y" - kexec handover is enabled 3027 3028 kho_scratch= [KEXEC,EARLY] 3029 Format: ll[KMG],mm[KMG],nn[KMG] | nn% 3030 Defines the size of the KHO scratch region. The KHO 3031 scratch regions are physically contiguous memory 3032 ranges that can only be used for non-kernel 3033 allocations. That way, even when memory is heavily 3034 fragmented with handed over memory, the kexeced 3035 kernel will always have enough contiguous ranges to 3036 bootstrap itself. 3037 3038 It is possible to specify the exact amount of 3039 memory in the form of "ll[KMG],mm[KMG],nn[KMG]" 3040 where the first parameter defines the size of a low 3041 memory scratch area, the second parameter defines 3042 the size of a global scratch area and the third 3043 parameter defines the size of additional per-node 3044 scratch areas. The form "nn%" defines scale factor 3045 (in percents) of memory that was used during boot. 3046 3047 kmac= [MIPS] Korina ethernet MAC address. 3048 Configure the RouterBoard 532 series on-chip 3049 Ethernet adapter MAC address. 3050 3051 kmemleak= [KNL,EARLY] Boot-time kmemleak enable/disable 3052 Valid arguments: on, off 3053 Default: on 3054 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 3055 the default is off. 3056 3057 kprobe_event=[probe-list] 3058 [FTRACE] Add kprobe events and enable at boot time. 3059 The probe-list is a semicolon delimited list of probe 3060 definitions. Each definition is same as kprobe_events 3061 interface, but the parameters are comma delimited. 3062 For example, to add a kprobe event on vfs_read with 3063 arg1 and arg2, add to the command line; 3064 3065 kprobe_event=p,vfs_read,$arg1,$arg2 3066 3067 See also Documentation/trace/kprobetrace.rst "Kernel 3068 Boot Parameter" section. 3069 3070 kpti= [ARM64,EARLY] Control page table isolation of 3071 user and kernel address spaces. 3072 Default: enabled on cores which need mitigation. 3073 0: force disabled 3074 1: force enabled 3075 3076 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires 3077 CONFIG_KUNIT to be set to be fully enabled. The 3078 default value can be overridden via 3079 KUNIT_DEFAULT_ENABLED. 3080 Default is 1 (enabled) 3081 3082 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 3083 Default is 0 (don't ignore, but inject #GP) 3084 3085 kvm.eager_page_split= 3086 [KVM,X86] Controls whether or not KVM will try to 3087 proactively split all huge pages during dirty logging. 3088 Eager page splitting reduces interruptions to vCPU 3089 execution by eliminating the write-protection faults 3090 and MMU lock contention that would otherwise be 3091 required to split huge pages lazily. 3092 3093 VM workloads that rarely perform writes or that write 3094 only to a small region of VM memory may benefit from 3095 disabling eager page splitting to allow huge pages to 3096 still be used for reads. 3097 3098 The behavior of eager page splitting depends on whether 3099 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If 3100 disabled, all huge pages in a memslot will be eagerly 3101 split when dirty logging is enabled on that memslot. If 3102 enabled, eager page splitting will be performed during 3103 the KVM_CLEAR_DIRTY ioctl, and only for the pages being 3104 cleared. 3105 3106 Eager page splitting is only supported when kvm.tdp_mmu=Y. 3107 3108 Default is Y (on). 3109 3110 kvm.enable_pmu=[KVM,X86] 3111 If enabled, KVM will virtualize PMU functionality based 3112 on the virtual CPU model defined by userspace. This 3113 can be overridden on a per-VM basis via 3114 KVM_CAP_PMU_CAPABILITY. 3115 3116 If disabled, KVM will not virtualize PMU functionality, 3117 e.g. MSRs, PMCs, PMIs, etc., even if userspace defines 3118 a virtual CPU model that contains PMU assets. 3119 3120 Note, KVM's vPMU support implicitly requires running 3121 with an in-kernel local APIC, e.g. to deliver PMIs to 3122 the guest. Running without an in-kernel local APIC is 3123 not supported, though KVM will allow such a combination 3124 (with severely degraded functionality). 3125 3126 See also enable_mediated_pmu. 3127 3128 Default is Y (on). 3129 3130 kvm.enable_virt_at_load=[KVM,ARM64,LOONGARCH,MIPS,RISCV,X86] 3131 If enabled, KVM will enable virtualization in hardware 3132 when KVM is loaded, and disable virtualization when KVM 3133 is unloaded (if KVM is built as a module). 3134 3135 If disabled, KVM will dynamically enable and disable 3136 virtualization on-demand when creating and destroying 3137 VMs, i.e. on the 0=>1 and 1=>0 transitions of the 3138 number of VMs. 3139 3140 Enabling virtualization at module load avoids potential 3141 latency for creation of the 0=>1 VM, as KVM serializes 3142 virtualization enabling across all online CPUs. The 3143 "cost" of enabling virtualization when KVM is loaded, 3144 is that doing so may interfere with using out-of-tree 3145 hypervisors that want to "own" virtualization hardware. 3146 3147 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 3148 Default is false (don't support). 3149 3150 kvm.nx_huge_pages= 3151 [KVM] Controls the software workaround for the 3152 X86_BUG_ITLB_MULTIHIT bug. 3153 force : Always deploy workaround. 3154 off : Never deploy workaround. 3155 auto : Deploy workaround based on the presence of 3156 X86_BUG_ITLB_MULTIHIT. 3157 3158 Default is 'auto'. 3159 3160 If the software workaround is enabled for the host, 3161 guests do need not to enable it for nested guests. 3162 3163 kvm.nx_huge_pages_recovery_ratio= 3164 [KVM] Controls how many 4KiB pages are periodically zapped 3165 back to huge pages. 0 disables the recovery, otherwise if 3166 the value is N KVM will zap 1/Nth of the 4KiB pages every 3167 period (see below). The default is 60. 3168 3169 kvm.nx_huge_pages_recovery_period_ms= 3170 [KVM] Controls the time period at which KVM zaps 4KiB pages 3171 back to huge pages. If the value is a non-zero N, KVM will 3172 zap a portion (see ratio above) of the pages every N msecs. 3173 If the value is 0 (the default), KVM will pick a period based 3174 on the ratio, such that a page is zapped after 1 hour on average. 3175 3176 kvm-{amd,intel}.enable_mediated_pmu=[KVM,AMD,INTEL] 3177 If enabled, KVM will provide a mediated virtual PMU, 3178 instead of the default perf-based virtual PMU (if 3179 kvm.enable_pmu is true and PMU is enumerated via the 3180 virtual CPU model). 3181 3182 With a perf-based vPMU, KVM operates as a user of perf, 3183 i.e. emulates guest PMU counters using perf events. 3184 KVM-created perf events are managed by perf as regular 3185 (guest-only) events, e.g. are scheduled in/out, contend 3186 for hardware resources, etc. Using a perf-based vPMU 3187 allows guest and host usage of the PMU to co-exist, but 3188 incurs non-trivial overhead and can result in silently 3189 dropped guest events (due to resource contention). 3190 3191 With a mediated vPMU, hardware PMU state is context 3192 switched around the world switch to/from the guest. 3193 KVM mediates which events the guest can utilize, but 3194 gives the guest direct access to all other PMU assets 3195 when possible (KVM may intercept some accesses if the 3196 virtual CPU model provides a subset of hardware PMU 3197 functionality). Using a mediated vPMU significantly 3198 reduces PMU virtualization overhead and eliminates lost 3199 guest events, but is mutually exclusive with using perf 3200 to profile KVM guests and adds latency to most VM-Exits 3201 (to context switch PMU state). 3202 3203 Default is N (off). 3204 3205 kvm-amd.nested= [KVM,AMD] Control nested virtualization feature in 3206 KVM/SVM. Default is 1 (enabled). 3207 3208 kvm-amd.npt= [KVM,AMD] Control KVM's use of Nested Page Tables, 3209 a.k.a. Two-Dimensional Page Tables. Default is 1 3210 (enabled). Disable by KVM if hardware lacks support 3211 for NPT. 3212 3213 kvm-amd.ciphertext_hiding_asids= 3214 [KVM,AMD] Ciphertext hiding prevents disallowed accesses 3215 to SNP private memory from reading ciphertext. Instead, 3216 reads will see constant default values (0xff). 3217 3218 If ciphertext hiding is enabled, the joint SEV-ES and 3219 SEV-SNP ASID space is partitioned into separate SEV-ES 3220 and SEV-SNP ASID ranges, with the SEV-SNP range being 3221 [1..max_snp_asid] and the SEV-ES range being 3222 (max_snp_asid..min_sev_asid), where min_sev_asid is 3223 enumerated by CPUID.0x.8000_001F[EDX]. 3224 3225 A non-zero value enables SEV-SNP ciphertext hiding and 3226 adjusts the ASID ranges for SEV-ES and SEV-SNP guests. 3227 KVM caps the number of SEV-SNP ASIDs at the maximum 3228 possible value, e.g. specifying -1u will assign all 3229 joint SEV-ES and SEV-SNP ASIDs to SEV-SNP. Note, 3230 assigning all joint ASIDs to SEV-SNP, i.e. configuring 3231 max_snp_asid == min_sev_asid-1, will effectively make 3232 SEV-ES unusable. 3233 3234 kvm-arm.mode= 3235 [KVM,ARM,EARLY] Select one of KVM/arm64's modes of 3236 operation. 3237 3238 none: Forcefully disable KVM. 3239 3240 nvhe: Standard nVHE-based mode, without support for 3241 protected guests. 3242 3243 protected: Mode with support for guests whose state is 3244 kept private from the host, using VHE or 3245 nVHE depending on HW support. 3246 3247 nested: VHE-based mode with support for nested 3248 virtualization. Requires at least ARMv8.4 3249 hardware (with FEAT_NV2). 3250 3251 Defaults to VHE/nVHE based on hardware support. Setting 3252 mode to "protected" will disable kexec and hibernation 3253 for the host. To force nVHE on VHE hardware, add 3254 "arm64_sw.hvhe=0 id_aa64mmfr1.vh=0" to the 3255 command-line. 3256 "nested" is experimental and should be used with 3257 extreme caution. 3258 3259 kvm-arm.vgic_v3_group0_trap= 3260 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-0 3261 system registers 3262 3263 kvm-arm.vgic_v3_group1_trap= 3264 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-1 3265 system registers 3266 3267 kvm-arm.vgic_v3_common_trap= 3268 [KVM,ARM,EARLY] Trap guest accesses to GICv3 common 3269 system registers 3270 3271 kvm-arm.vgic_v4_enable= 3272 [KVM,ARM,EARLY] Allow use of GICv4 for direct 3273 injection of LPIs. 3274 3275 kvm-arm.wfe_trap_policy= 3276 [KVM,ARM] Control when to set WFE instruction trap for 3277 KVM VMs. Traps are allowed but not guaranteed by the 3278 CPU architecture. 3279 3280 trap: set WFE instruction trap 3281 3282 notrap: clear WFE instruction trap 3283 3284 kvm-arm.wfi_trap_policy= 3285 [KVM,ARM] Control when to set WFI instruction trap for 3286 KVM VMs. Traps are allowed but not guaranteed by the 3287 CPU architecture. 3288 3289 trap: set WFI instruction trap 3290 3291 notrap: clear WFI instruction trap 3292 3293 kvm_cma_resv_ratio=n [PPC,EARLY] 3294 Reserves given percentage from system memory area for 3295 contiguous memory allocation for KVM hash pagetable 3296 allocation. 3297 By default it reserves 5% of total system memory. 3298 Format: <integer> 3299 Default: 5 3300 3301 kvm-intel.ept= [KVM,Intel] Control KVM's use of Extended Page Tables, 3302 a.k.a. Two-Dimensional Page Tables. Default is 1 3303 (enabled). Disable by KVM if hardware lacks support 3304 for EPT. 3305 3306 kvm-intel.emulate_invalid_guest_state= 3307 [KVM,Intel] Control whether to emulate invalid guest 3308 state. Ignored if kvm-intel.enable_unrestricted_guest=1, 3309 as guest state is never invalid for unrestricted 3310 guests. This param doesn't apply to nested guests (L2), 3311 as KVM never emulates invalid L2 guest state. 3312 Default is 1 (enabled). 3313 3314 kvm-intel.flexpriority= 3315 [KVM,Intel] Control KVM's use of FlexPriority feature 3316 (TPR shadow). Default is 1 (enabled). Disable by KVM if 3317 hardware lacks support for it. 3318 3319 kvm-intel.nested= 3320 [KVM,Intel] Control nested virtualization feature in 3321 KVM/VMX. Default is 1 (enabled). 3322 3323 kvm-intel.unrestricted_guest= 3324 [KVM,Intel] Control KVM's use of unrestricted guest 3325 feature (virtualized real and unpaged mode). Default 3326 is 1 (enabled). Disable by KVM if EPT is disabled or 3327 hardware lacks support for it. 3328 3329 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 3330 CVE-2018-3620. 3331 3332 Valid arguments: never, cond, always 3333 3334 always: L1D cache flush on every VMENTER. 3335 cond: Flush L1D on VMENTER only when the code between 3336 VMEXIT and VMENTER can leak host memory. 3337 never: Disables the mitigation 3338 3339 Default is cond (do L1 cache flush in specific instances) 3340 3341 kvm-intel.vpid= [KVM,Intel] Control KVM's use of Virtual Processor 3342 Identification feature (tagged TLBs). Default is 1 3343 (enabled). Disable by KVM if hardware lacks support 3344 for it. 3345 3346 l1d_flush= [X86,INTEL,EARLY] 3347 Control mitigation for L1D based snooping vulnerability. 3348 3349 Certain CPUs are vulnerable to an exploit against CPU 3350 internal buffers which can forward information to a 3351 disclosure gadget under certain conditions. 3352 3353 In vulnerable processors, the speculatively 3354 forwarded data can be used in a cache side channel 3355 attack, to access data to which the attacker does 3356 not have direct access. 3357 3358 This parameter controls the mitigation. The 3359 options are: 3360 3361 on - enable the interface for the mitigation 3362 3363 l1tf= [X86,EARLY] Control mitigation of the L1TF vulnerability on 3364 affected CPUs 3365 3366 The kernel PTE inversion protection is unconditionally 3367 enabled and cannot be disabled. 3368 3369 full 3370 Provides all available mitigations for the 3371 L1TF vulnerability. Disables SMT and 3372 enables all mitigations in the 3373 hypervisors, i.e. unconditional L1D flush. 3374 3375 SMT control and L1D flush control via the 3376 sysfs interface is still possible after 3377 boot. Hypervisors will issue a warning 3378 when the first VM is started in a 3379 potentially insecure configuration, 3380 i.e. SMT enabled or L1D flush disabled. 3381 3382 full,force 3383 Same as 'full', but disables SMT and L1D 3384 flush runtime control. Implies the 3385 'nosmt=force' command line option. 3386 (i.e. sysfs control of SMT is disabled.) 3387 3388 flush 3389 Leaves SMT enabled and enables the default 3390 hypervisor mitigation, i.e. conditional 3391 L1D flush. 3392 3393 SMT control and L1D flush control via the 3394 sysfs interface is still possible after 3395 boot. Hypervisors will issue a warning 3396 when the first VM is started in a 3397 potentially insecure configuration, 3398 i.e. SMT enabled or L1D flush disabled. 3399 3400 flush,nosmt 3401 3402 Disables SMT and enables the default 3403 hypervisor mitigation. 3404 3405 SMT control and L1D flush control via the 3406 sysfs interface is still possible after 3407 boot. Hypervisors will issue a warning 3408 when the first VM is started in a 3409 potentially insecure configuration, 3410 i.e. SMT enabled or L1D flush disabled. 3411 3412 flush,nowarn 3413 Same as 'flush', but hypervisors will not 3414 warn when a VM is started in a potentially 3415 insecure configuration. 3416 3417 off 3418 Disables hypervisor mitigations and doesn't 3419 emit any warnings. 3420 It also drops the swap size and available 3421 RAM limit restriction on both hypervisor and 3422 bare metal. 3423 3424 Default is 'flush'. 3425 3426 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 3427 3428 l2cr= [PPC] 3429 3430 l3cr= [PPC] 3431 3432 lapic [X86-32,APIC,EARLY] Enable the local APIC even if BIOS 3433 disabled it. 3434 3435 lapic= [X86,APIC] Do not use TSC deadline 3436 value for LAPIC timer one-shot implementation. Default 3437 back to the programmable timer unit in the LAPIC. 3438 Format: notscdeadline 3439 3440 lapic_timer_c2_ok [X86,APIC,EARLY] trust the local apic timer 3441 in C2 power state. 3442 3443 libata.dma= [LIBATA] DMA control 3444 libata.dma=0 Disable all PATA and SATA DMA 3445 libata.dma=1 PATA and SATA Disk DMA only 3446 libata.dma=2 ATAPI (CDROM) DMA only 3447 libata.dma=4 Compact Flash DMA only 3448 Combinations also work, so libata.dma=3 enables DMA 3449 for disks and CDROMs, but not CFs. 3450 3451 libata.ignore_hpa= [LIBATA] Ignore HPA limit 3452 libata.ignore_hpa=0 keep BIOS limits (default) 3453 libata.ignore_hpa=1 ignore limits, using full disk 3454 3455 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 3456 when set. 3457 Format: <int> 3458 3459 libata.force= [LIBATA] Force configurations. The format is a comma- 3460 separated list of "[ID:]VAL" where ID is PORT[.DEVICE]. 3461 PORT and DEVICE are decimal numbers matching port, link 3462 or device. Basically, it matches the ATA ID string 3463 printed on console by libata. If the whole ID part is 3464 omitted, the last PORT and DEVICE values are used. If 3465 ID hasn't been specified yet, the configuration applies 3466 to all ports, links and devices. 3467 3468 If only DEVICE is omitted, the parameter applies to 3469 the port and all links and devices behind it. DEVICE 3470 number of 0 either selects the first device or the 3471 first fan-out link behind PMP device. It does not 3472 select the host link. DEVICE number of 15 selects the 3473 host link and device attached to it. 3474 3475 The VAL specifies the configuration to force. As long 3476 as there is no ambiguity, shortcut notation is allowed. 3477 For example, both 1.5 and 1.5G would work for 1.5Gbps. 3478 The following configurations can be forced. 3479 3480 * Cable type: 40c, 80c, short40c, unk, ign or sata. 3481 Any ID with matching PORT is used. 3482 3483 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 3484 3485 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 3486 udma[/][16,25,33,44,66,100,133] notation is also 3487 allowed. 3488 3489 * nohrst, nosrst, norst: suppress hard, soft and both 3490 resets. 3491 3492 * rstonce: only attempt one reset during hot-unplug 3493 link recovery. 3494 3495 * [no]dbdelay: Enable or disable the extra 200ms delay 3496 before debouncing a link PHY and device presence 3497 detection. 3498 3499 * [no]ncq: Turn on or off NCQ. 3500 3501 * [no]ncqtrim: Enable or disable queued DSM TRIM. 3502 3503 * [no]ncqati: Enable or disable NCQ trim on ATI chipset. 3504 3505 * [no]trim: Enable or disable (unqueued) TRIM. 3506 3507 * trim_zero: Indicate that TRIM command zeroes data. 3508 3509 * max_trim_128m: Set 128M maximum trim size limit. 3510 3511 * [no]dma: Turn on or off DMA transfers. 3512 3513 * atapi_dmadir: Enable ATAPI DMADIR bridge support. 3514 3515 * atapi_mod16_dma: Enable the use of ATAPI DMA for 3516 commands that are not a multiple of 16 bytes. 3517 3518 * [no]dmalog: Enable or disable the use of the 3519 READ LOG DMA EXT command to access logs. 3520 3521 * [no]iddevlog: Enable or disable access to the 3522 identify device data log. 3523 3524 * [no]logdir: Enable or disable access to the general 3525 purpose log directory. 3526 3527 * max_sec=<sectors>: Set the transfer size limit, in 3528 number of 512-byte sectors, to the value specified in 3529 <sectors>. The value specified in <sectors> has to be 3530 a non-zero positive integer. 3531 3532 * max_sec_128: Set transfer size limit to 128 sectors. 3533 3534 * max_sec_1024: Set or clear transfer size limit to 3535 1024 sectors. 3536 3537 * max_sec_lba48: Set or clear transfer size limit to 3538 65535 sectors. 3539 3540 * external: Mark port as external (hotplug-capable). 3541 3542 * [no]lpm: Enable or disable link power management. 3543 3544 * [no]setxfer: Indicate if transfer speed mode setting 3545 should be skipped. 3546 3547 * [no]fua: Disable or enable FUA (Force Unit Access) 3548 support for devices supporting this feature. 3549 3550 * dump_id: Dump IDENTIFY data. 3551 3552 * disable: Disable this device. 3553 3554 If there are multiple matching configurations changing 3555 the same attribute, the last one is used. 3556 3557 liveupdate= [KNL,EARLY] 3558 Format: <bool> 3559 Enable Live Update Orchestrator (LUO). 3560 Default: off. 3561 3562 lockd.nlm_grace_period=P [NFS] Assign grace period. 3563 Format: <integer> 3564 3565 lockd.nlm_tcpport=N [NFS] Assign TCP port. 3566 Format: <integer> 3567 3568 lockd.nlm_timeout=T [NFS] Assign timeout value. 3569 Format: <integer> 3570 3571 lockd.nlm_udpport=M [NFS] Assign UDP port. 3572 Format: <integer> 3573 3574 lockdown= [SECURITY,EARLY] 3575 { integrity | confidentiality } 3576 Enable the kernel lockdown feature. If set to 3577 integrity, kernel features that allow userland to 3578 modify the running kernel are disabled. If set to 3579 confidentiality, kernel features that allow userland 3580 to extract confidential information from the kernel 3581 are also disabled. 3582 3583 locktorture.acq_writer_lim= [KNL] 3584 Set the time limit in jiffies for a lock 3585 acquisition. Acquisitions exceeding this limit 3586 will result in a splat once they do complete. 3587 3588 locktorture.bind_readers= [KNL] 3589 Specify the list of CPUs to which the readers are 3590 to be bound. 3591 3592 locktorture.bind_writers= [KNL] 3593 Specify the list of CPUs to which the writers are 3594 to be bound. 3595 3596 locktorture.call_rcu_chains= [KNL] 3597 Specify the number of self-propagating call_rcu() 3598 chains to set up. These are used to ensure that 3599 there is a high probability of an RCU grace period 3600 in progress at any given time. Defaults to 0, 3601 which disables these call_rcu() chains. 3602 3603 locktorture.long_hold= [KNL] 3604 Specify the duration in milliseconds for the 3605 occasional long-duration lock hold time. Defaults 3606 to 100 milliseconds. Select 0 to disable. 3607 3608 locktorture.nested_locks= [KNL] 3609 Specify the maximum lock nesting depth that 3610 locktorture is to exercise, up to a limit of 8 3611 (MAX_NESTED_LOCKS). Specify zero to disable. 3612 Note that this parameter is ineffective on types 3613 of locks that do not support nested acquisition. 3614 3615 locktorture.nreaders_stress= [KNL] 3616 Set the number of locking read-acquisition kthreads. 3617 Defaults to being automatically set based on the 3618 number of online CPUs. 3619 3620 locktorture.nwriters_stress= [KNL] 3621 Set the number of locking write-acquisition kthreads. 3622 3623 locktorture.onoff_holdoff= [KNL] 3624 Set time (s) after boot for CPU-hotplug testing. 3625 3626 locktorture.onoff_interval= [KNL] 3627 Set time (s) between CPU-hotplug operations, or 3628 zero to disable CPU-hotplug testing. 3629 3630 locktorture.rt_boost= [KNL] 3631 Do periodic testing of real-time lock priority 3632 boosting. Select 0 to disable, 1 to boost 3633 only rt_mutex, and 2 to boost unconditionally. 3634 Defaults to 2, which might seem to be an 3635 odd choice, but which should be harmless for 3636 non-real-time spinlocks, due to their disabling 3637 of preemption. Note that non-realtime mutexes 3638 disable boosting. 3639 3640 locktorture.rt_boost_factor= [KNL] 3641 Number that determines how often and for how 3642 long priority boosting is exercised. This is 3643 scaled down by the number of writers, so that the 3644 number of boosts per unit time remains roughly 3645 constant as the number of writers increases. 3646 On the other hand, the duration of each boost 3647 increases with the number of writers. 3648 3649 locktorture.shuffle_interval= [KNL] 3650 Set task-shuffle interval (jiffies). Shuffling 3651 tasks allows some CPUs to go into dyntick-idle 3652 mode during the locktorture test. 3653 3654 locktorture.shutdown_secs= [KNL] 3655 Set time (s) after boot system shutdown. This 3656 is useful for hands-off automated testing. 3657 3658 locktorture.stat_interval= [KNL] 3659 Time (s) between statistics printk()s. 3660 3661 locktorture.stutter= [KNL] 3662 Time (s) to stutter testing, for example, 3663 specifying five seconds causes the test to run for 3664 five seconds, wait for five seconds, and so on. 3665 This tests the locking primitive's ability to 3666 transition abruptly to and from idle. 3667 3668 locktorture.torture_type= [KNL] 3669 Specify the locking implementation to test. 3670 3671 locktorture.verbose= [KNL] 3672 Enable additional printk() statements. 3673 3674 locktorture.writer_fifo= [KNL] 3675 Run the write-side locktorture kthreads at 3676 sched_set_fifo() real-time priority. 3677 3678 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 3679 Format: <irq> 3680 3681 loglevel= [KNL,EARLY] 3682 All Kernel Messages with a loglevel smaller than the 3683 console loglevel will be printed to the console. It can 3684 also be changed with klogd or other programs. The 3685 loglevels are defined as follows: 3686 3687 0 (KERN_EMERG) system is unusable 3688 1 (KERN_ALERT) action must be taken immediately 3689 2 (KERN_CRIT) critical conditions 3690 3 (KERN_ERR) error conditions 3691 4 (KERN_WARNING) warning conditions 3692 5 (KERN_NOTICE) normal but significant condition 3693 6 (KERN_INFO) informational 3694 7 (KERN_DEBUG) debug-level messages 3695 3696 log_buf_len=n[KMG] [KNL,EARLY] 3697 Sets the size of the printk ring buffer, in bytes. 3698 n must be a power of two and greater than the 3699 minimal size. The minimal size is defined by 3700 LOG_BUF_SHIFT kernel config parameter. There 3701 is also CONFIG_LOG_CPU_MAX_BUF_SHIFT config 3702 parameter that allows to increase the default size 3703 depending on the number of CPUs. See init/Kconfig 3704 for more details. 3705 3706 logo.nologo [FB] Disables display of the built-in Linux logo. 3707 This may be used to provide more screen space for 3708 kernel log messages and is useful when debugging 3709 kernel boot problems. 3710 3711 lp=0 [LP] Specify parallel ports to use, e.g, 3712 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 3713 lp=reset first parallel port). 'lp=0' disables the 3714 lp=auto printer driver. 'lp=reset' (which can be 3715 specified in addition to the ports) causes 3716 attached printers to be reset. Using 3717 lp=port1,port2,... specifies the parallel ports 3718 to associate lp devices with, starting with 3719 lp0. A port specification may be 'none' to skip 3720 that lp device, or a parport name such as 3721 'parport0'. Specifying 'lp=auto' instead of a 3722 port specification list means that device IDs 3723 from each port should be examined, to see if 3724 an IEEE 1284-compliant printer is attached; if 3725 so, the driver will manage that printer. 3726 See also header of drivers/char/lp.c. 3727 3728 lpj=n [KNL] 3729 Sets loops_per_jiffy to given constant, thus avoiding 3730 time-consuming boot-time autodetection (up to 250 ms per 3731 CPU). 0 enables autodetection (default). To determine 3732 the correct value for your kernel, boot with normal 3733 autodetection and see what value is printed. Note that 3734 on SMP systems the preset will be applied to all CPUs, 3735 which is likely to cause problems if your CPUs need 3736 significantly divergent settings. An incorrect value 3737 will cause delays in the kernel to be wrong, leading to 3738 unpredictable I/O errors and other breakage. Although 3739 unlikely, in the extreme case this might damage your 3740 hardware. 3741 3742 lsm.debug [SECURITY] Enable LSM initialization debugging output. 3743 3744 lsm=lsm1,...,lsmN 3745 [SECURITY] Choose order of LSM initialization. This 3746 overrides CONFIG_LSM, and the "security=" parameter. 3747 3748 machtype= [Loongson] Share the same kernel image file between 3749 different yeeloong laptops. 3750 Example: machtype=lemote-yeeloong-2f-7inch 3751 3752 maxcpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 3753 will bring up during bootup. maxcpus=n : n >= 0 limits 3754 the kernel to bring up 'n' processors. Surely after 3755 bootup you can bring up the other plugged cpu by executing 3756 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 3757 only takes effect during system bootup. 3758 While n=0 is a special case, it is equivalent to "nosmp", 3759 which also disables the IO APIC. 3760 3761 max_loop= [LOOP] The number of loop block devices that get 3762 (loop.max_loop) unconditionally pre-created at init time. The default 3763 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 3764 of statically allocating a predefined number, loop 3765 devices can be requested on-demand with the 3766 /dev/loop-control interface. 3767 3768 mce= [X86-{32,64}] 3769 3770 Please see Documentation/arch/x86/x86_64/machinecheck.rst for sysfs runtime tunables. 3771 3772 off 3773 disable machine check 3774 3775 no_cmci 3776 disable CMCI(Corrected Machine Check Interrupt) that 3777 Intel processor supports. Usually this disablement is 3778 not recommended, but it might be handy if your 3779 hardware is misbehaving. 3780 3781 Note that you'll get more problems without CMCI than 3782 with due to the shared banks, i.e. you might get 3783 duplicated error logs. 3784 3785 dont_log_ce 3786 don't make logs for corrected errors. All events 3787 reported as corrected are silently cleared by OS. This 3788 option will be useful if you have no interest in any 3789 of corrected errors. 3790 3791 ignore_ce 3792 disable features for corrected errors, e.g. 3793 polling timer and CMCI. All events reported as 3794 corrected are not cleared by OS and remained in its 3795 error banks. 3796 3797 Usually this disablement is not recommended, however 3798 if there is an agent checking/clearing corrected 3799 errors (e.g. BIOS or hardware monitoring 3800 applications), conflicting with OS's error handling, 3801 and you cannot deactivate the agent, then this option 3802 will be a help. 3803 3804 no_lmce 3805 do not opt-in to Local MCE delivery. Use legacy method 3806 to broadcast MCEs. 3807 3808 bootlog 3809 enable logging of machine checks left over from 3810 booting. Disabled by default on AMD Fam10h and older 3811 because some BIOS leave bogus ones. 3812 3813 If your BIOS doesn't do that it's a good idea to 3814 enable though to make sure you log even machine check 3815 events that result in a reboot. On Intel systems it is 3816 enabled by default. 3817 3818 nobootlog 3819 disable boot machine check logging. 3820 3821 monarchtimeout (number) 3822 sets the time in us to wait for other CPUs on machine 3823 checks. 0 to disable. 3824 3825 bios_cmci_threshold 3826 don't overwrite the bios-set CMCI threshold. This boot 3827 option prevents Linux from overwriting the CMCI 3828 threshold set by the bios. Without this option, Linux 3829 always sets the CMCI threshold to 1. Enabling this may 3830 make memory predictive failure analysis less effective 3831 if the bios sets thresholds for memory errors since we 3832 will not see details for all errors. 3833 3834 recovery 3835 force-enable recoverable machine check code paths 3836 3837 Everything else is in sysfs now. 3838 3839 3840 md= [HW] RAID subsystems devices and level 3841 See Documentation/admin-guide/md.rst. 3842 3843 mdacon= [MDA] 3844 Format: <first>,<last> 3845 Specifies range of consoles to be captured by the MDA. 3846 3847 mds= [X86,INTEL,EARLY] 3848 Control mitigation for the Micro-architectural Data 3849 Sampling (MDS) vulnerability. 3850 3851 Certain CPUs are vulnerable to an exploit against CPU 3852 internal buffers which can forward information to a 3853 disclosure gadget under certain conditions. 3854 3855 In vulnerable processors, the speculatively 3856 forwarded data can be used in a cache side channel 3857 attack, to access data to which the attacker does 3858 not have direct access. 3859 3860 This parameter controls the MDS mitigation. The 3861 options are: 3862 3863 full - Enable MDS mitigation on vulnerable CPUs 3864 full,nosmt - Enable MDS mitigation and disable 3865 SMT on vulnerable CPUs 3866 off - Unconditionally disable MDS mitigation 3867 3868 On TAA-affected machines, mds=off can be prevented by 3869 an active TAA mitigation as both vulnerabilities are 3870 mitigated with the same mechanism so in order to disable 3871 this mitigation, you need to specify tsx_async_abort=off 3872 too. 3873 3874 Not specifying this option is equivalent to 3875 mds=full. 3876 3877 For details see: Documentation/admin-guide/hw-vuln/mds.rst 3878 3879 mem=nn[KMG] [HEXAGON,EARLY] Set the memory size. 3880 Must be specified, otherwise memory size will be 0. 3881 3882 mem=nn[KMG] [KNL,BOOT,EARLY] Force usage of a specific amount 3883 of memory Amount of memory to be used in cases 3884 as follows: 3885 3886 1 for test; 3887 2 when the kernel is not able to see the whole system memory; 3888 3 memory that lies after 'mem=' boundary is excluded from 3889 the hypervisor, then assigned to KVM guests. 3890 4 to limit the memory available for kdump kernel. 3891 3892 [ARC,MICROBLAZE] - the limit applies only to low memory, 3893 high memory is not affected. 3894 3895 [ARM64] - only limits memory covered by the linear 3896 mapping. The NOMAP regions are not affected. 3897 3898 [X86] Work as limiting max address. Use together 3899 with memmap= to avoid physical address space collisions. 3900 Without memmap= PCI devices could be placed at addresses 3901 belonging to unused RAM. 3902 3903 Note that this only takes effects during boot time since 3904 in above case 3, memory may need be hot added after boot 3905 if system memory of hypervisor is not sufficient. 3906 3907 mem=nn[KMG]@ss[KMG] 3908 [ARM,MIPS,EARLY] - override the memory layout 3909 reported by firmware. 3910 Define a memory region of size nn[KMG] starting at 3911 ss[KMG]. 3912 Multiple different regions can be specified with 3913 multiple mem= parameters on the command line. 3914 3915 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 3916 memory. 3917 3918 memblock=debug [KNL,EARLY] Enable memblock debug messages. 3919 3920 memchunk=nn[KMG] 3921 [KNL,SH] Allow user to override the default size for 3922 per-device physically contiguous DMA buffers. 3923 3924 memhp_default_state=online/offline/online_kernel/online_movable 3925 [KNL] Set the initial state for the memory hotplug 3926 onlining policy. If not specified, the default value is 3927 set according to the 3928 CONFIG_MHP_DEFAULT_ONLINE_TYPE kernel config 3929 options. 3930 See Documentation/admin-guide/mm/memory-hotplug.rst. 3931 3932 memmap=exactmap [KNL,X86,EARLY] Enable setting of an exact 3933 E820 memory map, as specified by the user. 3934 Such memmap=exactmap lines can be constructed based on 3935 BIOS output or other requirements. See the memmap=nn@ss 3936 option description. 3937 3938 memmap=nn[KMG]@ss[KMG] 3939 [KNL, X86,MIPS,XTENSA,EARLY] Force usage of a specific region of memory. 3940 Region of memory to be used is from ss to ss+nn. 3941 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 3942 which limits max address to nn[KMG]. 3943 Multiple different regions can be specified, 3944 comma delimited. 3945 Example: 3946 memmap=100M@2G,100M#3G,1G!1024G 3947 3948 memmap=nn[KMG]#ss[KMG] 3949 [KNL,ACPI,EARLY] Mark specific memory as ACPI data. 3950 Region of memory to be marked is from ss to ss+nn. 3951 3952 memmap=nn[KMG]$ss[KMG] 3953 [KNL,ACPI,EARLY] Mark specific memory as reserved. 3954 Region of memory to be reserved is from ss to ss+nn. 3955 Example: Exclude memory from 0x18690000-0x1869ffff 3956 memmap=64K$0x18690000 3957 or 3958 memmap=0x10000$0x18690000 3959 Some bootloaders may need an escape character before '$', 3960 like Grub2, otherwise '$' and the following number 3961 will be eaten. 3962 3963 memmap=nn[KMG]!ss[KMG,EARLY] 3964 [KNL,X86] Mark specific memory as protected. 3965 Region of memory to be used, from ss to ss+nn. 3966 The memory region may be marked as e820 type 12 (0xc) 3967 and is NVDIMM or ADR memory. 3968 3969 memmap=<size>%<offset>-<oldtype>+<newtype> 3970 [KNL,ACPI,EARLY] Convert memory within the specified region 3971 from <oldtype> to <newtype>. If "-<oldtype>" is left 3972 out, the whole region will be marked as <newtype>, 3973 even if previously unavailable. If "+<newtype>" is left 3974 out, matching memory will be removed. Types are 3975 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 3976 3 = ACPI, 12 = PRAM. 3977 3978 memory_corruption_check=0/1 [X86,EARLY] 3979 Some BIOSes seem to corrupt the first 64k of 3980 memory when doing things like suspend/resume. 3981 Setting this option will scan the memory 3982 looking for corruption. Enabling this will 3983 both detect corruption and prevent the kernel 3984 from using the memory being corrupted. 3985 However, it's intended as a diagnostic tool; if 3986 repeatable BIOS-originated corruption always 3987 affects the same memory, you can use memmap= 3988 to prevent the kernel from using that memory. 3989 3990 memory_corruption_check_size=size [X86,EARLY] 3991 By default it checks for corruption in the low 3992 64k, making this memory unavailable for normal 3993 use. Use this parameter to scan for 3994 corruption in more or less memory. 3995 3996 memory_corruption_check_period=seconds [X86,EARLY] 3997 By default it checks for corruption every 60 3998 seconds. Use this parameter to check at some 3999 other rate. 0 disables periodic checking. 4000 4001 memory_hotplug.memmap_on_memory 4002 [KNL,X86,ARM] Boolean flag to enable this feature. 4003 Format: {on | off (default)} 4004 When enabled, runtime hotplugged memory will 4005 allocate its internal metadata (struct pages, 4006 those vmemmap pages cannot be optimized even 4007 if hugetlb_free_vmemmap is enabled) from the 4008 hotadded memory which will allow to hotadd a 4009 lot of memory without requiring additional 4010 memory to do so. 4011 This feature is disabled by default because it 4012 has some implication on large (e.g. GB) 4013 allocations in some configurations (e.g. small 4014 memory blocks). 4015 The state of the flag can be read in 4016 /sys/module/memory_hotplug/parameters/memmap_on_memory. 4017 Note that even when enabled, there are a few cases where 4018 the feature is not effective. 4019 4020 memtest= [KNL,X86,ARM,M68K,PPC,RISCV,EARLY] Enable memtest 4021 Format: <integer> 4022 default : 0 <disable> 4023 Specifies the number of memtest passes to be 4024 performed. Each pass selects another test 4025 pattern from a given set of patterns. Memtest 4026 fills the memory with this pattern, validates 4027 memory contents and reserves bad memory 4028 regions that are detected. 4029 4030 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 4031 Valid arguments: on, off 4032 Default: off 4033 mem_encrypt=on: Activate SME 4034 mem_encrypt=off: Do not activate SME 4035 4036 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst 4037 for details on when memory encryption can be activated. 4038 4039 mem_sleep_default= [SUSPEND] Default system suspend mode: 4040 s2idle - Suspend-To-Idle 4041 shallow - Power-On Suspend or equivalent (if supported) 4042 deep - Suspend-To-RAM or equivalent (if supported) 4043 See Documentation/admin-guide/pm/sleep-states.rst. 4044 4045 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 4046 the BIOS has incorrectly applied a workaround. TinyBIOS 4047 version 0.98 is known to be affected, 0.99 fixes the 4048 problem by letting the user disable the workaround. 4049 4050 mga= [HW,DRM] 4051 4052 microcode= [X86] Control the behavior of the microcode loader. 4053 Available options, comma separated: 4054 4055 base_rev=X - with <X> with format: <u32> 4056 Set the base microcode revision of each thread when in 4057 debug mode. 4058 4059 dis_ucode_ldr: disable the microcode loader 4060 4061 force_minrev: 4062 Enable or disable the microcode minimal revision 4063 enforcement for the runtime microcode loader. 4064 4065 mini2440= [ARM,HW,KNL] 4066 Format:[0..2][b][c][t] 4067 Default: "0tb" 4068 MINI2440 configuration specification: 4069 0 - The attached screen is the 3.5" TFT 4070 1 - The attached screen is the 7" TFT 4071 2 - The VGA Shield is attached (1024x768) 4072 Leaving out the screen size parameter will not load 4073 the TFT driver, and the framebuffer will be left 4074 unconfigured. 4075 b - Enable backlight. The TFT backlight pin will be 4076 linked to the kernel VESA blanking code and a GPIO 4077 LED. This parameter is not necessary when using the 4078 VGA shield. 4079 c - Enable the s3c camera interface. 4080 t - Reserved for enabling touchscreen support. The 4081 touchscreen support is not enabled in the mainstream 4082 kernel as of 2.6.30, a preliminary port can be found 4083 in the "bleeding edge" mini2440 support kernel at 4084 https://repo.or.cz/w/linux-2.6/mini2440.git 4085 4086 mitigations= 4087 [X86,PPC,S390,ARM64,EARLY] Control optional mitigations for 4088 CPU vulnerabilities. This is a set of curated, 4089 arch-independent options, each of which is an 4090 aggregation of existing arch-specific options. 4091 4092 Note, "mitigations" is supported if and only if the 4093 kernel was built with CPU_MITIGATIONS=y. 4094 4095 off 4096 Disable all optional CPU mitigations. This 4097 improves system performance, but it may also 4098 expose users to several CPU vulnerabilities. 4099 Equivalent to: if nokaslr then kpti=0 [ARM64] 4100 gather_data_sampling=off [X86] 4101 indirect_target_selection=off [X86] 4102 kvm.nx_huge_pages=off [X86] 4103 l1tf=off [X86] 4104 mds=off [X86] 4105 mmio_stale_data=off [X86] 4106 no_entry_flush [PPC] 4107 no_uaccess_flush [PPC] 4108 nobp=0 [S390] 4109 nopti [X86,PPC] 4110 nospectre_bhb [ARM64] 4111 nospectre_v1 [X86,PPC] 4112 nospectre_v2 [X86,PPC,S390,ARM64] 4113 reg_file_data_sampling=off [X86] 4114 retbleed=off [X86] 4115 spec_rstack_overflow=off [X86] 4116 spec_store_bypass_disable=off [X86,PPC] 4117 spectre_bhi=off [X86] 4118 spectre_v2_user=off [X86] 4119 srbds=off [X86,INTEL] 4120 ssbd=force-off [ARM64] 4121 tsa=off [X86,AMD] 4122 tsx_async_abort=off [X86] 4123 vmscape=off [X86] 4124 4125 Exceptions: 4126 This does not have any effect on 4127 kvm.nx_huge_pages when 4128 kvm.nx_huge_pages=force. 4129 4130 auto (default) 4131 Mitigate all CPU vulnerabilities, but leave SMT 4132 enabled, even if it's vulnerable. This is for 4133 users who don't want to be surprised by SMT 4134 getting disabled across kernel upgrades, or who 4135 have other ways of avoiding SMT-based attacks. 4136 Equivalent to: (default behavior) 4137 4138 auto,nosmt 4139 Mitigate all CPU vulnerabilities, disabling SMT 4140 if needed. This is for users who always want to 4141 be fully mitigated, even if it means losing SMT. 4142 Equivalent to: l1tf=flush,nosmt [X86] 4143 mds=full,nosmt [X86] 4144 tsx_async_abort=full,nosmt [X86] 4145 mmio_stale_data=full,nosmt [X86] 4146 retbleed=auto,nosmt [X86] 4147 4148 [X86] After one of the above options, additionally 4149 supports attack-vector based controls as documented in 4150 Documentation/admin-guide/hw-vuln/attack_vector_controls.rst 4151 4152 mminit_loglevel= 4153 [KNL,EARLY] When CONFIG_DEBUG_MEMORY_INIT is set, this 4154 parameter allows control of the logging verbosity for 4155 the additional memory initialisation checks. A value 4156 of 0 disables mminit logging and a level of 4 will 4157 log everything. Information is printed at KERN_DEBUG 4158 so loglevel=8 may also need to be specified. 4159 4160 mmio_stale_data= 4161 [X86,INTEL,EARLY] Control mitigation for the Processor 4162 MMIO Stale Data vulnerabilities. 4163 4164 Processor MMIO Stale Data is a class of 4165 vulnerabilities that may expose data after an MMIO 4166 operation. Exposed data could originate or end in 4167 the same CPU buffers as affected by MDS and TAA. 4168 Therefore, similar to MDS and TAA, the mitigation 4169 is to clear the affected CPU buffers. 4170 4171 This parameter controls the mitigation. The 4172 options are: 4173 4174 full - Enable mitigation on vulnerable CPUs 4175 4176 full,nosmt - Enable mitigation and disable SMT on 4177 vulnerable CPUs. 4178 4179 off - Unconditionally disable mitigation 4180 4181 On MDS or TAA affected machines, 4182 mmio_stale_data=off can be prevented by an active 4183 MDS or TAA mitigation as these vulnerabilities are 4184 mitigated with the same mechanism so in order to 4185 disable this mitigation, you need to specify 4186 mds=off and tsx_async_abort=off too. 4187 4188 Not specifying this option is equivalent to 4189 mmio_stale_data=full. 4190 4191 For details see: 4192 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst 4193 4194 <module>.async_probe[=<bool>] [KNL] 4195 If no <bool> value is specified or if the value 4196 specified is not a valid <bool>, enable asynchronous 4197 probe on this module. Otherwise, enable/disable 4198 asynchronous probe on this module as indicated by the 4199 <bool> value. See also: module.async_probe 4200 4201 module.async_probe=<bool> 4202 [KNL] When set to true, modules will use async probing 4203 by default. To enable/disable async probing for a 4204 specific module, use the module specific control that 4205 is documented under <module>.async_probe. When both 4206 module.async_probe and <module>.async_probe are 4207 specified, <module>.async_probe takes precedence for 4208 the specific module. 4209 4210 module.enable_dups_trace 4211 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set, 4212 this means that duplicate request_module() calls will 4213 trigger a WARN_ON() instead of a pr_warn(). Note that 4214 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s 4215 will always be issued and this option does nothing. 4216 module.sig_enforce 4217 [KNL] When CONFIG_MODULE_SIG is set, this means that 4218 modules without (valid) signatures will fail to load. 4219 Note that if CONFIG_MODULE_SIG_FORCE is set, that 4220 is always true, so this option does nothing. 4221 4222 module_blacklist= [KNL] Do not load a comma-separated list of 4223 modules. Useful for debugging problem modules. 4224 4225 mousedev.tap_time= 4226 [MOUSE] Maximum time between finger touching and 4227 leaving touchpad surface for touch to be considered 4228 a tap and be reported as a left button click (for 4229 touchpads working in absolute mode only). 4230 Format: <msecs> 4231 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 4232 reporting absolute coordinates, such as tablets 4233 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 4234 reporting absolute coordinates, such as tablets 4235 4236 movablecore= [KNL,X86,PPC,EARLY] 4237 Format: nn[KMGTPE] | nn% 4238 This parameter is the complement to kernelcore=, it 4239 specifies the amount of memory used for migratable 4240 allocations. If both kernelcore and movablecore is 4241 specified, then kernelcore will be at *least* the 4242 specified value but may be more. If movablecore on its 4243 own is specified, the administrator must be careful 4244 that the amount of memory usable for all allocations 4245 is not too small. 4246 4247 movable_node [KNL,EARLY] Boot-time switch to make hotplugable memory 4248 NUMA nodes to be movable. This means that the memory 4249 of such nodes will be usable only for movable 4250 allocations which rules out almost all kernel 4251 allocations. Use with caution! 4252 4253 MTD_Partition= [MTD] 4254 Format: <name>,<region-number>,<size>,<offset> 4255 4256 MTD_Region= [MTD] Format: 4257 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 4258 4259 mtdparts= [MTD] 4260 See drivers/mtd/parsers/cmdlinepart.c 4261 4262 mtouchusb.raw_coordinates= 4263 [HW] Make the MicroTouch USB driver use raw coordinates 4264 ('y', default) or cooked coordinates ('n') 4265 4266 mtrr=debug [X86,EARLY] 4267 Enable printing debug information related to MTRR 4268 registers at boot time. 4269 4270 mtrr_chunk_size=nn[KMG,X86,EARLY] 4271 used for mtrr cleanup. It is largest continuous chunk 4272 that could hold holes aka. UC entries. 4273 4274 mtrr_gran_size=nn[KMG,X86,EARLY] 4275 Used for mtrr cleanup. It is granularity of mtrr block. 4276 Default is 1. 4277 Large value could prevent small alignment from 4278 using up MTRRs. 4279 4280 mtrr_spare_reg_nr=n [X86,EARLY] 4281 Format: <integer> 4282 Range: 0,7 : spare reg number 4283 Default : 1 4284 Used for mtrr cleanup. It is spare mtrr entries number. 4285 Set to 2 or more if your graphical card needs more. 4286 4287 multitce=off [PPC] This parameter disables the use of the pSeries 4288 firmware feature for updating multiple TCE entries 4289 at a time. 4290 4291 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 4292 4293 netdev= [NET] Network devices parameters 4294 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 4295 Note that mem_start is often overloaded to mean 4296 something different and driver-specific. 4297 This usage is only documented in each driver source 4298 file if at all. 4299 4300 netpoll.carrier_timeout= 4301 [NET] Specifies amount of time (in seconds) that 4302 netpoll should wait for a carrier. By default netpoll 4303 waits 4 seconds. 4304 4305 nf_conntrack.acct= 4306 [NETFILTER] Enable connection tracking flow accounting 4307 0 to disable accounting 4308 1 to enable accounting 4309 Default value is 0. 4310 4311 nfs.cache_getent= 4312 [NFS] sets the pathname to the program which is used 4313 to update the NFS client cache entries. 4314 4315 nfs.cache_getent_timeout= 4316 [NFS] sets the timeout after which an attempt to 4317 update a cache entry is deemed to have failed. 4318 4319 nfs.callback_nr_threads= 4320 [NFSv4] set the total number of threads that the 4321 NFS client will assign to service NFSv4 callback 4322 requests. 4323 4324 nfs.callback_tcpport= 4325 [NFS] set the TCP port on which the NFSv4 callback 4326 channel should listen. 4327 4328 nfs.delay_retrans= 4329 [NFS] specifies the number of times the NFSv4 client 4330 retries the request before returning an EAGAIN error, 4331 after a reply of NFS4ERR_DELAY from the server. 4332 Only applies if the softerr mount option is enabled, 4333 and the specified value is >= 0. 4334 4335 nfs.enable_ino64= 4336 [NFS] enable 64-bit inode numbers. 4337 If zero, the NFS client will fake up a 32-bit inode 4338 number for the readdir() and stat() syscalls instead 4339 of returning the full 64-bit number. 4340 The default is to return 64-bit inode numbers. 4341 4342 nfs.idmap_cache_timeout= 4343 [NFS] set the maximum lifetime for idmapper cache 4344 entries. 4345 4346 nfs.max_session_cb_slots= 4347 [NFSv4.1] Sets the maximum number of session 4348 slots the client will assign to the callback 4349 channel. This determines the maximum number of 4350 callbacks the client will process in parallel for 4351 a particular server. 4352 4353 nfs.max_session_slots= 4354 [NFSv4.1] Sets the maximum number of session slots 4355 the client will attempt to negotiate with the server. 4356 This limits the number of simultaneous RPC requests 4357 that the client can send to the NFSv4.1 server. 4358 Note that there is little point in setting this 4359 value higher than the max_tcp_slot_table_limit. 4360 4361 nfs.nfs4_disable_idmapping= 4362 [NFSv4] When set to the default of '1', this option 4363 ensures that both the RPC level authentication 4364 scheme and the NFS level operations agree to use 4365 numeric uids/gids if the mount is using the 4366 'sec=sys' security flavour. In effect it is 4367 disabling idmapping, which can make migration from 4368 legacy NFSv2/v3 systems to NFSv4 easier. 4369 Servers that do not support this mode of operation 4370 will be autodetected by the client, and it will fall 4371 back to using the idmapper. 4372 To turn off this behaviour, set the value to '0'. 4373 4374 nfs.nfs4_unique_id= 4375 [NFS4] Specify an additional fixed unique ident- 4376 ification string that NFSv4 clients can insert into 4377 their nfs_client_id4 string. This is typically a 4378 UUID that is generated at system install time. 4379 4380 nfs.recover_lost_locks= 4381 [NFSv4] Attempt to recover locks that were lost due 4382 to a lease timeout on the server. Please note that 4383 doing this risks data corruption, since there are 4384 no guarantees that the file will remain unchanged 4385 after the locks are lost. 4386 If you want to enable the kernel legacy behaviour of 4387 attempting to recover these locks, then set this 4388 parameter to '1'. 4389 The default parameter value of '0' causes the kernel 4390 not to attempt recovery of lost locks. 4391 4392 nfs.send_implementation_id= 4393 [NFSv4.1] Send client implementation identification 4394 information in exchange_id requests. 4395 If zero, no implementation identification information 4396 will be sent. 4397 The default is to send the implementation identification 4398 information. 4399 4400 nfs4.layoutstats_timer= 4401 [NFSv4.2] Change the rate at which the kernel sends 4402 layoutstats to the pNFS metadata server. 4403 4404 Setting this to value to 0 causes the kernel to use 4405 whatever value is the default set by the layout 4406 driver. A non-zero value sets the minimum interval 4407 in seconds between layoutstats transmissions. 4408 4409 nfsd.inter_copy_offload_enable= 4410 [NFSv4.2] When set to 1, the server will support 4411 server-to-server copies for which this server is 4412 the destination of the copy. 4413 4414 nfsd.nfs4_disable_idmapping= 4415 [NFSv4] When set to the default of '1', the NFSv4 4416 server will return only numeric uids and gids to 4417 clients using auth_sys, and will accept numeric uids 4418 and gids from such clients. This is intended to ease 4419 migration from NFSv2/v3. 4420 4421 nfsd.nfsd4_ssc_umount_timeout= 4422 [NFSv4.2] When used as the destination of a 4423 server-to-server copy, knfsd temporarily mounts 4424 the source server. It caches the mount in case 4425 it will be needed again, and discards it if not 4426 used for the number of milliseconds specified by 4427 this parameter. 4428 4429 nfsaddrs= [NFS] Deprecated. Use ip= instead. 4430 See Documentation/admin-guide/nfs/nfsroot.rst. 4431 4432 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 4433 See Documentation/admin-guide/nfs/nfsroot.rst. 4434 4435 nfsrootdebug [NFS] enable nfsroot debugging messages. 4436 See Documentation/admin-guide/nfs/nfsroot.rst. 4437 4438 nmi_backtrace.backtrace_idle [KNL] 4439 Dump stacks even of idle CPUs in response to an 4440 NMI stack-backtrace request. 4441 4442 nmi_debug= [KNL,SH] Specify one or more actions to take 4443 when a NMI is triggered. 4444 Format: [state][,regs][,debounce][,die] 4445 4446 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 4447 Format: [panic,][nopanic,][rNNN,][num] 4448 Valid num: 0 or 1 4449 0 - turn hardlockup detector in nmi_watchdog off 4450 1 - turn hardlockup detector in nmi_watchdog on 4451 rNNN - configure the watchdog with raw perf event 0xNNN 4452 4453 When panic is specified, panic when an NMI watchdog 4454 timeout occurs (or 'nopanic' to not panic on an NMI 4455 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 4456 To disable both hard and soft lockup detectors, 4457 please see 'nowatchdog'. 4458 This is useful when you use a panic=... timeout and 4459 need the box quickly up again. 4460 4461 These settings can be accessed at runtime via 4462 the nmi_watchdog and hardlockup_panic sysctls. 4463 4464 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 4465 emulation library even if a 387 maths coprocessor 4466 is present. 4467 4468 no4lvl [RISCV,EARLY] Disable 4-level and 5-level paging modes. 4469 Forces kernel to use 3-level paging instead. 4470 4471 no5lvl [X86-64,RISCV,EARLY] Disable 5-level paging mode. Forces 4472 kernel to use 4-level paging instead. 4473 4474 noalign [KNL,ARM] 4475 4476 noapic [SMP,APIC,EARLY] Tells the kernel to not make use of any 4477 IOAPICs that may be present in the system. 4478 4479 noapictimer [APIC,X86] Don't set up the APIC timer 4480 4481 noautogroup Disable scheduler automatic task group creation. 4482 4483 nocache [ARM,EARLY] 4484 4485 no_console_suspend 4486 [HW] Never suspend the console 4487 Disable suspending of consoles during suspend and 4488 hibernate operations. Once disabled, debugging 4489 messages can reach various consoles while the rest 4490 of the system is being put to sleep (ie, while 4491 debugging driver suspend/resume hooks). This may 4492 not work reliably with all consoles, but is known 4493 to work with serial and VGA consoles. 4494 To facilitate more flexible debugging, we also add 4495 console_suspend, a printk module parameter to control 4496 it. Users could use console_suspend (usually 4497 /sys/module/printk/parameters/console_suspend) to 4498 turn on/off it dynamically. 4499 4500 no_debug_objects 4501 [KNL,EARLY] Disable object debugging 4502 4503 nodsp [SH] Disable hardware DSP at boot time. 4504 4505 noefi [EFI,EARLY] Disable EFI runtime services support. 4506 4507 no_entry_flush [PPC,EARLY] Don't flush the L1-D cache when entering the kernel. 4508 4509 noexec32 [X86-64] 4510 This affects only 32-bit executables. 4511 noexec32=on: enable non-executable mappings (default) 4512 read doesn't imply executable mappings 4513 noexec32=off: disable non-executable mappings 4514 read implies executable mappings 4515 4516 no_file_caps Tells the kernel not to honor file capabilities. The 4517 only way then for a file to be executed with privilege 4518 is to be setuid root or executed by root. 4519 4520 nofpu [MIPS,SH] Disable hardware FPU at boot time. 4521 4522 nofsgsbase [X86] Disables FSGSBASE instructions. 4523 4524 nofxsr [BUGS=X86-32] Disables x86 floating point extended 4525 register save and restore. The kernel will only save 4526 legacy floating-point registers on task switch. 4527 4528 nogbpages [X86] Do not use GB pages for kernel direct mappings. 4529 4530 no_hash_pointers 4531 [KNL,EARLY] 4532 Alias for "hash_pointers=never". 4533 4534 nohibernate [HIBERNATION] Disable hibernation and resume. 4535 4536 nohlt [ARM,ARM64,MICROBLAZE,MIPS,PPC,RISCV,SH] Forces the kernel to 4537 busy wait in do_idle() and not use the arch_cpu_idle() 4538 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP 4539 to be effective. This is useful on platforms where the 4540 sleep(SH) or wfi(ARM,ARM64) instructions do not work 4541 correctly or when doing power measurements to evaluate 4542 the impact of the sleep instructions. This is also 4543 useful when using JTAG debugger. 4544 4545 nohpet [X86] Don't use the HPET timer. 4546 4547 nohugeiomap [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge I/O mappings. 4548 4549 nohugevmalloc [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge vmalloc mappings. 4550 4551 nohz= [KNL] Boottime enable/disable dynamic ticks 4552 Valid arguments: on, off 4553 Default: on 4554 4555 nohz_full= [KNL,BOOT,SMP,ISOL] 4556 The argument is a cpu list, as described above. 4557 In kernels built with CONFIG_NO_HZ_FULL=y, set 4558 the specified list of CPUs whose tick will be stopped 4559 whenever possible. The boot CPU will be forced outside 4560 the range to maintain the timekeeping. Any CPUs 4561 in this list will have their RCU callbacks offloaded, 4562 just as if they had also been called out in the 4563 rcu_nocbs= boot parameter. 4564 4565 Note that this argument takes precedence over 4566 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 4567 4568 noinitrd [Deprecated,RAM] Tells the kernel not to load any configured 4569 initial RAM disk. Currently this parameter applies to 4570 initrd only, not to initramfs. But it applies to both 4571 in EFI mode. 4572 4573 nointremap [X86-64,Intel-IOMMU,EARLY] Do not enable interrupt 4574 remapping. 4575 [Deprecated - use intremap=off] 4576 4577 noinvpcid [X86,EARLY] Disable the INVPCID cpu feature. 4578 4579 noiotrap [SH] Disables trapped I/O port accesses. 4580 4581 noirqdebug [X86-32] Disables the code which attempts to detect and 4582 disable unhandled interrupt sources. 4583 4584 noisapnp [ISAPNP] Disables ISA PnP code. 4585 4586 nokaslr [KNL,EARLY] 4587 When CONFIG_RANDOMIZE_BASE is set, this disables 4588 kernel and module base offset ASLR (Address Space 4589 Layout Randomization). 4590 4591 no-kvmapf [X86,KVM,EARLY] Disable paravirtualized asynchronous page 4592 fault handling. 4593 4594 no-kvmclock [X86,KVM,EARLY] Disable paravirtualized KVM clock driver 4595 4596 nolapic [X86-32,APIC,EARLY] Do not enable or use the local APIC. 4597 4598 nolapic_timer [X86-32,APIC,EARLY] Do not use the local APIC timer. 4599 4600 nomce [X86-32] Disable Machine Check Exception 4601 4602 nomfgpt [X86-32] Disable Multi-Function General Purpose 4603 Timer usage (for AMD Geode machines). 4604 4605 nomodeset Disable kernel modesetting. Most systems' firmware 4606 sets up a display mode and provides framebuffer memory 4607 for output. With nomodeset, DRM and fbdev drivers will 4608 not load if they could possibly displace the pre- 4609 initialized output. Only the system framebuffer will 4610 be available for use. The respective drivers will not 4611 perform display-mode changes or accelerated rendering. 4612 4613 Useful as error fallback, or for testing and debugging. 4614 4615 nomodule Disable module load 4616 4617 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 4618 shutdown the other cpus. Instead use the REBOOT_VECTOR 4619 irq. 4620 4621 nopat [X86,EARLY] Disable PAT (page attribute table extension of 4622 pagetables) support. 4623 4624 nopcid [X86-64,EARLY] Disable the PCID cpu feature. 4625 4626 nopku [X86] Disable Memory Protection Keys CPU feature found 4627 in some Intel CPUs. 4628 4629 nopti [X86-64,EARLY] 4630 Equivalent to pti=off 4631 4632 nopv= [X86,XEN,KVM,HYPER_V,VMWARE,EARLY] 4633 Disables the PV optimizations forcing the guest to run 4634 as generic guest with no PV drivers. Currently support 4635 XEN HVM, KVM, HYPER_V and VMWARE guest. 4636 4637 nopvspin [X86,XEN,KVM,EARLY] 4638 Disables the qspinlock slow path using PV optimizations 4639 which allow the hypervisor to 'idle' the guest on lock 4640 contention. 4641 4642 norandmaps Don't use address space randomization. Equivalent to 4643 echo 0 > /proc/sys/kernel/randomize_va_space 4644 4645 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 4646 with UP alternatives 4647 4648 noresume [SWSUSP] Disables resume and restores original swap 4649 space. 4650 4651 no-scroll [VGA] Disables scrollback. 4652 This is required for the Braillex ib80-piezo Braille 4653 reader made by F.H. Papenmeier (Germany). 4654 4655 nosgx [X86-64,SGX,EARLY] Disables Intel SGX kernel support. 4656 4657 nosmap [PPC,EARLY] 4658 Disable SMAP (Supervisor Mode Access Prevention) 4659 even if it is supported by processor. 4660 4661 nosmep [PPC64s,EARLY] 4662 Disable SMEP (Supervisor Mode Execution Prevention) 4663 even if it is supported by processor. 4664 4665 nosmp [SMP,EARLY] Tells an SMP kernel to act as a UP kernel, 4666 and disable the IO APIC. legacy for "maxcpus=0". 4667 4668 nosmt [KNL,MIPS,PPC,EARLY] Disable symmetric multithreading (SMT). 4669 Equivalent to smt=1. 4670 4671 [KNL,LOONGARCH,X86,PPC,S390] Disable symmetric multithreading (SMT). 4672 nosmt=force: Force disable SMT, cannot be undone 4673 via the sysfs control file. 4674 4675 nosoftlockup [KNL] Disable the soft-lockup detector. 4676 4677 nospec_store_bypass_disable 4678 [HW,EARLY] Disable all mitigations for the Speculative 4679 Store Bypass vulnerability 4680 4681 nospectre_bhb [ARM64,EARLY] Disable all mitigations for Spectre-BHB (branch 4682 history injection) vulnerability. System may allow data leaks 4683 with this option. 4684 4685 nospectre_v1 [X86,PPC,EARLY] Disable mitigations for Spectre Variant 1 4686 (bounds check bypass). With this option data leaks are 4687 possible in the system. 4688 4689 nospectre_v2 [X86,PPC_E500,ARM64,EARLY] Disable all mitigations 4690 for the Spectre variant 2 (indirect branch 4691 prediction) vulnerability. System may allow data 4692 leaks with this option. 4693 4694 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES,RISCV,LOONGARCH,EARLY] 4695 Disable paravirtualized steal time accounting. steal time 4696 is computed, but won't influence scheduler behaviour 4697 4698 nosync [HW,M68K] Disables sync negotiation for all devices. 4699 4700 no_timer_check [X86,APIC] Disables the code which tests for broken 4701 timer IRQ sources, i.e., the IO-APIC timer. This can 4702 work around problems with incorrect timer 4703 initialization on some boards. 4704 4705 no_uaccess_flush 4706 [PPC,EARLY] Don't flush the L1-D cache after accessing user data. 4707 4708 novmcoredd [KNL,KDUMP] 4709 Disable device dump. Device dump allows drivers to 4710 append dump data to vmcore so you can collect driver 4711 specified debug info. Drivers can append the data 4712 without any limit and this data is stored in memory, 4713 so this may cause significant memory stress. Disabling 4714 device dump can help save memory but the driver debug 4715 data will be no longer available. This parameter 4716 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 4717 is set. 4718 4719 no-vmw-sched-clock 4720 [X86,PV_OPS,EARLY] Disable paravirtualized VMware 4721 scheduler clock and use the default one. 4722 4723 nowatchdog [KNL] Disable both lockup detectors, i.e. 4724 soft-lockup and NMI watchdog (hard-lockup). 4725 4726 nowb [ARM,EARLY] 4727 4728 nox2apic [X86-64,APIC,EARLY] Do not enable x2APIC mode. 4729 4730 NOTE: this parameter will be ignored on systems with the 4731 LEGACY_XAPIC_DISABLED bit set in the 4732 IA32_XAPIC_DISABLE_STATUS MSR. 4733 4734 noxsave [BUGS=X86] Disables x86 extended register state save 4735 and restore using xsave. The kernel will fallback to 4736 enabling legacy floating-point and sse state. 4737 4738 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 4739 register states. The kernel will fall back to use 4740 xsave to save the states. By using this parameter, 4741 performance of saving the states is degraded because 4742 xsave doesn't support modified optimization while 4743 xsaveopt supports it on xsaveopt enabled systems. 4744 4745 noxsaves [X86] Disables xsaves and xrstors used in saving and 4746 restoring x86 extended register state in compacted 4747 form of xsave area. The kernel will fall back to use 4748 xsaveopt and xrstor to save and restore the states 4749 in standard form of xsave area. By using this 4750 parameter, xsave area per process might occupy more 4751 memory on xsaves enabled systems. 4752 4753 nr_cpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 4754 could support. nr_cpus=n : n >= 1 limits the kernel to 4755 support 'n' processors. It could be larger than the 4756 number of already plugged CPU during bootup, later in 4757 runtime you can physically add extra cpu until it reaches 4758 n. So during boot up some boot time memory for per-cpu 4759 variables need be pre-allocated for later physical cpu 4760 hot plugging. 4761 4762 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 4763 4764 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86, EARLY] 4765 Disable NUMA, Only set up a single NUMA node 4766 spanning all memory. 4767 4768 numa=fake=<size>[MG] 4769 [KNL, ARM64, RISCV, X86, EARLY] 4770 If given as a memory unit, fills all system RAM with 4771 nodes of size interleaved over physical nodes. 4772 4773 numa=fake=<N> 4774 [KNL, ARM64, RISCV, X86, EARLY] 4775 If given as an integer, fills all system RAM with N 4776 fake nodes interleaved over physical nodes. 4777 4778 numa=fake=<N>U 4779 [KNL, ARM64, RISCV, X86, EARLY] 4780 If given as an integer followed by 'U', it will 4781 divide each physical node into N emulated nodes. 4782 4783 numa=noacpi [X86] Don't parse the SRAT table for NUMA setup 4784 4785 numa=nohmat [X86] Don't parse the HMAT table for NUMA setup, or 4786 soft-reserved memory partitioning. 4787 4788 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic 4789 NUMA balancing. 4790 Allowed values are enable and disable 4791 4792 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 4793 'node', 'default' can be specified 4794 This can be set from sysctl after boot. 4795 See Documentation/admin-guide/sysctl/vm.rst for details. 4796 4797 nvme.quirks= [NVME] A list of quirk entries to augment the built-in 4798 nvme quirk list. List entries are separated by a 4799 '-' character. 4800 Each entry has the form VendorID:ProductID:quirk_names. 4801 The IDs are 4-digits hex numbers and quirk_names is a 4802 list of quirk names separated by commas. A quirk name 4803 can be prefixed by '^', meaning that the specified 4804 quirk must be disabled. 4805 4806 Example: 4807 nvme.quirks=7710:2267:bogus_nid,^identify_cns-9900:7711:broken_msi 4808 4809 ohci1394_dma=early [HW,EARLY] enable debugging via the ohci1394 driver. 4810 See Documentation/core-api/debugging-via-ohci1394.rst for more 4811 info. 4812 4813 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 4814 Rather than timing out after 20 ms if an EC 4815 command is not properly ACKed, override the length 4816 of the timeout. We have interrupts disabled while 4817 waiting for the ACK, so if this is set too high 4818 interrupts *may* be lost! 4819 4820 omap_mux= [OMAP] Override bootloader pin multiplexing. 4821 Format: <mux_mode0.mode_name=value>... 4822 For example, to override I2C bus2: 4823 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 4824 4825 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 4826 4827 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 4828 4829 boundary - index of last SLC block on Flex-OneNAND. 4830 The remaining blocks are configured as MLC blocks. 4831 lock - Configure if Flex-OneNAND boundary should be locked. 4832 Once locked, the boundary cannot be changed. 4833 1 indicates lock status, 0 indicates unlock status. 4834 4835 oops=panic [KNL,EARLY] 4836 Always panic on oopses. Default is to just kill the 4837 process, but there is a small probability of 4838 deadlocking the machine. 4839 This will also cause panics on machine check exceptions. 4840 Useful together with panic=30 to trigger a reboot. 4841 4842 page_alloc.shuffle= 4843 [KNL] Boolean flag to control whether the page allocator 4844 should randomize its free lists. This parameter can be 4845 used to enable/disable page randomization. The state of 4846 the flag can be read from sysfs at: 4847 /sys/module/page_alloc/parameters/shuffle. 4848 This parameter is only available if CONFIG_SHUFFLE_PAGE_ALLOCATOR=y. 4849 4850 page_owner= [KNL,EARLY] Boot-time page_owner enabling option. 4851 Storage of the information about who allocated 4852 each page is disabled in default. With this switch, 4853 we can turn it on. 4854 on: enable the feature 4855 4856 page_poison= [KNL,EARLY] Boot-time parameter changing the state of 4857 poisoning on the buddy allocator, available with 4858 CONFIG_PAGE_POISONING=y. 4859 off: turn off poisoning (default) 4860 on: turn on poisoning 4861 4862 page_reporting.page_reporting_order= 4863 [KNL] Minimal page reporting order 4864 Format: <integer> 4865 Adjust the minimal page reporting order. The page 4866 reporting is disabled when it exceeds MAX_PAGE_ORDER. 4867 4868 panic= [KNL] Kernel behaviour on panic: delay <timeout> 4869 timeout > 0: seconds before rebooting 4870 timeout = 0: wait forever 4871 timeout < 0: reboot immediately 4872 Format: <timeout> 4873 4874 panic_on_taint= [KNL,EARLY] 4875 Bitmask for conditionally calling panic() in add_taint() 4876 Format: <hex>[,nousertaint] 4877 Hexadecimal bitmask representing the set of TAINT flags 4878 that will cause the kernel to panic when add_taint() is 4879 called with any of the flags in this set. 4880 The optional switch "nousertaint" can be utilized to 4881 prevent userspace forced crashes by writing to sysctl 4882 /proc/sys/kernel/tainted any flagset matching with the 4883 bitmask set on panic_on_taint. 4884 See Documentation/admin-guide/tainted-kernels.rst for 4885 extra details on the taint flags that users can pick 4886 to compose the bitmask to assign to panic_on_taint. 4887 4888 panic_on_warn=1 panic() instead of WARN(). Useful to cause kdump 4889 on a WARN(). 4890 4891 panic_force_cpu= 4892 [KNL,SMP] Force panic handling to execute on a specific CPU. 4893 Format: <cpu number> 4894 Some platforms require panic handling to occur on a 4895 specific CPU for the crash kernel to function correctly. 4896 This can be due to firmware limitations, interrupt routing 4897 constraints, or platform-specific requirements where only 4898 a particular CPU can safely enter the crash kernel. 4899 When set, panic() will redirect execution to the specified 4900 CPU before proceeding with the normal panic and kexec flow. 4901 If the target CPU is offline or unavailable, panic proceeds 4902 on the current CPU. 4903 This option should only be used for systems with the above 4904 constraints as it might cause the panic operation to be less reliable. 4905 4906 panic_print= Bitmask for printing system info when panic happens. 4907 User can chose combination of the following bits: 4908 bit 0: print all tasks info 4909 bit 1: print system memory info 4910 bit 2: print timer info 4911 bit 3: print locks info if CONFIG_LOCKDEP is on 4912 bit 4: print ftrace buffer 4913 bit 5: replay all kernel messages on consoles at the end of panic 4914 bit 6: print all CPUs backtrace (if available in the arch) 4915 bit 7: print only tasks in uninterruptible (blocked) state 4916 *Be aware* that this option may print a _lot_ of lines, 4917 so there are risks of losing older messages in the log. 4918 Use this option carefully, maybe worth to setup a 4919 bigger log buffer with "log_buf_len" along with this. 4920 4921 panic_sys_info= A comma separated list of extra information to be dumped 4922 on panic. 4923 Format: val[,val...] 4924 Where @val can be any of the following: 4925 4926 tasks: print all tasks info 4927 mem: print system memory info 4928 timers: print timers info 4929 locks: print locks info if CONFIG_LOCKDEP is on 4930 ftrace: print ftrace buffer 4931 all_bt: print all CPUs backtrace (if available in the arch) 4932 blocked_tasks: print only tasks in uninterruptible (blocked) state 4933 4934 This is a human readable alternative to the 'panic_print' option. 4935 4936 panic_console_replay 4937 When panic happens, replay all kernel messages on 4938 consoles at the end of panic. 4939 4940 parkbd.port= [HW] Parallel port number the keyboard adapter is 4941 connected to, default is 0. 4942 Format: <parport#> 4943 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 4944 0 for XT, 1 for AT (default is AT). 4945 Format: <mode> 4946 4947 parport= [HW,PPT] Specify parallel ports. 0 disables. 4948 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 4949 Use 'auto' to force the driver to use any 4950 IRQ/DMA settings detected (the default is to 4951 ignore detected IRQ/DMA settings because of 4952 possible conflicts). You can specify the base 4953 address, IRQ, and DMA settings; IRQ and DMA 4954 should be numbers, or 'auto' (for using detected 4955 settings on that particular port), or 'nofifo' 4956 (to avoid using a FIFO even if it is detected). 4957 Parallel ports are assigned in the order they 4958 are specified on the command line, starting 4959 with parport0. 4960 4961 parport_init_mode= [HW,PPT] 4962 Configure VIA parallel port to operate in 4963 a specific mode. This is necessary on Pegasos 4964 computer where firmware has no options for setting 4965 up parallel port mode and sets it to spp. 4966 Currently this function knows 686a and 8231 chips. 4967 Format: [spp|ps2|epp|ecp|ecpepp] 4968 4969 pata_legacy.all= [HW,LIBATA] 4970 Format: <int> 4971 Set to non-zero to probe primary and secondary ISA 4972 port ranges on PCI systems where no PCI PATA device 4973 has been found at either range. Disabled by default. 4974 4975 pata_legacy.autospeed= [HW,LIBATA] 4976 Format: <int> 4977 Set to non-zero if a chip is present that snoops speed 4978 changes. Disabled by default. 4979 4980 pata_legacy.ht6560a= [HW,LIBATA] 4981 Format: <int> 4982 Set to 1, 2, or 3 for HT 6560A on the primary channel, 4983 the secondary channel, or both channels respectively. 4984 Disabled by default. 4985 4986 pata_legacy.ht6560b= [HW,LIBATA] 4987 Format: <int> 4988 Set to 1, 2, or 3 for HT 6560B on the primary channel, 4989 the secondary channel, or both channels respectively. 4990 Disabled by default. 4991 4992 pata_legacy.iordy_mask= [HW,LIBATA] 4993 Format: <int> 4994 IORDY enable mask. Set individual bits to allow IORDY 4995 for the respective channel. Bit 0 is for the first 4996 legacy channel handled by this driver, bit 1 is for 4997 the second channel, and so on. The sequence will often 4998 correspond to the primary legacy channel, the secondary 4999 legacy channel, and so on, but the handling of a PCI 5000 bus and the use of other driver options may interfere 5001 with the sequence. By default IORDY is allowed across 5002 all channels. 5003 5004 pata_legacy.opti82c46x= [HW,LIBATA] 5005 Format: <int> 5006 Set to 1, 2, or 3 for Opti 82c611A on the primary 5007 channel, the secondary channel, or both channels 5008 respectively. Disabled by default. 5009 5010 pata_legacy.opti82c611a= [HW,LIBATA] 5011 Format: <int> 5012 Set to 1, 2, or 3 for Opti 82c465MV on the primary 5013 channel, the secondary channel, or both channels 5014 respectively. Disabled by default. 5015 5016 pata_legacy.pio_mask= [HW,LIBATA] 5017 Format: <int> 5018 PIO mode mask for autospeed devices. Set individual 5019 bits to allow the use of the respective PIO modes. 5020 Bit 0 is for mode 0, bit 1 is for mode 1, and so on. 5021 All modes allowed by default. 5022 5023 pata_legacy.probe_all= [HW,LIBATA] 5024 Format: <int> 5025 Set to non-zero to probe tertiary and further ISA 5026 port ranges on PCI systems. Disabled by default. 5027 5028 pata_legacy.probe_mask= [HW,LIBATA] 5029 Format: <int> 5030 Probe mask for legacy ISA PATA ports. Depending on 5031 platform configuration and the use of other driver 5032 options up to 6 legacy ports are supported: 0x1f0, 5033 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing 5034 of individual ports can be disabled by setting the 5035 corresponding bits in the mask to 1. Bit 0 is for 5036 the first port in the list above (0x1f0), and so on. 5037 By default all supported ports are probed. 5038 5039 pata_legacy.qdi= [HW,LIBATA] 5040 Format: <int> 5041 Set to non-zero to probe QDI controllers. By default 5042 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise. 5043 5044 pata_legacy.winbond= [HW,LIBATA] 5045 Format: <int> 5046 Set to non-zero to probe Winbond controllers. Use 5047 the standard I/O port (0x130) if 1, otherwise the 5048 value given is the I/O port to use (typically 0x1b0). 5049 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE, 5050 0 otherwise. 5051 5052 pata_platform.pio_mask= [HW,LIBATA] 5053 Format: <int> 5054 Supported PIO mode mask. Set individual bits to allow 5055 the use of the respective PIO modes. Bit 0 is for 5056 mode 0, bit 1 is for mode 1, and so on. Mode 0 only 5057 allowed by default. 5058 5059 pause_on_oops=<int> 5060 Halt all CPUs after the first oops has been printed for 5061 the specified number of seconds. This is to be used if 5062 your oopses keep scrolling off the screen. 5063 5064 pcbit= [HW,ISDN] 5065 5066 pci=option[,option...] [PCI,EARLY] various PCI subsystem options. 5067 5068 Some options herein operate on a specific device 5069 or a set of devices (<pci_dev>). These are 5070 specified in one of the following formats: 5071 5072 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 5073 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 5074 5075 Note: the first format specifies a PCI 5076 bus/device/function address which may change 5077 if new hardware is inserted, if motherboard 5078 firmware changes, or due to changes caused 5079 by other kernel parameters. If the 5080 domain is left unspecified, it is 5081 taken to be zero. Optionally, a path 5082 to a device through multiple device/function 5083 addresses can be specified after the base 5084 address (this is more robust against 5085 renumbering issues). The second format 5086 selects devices using IDs from the 5087 configuration space which may match multiple 5088 devices in the system. 5089 5090 earlydump dump PCI config space before the kernel 5091 changes anything 5092 off [X86] don't probe for the PCI bus 5093 bios [X86-32] force use of PCI BIOS, don't access 5094 the hardware directly. Use this if your machine 5095 has a non-standard PCI host bridge. 5096 nobios [X86-32] disallow use of PCI BIOS, only direct 5097 hardware access methods are allowed. Use this 5098 if you experience crashes upon bootup and you 5099 suspect they are caused by the BIOS. 5100 conf1 [X86] Force use of PCI Configuration Access 5101 Mechanism 1 (config address in IO port 0xCF8, 5102 data in IO port 0xCFC, both 32-bit). 5103 conf2 [X86] Force use of PCI Configuration Access 5104 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 5105 the function, IO port 0xCFA, also 8-bit, sets 5106 bus number. The config space is then accessed 5107 through ports 0xC000-0xCFFF). 5108 See http://wiki.osdev.org/PCI for more info 5109 on the configuration access mechanisms. 5110 noaer [PCIE] If the PCIEAER kernel config parameter is 5111 enabled, this kernel boot option can be used to 5112 disable the use of PCIE advanced error reporting. 5113 nodomains [PCI] Disable support for multiple PCI 5114 root domains (aka PCI segments, in ACPI-speak). 5115 nommconf [X86] Disable use of MMCONFIG for PCI 5116 Configuration 5117 check_enable_amd_mmconf [X86] check for and enable 5118 properly configured MMIO access to PCI 5119 config space on AMD family 10h CPU 5120 nomsi [MSI] If the PCI_MSI kernel config parameter is 5121 enabled, this kernel boot option can be used to 5122 disable the use of MSI interrupts system-wide. 5123 noioapicquirk [APIC] Disable all boot interrupt quirks. 5124 Safety option to keep boot IRQs enabled. This 5125 should never be necessary. 5126 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 5127 primary IO-APIC for bridges that cannot disable 5128 boot IRQs. This fixes a source of spurious IRQs 5129 when the system masks IRQs. 5130 noioapicreroute [APIC] Disable workaround that uses the 5131 boot IRQ equivalent of an IRQ that connects to 5132 a chipset where boot IRQs cannot be disabled. 5133 The opposite of ioapicreroute. 5134 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 5135 routing table. These calls are known to be buggy 5136 on several machines and they hang the machine 5137 when used, but on other computers it's the only 5138 way to get the interrupt routing table. Try 5139 this option if the kernel is unable to allocate 5140 IRQs or discover secondary PCI buses on your 5141 motherboard. 5142 rom [X86] Assign address space to expansion ROMs. 5143 Use with caution as certain devices share 5144 address decoders between ROMs and other 5145 resources. 5146 norom [X86] Do not assign address space to 5147 expansion ROMs that do not already have 5148 BIOS assigned address ranges. 5149 nobar [X86] Do not assign address space to the 5150 BARs that weren't assigned by the BIOS. 5151 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 5152 assigned automatically to PCI devices. You can 5153 make the kernel exclude IRQs of your ISA cards 5154 this way. 5155 pirqaddr=0xAAAAA [X86] Specify the physical address 5156 of the PIRQ table (normally generated 5157 by the BIOS) if it is outside the 5158 F0000h-100000h range. 5159 lastbus=N [X86] Scan all buses thru bus #N. Can be 5160 useful if the kernel is unable to find your 5161 secondary buses and you want to tell it 5162 explicitly which ones they are. 5163 assign-busses [X86] Always assign all PCI bus 5164 numbers ourselves, overriding 5165 whatever the firmware may have done. 5166 usepirqmask [X86] Honor the possible IRQ mask stored 5167 in the BIOS $PIR table. This is needed on 5168 some systems with broken BIOSes, notably 5169 some HP Pavilion N5400 and Omnibook XE3 5170 notebooks. This will have no effect if ACPI 5171 IRQ routing is enabled. 5172 noacpi [X86] Do not use ACPI for IRQ routing 5173 or for PCI scanning. 5174 use_crs [X86] Use PCI host bridge window information 5175 from ACPI. On BIOSes from 2008 or later, this 5176 is enabled by default. If you need to use this, 5177 please report a bug. 5178 nocrs [X86] Ignore PCI host bridge windows from ACPI. 5179 If you need to use this, please report a bug. 5180 use_e820 [X86] Use E820 reservations to exclude parts of 5181 PCI host bridge windows. This is a workaround 5182 for BIOS defects in host bridge _CRS methods. 5183 If you need to use this, please report a bug to 5184 <linux-pci@vger.kernel.org>. 5185 no_e820 [X86] Ignore E820 reservations for PCI host 5186 bridge windows. This is the default on modern 5187 hardware. If you need to use this, please report 5188 a bug to <linux-pci@vger.kernel.org>. 5189 routeirq Do IRQ routing for all PCI devices. 5190 This is normally done in pci_enable_device(), 5191 so this option is a temporary workaround 5192 for broken drivers that don't call it. 5193 skip_isa_align [X86] do not align io start addr, so can 5194 handle more pci cards 5195 noearly [X86] Don't do any early type 1 scanning. 5196 This might help on some broken boards which 5197 machine check when some devices' config space 5198 is read. But various workarounds are disabled 5199 and some IOMMU drivers will not work. 5200 bfsort Sort PCI devices into breadth-first order. 5201 This sorting is done to get a device 5202 order compatible with older (<= 2.4) kernels. 5203 nobfsort Don't sort PCI devices into breadth-first order. 5204 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 5205 tuning and use the BIOS-configured MPS defaults. 5206 pcie_bus_safe Set every device's MPS to the largest value 5207 supported by all devices below the root complex. 5208 pcie_bus_perf Set device MPS to the largest allowable MPS 5209 based on its parent bus. Also set MRRS (Max 5210 Read Request Size) to the largest supported 5211 value (no larger than the MPS that the device 5212 or bus can support) for best performance. 5213 pcie_bus_peer2peer Set every device's MPS to 128B, which 5214 every device is guaranteed to support. This 5215 configuration allows peer-to-peer DMA between 5216 any pair of devices, possibly at the cost of 5217 reduced performance. This also guarantees 5218 that hot-added devices will work. 5219 cbiosize=nn[KMG] The fixed amount of bus space which is 5220 reserved for the CardBus bridge's IO window. 5221 The default value is 256 bytes. 5222 cbmemsize=nn[KMG] The fixed amount of bus space which is 5223 reserved for the CardBus bridge's memory 5224 window. The default value is 64 megabytes. 5225 resource_alignment= 5226 Format: 5227 [<order of align>@]<pci_dev>[; ...] 5228 Specifies alignment and device to reassign 5229 aligned memory resources. How to 5230 specify the device is described above. 5231 If <order of align> is not specified, 5232 PAGE_SIZE is used as alignment. 5233 A PCI-PCI bridge can be specified if resource 5234 windows need to be expanded. 5235 To specify the alignment for several 5236 instances of a device, the PCI vendor, 5237 device, subvendor, and subdevice may be 5238 specified, e.g., 12@pci:8086:9c22:103c:198f 5239 for 4096-byte alignment. 5240 ecrc= Enable/disable PCIe ECRC (transaction layer 5241 end-to-end CRC checking). Only effective if 5242 OS has native AER control (either granted by 5243 ACPI _OSC or forced via "pcie_ports=native") 5244 bios: Use BIOS/firmware settings. This is the 5245 the default. 5246 off: Turn ECRC off 5247 on: Turn ECRC on. 5248 hpiosize=nn[KMG] The fixed amount of bus space which is 5249 reserved for hotplug bridge's IO window. 5250 Default size is 256 bytes. 5251 hpmmiosize=nn[KMG] The fixed amount of bus space which is 5252 reserved for hotplug bridge's MMIO window. 5253 Default size is 2 megabytes. 5254 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 5255 reserved for hotplug bridge's MMIO_PREF window. 5256 Default size is 2 megabytes. 5257 hpmemsize=nn[KMG] The fixed amount of bus space which is 5258 reserved for hotplug bridge's MMIO and 5259 MMIO_PREF window. 5260 Default size is 2 megabytes. 5261 hpbussize=nn The minimum amount of additional bus numbers 5262 reserved for buses below a hotplug bridge. 5263 Default is 1. 5264 realloc= Enable/disable reallocating PCI bridge resources 5265 if allocations done by BIOS are too small to 5266 accommodate resources required by all child 5267 devices. 5268 off: Turn realloc off 5269 on: Turn realloc on 5270 realloc same as realloc=on 5271 noari do not use PCIe ARI. 5272 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 5273 do not use PCIe ATS (and IOMMU device IOTLB). 5274 pcie_scan_all Scan all possible PCIe devices. Otherwise we 5275 only look for one device below a PCIe downstream 5276 port. 5277 big_root_window Try to add a big 64bit memory window to the PCIe 5278 root complex on AMD CPUs. Some GFX hardware 5279 can resize a BAR to allow access to all VRAM. 5280 Adding the window is slightly risky (it may 5281 conflict with unreported devices), so this 5282 taints the kernel. 5283 disable_acs_redir=<pci_dev>[; ...] 5284 Specify one or more PCI devices (in the format 5285 specified above) separated by semicolons. 5286 Each device specified will have the PCI ACS 5287 redirect capabilities forced off which will 5288 allow P2P traffic between devices through 5289 bridges without forcing it upstream. Note: 5290 this removes isolation between devices and 5291 may put more devices in an IOMMU group. 5292 config_acs= 5293 Format: 5294 <ACS flags>@<pci_dev>[; ...] 5295 Specify one or more PCI devices (in the format 5296 specified above) optionally prepended with flags 5297 and separated by semicolons. The respective 5298 capabilities will be enabled, disabled or 5299 unchanged based on what is specified in 5300 flags. 5301 5302 ACS Flags is defined as follows: 5303 bit-0 : ACS Source Validation 5304 bit-1 : ACS Translation Blocking 5305 bit-2 : ACS P2P Request Redirect 5306 bit-3 : ACS P2P Completion Redirect 5307 bit-4 : ACS Upstream Forwarding 5308 bit-5 : ACS P2P Egress Control 5309 bit-6 : ACS Direct Translated P2P 5310 Each bit can be marked as: 5311 '0' – force disabled 5312 '1' – force enabled 5313 'x' – unchanged 5314 For example, 5315 pci=config_acs=10x@pci:0:0 5316 would configure all devices that support 5317 ACS to enable P2P Request Redirect, disable 5318 Translation Blocking, and leave Source 5319 Validation unchanged from whatever power-up 5320 or firmware set it to. 5321 5322 Note: this may remove isolation between devices 5323 and may put more devices in an IOMMU group. 5324 force_floating [S390] Force usage of floating interrupts. 5325 nomio [S390] Do not use MIO instructions. 5326 norid [S390] ignore the RID field and force use of 5327 one PCI domain per PCI function 5328 notph [PCIE] If the PCIE_TPH kernel config parameter 5329 is enabled, this kernel boot option can be used 5330 to disable PCIe TLP Processing Hints support 5331 system-wide. 5332 5333 pcie_aspm= [PCIE] Forcibly enable or ignore PCIe Active State Power 5334 Management. 5335 off Don't touch ASPM configuration at all. Leave any 5336 configuration done by firmware unchanged. 5337 force Enable ASPM even on devices that claim not to support it. 5338 WARNING: Forcing ASPM on may cause system lockups. 5339 5340 pcie_ports= [PCIE] PCIe port services handling: 5341 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 5342 even if the platform doesn't give the OS permission to 5343 use them. This may cause conflicts if the platform 5344 also tries to use these services. 5345 dpc-native Use native PCIe service for DPC only. May 5346 cause conflicts if firmware uses AER or DPC. 5347 compat Disable native PCIe services (PME, AER, DPC, PCIe 5348 hotplug). 5349 5350 pcie_port_pm= [PCIE] PCIe port power management handling: 5351 off Disable power management of all PCIe ports 5352 force Forcibly enable power management of all PCIe ports 5353 5354 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 5355 nomsi Do not use MSI for native PCIe PME signaling (this makes 5356 all PCIe root ports use INTx for all services). 5357 5358 pcmv= [HW,PCMCIA] BadgePAD 4 5359 5360 pd_ignore_unused 5361 [PM] 5362 Keep all power-domains already enabled by bootloader on, 5363 even if no driver has claimed them. This is useful 5364 for debug and development, but should not be 5365 needed on a platform with proper driver support. 5366 5367 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 5368 boot time. 5369 Format: { 0 | 1 } 5370 See arch/parisc/kernel/pdc_chassis.c 5371 5372 percpu_alloc= [MM,EARLY] 5373 Select which percpu first chunk allocator to use. 5374 Currently supported values are "embed" and "page". 5375 Archs may support subset or none of the selections. 5376 See comments in mm/percpu.c for details on each 5377 allocator. This parameter is primarily for debugging 5378 and performance comparison. 5379 5380 pirq= [SMP,APIC] Manual mp-table setup 5381 See Documentation/arch/x86/i386/IO-APIC.rst. 5382 5383 plip= [PPT,NET] Parallel port network link 5384 Format: { parport<nr> | timid | 0 } 5385 See also Documentation/admin-guide/parport.rst. 5386 5387 pmtmr= [X86] Manual setup of pmtmr I/O Port. 5388 Override pmtimer IOPort with a hex value. 5389 e.g. pmtmr=0x508 5390 5391 pmu_override= [PPC] Override the PMU. 5392 This option takes over the PMU facility, so it is no 5393 longer usable by perf. Setting this option starts the 5394 PMU counters by setting MMCR0 to 0 (the FC bit is 5395 cleared). If a number is given, then MMCR1 is set to 5396 that number, otherwise (e.g., 'pmu_override=on'), MMCR1 5397 remains 0. 5398 5399 pm_async= [PM] 5400 Format: off 5401 This parameter sets the initial value of the 5402 /sys/power/pm_async sysfs knob at boot time. 5403 If set to "off", disables asynchronous suspend and 5404 resume of devices during system-wide power transitions. 5405 This can be useful on platforms where device 5406 dependencies are not well-defined, or for debugging 5407 power management issues. Asynchronous operations are 5408 enabled by default. 5409 5410 5411 pm_debug_messages [SUSPEND,KNL] 5412 Enable suspend/resume debug messages during boot up. 5413 5414 pnp.debug=1 [PNP] 5415 Enable PNP debug messages (depends on the 5416 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 5417 via /sys/module/pnp/parameters/debug. We always show 5418 current resource usage; turning this on also shows 5419 possible settings and some assignment information. 5420 5421 pnpacpi= [ACPI] 5422 { off } 5423 5424 pnpbios= [ISAPNP] 5425 { on | off | curr | res | no-curr | no-res } 5426 5427 pnp_reserve_irq= 5428 [ISAPNP] Exclude IRQs for the autoconfiguration 5429 5430 pnp_reserve_dma= 5431 [ISAPNP] Exclude DMAs for the autoconfiguration 5432 5433 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 5434 Ranges are in pairs (I/O port base and size). 5435 5436 pnp_reserve_mem= 5437 [ISAPNP] Exclude memory regions for the 5438 autoconfiguration. 5439 Ranges are in pairs (memory base and size). 5440 5441 ports= [IP_VS_FTP] IPVS ftp helper module 5442 Default is 21. 5443 Up to 8 (IP_VS_APP_MAX_PORTS) ports 5444 may be specified. 5445 Format: <port>,<port>.... 5446 5447 possible_cpus= [SMP,S390,X86] 5448 Format: <unsigned int> 5449 Set the number of possible CPUs, overriding the 5450 regular discovery mechanisms (such as ACPI/FW, etc). 5451 5452 powersave=off [PPC] This option disables power saving features. 5453 It specifically disables cpuidle and sets the 5454 platform machine description specific power_save 5455 function to NULL. On Idle the CPU just reduces 5456 execution priority. 5457 5458 ppc_strict_facility_enable 5459 [PPC,ENABLE] This option catches any kernel floating point, 5460 Altivec, VSX and SPE outside of regions specifically 5461 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 5462 There is some performance impact when enabling this. 5463 5464 ppc_tm= [PPC,EARLY] 5465 Format: {"off"} 5466 Disable Hardware Transactional Memory 5467 5468 preempt= [KNL] 5469 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC 5470 none - Limited to cond_resched() calls 5471 voluntary - Limited to cond_resched() and might_sleep() calls 5472 full - Any section that isn't explicitly preempt disabled 5473 can be preempted anytime. Tasks will also yield 5474 contended spinlocks (if the critical section isn't 5475 explicitly preempt disabled beyond the lock itself). 5476 lazy - Scheduler controlled. Similar to full but instead 5477 of preempting the task immediately, the task gets 5478 one HZ tick time to yield itself before the 5479 preemption will be forced. One preemption is when the 5480 task returns to user space. 5481 5482 print-fatal-signals= 5483 [KNL] debug: print fatal signals 5484 5485 If enabled, warn about various signal handling 5486 related application anomalies: too many signals, 5487 too many POSIX.1 timers, fatal signals causing a 5488 coredump - etc. 5489 5490 If you hit the warning due to signal overflow, 5491 you might want to try "ulimit -i unlimited". 5492 5493 default: off. 5494 5495 printk.always_kmsg_dump= 5496 Trigger kmsg_dump for cases other than kernel oops or 5497 panics 5498 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5499 default: disabled 5500 5501 printk.console_no_auto_verbose= 5502 Disable console loglevel raise on oops, panic 5503 or lockdep-detected issues (only if lock debug is on). 5504 With an exception to setups with low baudrate on 5505 serial console, keeping this 0 is a good choice 5506 in order to provide more debug information. 5507 Format: <bool> 5508 default: 0 (auto_verbose is enabled) 5509 5510 printk.debug_non_panic_cpus= 5511 Allows storing messages from non-panic CPUs into 5512 the printk log buffer during panic(). They are 5513 flushed to consoles by the panic-CPU on 5514 a best-effort basis. 5515 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5516 Default: disabled 5517 5518 printk.devkmsg={on,off,ratelimit} 5519 Control writing to /dev/kmsg. 5520 on - unlimited logging to /dev/kmsg from userspace 5521 off - logging to /dev/kmsg disabled 5522 ratelimit - ratelimit the logging 5523 Default: ratelimit 5524 5525 printk.time= Show timing data prefixed to each printk message line 5526 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5527 5528 proc_mem.force_override= [KNL] 5529 Format: {always | ptrace | never} 5530 Traditionally /proc/pid/mem allows memory permissions to be 5531 overridden without restrictions. This option may be set to 5532 restrict that. Can be one of: 5533 - 'always': traditional behavior always allows mem overrides. 5534 - 'ptrace': only allow mem overrides for active ptracers. 5535 - 'never': never allow mem overrides. 5536 If not specified, default is the CONFIG_PROC_MEM_* choice. 5537 5538 processor.max_cstate= [HW,ACPI] 5539 Limit processor to maximum C-state 5540 max_cstate=9 overrides any DMI blacklist limit. 5541 5542 processor.nocst [HW,ACPI] 5543 Ignore the _CST method to determine C-states, 5544 instead using the legacy FADT method 5545 5546 profile= [KNL] Enable kernel profiling via /proc/profile 5547 Format: [<profiletype>,]<number> 5548 Param: <profiletype>: "schedule" or "kvm" 5549 [defaults to kernel profiling] 5550 Param: "schedule" - profile schedule points. 5551 Param: "kvm" - profile VM exits. 5552 Param: <number> - step/bucket size as a power of 2 for 5553 statistical time based profiling. 5554 5555 prot_virt= [S390] enable hosting protected virtual machines 5556 isolated from the hypervisor (if hardware supports 5557 that). If enabled, the default kernel base address 5558 might be overridden even when Kernel Address Space 5559 Layout Randomization is disabled. 5560 Format: <bool> 5561 5562 psi= [KNL] Enable or disable pressure stall information 5563 tracking. 5564 Format: <bool> 5565 5566 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 5567 probe for; one of (bare|imps|exps|lifebook|any). 5568 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 5569 per second. 5570 psmouse.resetafter= [HW,MOUSE] 5571 Try to reset the device after so many bad packets 5572 (0 = never). 5573 psmouse.resolution= 5574 [HW,MOUSE] Set desired mouse resolution, in dpi. 5575 psmouse.smartscroll= 5576 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 5577 0 = disabled, 1 = enabled (default). 5578 5579 pstore.backend= Specify the name of the pstore backend to use 5580 5581 pti= [X86-64] Control Page Table Isolation of user and 5582 kernel address spaces. Disabling this feature 5583 removes hardening, but improves performance of 5584 system calls and interrupts. 5585 5586 on - unconditionally enable 5587 off - unconditionally disable 5588 auto - kernel detects whether your CPU model is 5589 vulnerable to issues that PTI mitigates 5590 5591 Not specifying this option is equivalent to pti=auto. 5592 5593 pty.legacy_count= 5594 [KNL] Number of legacy pty's. Overwrites compiled-in 5595 default number. 5596 5597 quiet [KNL,EARLY] Disable most log messages 5598 5599 r128= [HW,DRM] 5600 5601 radix_hcall_invalidate=on [PPC/PSERIES] 5602 Disable RADIX GTSE feature and use hcall for TLB 5603 invalidate. 5604 5605 raid= [HW,RAID] 5606 See Documentation/admin-guide/md.rst. 5607 5608 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 5609 See Documentation/admin-guide/blockdev/ramdisk.rst. 5610 5611 ramdisk_start= [Deprecated,RAM] RAM disk image start address 5612 5613 random.trust_cpu=off 5614 [KNL,EARLY] Disable trusting the use of the CPU's 5615 random number generator (if available) to 5616 initialize the kernel's RNG. 5617 5618 random.trust_bootloader=off 5619 [KNL,EARLY] Disable trusting the use of the a seed 5620 passed by the bootloader (if available) to 5621 initialize the kernel's RNG. 5622 5623 randomize_kstack_offset= 5624 [KNL,EARLY] Enable or disable kernel stack offset 5625 randomization, which provides roughly 5 bits of 5626 entropy, frustrating memory corruption attacks 5627 that depend on stack address determinism or 5628 cross-syscall address exposures. This is only 5629 available on architectures that have defined 5630 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET. 5631 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5632 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT. 5633 5634 ras=option[,option,...] [KNL] RAS-specific options 5635 5636 cec_disable [X86] 5637 Disable the Correctable Errors Collector, 5638 see CONFIG_RAS_CEC help text. 5639 5640 rcu_nocbs[=cpu-list] 5641 [KNL] The optional argument is a cpu list, 5642 as described above. 5643 5644 In kernels built with CONFIG_RCU_NOCB_CPU=y, 5645 enable the no-callback CPU mode, which prevents 5646 such CPUs' callbacks from being invoked in 5647 softirq context. Invocation of such CPUs' RCU 5648 callbacks will instead be offloaded to "rcuox/N" 5649 kthreads created for that purpose, where "x" is 5650 "p" for RCU-preempt, "s" for RCU-sched, and "g" 5651 for the kthreads that mediate grace periods; and 5652 "N" is the CPU number. This reduces OS jitter on 5653 the offloaded CPUs, which can be useful for HPC 5654 and real-time workloads. It can also improve 5655 energy efficiency for asymmetric multiprocessors. 5656 5657 If a cpulist is passed as an argument, the specified 5658 list of CPUs is set to no-callback mode from boot. 5659 5660 Otherwise, if the '=' sign and the cpulist 5661 arguments are omitted, no CPU will be set to 5662 no-callback mode from boot but the mode may be 5663 toggled at runtime via cpusets. 5664 5665 Note that this argument takes precedence over 5666 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 5667 5668 rcu_nocb_poll [KNL] 5669 Rather than requiring that offloaded CPUs 5670 (specified by rcu_nocbs= above) explicitly 5671 awaken the corresponding "rcuoN" kthreads, 5672 make these kthreads poll for callbacks. 5673 This improves the real-time response for the 5674 offloaded CPUs by relieving them of the need to 5675 wake up the corresponding kthread, but degrades 5676 energy efficiency by requiring that the kthreads 5677 periodically wake up to do the polling. 5678 5679 rcutree.blimit= [KNL] 5680 Set maximum number of finished RCU callbacks to 5681 process in one batch. 5682 5683 rcutree.csd_lock_suppress_rcu_stall= [KNL] 5684 Do only a one-line RCU CPU stall warning when 5685 there is an ongoing too-long CSD-lock wait. 5686 5687 rcutree.do_rcu_barrier= [KNL] 5688 Request a call to rcu_barrier(). This is 5689 throttled so that userspace tests can safely 5690 hammer on the sysfs variable if they so choose. 5691 If triggered before the RCU grace-period machinery 5692 is fully active, this will error out with EAGAIN. 5693 5694 rcutree.dump_tree= [KNL] 5695 Dump the structure of the rcu_node combining tree 5696 out at early boot. This is used for diagnostic 5697 purposes, to verify correct tree setup. 5698 5699 rcutree.gp_cleanup_delay= [KNL] 5700 Set the number of jiffies to delay each step of 5701 RCU grace-period cleanup. 5702 5703 rcutree.gp_init_delay= [KNL] 5704 Set the number of jiffies to delay each step of 5705 RCU grace-period initialization. 5706 5707 rcutree.gp_preinit_delay= [KNL] 5708 Set the number of jiffies to delay each step of 5709 RCU grace-period pre-initialization, that is, 5710 the propagation of recent CPU-hotplug changes up 5711 the rcu_node combining tree. 5712 5713 rcutree.jiffies_till_first_fqs= [KNL] 5714 Set delay from grace-period initialization to 5715 first attempt to force quiescent states. 5716 Units are jiffies, minimum value is zero, 5717 and maximum value is HZ. 5718 5719 rcutree.jiffies_till_next_fqs= [KNL] 5720 Set delay between subsequent attempts to force 5721 quiescent states. Units are jiffies, minimum 5722 value is one, and maximum value is HZ. 5723 5724 rcutree.jiffies_till_sched_qs= [KNL] 5725 Set required age in jiffies for a 5726 given grace period before RCU starts 5727 soliciting quiescent-state help from 5728 rcu_note_context_switch() and cond_resched(). 5729 If not specified, the kernel will calculate 5730 a value based on the most recent settings 5731 of rcutree.jiffies_till_first_fqs 5732 and rcutree.jiffies_till_next_fqs. 5733 This calculated value may be viewed in 5734 rcutree.jiffies_to_sched_qs. Any attempt to set 5735 rcutree.jiffies_to_sched_qs will be cheerfully 5736 overwritten. 5737 5738 rcutree.kthread_prio= [KNL,BOOT] 5739 Set the SCHED_FIFO priority of the RCU per-CPU 5740 kthreads (rcuc/N). This value is also used for 5741 the priority of the RCU boost threads (rcub/N) 5742 and for the RCU grace-period kthreads (rcu_bh, 5743 rcu_preempt, and rcu_sched). If RCU_BOOST is 5744 set, valid values are 1-99 and the default is 1 5745 (the least-favored priority). Otherwise, when 5746 RCU_BOOST is not set, valid values are 0-99 and 5747 the default is zero (non-realtime operation). 5748 When RCU_NOCB_CPU is set, also adjust the 5749 priority of NOCB callback kthreads. 5750 5751 rcutree.nocb_nobypass_lim_per_jiffy= [KNL] 5752 On callback-offloaded (rcu_nocbs) CPUs, 5753 RCU reduces the lock contention that would 5754 otherwise be caused by callback floods through 5755 use of the ->nocb_bypass list. However, in the 5756 common non-flooded case, RCU queues directly to 5757 the main ->cblist in order to avoid the extra 5758 overhead of the ->nocb_bypass list and its lock. 5759 But if there are too many callbacks queued during 5760 a single jiffy, RCU pre-queues the callbacks into 5761 the ->nocb_bypass queue. The definition of "too 5762 many" is supplied by this kernel boot parameter. 5763 5764 rcutree.nohz_full_patience_delay= [KNL] 5765 On callback-offloaded (rcu_nocbs) CPUs, avoid 5766 disturbing RCU unless the grace period has 5767 reached the specified age in milliseconds. 5768 Defaults to zero. Large values will be capped 5769 at five seconds. All values will be rounded down 5770 to the nearest value representable by jiffies. 5771 5772 rcutree.qhimark= [KNL] 5773 Set threshold of queued RCU callbacks beyond which 5774 batch limiting is disabled. 5775 5776 rcutree.qlowmark= [KNL] 5777 Set threshold of queued RCU callbacks below which 5778 batch limiting is re-enabled. 5779 5780 rcutree.qovld= [KNL] 5781 Set threshold of queued RCU callbacks beyond which 5782 RCU's force-quiescent-state scan will aggressively 5783 enlist help from cond_resched() and sched IPIs to 5784 help CPUs more quickly reach quiescent states. 5785 Set to less than zero to make this be set based 5786 on rcutree.qhimark at boot time and to zero to 5787 disable more aggressive help enlistment. 5788 5789 rcutree.rcu_delay_page_cache_fill_msec= [KNL] 5790 Set the page-cache refill delay (in milliseconds) 5791 in response to low-memory conditions. The range 5792 of permitted values is in the range 0:100000. 5793 5794 rcutree.rcu_divisor= [KNL] 5795 Set the shift-right count to use to compute 5796 the callback-invocation batch limit bl from 5797 the number of callbacks queued on this CPU. 5798 The result will be bounded below by the value of 5799 the rcutree.blimit kernel parameter. Every bl 5800 callbacks, the softirq handler will exit in 5801 order to allow the CPU to do other work. 5802 5803 Please note that this callback-invocation batch 5804 limit applies only to non-offloaded callback 5805 invocation. Offloaded callbacks are instead 5806 invoked in the context of an rcuoc kthread, which 5807 scheduler will preempt as it does any other task. 5808 5809 rcutree.rcu_fanout_exact= [KNL] 5810 Disable autobalancing of the rcu_node combining 5811 tree. This is used by rcutorture, and might 5812 possibly be useful for architectures having high 5813 cache-to-cache transfer latencies. 5814 5815 rcutree.rcu_fanout_leaf= [KNL] 5816 Change the number of CPUs assigned to each 5817 leaf rcu_node structure. Useful for very 5818 large systems, which will choose the value 64, 5819 and for NUMA systems with large remote-access 5820 latencies, which will choose a value aligned 5821 with the appropriate hardware boundaries. 5822 5823 rcutree.rcu_min_cached_objs= [KNL] 5824 Minimum number of objects which are cached and 5825 maintained per one CPU. Object size is equal 5826 to PAGE_SIZE. The cache allows to reduce the 5827 pressure to page allocator, also it makes the 5828 whole algorithm to behave better in low memory 5829 condition. 5830 5831 rcutree.rcu_nocb_gp_stride= [KNL] 5832 Set the number of NOCB callback kthreads in 5833 each group, which defaults to the square root 5834 of the number of CPUs. Larger numbers reduce 5835 the wakeup overhead on the global grace-period 5836 kthread, but increases that same overhead on 5837 each group's NOCB grace-period kthread. 5838 5839 rcutree.rcu_kick_kthreads= [KNL] 5840 Cause the grace-period kthread to get an extra 5841 wake_up() if it sleeps three times longer than 5842 it should at force-quiescent-state time. 5843 This wake_up() will be accompanied by a 5844 WARN_ONCE() splat and an ftrace_dump(). 5845 5846 rcutree.rcu_resched_ns= [KNL] 5847 Limit the time spend invoking a batch of RCU 5848 callbacks to the specified number of nanoseconds. 5849 By default, this limit is checked only once 5850 every 32 callbacks in order to limit the pain 5851 inflicted by local_clock() overhead. 5852 5853 rcutree.rcu_unlock_delay= [KNL] 5854 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 5855 this specifies an rcu_read_unlock()-time delay 5856 in microseconds. This defaults to zero. 5857 Larger delays increase the probability of 5858 catching RCU pointer leaks, that is, buggy use 5859 of RCU-protected pointers after the relevant 5860 rcu_read_unlock() has completed. 5861 5862 rcutree.sysrq_rcu= [KNL] 5863 Commandeer a sysrq key to dump out Tree RCU's 5864 rcu_node tree with an eye towards determining 5865 why a new grace period has not yet started. 5866 5867 rcutree.use_softirq= [KNL] 5868 If set to zero, move all RCU_SOFTIRQ processing to 5869 per-CPU rcuc kthreads. Defaults to a non-zero 5870 value, meaning that RCU_SOFTIRQ is used by default. 5871 Specify rcutree.use_softirq=0 to use rcuc kthreads. 5872 5873 But note that CONFIG_PREEMPT_RT=y kernels disable 5874 this kernel boot parameter, forcibly setting it 5875 to zero. 5876 5877 rcutree.enable_rcu_lazy= [KNL] 5878 To save power, batch RCU callbacks and flush after 5879 delay, memory pressure or callback list growing too 5880 big. 5881 5882 rcutree.rcu_normal_wake_from_gp= [KNL] 5883 Reduces a latency of synchronize_rcu() call. This approach 5884 maintains its own track of synchronize_rcu() callers, so it 5885 does not interact with regular callbacks because it does not 5886 use a call_rcu[_hurry]() path. Please note, this is for a 5887 normal grace period. 5888 5889 How to enable it: 5890 5891 echo 1 > /sys/module/rcutree/parameters/rcu_normal_wake_from_gp 5892 or pass a boot parameter "rcutree.rcu_normal_wake_from_gp=1" 5893 5894 Default is 1 if num_possible_cpus() <= 16 and it is not explicitly 5895 disabled by the boot parameter passing 0. 5896 5897 rcuscale.gp_async= [KNL] 5898 Measure performance of asynchronous 5899 grace-period primitives such as call_rcu(). 5900 5901 rcuscale.gp_async_max= [KNL] 5902 Specify the maximum number of outstanding 5903 callbacks per writer thread. When a writer 5904 thread exceeds this limit, it invokes the 5905 corresponding flavor of rcu_barrier() to allow 5906 previously posted callbacks to drain. 5907 5908 rcuscale.gp_exp= [KNL] 5909 Measure performance of expedited synchronous 5910 grace-period primitives. 5911 5912 rcuscale.holdoff= [KNL] 5913 Set test-start holdoff period. The purpose of 5914 this parameter is to delay the start of the 5915 test until boot completes in order to avoid 5916 interference. 5917 5918 rcuscale.kfree_by_call_rcu= [KNL] 5919 In kernels built with CONFIG_RCU_LAZY=y, test 5920 call_rcu() instead of kfree_rcu(). 5921 5922 rcuscale.kfree_mult= [KNL] 5923 Instead of allocating an object of size kfree_obj, 5924 allocate one of kfree_mult * sizeof(kfree_obj). 5925 Defaults to 1. 5926 5927 rcuscale.kfree_rcu_test= [KNL] 5928 Set to measure performance of kfree_rcu() flooding. 5929 5930 rcuscale.kfree_rcu_test_double= [KNL] 5931 Test the double-argument variant of kfree_rcu(). 5932 If this parameter has the same value as 5933 rcuscale.kfree_rcu_test_single, both the single- 5934 and double-argument variants are tested. 5935 5936 rcuscale.kfree_rcu_test_single= [KNL] 5937 Test the single-argument variant of kfree_rcu(). 5938 If this parameter has the same value as 5939 rcuscale.kfree_rcu_test_double, both the single- 5940 and double-argument variants are tested. 5941 5942 rcuscale.kfree_nthreads= [KNL] 5943 The number of threads running loops of kfree_rcu(). 5944 5945 rcuscale.kfree_alloc_num= [KNL] 5946 Number of allocations and frees done in an iteration. 5947 5948 rcuscale.kfree_loops= [KNL] 5949 Number of loops doing rcuscale.kfree_alloc_num number 5950 of allocations and frees. 5951 5952 rcuscale.minruntime= [KNL] 5953 Set the minimum test run time in seconds. This 5954 does not affect the data-collection interval, 5955 but instead allows better measurement of things 5956 like CPU consumption. 5957 5958 rcuscale.nreaders= [KNL] 5959 Set number of RCU readers. The value -1 selects 5960 N, where N is the number of CPUs. A value 5961 "n" less than -1 selects N-n+1, where N is again 5962 the number of CPUs. For example, -2 selects N 5963 (the number of CPUs), -3 selects N+1, and so on. 5964 A value of "n" less than or equal to -N selects 5965 a single reader. 5966 5967 rcuscale.nwriters= [KNL] 5968 Set number of RCU writers. The values operate 5969 the same as for rcuscale.nreaders. 5970 N, where N is the number of CPUs 5971 5972 rcuscale.scale_type= [KNL] 5973 Specify the RCU implementation to test. 5974 5975 rcuscale.shutdown= [KNL] 5976 Shut the system down after performance tests 5977 complete. This is useful for hands-off automated 5978 testing. 5979 5980 rcuscale.verbose= [KNL] 5981 Enable additional printk() statements. 5982 5983 rcuscale.writer_holdoff= [KNL] 5984 Write-side holdoff between grace periods, 5985 in microseconds. The default of zero says 5986 no holdoff. 5987 5988 rcuscale.writer_holdoff_jiffies= [KNL] 5989 Additional write-side holdoff between grace 5990 periods, but in jiffies. The default of zero 5991 says no holdoff. 5992 5993 rcutorture.fqs_duration= [KNL] 5994 Set duration of force_quiescent_state bursts 5995 in microseconds. 5996 5997 rcutorture.fqs_holdoff= [KNL] 5998 Set holdoff time within force_quiescent_state bursts 5999 in microseconds. 6000 6001 rcutorture.fqs_stutter= [KNL] 6002 Set wait time between force_quiescent_state bursts 6003 in seconds. 6004 6005 rcutorture.fwd_progress= [KNL] 6006 Specifies the number of kthreads to be used 6007 for RCU grace-period forward-progress testing 6008 for the types of RCU supporting this notion. 6009 Defaults to 1 kthread, values less than zero or 6010 greater than the number of CPUs cause the number 6011 of CPUs to be used. 6012 6013 rcutorture.fwd_progress_div= [KNL] 6014 Specify the fraction of a CPU-stall-warning 6015 period to do tight-loop forward-progress testing. 6016 6017 rcutorture.fwd_progress_holdoff= [KNL] 6018 Number of seconds to wait between successive 6019 forward-progress tests. 6020 6021 rcutorture.fwd_progress_need_resched= [KNL] 6022 Enclose cond_resched() calls within checks for 6023 need_resched() during tight-loop forward-progress 6024 testing. 6025 6026 rcutorture.gp_cond= [KNL] 6027 Use conditional/asynchronous update-side 6028 normal-grace-period primitives, if available. 6029 6030 rcutorture.gp_cond_exp= [KNL] 6031 Use conditional/asynchronous update-side 6032 expedited-grace-period primitives, if available. 6033 6034 rcutorture.gp_cond_full= [KNL] 6035 Use conditional/asynchronous update-side 6036 normal-grace-period primitives that also take 6037 concurrent expedited grace periods into account, 6038 if available. 6039 6040 rcutorture.gp_cond_exp_full= [KNL] 6041 Use conditional/asynchronous update-side 6042 expedited-grace-period primitives that also take 6043 concurrent normal grace periods into account, 6044 if available. 6045 6046 rcutorture.gp_cond_wi= [KNL] 6047 Nominal wait interval for normal conditional 6048 grace periods (specified by rcutorture's 6049 gp_cond and gp_cond_full module parameters), 6050 in microseconds. The actual wait interval will 6051 be randomly selected to nanosecond granularity up 6052 to this wait interval. Defaults to 16 jiffies, 6053 for example, 16,000 microseconds on a system 6054 with HZ=1000. 6055 6056 rcutorture.gp_cond_wi_exp= [KNL] 6057 Nominal wait interval for expedited conditional 6058 grace periods (specified by rcutorture's 6059 gp_cond_exp and gp_cond_exp_full module 6060 parameters), in microseconds. The actual wait 6061 interval will be randomly selected to nanosecond 6062 granularity up to this wait interval. Defaults to 6063 128 microseconds. 6064 6065 rcutorture.gp_exp= [KNL] 6066 Use expedited update-side primitives, if available. 6067 6068 rcutorture.gp_normal= [KNL] 6069 Use normal (non-expedited) asynchronous 6070 update-side primitives, if available. 6071 6072 rcutorture.gp_poll= [KNL] 6073 Use polled update-side normal-grace-period 6074 primitives, if available. 6075 6076 rcutorture.gp_poll_exp= [KNL] 6077 Use polled update-side expedited-grace-period 6078 primitives, if available. 6079 6080 rcutorture.gp_poll_full= [KNL] 6081 Use polled update-side normal-grace-period 6082 primitives that also take concurrent expedited 6083 grace periods into account, if available. 6084 6085 rcutorture.gp_poll_exp_full= [KNL] 6086 Use polled update-side expedited-grace-period 6087 primitives that also take concurrent normal 6088 grace periods into account, if available. 6089 6090 rcutorture.gp_poll_wi= [KNL] 6091 Nominal wait interval for normal conditional 6092 grace periods (specified by rcutorture's 6093 gp_poll and gp_poll_full module parameters), 6094 in microseconds. The actual wait interval will 6095 be randomly selected to nanosecond granularity up 6096 to this wait interval. Defaults to 16 jiffies, 6097 for example, 16,000 microseconds on a system 6098 with HZ=1000. 6099 6100 rcutorture.gp_poll_wi_exp= [KNL] 6101 Nominal wait interval for expedited conditional 6102 grace periods (specified by rcutorture's 6103 gp_poll_exp and gp_poll_exp_full module 6104 parameters), in microseconds. The actual wait 6105 interval will be randomly selected to nanosecond 6106 granularity up to this wait interval. Defaults to 6107 128 microseconds. 6108 6109 rcutorture.gp_sync= [KNL] 6110 Use normal (non-expedited) synchronous 6111 update-side primitives, if available. If all 6112 of rcutorture.gp_cond=, rcutorture.gp_exp=, 6113 rcutorture.gp_normal=, and rcutorture.gp_sync= 6114 are zero, rcutorture acts as if is interpreted 6115 they are all non-zero. 6116 6117 rcutorture.gpwrap_lag= [KNL] 6118 Enable grace-period wrap lag testing. Setting 6119 to false prevents the gpwrap lag test from 6120 running. Default is true. 6121 6122 rcutorture.gpwrap_lag_gps= [KNL] 6123 Set the value for grace-period wrap lag during 6124 active lag testing periods. This controls how many 6125 grace periods differences we tolerate between 6126 rdp and rnp's gp_seq before setting overflow flag. 6127 The default is always set to 8. 6128 6129 rcutorture.gpwrap_lag_cycle_mins= [KNL] 6130 Set the total cycle duration for gpwrap lag 6131 testing in minutes. This is the total time for 6132 one complete cycle of active and inactive 6133 testing periods. Default is 30 minutes. 6134 6135 rcutorture.gpwrap_lag_active_mins= [KNL] 6136 Set the duration for which gpwrap lag is active 6137 within each cycle, in minutes. During this time, 6138 the grace-period wrap lag will be set to the 6139 value specified by gpwrap_lag_gps. Default is 6140 5 minutes. 6141 6142 rcutorture.irqreader= [KNL] 6143 Run RCU readers from irq handlers, or, more 6144 accurately, from a timer handler. Not all RCU 6145 flavors take kindly to this sort of thing. 6146 6147 rcutorture.leakpointer= [KNL] 6148 Leak an RCU-protected pointer out of the reader. 6149 This can of course result in splats, and is 6150 intended to test the ability of things like 6151 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 6152 such leaks. 6153 6154 rcutorture.n_barrier_cbs= [KNL] 6155 Set callbacks/threads for rcu_barrier() testing. 6156 6157 rcutorture.nfakewriters= [KNL] 6158 Set number of concurrent RCU writers. These just 6159 stress RCU, they don't participate in the actual 6160 test, hence the "fake". 6161 6162 rcutorture.nocbs_nthreads= [KNL] 6163 Set number of RCU callback-offload togglers. 6164 Zero (the default) disables toggling. 6165 6166 rcutorture.nocbs_toggle= [KNL] 6167 Set the delay in milliseconds between successive 6168 callback-offload toggling attempts. 6169 6170 rcutorture.nreaders= [KNL] 6171 Set number of RCU readers. The value -1 selects 6172 N-1, where N is the number of CPUs. A value 6173 "n" less than -1 selects N-n-2, where N is again 6174 the number of CPUs. For example, -2 selects N 6175 (the number of CPUs), -3 selects N+1, and so on. 6176 6177 rcutorture.object_debug= [KNL] 6178 Enable debug-object double-call_rcu() testing. 6179 6180 rcutorture.onoff_holdoff= [KNL] 6181 Set time (s) after boot for CPU-hotplug testing. 6182 6183 rcutorture.onoff_interval= [KNL] 6184 Set time (jiffies) between CPU-hotplug operations, 6185 or zero to disable CPU-hotplug testing. 6186 6187 rcutorture.preempt_duration= [KNL] 6188 Set duration (in milliseconds) of preemptions 6189 by a high-priority FIFO real-time task. Set to 6190 zero (the default) to disable. The CPUs to 6191 preempt are selected randomly from the set that 6192 are online at a given point in time. Races with 6193 CPUs going offline are ignored, with that attempt 6194 at preemption skipped. 6195 6196 rcutorture.preempt_interval= [KNL] 6197 Set interval (in milliseconds, defaulting to one 6198 second) between preemptions by a high-priority 6199 FIFO real-time task. This delay is mediated 6200 by an hrtimer and is further fuzzed to avoid 6201 inadvertent synchronizations. 6202 6203 rcutorture.read_exit_burst= [KNL] 6204 The number of times in a given read-then-exit 6205 episode that a set of read-then-exit kthreads 6206 is spawned. 6207 6208 rcutorture.read_exit_delay= [KNL] 6209 The delay, in seconds, between successive 6210 read-then-exit testing episodes. 6211 6212 rcutorture.reader_flavor= [KNL] 6213 A bit mask indicating which readers to use. 6214 If there is more than one bit set, the readers 6215 are entered from low-order bit up, and are 6216 exited in the opposite order. For SRCU, the 6217 0x1 bit is normal readers, 0x2 NMI-safe readers, 6218 and 0x4 light-weight readers. 6219 6220 rcutorture.shuffle_interval= [KNL] 6221 Set task-shuffle interval (s). Shuffling tasks 6222 allows some CPUs to go into dyntick-idle mode 6223 during the rcutorture test. 6224 6225 rcutorture.shutdown_secs= [KNL] 6226 Set time (s) after boot system shutdown. This 6227 is useful for hands-off automated testing. 6228 6229 rcutorture.stall_cpu= [KNL] 6230 Duration of CPU stall (s) to test RCU CPU stall 6231 warnings, zero to disable. 6232 6233 rcutorture.stall_cpu_block= [KNL] 6234 Sleep while stalling if set. This will result 6235 in warnings from preemptible RCU in addition to 6236 any other stall-related activity. Note that 6237 in kernels built with CONFIG_PREEMPTION=n and 6238 CONFIG_PREEMPT_COUNT=y, this parameter will 6239 cause the CPU to pass through a quiescent state. 6240 Given CONFIG_PREEMPTION=n, this will suppress 6241 RCU CPU stall warnings, but will instead result 6242 in scheduling-while-atomic splats. 6243 6244 Use of this module parameter results in splats. 6245 6246 6247 rcutorture.stall_cpu_holdoff= [KNL] 6248 Time to wait (s) after boot before inducing stall. 6249 6250 rcutorture.stall_cpu_irqsoff= [KNL] 6251 Disable interrupts while stalling if set, but only 6252 on the first stall in the set. 6253 6254 rcutorture.stall_cpu_repeat= [KNL] 6255 Number of times to repeat the stall sequence, 6256 so that rcutorture.stall_cpu_repeat=3 will result 6257 in four stall sequences. 6258 6259 rcutorture.stall_gp_kthread= [KNL] 6260 Duration (s) of forced sleep within RCU 6261 grace-period kthread to test RCU CPU stall 6262 warnings, zero to disable. If both stall_cpu 6263 and stall_gp_kthread are specified, the 6264 kthread is starved first, then the CPU. 6265 6266 rcutorture.stat_interval= [KNL] 6267 Time (s) between statistics printk()s. 6268 6269 rcutorture.stutter= [KNL] 6270 Time (s) to stutter testing, for example, specifying 6271 five seconds causes the test to run for five seconds, 6272 wait for five seconds, and so on. This tests RCU's 6273 ability to transition abruptly to and from idle. 6274 6275 rcutorture.test_boost= [KNL] 6276 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 6277 "Maybe" means test if the RCU implementation 6278 under test support RCU priority boosting. 6279 6280 rcutorture.test_boost_duration= [KNL] 6281 Duration (s) of each individual boost test. 6282 6283 rcutorture.test_boost_holdoff= [KNL] 6284 Holdoff time (s) from start of test to the start 6285 of RCU priority-boost testing. Defaults to zero, 6286 that is, no holdoff. 6287 6288 rcutorture.test_boost_interval= [KNL] 6289 Interval (s) between each boost test. 6290 6291 rcutorture.test_no_idle_hz= [KNL] 6292 Test RCU's dyntick-idle handling. See also the 6293 rcutorture.shuffle_interval parameter. 6294 6295 rcutorture.torture_type= [KNL] 6296 Specify the RCU implementation to test. 6297 6298 rcutorture.verbose= [KNL] 6299 Enable additional printk() statements. 6300 6301 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 6302 Dump ftrace buffer after reporting RCU CPU 6303 stall warning. 6304 6305 rcupdate.rcu_cpu_stall_notifiers= [KNL] 6306 Provide RCU CPU stall notifiers, but see the 6307 warnings in the RCU_CPU_STALL_NOTIFIER Kconfig 6308 option's help text. TL;DR: You almost certainly 6309 do not want rcupdate.rcu_cpu_stall_notifiers. 6310 6311 rcupdate.rcu_cpu_stall_suppress= [KNL] 6312 Suppress RCU CPU stall warning messages. 6313 6314 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 6315 Suppress RCU CPU stall warning messages and 6316 rcutorture writer stall warnings that occur 6317 during early boot, that is, during the time 6318 before the init task is spawned. 6319 6320 rcupdate.rcu_cpu_stall_timeout= [KNL] 6321 Set timeout for RCU CPU stall warning messages. 6322 The value is in seconds and the maximum allowed 6323 value is 300 seconds. 6324 6325 rcupdate.rcu_exp_cpu_stall_timeout= [KNL] 6326 Set timeout for expedited RCU CPU stall warning 6327 messages. The value is in milliseconds 6328 and the maximum allowed value is 21000 6329 milliseconds. Please note that this value is 6330 adjusted to an arch timer tick resolution. 6331 Setting this to zero causes the value from 6332 rcupdate.rcu_cpu_stall_timeout to be used (after 6333 conversion from seconds to milliseconds). 6334 6335 rcupdate.rcu_cpu_stall_cputime= [KNL] 6336 Provide statistics on the cputime and count of 6337 interrupts and tasks during the sampling period. For 6338 multiple continuous RCU stalls, all sampling periods 6339 begin at half of the first RCU stall timeout. 6340 6341 rcupdate.rcu_exp_stall_task_details= [KNL] 6342 Print stack dumps of any tasks blocking the 6343 current expedited RCU grace period during an 6344 expedited RCU CPU stall warning. 6345 6346 rcupdate.rcu_expedited= [KNL] 6347 Use expedited grace-period primitives, for 6348 example, synchronize_rcu_expedited() instead 6349 of synchronize_rcu(). This reduces latency, 6350 but can increase CPU utilization, degrade 6351 real-time latency, and degrade energy efficiency. 6352 No effect on CONFIG_TINY_RCU kernels. 6353 6354 rcupdate.rcu_normal= [KNL] 6355 Use only normal grace-period primitives, 6356 for example, synchronize_rcu() instead of 6357 synchronize_rcu_expedited(). This improves 6358 real-time latency, CPU utilization, and 6359 energy efficiency, but can expose users to 6360 increased grace-period latency. This parameter 6361 overrides rcupdate.rcu_expedited. No effect on 6362 CONFIG_TINY_RCU kernels. 6363 6364 rcupdate.rcu_normal_after_boot= [KNL] 6365 Once boot has completed (that is, after 6366 rcu_end_inkernel_boot() has been invoked), use 6367 only normal grace-period primitives. No effect 6368 on CONFIG_TINY_RCU kernels. 6369 6370 But note that CONFIG_PREEMPT_RT=y kernels enables 6371 this kernel boot parameter, forcibly setting 6372 it to the value one, that is, converting any 6373 post-boot attempt at an expedited RCU grace 6374 period to instead use normal non-expedited 6375 grace-period processing. 6376 6377 rcupdate.rcu_task_collapse_lim= [KNL] 6378 Set the maximum number of callbacks present 6379 at the beginning of a grace period that allows 6380 the RCU Tasks flavors to collapse back to using 6381 a single callback queue. This switching only 6382 occurs when rcupdate.rcu_task_enqueue_lim is 6383 set to the default value of -1. 6384 6385 rcupdate.rcu_task_contend_lim= [KNL] 6386 Set the minimum number of callback-queuing-time 6387 lock-contention events per jiffy required to 6388 cause the RCU Tasks flavors to switch to per-CPU 6389 callback queuing. This switching only occurs 6390 when rcupdate.rcu_task_enqueue_lim is set to 6391 the default value of -1. 6392 6393 rcupdate.rcu_task_enqueue_lim= [KNL] 6394 Set the number of callback queues to use for the 6395 RCU Tasks family of RCU flavors. The default 6396 of -1 allows this to be automatically (and 6397 dynamically) adjusted. This parameter is intended 6398 for use in testing. 6399 6400 rcupdate.rcu_task_lazy_lim= [KNL] 6401 Number of callbacks on a given CPU that will 6402 cancel laziness on that CPU. Use -1 to disable 6403 cancellation of laziness, but be advised that 6404 doing so increases the danger of OOM due to 6405 callback flooding. 6406 6407 rcupdate.rcu_task_stall_info= [KNL] 6408 Set initial timeout in jiffies for RCU task stall 6409 informational messages, which give some indication 6410 of the problem for those not patient enough to 6411 wait for ten minutes. Informational messages are 6412 only printed prior to the stall-warning message 6413 for a given grace period. Disable with a value 6414 less than or equal to zero. Defaults to ten 6415 seconds. A change in value does not take effect 6416 until the beginning of the next grace period. 6417 6418 rcupdate.rcu_task_stall_info_mult= [KNL] 6419 Multiplier for time interval between successive 6420 RCU task stall informational messages for a given 6421 RCU tasks grace period. This value is clamped 6422 to one through ten, inclusive. It defaults to 6423 the value three, so that the first informational 6424 message is printed 10 seconds into the grace 6425 period, the second at 40 seconds, the third at 6426 160 seconds, and then the stall warning at 600 6427 seconds would prevent a fourth at 640 seconds. 6428 6429 rcupdate.rcu_task_stall_timeout= [KNL] 6430 Set timeout in jiffies for RCU task stall 6431 warning messages. Disable with a value less 6432 than or equal to zero. Defaults to ten minutes. 6433 A change in value does not take effect until 6434 the beginning of the next grace period. 6435 6436 rcupdate.rcu_tasks_lazy_ms= [KNL] 6437 Set timeout in milliseconds RCU Tasks asynchronous 6438 callback batching for call_rcu_tasks(). 6439 A negative value will take the default. A value 6440 of zero will disable batching. Batching is 6441 always disabled for synchronize_rcu_tasks(). 6442 6443 rcupdate.rcu_self_test= [KNL] 6444 Run the RCU early boot self tests 6445 6446 rdinit= [KNL] 6447 Format: <full_path> 6448 Run specified binary instead of /init from the ramdisk, 6449 used for early userspace startup. See initrd. 6450 6451 rdrand= [X86,EARLY] 6452 force - Override the decision by the kernel to hide the 6453 advertisement of RDRAND support (this affects 6454 certain AMD processors because of buggy BIOS 6455 support, specifically around the suspend/resume 6456 path). 6457 6458 rdt= [HW,X86,RDT] 6459 Turn on/off individual RDT features. List is: 6460 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 6461 mba, smba, bmec, abmc, sdciae, energy[:guid], 6462 perf[:guid]. 6463 E.g. to turn on cmt and turn off mba use: 6464 rdt=cmt,!mba 6465 To turn off all energy telemetry monitoring and ensure that 6466 perf telemetry monitoring associated with guid 0x12345 6467 is enabled use: 6468 rdt=!energy,perf:0x12345 6469 6470 reboot= [KNL] 6471 Format (x86 or x86_64): 6472 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \ 6473 [[,]s[mp]#### \ 6474 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 6475 [[,]f[orce] 6476 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 6477 (prefix with 'panic_' to set mode for panic 6478 reboot only), 6479 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 6480 reboot_force is either force or not specified, 6481 reboot_cpu is s[mp]#### with #### being the processor 6482 to be used for rebooting. 6483 6484 acpi 6485 Use the ACPI RESET_REG in the FADT. If ACPI is not 6486 configured or the ACPI reset does not work, the reboot 6487 path attempts the reset using the keyboard controller. 6488 6489 bios 6490 Use the CPU reboot vector for warm reset 6491 6492 cold 6493 Set the cold reboot flag 6494 6495 default 6496 There are some built-in platform specific "quirks" 6497 - you may see: "reboot: <name> series board detected. 6498 Selecting <type> for reboots." In the case where you 6499 think the quirk is in error (e.g. you have newer BIOS, 6500 or newer board) using this option will ignore the 6501 built-in quirk table, and use the generic default 6502 reboot actions. 6503 6504 efi 6505 Use efi reset_system runtime service. If EFI is not 6506 configured or the EFI reset does not work, the reboot 6507 path attempts the reset using the keyboard controller. 6508 6509 force 6510 Don't stop other CPUs on reboot. This can make reboot 6511 more reliable in some cases. 6512 6513 kbd 6514 Use the keyboard controller. cold reset (default) 6515 6516 pci 6517 Use a write to the PCI config space register 0xcf9 to 6518 trigger reboot. 6519 6520 triple 6521 Force a triple fault (init) 6522 6523 warm 6524 Don't set the cold reboot flag 6525 6526 Using warm reset will be much faster especially on big 6527 memory systems because the BIOS will not go through 6528 the memory check. Disadvantage is that not all 6529 hardware will be completely reinitialized on reboot so 6530 there may be boot problems on some systems. 6531 6532 6533 refscale.holdoff= [KNL] 6534 Set test-start holdoff period. The purpose of 6535 this parameter is to delay the start of the 6536 test until boot completes in order to avoid 6537 interference. 6538 6539 refscale.lookup_instances= [KNL] 6540 Number of data elements to use for the forms of 6541 SLAB_TYPESAFE_BY_RCU testing. A negative number 6542 is negated and multiplied by nr_cpu_ids, while 6543 zero specifies nr_cpu_ids. 6544 6545 refscale.loops= [KNL] 6546 Set the number of loops over the synchronization 6547 primitive under test. Increasing this number 6548 reduces noise due to loop start/end overhead, 6549 but the default has already reduced the per-pass 6550 noise to a handful of picoseconds on ca. 2020 6551 x86 laptops. 6552 6553 refscale.nreaders= [KNL] 6554 Set number of readers. The default value of -1 6555 selects N, where N is roughly 75% of the number 6556 of CPUs. A value of zero is an interesting choice. 6557 6558 refscale.nruns= [KNL] 6559 Set number of runs, each of which is dumped onto 6560 the console log. 6561 6562 refscale.readdelay= [KNL] 6563 Set the read-side critical-section duration, 6564 measured in microseconds. 6565 6566 refscale.scale_type= [KNL] 6567 Specify the read-protection implementation to test. 6568 6569 refscale.shutdown= [KNL] 6570 Shut down the system at the end of the performance 6571 test. This defaults to 1 (shut it down) when 6572 refscale is built into the kernel and to 0 (leave 6573 it running) when refscale is built as a module. 6574 6575 refscale.verbose= [KNL] 6576 Enable additional printk() statements. 6577 6578 refscale.verbose_batched= [KNL] 6579 Batch the additional printk() statements. If zero 6580 (the default) or negative, print everything. Otherwise, 6581 print every Nth verbose statement, where N is the value 6582 specified. 6583 6584 regulator_ignore_unused 6585 [REGULATOR] 6586 Prevents regulator framework from disabling regulators 6587 that are unused, due no driver claiming them. This may 6588 be useful for debug and development, but should not be 6589 needed on a platform with proper driver support. 6590 6591 relax_domain_level= 6592 [KNL, SMP] Set scheduler's default relax_domain_level. 6593 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 6594 6595 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 6596 Format: <base1>,<size1>[,<base2>,<size2>,...] 6597 Reserve I/O ports or memory so the kernel won't use 6598 them. If <base> is less than 0x10000, the region 6599 is assumed to be I/O ports; otherwise it is memory. 6600 6601 reserve_mem= [RAM] 6602 Format: nn[KMG]:<align>:<label> 6603 Reserve physical memory and label it with a name that 6604 other subsystems can use to access it. This is typically 6605 used for systems that do not wipe the RAM, and this command 6606 line will try to reserve the same physical memory on 6607 soft reboots. Note, it is not guaranteed to be the same 6608 location. For example, if anything about the system changes 6609 or if booting a different kernel. It can also fail if KASLR 6610 places the kernel at the location of where the RAM reservation 6611 was from a previous boot, the new reservation will be at a 6612 different location. 6613 Any subsystem using this feature must add a way to verify 6614 that the contents of the physical memory is from a previous 6615 boot, as there may be cases where the memory will not be 6616 located at the same location. 6617 6618 The format is size:align:label for example, to request 6619 12 megabytes of 4096 alignment for ramoops: 6620 6621 reserve_mem=12M:4096:oops ramoops.mem_name=oops 6622 6623 reservetop= [X86-32,EARLY] 6624 Format: nn[KMG] 6625 Reserves a hole at the top of the kernel virtual 6626 address space. 6627 6628 reset_devices [KNL] Force drivers to reset the underlying device 6629 during initialization. 6630 6631 resume= [SWSUSP] 6632 Specify the partition device for software suspend 6633 Format: 6634 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 6635 6636 resume_offset= [SWSUSP] 6637 Specify the offset from the beginning of the partition 6638 given by "resume=" at which the swap header is located, 6639 in <PAGE_SIZE> units (needed only for swap files). 6640 See Documentation/power/swsusp-and-swap-files.rst 6641 6642 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 6643 read the resume files 6644 6645 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 6646 Useful for devices that are detected asynchronously 6647 (e.g. USB and MMC devices). 6648 6649 retain_initrd [RAM] Keep initrd memory after extraction. After boot, it will 6650 be accessible via /sys/firmware/initrd. 6651 6652 retbleed= [X86] Control mitigation of RETBleed (Arbitrary 6653 Speculative Code Execution with Return Instructions) 6654 vulnerability. 6655 6656 AMD-based UNRET and IBPB mitigations alone do not stop 6657 sibling threads from influencing the predictions of other 6658 sibling threads. For that reason, STIBP is used on pro- 6659 cessors that support it, and mitigate SMT on processors 6660 that don't. 6661 6662 off - no mitigation 6663 auto - automatically select a mitigation 6664 auto,nosmt - automatically select a mitigation, 6665 disabling SMT if necessary for 6666 the full mitigation (only on Zen1 6667 and older without STIBP). 6668 ibpb - On AMD, mitigate short speculation 6669 windows on basic block boundaries too. 6670 Safe, highest perf impact. It also 6671 enables STIBP if present. Not suitable 6672 on Intel. 6673 ibpb,nosmt - Like "ibpb" above but will disable SMT 6674 when STIBP is not available. This is 6675 the alternative for systems which do not 6676 have STIBP. 6677 unret - Force enable untrained return thunks, 6678 only effective on AMD f15h-f17h based 6679 systems. 6680 unret,nosmt - Like unret, but will disable SMT when STIBP 6681 is not available. This is the alternative for 6682 systems which do not have STIBP. 6683 6684 Selecting 'auto' will choose a mitigation method at run 6685 time according to the CPU. 6686 6687 Not specifying this option is equivalent to retbleed=auto. 6688 6689 rfkill.default_state= 6690 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 6691 etc. communication is blocked by default. 6692 1 Unblocked. 6693 6694 rfkill.master_switch_mode= 6695 0 The "airplane mode" button does nothing. 6696 1 The "airplane mode" button toggles between everything 6697 blocked and the previous configuration. 6698 2 The "airplane mode" button toggles between everything 6699 blocked and everything unblocked. 6700 6701 ring3mwait=disable 6702 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 6703 CPUs. 6704 6705 riscv_isa_fallback [RISCV,EARLY] 6706 When CONFIG_RISCV_ISA_FALLBACK is not enabled, permit 6707 falling back to detecting extension support by parsing 6708 "riscv,isa" property on devicetree systems when the 6709 replacement properties are not found. See the Kconfig 6710 entry for RISCV_ISA_FALLBACK. 6711 6712 riscv_nousercfi= 6713 all Disable user CFI ABI to userspace even if cpu extension 6714 are available. 6715 bcfi Disable user backward CFI ABI to userspace even if 6716 the shadow stack extension is available. 6717 fcfi Disable user forward CFI ABI to userspace even if the 6718 landing pad extension is available. 6719 6720 ro [KNL] Mount root device read-only on boot 6721 6722 rodata= [KNL,EARLY] 6723 on Mark read-only kernel memory as read-only (default). 6724 off Leave read-only kernel memory writable for debugging. 6725 noalias Mark read-only kernel memory as read-only but retain 6726 writable aliases in the direct map for regions outside 6727 of the kernel image. [arm64] 6728 6729 rockchip.usb_uart 6730 [EARLY] 6731 Enable the uart passthrough on the designated usb port 6732 on Rockchip SoCs. When active, the signals of the 6733 debug-uart get routed to the D+ and D- pins of the usb 6734 port and the regular usb controller gets disabled. 6735 6736 root= [KNL] Root filesystem 6737 Usually this is a block device specifier of some kind, 6738 see the early_lookup_bdev comment in 6739 block/early-lookup.c for details. 6740 Alternatively this can be "ram" for the legacy initial 6741 ramdisk, "nfs" and "cifs" for root on a network file 6742 system, or "mtd" and "ubi" for mounting from raw flash. 6743 6744 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 6745 mount the root filesystem 6746 6747 rootflags= [KNL] Set root filesystem mount option string 6748 6749 rseq_slice_ext= [KNL] RSEQ based time slice extension 6750 Format: boolean 6751 Control enablement of RSEQ based time slice extension. 6752 Default is 'on'. 6753 6754 initramfs_options= [KNL] 6755 Specify mount options for for the initramfs mount. 6756 6757 rootfstype= [KNL] Set root filesystem type 6758 6759 rootwait [KNL] Wait (indefinitely) for root device to show up. 6760 Useful for devices that are detected asynchronously 6761 (e.g. USB and MMC devices). 6762 6763 rootwait= [KNL] Maximum time (in seconds) to wait for root device 6764 to show up before attempting to mount the root 6765 filesystem. 6766 6767 rproc_mem=nn[KMG][@address] 6768 [KNL,ARM,CMA] Remoteproc physical memory block. 6769 Memory area to be used by remote processor image, 6770 managed by CMA. 6771 6772 rseq_debug= [KNL] Enable or disable restartable sequence 6773 debug mode. Defaults to CONFIG_RSEQ_DEBUG_DEFAULT_ENABLE. 6774 Format: <bool> 6775 6776 rt_group_sched= [KNL] Enable or disable SCHED_RR/FIFO group scheduling 6777 when CONFIG_RT_GROUP_SCHED=y. Defaults to 6778 !CONFIG_RT_GROUP_SCHED_DEFAULT_DISABLED. 6779 Format: <bool> 6780 6781 rw [KNL] Mount root device read-write on boot 6782 6783 S [KNL] Run init in single mode 6784 6785 s390_iommu= [HW,S390] 6786 Set s390 IOTLB flushing mode 6787 strict 6788 With strict flushing every unmap operation will result 6789 in an IOTLB flush. Default is lazy flushing before 6790 reuse, which is faster. Deprecated, equivalent to 6791 iommu.strict=1. 6792 6793 s390_iommu_aperture= [KNL,S390] 6794 Specifies the size of the per device DMA address space 6795 accessible through the DMA and IOMMU APIs as a decimal 6796 factor of the size of main memory. 6797 The default is 1 meaning that one can concurrently use 6798 as many DMA addresses as physical memory is installed, 6799 if supported by hardware, and thus map all of memory 6800 once. With a value of 2 one can map all of memory twice 6801 and so on. As a special case a factor of 0 imposes no 6802 restrictions other than those given by hardware at the 6803 cost of significant additional memory use for tables. 6804 6805 sa1100ir [NET] 6806 See drivers/net/irda/sa1100_ir.c. 6807 6808 sched_proxy_exec= [KNL] 6809 Enables or disables "proxy execution" style 6810 solution to mutex-based priority inversion. 6811 Format: <bool> 6812 6813 sched_verbose [KNL,EARLY] Enables verbose scheduler debug messages. 6814 6815 schedstats= [KNL,X86] Enable or disable scheduled statistics. 6816 Allowed values are enable and disable. This feature 6817 incurs a small amount of overhead in the scheduler 6818 but is useful for debugging and performance tuning. 6819 6820 sched_thermal_decay_shift= 6821 [Deprecated] 6822 [KNL, SMP] Set a decay shift for scheduler thermal 6823 pressure signal. Thermal pressure signal follows the 6824 default decay period of other scheduler pelt 6825 signals(usually 32 ms but configurable). Setting 6826 sched_thermal_decay_shift will left shift the decay 6827 period for the thermal pressure signal by the shift 6828 value. 6829 i.e. with the default pelt decay period of 32 ms 6830 sched_thermal_decay_shift thermal pressure decay pr 6831 1 64 ms 6832 2 128 ms 6833 and so on. 6834 Format: integer between 0 and 10 6835 Default is 0. 6836 6837 scftorture.holdoff= [KNL] 6838 Number of seconds to hold off before starting 6839 test. Defaults to zero for module insertion and 6840 to 10 seconds for built-in smp_call_function() 6841 tests. 6842 6843 scftorture.longwait= [KNL] 6844 Request ridiculously long waits randomly selected 6845 up to the chosen limit in seconds. Zero (the 6846 default) disables this feature. Please note 6847 that requesting even small non-zero numbers of 6848 seconds can result in RCU CPU stall warnings, 6849 softlockup complaints, and so on. 6850 6851 scftorture.nthreads= [KNL] 6852 Number of kthreads to spawn to invoke the 6853 smp_call_function() family of functions. 6854 The default of -1 specifies a number of kthreads 6855 equal to the number of CPUs. 6856 6857 scftorture.onoff_holdoff= [KNL] 6858 Number seconds to wait after the start of the 6859 test before initiating CPU-hotplug operations. 6860 6861 scftorture.onoff_interval= [KNL] 6862 Number seconds to wait between successive 6863 CPU-hotplug operations. Specifying zero (which 6864 is the default) disables CPU-hotplug operations. 6865 6866 scftorture.shutdown_secs= [KNL] 6867 The number of seconds following the start of the 6868 test after which to shut down the system. The 6869 default of zero avoids shutting down the system. 6870 Non-zero values are useful for automated tests. 6871 6872 scftorture.stat_interval= [KNL] 6873 The number of seconds between outputting the 6874 current test statistics to the console. A value 6875 of zero disables statistics output. 6876 6877 scftorture.stutter_cpus= [KNL] 6878 The number of jiffies to wait between each change 6879 to the set of CPUs under test. 6880 6881 scftorture.use_cpus_read_lock= [KNL] 6882 Use use_cpus_read_lock() instead of the default 6883 preempt_disable() to disable CPU hotplug 6884 while invoking one of the smp_call_function*() 6885 functions. 6886 6887 scftorture.verbose= [KNL] 6888 Enable additional printk() statements. 6889 6890 scftorture.weight_single= [KNL] 6891 The probability weighting to use for the 6892 smp_call_function_single() function with a zero 6893 "wait" parameter. A value of -1 selects the 6894 default if all other weights are -1. However, 6895 if at least one weight has some other value, a 6896 value of -1 will instead select a weight of zero. 6897 6898 scftorture.weight_single_wait= [KNL] 6899 The probability weighting to use for the 6900 smp_call_function_single() function with a 6901 non-zero "wait" parameter. See weight_single. 6902 6903 scftorture.weight_many= [KNL] 6904 The probability weighting to use for the 6905 smp_call_function_many() function with a zero 6906 "wait" parameter. See weight_single. 6907 Note well that setting a high probability for 6908 this weighting can place serious IPI load 6909 on the system. 6910 6911 scftorture.weight_many_wait= [KNL] 6912 The probability weighting to use for the 6913 smp_call_function_many() function with a 6914 non-zero "wait" parameter. See weight_single 6915 and weight_many. 6916 6917 scftorture.weight_all= [KNL] 6918 The probability weighting to use for the 6919 smp_call_function_all() function with a zero 6920 "wait" parameter. See weight_single and 6921 weight_many. 6922 6923 scftorture.weight_all_wait= [KNL] 6924 The probability weighting to use for the 6925 smp_call_function_all() function with a 6926 non-zero "wait" parameter. See weight_single 6927 and weight_many. 6928 6929 sdw_mclk_divider=[SDW] 6930 Specify the MCLK divider for Intel SoundWire buses in 6931 case the BIOS does not provide the clock rate properly. 6932 6933 skew_tick= [KNL,EARLY] Offset the periodic timer tick per cpu to mitigate 6934 xtime_lock contention on larger systems, and/or RCU lock 6935 contention on all systems with CONFIG_MAXSMP set. 6936 Format: { "0" | "1" } 6937 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 6938 1 -- enable. 6939 Note: increases power consumption, thus should only be 6940 enabled if running jitter sensitive (HPC/RT) workloads. 6941 6942 security= [SECURITY] Choose a legacy "major" security module to 6943 enable at boot. This has been deprecated by the 6944 "lsm=" parameter. 6945 6946 selinux= [SELINUX] Disable or enable SELinux at boot time. 6947 Format: { "0" | "1" } 6948 See security/selinux/Kconfig help text. 6949 0 -- disable. 6950 1 -- enable. 6951 Default value is 1. 6952 6953 serialnumber [BUGS=X86-32] 6954 6955 sev=option[,option...] [X86-64] 6956 6957 debug 6958 Enable debug messages. 6959 6960 nosnp 6961 Do not enable SEV-SNP (applies to host/hypervisor 6962 only). Setting 'nosnp' avoids the RMP check overhead 6963 in memory accesses when users do not want to run 6964 SEV-SNP guests. 6965 6966 shapers= [NET] 6967 Maximal number of shapers. 6968 6969 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 6970 Limit apic dumping. The parameter defines the maximal 6971 number of local apics being dumped. Also it is possible 6972 to set it to "all" by meaning -- no limit here. 6973 Format: { 1 (default) | 2 | ... | all }. 6974 The parameter valid if only apic=debug or 6975 apic=verbose is specified. 6976 Example: apic=debug show_lapic=all 6977 6978 slab_debug[=options[,slabs][;[options[,slabs]]...] [MM] 6979 Enabling slab_debug allows one to determine the 6980 culprit if slab objects become corrupted. Enabling 6981 slab_debug can create guard zones around objects and 6982 may poison objects when not in use. Also tracks the 6983 last alloc / free. For more information see 6984 Documentation/admin-guide/mm/slab.rst. 6985 (slub_debug legacy name also accepted for now) 6986 6987 Using this option implies the "no_hash_pointers" 6988 option which can be undone by adding the 6989 "hash_pointers=always" option. 6990 6991 slab_max_order= [MM] 6992 Determines the maximum allowed order for slabs. 6993 A high setting may cause OOMs due to memory 6994 fragmentation. For more information see 6995 Documentation/admin-guide/mm/slab.rst. 6996 (slub_max_order legacy name also accepted for now) 6997 6998 slab_merge [MM] 6999 Enable merging of slabs with similar size when the 7000 kernel is built without CONFIG_SLAB_MERGE_DEFAULT. 7001 (slub_merge legacy name also accepted for now) 7002 7003 slab_min_objects= [MM] 7004 The minimum number of objects per slab. SLUB will 7005 increase the slab order up to slab_max_order to 7006 generate a sufficiently large slab able to contain 7007 the number of objects indicated. The higher the number 7008 of objects the smaller the overhead of tracking slabs 7009 and the less frequently locks need to be acquired. 7010 For more information see 7011 Documentation/admin-guide/mm/slab.rst. 7012 (slub_min_objects legacy name also accepted for now) 7013 7014 slab_min_order= [MM] 7015 Determines the minimum page order for slabs. Must be 7016 lower or equal to slab_max_order. For more information see 7017 Documentation/admin-guide/mm/slab.rst. 7018 (slub_min_order legacy name also accepted for now) 7019 7020 slab_nomerge [MM] 7021 Disable merging of slabs with similar size. May be 7022 necessary if there is some reason to distinguish 7023 allocs to different slabs, especially in hardened 7024 environments where the risk of heap overflows and 7025 layout control by attackers can usually be 7026 frustrated by disabling merging. This will reduce 7027 most of the exposure of a heap attack to a single 7028 cache (risks via metadata attacks are mostly 7029 unchanged). Debug options disable merging on their 7030 own. 7031 For more information see 7032 Documentation/admin-guide/mm/slab.rst. 7033 (slub_nomerge legacy name also accepted for now) 7034 7035 slab_strict_numa [MM] 7036 Support memory policies on a per object level 7037 in the slab allocator. The default is for memory 7038 policies to be applied at the folio level when 7039 a new folio is needed or a partial folio is 7040 retrieved from the lists. Increases overhead 7041 in the slab fastpaths but gains more accurate 7042 NUMA kernel object placement which helps with slow 7043 interconnects in NUMA systems. 7044 7045 slram= [HW,MTD] 7046 7047 smart2= [HW] 7048 Format: <io1>[,<io2>[,...,<io8>]] 7049 7050 smp.csd_lock_timeout= [KNL] 7051 Specify the period of time in milliseconds 7052 that smp_call_function() and friends will wait 7053 for a CPU to release the CSD lock. This is 7054 useful when diagnosing bugs involving CPUs 7055 disabling interrupts for extended periods 7056 of time. Defaults to 5,000 milliseconds, and 7057 setting a value of zero disables this feature. 7058 This feature may be more efficiently disabled 7059 using the csdlock_debug- kernel parameter. 7060 7061 smp.panic_on_ipistall= [KNL] 7062 If a csd_lock_timeout extends for more than 7063 the specified number of milliseconds, panic the 7064 system. By default, let CSD-lock acquisition 7065 take as long as they take. Specifying 300,000 7066 for this value provides a 5-minute timeout. 7067 7068 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 7069 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 7070 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 7071 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 7072 smsc-ircc2.ircc_irq= [HW] IRQ line 7073 smsc-ircc2.ircc_dma= [HW] DMA channel 7074 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 7075 0: Toshiba Satellite 1800 (GP data pin select) 7076 1: Fast pin select (default) 7077 2: ATC IRMode 7078 7079 smt= [KNL,MIPS,S390,EARLY] Set the maximum number of threads 7080 (logical CPUs) to use per physical CPU on systems 7081 capable of symmetric multithreading (SMT). Will 7082 be capped to the actual hardware limit. 7083 Format: <integer> 7084 Default: -1 (no limit) 7085 7086 softlockup_panic= 7087 [KNL] Should the soft-lockup detector generate panics. 7088 Format: <int> 7089 7090 A value of non-zero instructs the soft-lockup detector 7091 to panic the machine when a soft-lockup duration exceeds 7092 N thresholds. It is also controlled by the kernel.softlockup_panic 7093 sysctl and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 7094 respective build-time switch to that functionality. 7095 7096 softlockup_all_cpu_backtrace= 7097 [KNL] Should the soft-lockup detector generate 7098 backtraces on all cpus. 7099 Format: 0 | 1 7100 7101 sonypi.*= [HW] Sony Programmable I/O Control Device driver 7102 See Documentation/admin-guide/laptops/sonypi.rst 7103 7104 spectre_bhi= [X86] Control mitigation of Branch History Injection 7105 (BHI) vulnerability. This setting affects the 7106 deployment of the HW BHI control and the SW BHB 7107 clearing sequence. 7108 7109 on - (default) Enable the HW or SW mitigation as 7110 needed. This protects the kernel from 7111 both syscalls and VMs. 7112 vmexit - On systems which don't have the HW mitigation 7113 available, enable the SW mitigation on vmexit 7114 ONLY. On such systems, the host kernel is 7115 protected from VM-originated BHI attacks, but 7116 may still be vulnerable to syscall attacks. 7117 off - Disable the mitigation. 7118 7119 spectre_v2= [X86,EARLY] Control mitigation of Spectre variant 2 7120 (indirect branch speculation) vulnerability. 7121 The default operation protects the kernel from 7122 user space attacks. 7123 7124 on - unconditionally enable, implies 7125 spectre_v2_user=on 7126 off - unconditionally disable, implies 7127 spectre_v2_user=off 7128 auto - kernel detects whether your CPU model is 7129 vulnerable 7130 7131 Selecting 'on' will, and 'auto' may, choose a 7132 mitigation method at run time according to the 7133 CPU, the available microcode, the setting of the 7134 CONFIG_MITIGATION_RETPOLINE configuration option, 7135 and the compiler with which the kernel was built. 7136 7137 Selecting 'on' will also enable the mitigation 7138 against user space to user space task attacks. 7139 Selecting specific mitigation does not force enable 7140 user mitigations. 7141 7142 Selecting 'off' will disable both the kernel and 7143 the user space protections. 7144 7145 Specific mitigations can also be selected manually: 7146 7147 retpoline - replace indirect branches 7148 retpoline,generic - Retpolines 7149 retpoline,lfence - LFENCE; indirect branch 7150 retpoline,amd - alias for retpoline,lfence 7151 eibrs - Enhanced/Auto IBRS 7152 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines 7153 eibrs,lfence - Enhanced/Auto IBRS + LFENCE 7154 ibrs - use IBRS to protect kernel 7155 7156 Not specifying this option is equivalent to 7157 spectre_v2=auto. 7158 7159 spectre_v2_user= 7160 [X86] Control mitigation of Spectre variant 2 7161 (indirect branch speculation) vulnerability between 7162 user space tasks 7163 7164 on - Unconditionally enable mitigations. Is 7165 enforced by spectre_v2=on 7166 7167 off - Unconditionally disable mitigations. Is 7168 enforced by spectre_v2=off 7169 7170 prctl - Indirect branch speculation is enabled, 7171 but mitigation can be enabled via prctl 7172 per thread. The mitigation control state 7173 is inherited on fork. 7174 7175 prctl,ibpb 7176 - Like "prctl" above, but only STIBP is 7177 controlled per thread. IBPB is issued 7178 always when switching between different user 7179 space processes. 7180 7181 seccomp 7182 - Same as "prctl" above, but all seccomp 7183 threads will enable the mitigation unless 7184 they explicitly opt out. 7185 7186 seccomp,ibpb 7187 - Like "seccomp" above, but only STIBP is 7188 controlled per thread. IBPB is issued 7189 always when switching between different 7190 user space processes. 7191 7192 auto - Kernel selects the mitigation depending on 7193 the available CPU features and vulnerability. 7194 7195 Default mitigation: "prctl" 7196 7197 Not specifying this option is equivalent to 7198 spectre_v2_user=auto. 7199 7200 spec_rstack_overflow= 7201 [X86,EARLY] Control RAS overflow mitigation on AMD Zen CPUs 7202 7203 off - Disable mitigation 7204 microcode - Enable microcode mitigation only 7205 safe-ret - Enable sw-only safe RET mitigation (default) 7206 ibpb - Enable mitigation by issuing IBPB on 7207 kernel entry 7208 ibpb-vmexit - Issue IBPB only on VMEXIT 7209 (cloud-specific mitigation) 7210 7211 spec_store_bypass_disable= 7212 [HW,EARLY] Control Speculative Store Bypass (SSB) Disable mitigation 7213 (Speculative Store Bypass vulnerability) 7214 7215 Certain CPUs are vulnerable to an exploit against a 7216 a common industry wide performance optimization known 7217 as "Speculative Store Bypass" in which recent stores 7218 to the same memory location may not be observed by 7219 later loads during speculative execution. The idea 7220 is that such stores are unlikely and that they can 7221 be detected prior to instruction retirement at the 7222 end of a particular speculation execution window. 7223 7224 In vulnerable processors, the speculatively forwarded 7225 store can be used in a cache side channel attack, for 7226 example to read memory to which the attacker does not 7227 directly have access (e.g. inside sandboxed code). 7228 7229 This parameter controls whether the Speculative Store 7230 Bypass optimization is used. 7231 7232 On x86 the options are: 7233 7234 on - Unconditionally disable Speculative Store Bypass 7235 off - Unconditionally enable Speculative Store Bypass 7236 auto - Kernel detects whether the CPU model contains an 7237 implementation of Speculative Store Bypass and 7238 picks the most appropriate mitigation. If the 7239 CPU is not vulnerable, "off" is selected. If the 7240 CPU is vulnerable the default mitigation is 7241 architecture and Kconfig dependent. See below. 7242 prctl - Control Speculative Store Bypass per thread 7243 via prctl. Speculative Store Bypass is enabled 7244 for a process by default. The state of the control 7245 is inherited on fork. 7246 seccomp - Same as "prctl" above, but all seccomp threads 7247 will disable SSB unless they explicitly opt out. 7248 7249 Default mitigations: 7250 X86: "prctl" 7251 7252 On powerpc the options are: 7253 7254 on,auto - On Power8 and Power9 insert a store-forwarding 7255 barrier on kernel entry and exit. On Power7 7256 perform a software flush on kernel entry and 7257 exit. 7258 off - No action. 7259 7260 Not specifying this option is equivalent to 7261 spec_store_bypass_disable=auto. 7262 7263 split_lock_detect= 7264 [X86] Enable split lock detection or bus lock detection 7265 7266 When enabled (and if hardware support is present), atomic 7267 instructions that access data across cache line 7268 boundaries will result in an alignment check exception 7269 for split lock detection or a debug exception for 7270 bus lock detection. 7271 7272 off - not enabled 7273 7274 warn - the kernel will emit rate-limited warnings 7275 about applications triggering the #AC 7276 exception or the #DB exception. This mode is 7277 the default on CPUs that support split lock 7278 detection or bus lock detection. Default 7279 behavior is by #AC if both features are 7280 enabled in hardware. 7281 7282 fatal - the kernel will send SIGBUS to applications 7283 that trigger the #AC exception or the #DB 7284 exception. Default behavior is by #AC if 7285 both features are enabled in hardware. 7286 7287 ratelimit:N - 7288 Set system wide rate limit to N bus locks 7289 per second for bus lock detection. 7290 0 < N <= 1000. 7291 7292 N/A for split lock detection. 7293 7294 7295 If an #AC exception is hit in the kernel or in 7296 firmware (i.e. not while executing in user mode) 7297 the kernel will oops in either "warn" or "fatal" 7298 mode. 7299 7300 #DB exception for bus lock is triggered only when 7301 CPL > 0. 7302 7303 srbds= [X86,INTEL,EARLY] 7304 Control the Special Register Buffer Data Sampling 7305 (SRBDS) mitigation. 7306 7307 Certain CPUs are vulnerable to an MDS-like 7308 exploit which can leak bits from the random 7309 number generator. 7310 7311 By default, this issue is mitigated by 7312 microcode. However, the microcode fix can cause 7313 the RDRAND and RDSEED instructions to become 7314 much slower. Among other effects, this will 7315 result in reduced throughput from /dev/urandom. 7316 7317 The microcode mitigation can be disabled with 7318 the following option: 7319 7320 off: Disable mitigation and remove 7321 performance impact to RDRAND and RDSEED 7322 7323 srcutree.big_cpu_lim [KNL] 7324 Specifies the number of CPUs constituting a 7325 large system, such that srcu_struct structures 7326 should immediately allocate an srcu_node array. 7327 This kernel-boot parameter defaults to 128, 7328 but takes effect only when the low-order four 7329 bits of srcutree.convert_to_big is equal to 3 7330 (decide at boot). 7331 7332 srcutree.convert_to_big [KNL] 7333 Specifies under what conditions an SRCU tree 7334 srcu_struct structure will be converted to big 7335 form, that is, with an rcu_node tree: 7336 7337 0: Never. 7338 1: At init_srcu_struct() time. 7339 2: When rcutorture decides to. 7340 3: Decide at boot time (default). 7341 0x1X: Above plus if high contention. 7342 7343 Either way, the srcu_node tree will be sized based 7344 on the actual runtime number of CPUs (nr_cpu_ids) 7345 instead of the compile-time CONFIG_NR_CPUS. 7346 7347 srcutree.counter_wrap_check [KNL] 7348 Specifies how frequently to check for 7349 grace-period sequence counter wrap for the 7350 srcu_data structure's ->srcu_gp_seq_needed field. 7351 The greater the number of bits set in this kernel 7352 parameter, the less frequently counter wrap will 7353 be checked for. Note that the bottom two bits 7354 are ignored. 7355 7356 srcutree.exp_holdoff [KNL] 7357 Specifies how many nanoseconds must elapse 7358 since the end of the last SRCU grace period for 7359 a given srcu_struct until the next normal SRCU 7360 grace period will be considered for automatic 7361 expediting. Set to zero to disable automatic 7362 expediting. 7363 7364 srcutree.srcu_max_nodelay [KNL] 7365 Specifies the number of no-delay instances 7366 per jiffy for which the SRCU grace period 7367 worker thread will be rescheduled with zero 7368 delay. Beyond this limit, worker thread will 7369 be rescheduled with a sleep delay of one jiffy. 7370 7371 srcutree.srcu_max_nodelay_phase [KNL] 7372 Specifies the per-grace-period phase, number of 7373 non-sleeping polls of readers. Beyond this limit, 7374 grace period worker thread will be rescheduled 7375 with a sleep delay of one jiffy, between each 7376 rescan of the readers, for a grace period phase. 7377 7378 srcutree.srcu_retry_check_delay [KNL] 7379 Specifies number of microseconds of non-sleeping 7380 delay between each non-sleeping poll of readers. 7381 7382 srcutree.small_contention_lim [KNL] 7383 Specifies the number of update-side contention 7384 events per jiffy will be tolerated before 7385 initiating a conversion of an srcu_struct 7386 structure to big form. Note that the value of 7387 srcutree.convert_to_big must have the 0x10 bit 7388 set for contention-based conversions to occur. 7389 7390 ssbd= [ARM64,HW,EARLY] 7391 Speculative Store Bypass Disable control 7392 7393 On CPUs that are vulnerable to the Speculative 7394 Store Bypass vulnerability and offer a 7395 firmware based mitigation, this parameter 7396 indicates how the mitigation should be used: 7397 7398 force-on: Unconditionally enable mitigation for 7399 for both kernel and userspace 7400 force-off: Unconditionally disable mitigation for 7401 for both kernel and userspace 7402 kernel: Always enable mitigation in the 7403 kernel, and offer a prctl interface 7404 to allow userspace to register its 7405 interest in being mitigated too. 7406 7407 stack_guard_gap= [MM] 7408 override the default stack gap protection. The value 7409 is in page units and it defines how many pages prior 7410 to (for stacks growing down) resp. after (for stacks 7411 growing up) the main stack are reserved for no other 7412 mapping. Default value is 256 pages. 7413 7414 stack_depot_disable= [KNL,EARLY] 7415 Setting this to true through kernel command line will 7416 disable the stack depot thereby saving the static memory 7417 consumed by the stack hash table. By default this is set 7418 to false. 7419 7420 stack_depot_max_pools= [KNL,EARLY] 7421 Specify the maximum number of pools to use for storing 7422 stack traces. Pools are allocated on-demand up to this 7423 limit. Default value is 8191 pools. 7424 7425 stacktrace [FTRACE] 7426 Enable the stack tracer on boot up. 7427 7428 stacktrace_filter=[function-list] 7429 [FTRACE] Limit the functions that the stack tracer 7430 will trace at boot up. function-list is a comma-separated 7431 list of functions. This list can be changed at run 7432 time by the stack_trace_filter file in the debugfs 7433 tracing directory. Note, this enables stack tracing 7434 and the stacktrace above is not needed. 7435 7436 sti= [PARISC,HW] 7437 Format: <num> 7438 Set the STI (builtin display/keyboard on the HP-PARISC 7439 machines) console (graphic card) which should be used 7440 as the initial boot-console. 7441 See also comment in drivers/video/console/sticore.c. 7442 7443 sti_font= [HW] 7444 See comment in drivers/video/console/sticore.c. 7445 7446 stifb= [HW] 7447 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 7448 7449 strict_sas_size= 7450 [X86] 7451 Format: <bool> 7452 Enable or disable strict sigaltstack size checks 7453 against the required signal frame size which 7454 depends on the supported FPU features. This can 7455 be used to filter out binaries which have 7456 not yet been made aware of AT_MINSIGSTKSZ. 7457 7458 stress_hpt [PPC,EARLY] 7459 Limits the number of kernel HPT entries in the hash 7460 page table to increase the rate of hash page table 7461 faults on kernel addresses. 7462 7463 stress_slb [PPC,EARLY] 7464 Limits the number of kernel SLB entries, and flushes 7465 them frequently to increase the rate of SLB faults 7466 on kernel addresses. 7467 7468 no_slb_preload [PPC,EARLY] 7469 Disables slb preloading for userspace. 7470 7471 sunrpc.min_resvport= 7472 sunrpc.max_resvport= 7473 [NFS,SUNRPC] 7474 SunRPC servers often require that client requests 7475 originate from a privileged port (i.e. a port in the 7476 range 0 < portnr < 1024). 7477 An administrator who wishes to reserve some of these 7478 ports for other uses may adjust the range that the 7479 kernel's sunrpc client considers to be privileged 7480 using these two parameters to set the minimum and 7481 maximum port values. 7482 7483 sunrpc.svc_rpc_per_connection_limit= 7484 [NFS,SUNRPC] 7485 Limit the number of requests that the server will 7486 process in parallel from a single connection. 7487 The default value is 0 (no limit). 7488 7489 sunrpc.pool_mode= 7490 [NFS] 7491 Control how the NFS server code allocates CPUs to 7492 service thread pools. Depending on how many NICs 7493 you have and where their interrupts are bound, this 7494 option will affect which CPUs will do NFS serving. 7495 Note: this parameter cannot be changed while the 7496 NFS server is running. 7497 7498 auto the server chooses an appropriate mode 7499 automatically using heuristics 7500 global a single global pool contains all CPUs 7501 percpu one pool for each CPU 7502 pernode one pool for each NUMA node (equivalent 7503 to global on non-NUMA machines) 7504 7505 sunrpc.tcp_slot_table_entries= 7506 sunrpc.udp_slot_table_entries= 7507 [NFS,SUNRPC] 7508 Sets the upper limit on the number of simultaneous 7509 RPC calls that can be sent from the client to a 7510 server. Increasing these values may allow you to 7511 improve throughput, but will also increase the 7512 amount of memory reserved for use by the client. 7513 7514 suspend.pm_test_delay= 7515 [SUSPEND] 7516 Sets the number of seconds to remain in a suspend test 7517 mode before resuming the system (see 7518 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 7519 is set. Default value is 5. 7520 7521 svm= [PPC] 7522 Format: { on | off | y | n | 1 | 0 } 7523 This parameter controls use of the Protected 7524 Execution Facility on pSeries. 7525 7526 swiotlb= [ARM,PPC,MIPS,X86,S390,EARLY] 7527 Format: { <int> [,<int>] | force | noforce } 7528 <int> -- Number of I/O TLB slabs 7529 <int> -- Second integer after comma. Number of swiotlb 7530 areas with their own lock. Will be rounded up 7531 to a power of 2. 7532 force -- force using of bounce buffers even if they 7533 wouldn't be automatically used by the kernel 7534 noforce -- Never use bounce buffers (for debugging) 7535 7536 switches= [HW,M68k,EARLY] 7537 7538 sysctl.*= [KNL] 7539 Set a sysctl parameter, right before loading the init 7540 process, as if the value was written to the respective 7541 /proc/sys/... file. Both '.' and '/' are recognized as 7542 separators. Unrecognized parameters and invalid values 7543 are reported in the kernel log. Sysctls registered 7544 later by a loaded module cannot be set this way. 7545 Example: sysctl.vm.swappiness=40 7546 7547 sysrq_always_enabled 7548 [KNL] 7549 Ignore sysrq setting - this boot parameter will 7550 neutralize any effect of /proc/sys/kernel/sysrq. 7551 Useful for debugging. 7552 7553 tcpmhash_entries= [KNL,NET] 7554 Set the number of tcp_metrics_hash slots. 7555 Default value is 8192 or 16384 depending on total 7556 ram pages. This is used to specify the TCP metrics 7557 cache size. See Documentation/networking/ip-sysctl.rst 7558 "tcp_no_metrics_save" section for more details. 7559 7560 tdfx= [HW,DRM] 7561 7562 test_suspend= [SUSPEND] 7563 Format: { "mem" | "standby" | "freeze" }[,N] 7564 Specify "mem" (for Suspend-to-RAM) or "standby" (for 7565 standby suspend) or "freeze" (for suspend type freeze) 7566 as the system sleep state during system startup with 7567 the optional capability to repeat N number of times. 7568 The system is woken from this state using a 7569 wakeup-capable RTC alarm. 7570 7571 thash_entries= [KNL,NET] 7572 Set number of hash buckets for TCP connection 7573 7574 thermal.act= [HW,ACPI] 7575 -1: disable all active trip points in all thermal zones 7576 <degrees C>: override all lowest active trip points 7577 7578 thermal.crt= [HW,ACPI] 7579 -1: disable all critical trip points in all thermal zones 7580 <degrees C>: override all critical trip points 7581 7582 thermal.off= [HW,ACPI] 7583 1: disable ACPI thermal control 7584 7585 thermal.psv= [HW,ACPI] 7586 -1: disable all passive trip points 7587 <degrees C>: override all passive trip points to this 7588 value 7589 7590 thermal.tzp= [HW,ACPI] 7591 Specify global default ACPI thermal zone polling rate 7592 <deci-seconds>: poll all this frequency 7593 0: no polling (default) 7594 7595 thp_anon= [KNL] 7596 Format: <size>[KMG],<size>[KMG]:<state>;<size>[KMG]-<size>[KMG]:<state> 7597 state is one of "always", "madvise", "never" or "inherit". 7598 Control the default behavior of the system with respect 7599 to anonymous transparent hugepages. 7600 Can be used multiple times for multiple anon THP sizes. 7601 See Documentation/admin-guide/mm/transhuge.rst for more 7602 details. 7603 7604 threadirqs [KNL,EARLY] 7605 Force threading of all interrupt handlers except those 7606 marked explicitly IRQF_NO_THREAD. 7607 7608 thp_shmem= [KNL] 7609 Format: <size>[KMG],<size>[KMG]:<policy>;<size>[KMG]-<size>[KMG]:<policy> 7610 Control the default policy of each hugepage size for the 7611 internal shmem mount. <policy> is one of policies available 7612 for the shmem mount ("always", "inherit", "never", "within_size", 7613 and "advise"). 7614 It can be used multiple times for multiple shmem THP sizes. 7615 See Documentation/admin-guide/mm/transhuge.rst for more 7616 details. 7617 7618 topology= [S390,EARLY] 7619 Format: {off | on} 7620 Specify if the kernel should make use of the cpu 7621 topology information if the hardware supports this. 7622 The scheduler will make use of this information and 7623 e.g. base its process migration decisions on it. 7624 Default is on. 7625 7626 torture.disable_onoff_at_boot= [KNL] 7627 Prevent the CPU-hotplug component of torturing 7628 until after init has spawned. 7629 7630 torture.ftrace_dump_at_shutdown= [KNL] 7631 Dump the ftrace buffer at torture-test shutdown, 7632 even if there were no errors. This can be a 7633 very costly operation when many torture tests 7634 are running concurrently, especially on systems 7635 with rotating-rust storage. 7636 7637 torture.verbose_sleep_frequency= [KNL] 7638 Specifies how many verbose printk()s should be 7639 emitted between each sleep. The default of zero 7640 disables verbose-printk() sleeping. 7641 7642 torture.verbose_sleep_duration= [KNL] 7643 Duration of each verbose-printk() sleep in jiffies. 7644 7645 tpm.disable_pcr_integrity= [HW,TPM] 7646 Do not protect PCR registers from unintended physical 7647 access, or interposers in the bus by the means of 7648 having an integrity protected session wrapped around 7649 TPM2_PCR_Extend command. Consider this in a situation 7650 where TPM is heavily utilized by IMA, thus protection 7651 causing a major performance hit, and the space where 7652 machines are deployed is by other means guarded. 7653 7654 tpm_crb_ffa.busy_timeout_ms= [ARM64,TPM] 7655 Maximum time in milliseconds to retry sending a message 7656 to the TPM service before giving up. This parameter controls 7657 how long the system will continue retrying when the TPM 7658 service is busy. 7659 Format: <unsigned int> 7660 Default: 2000 (2 seconds) 7661 7662 tpm_suspend_pcr=[HW,TPM] 7663 Format: integer pcr id 7664 Specify that at suspend time, the tpm driver 7665 should extend the specified pcr with zeros, 7666 as a workaround for some chips which fail to 7667 flush the last written pcr on TPM_SaveState. 7668 This will guarantee that all the other pcrs 7669 are saved. 7670 7671 tpm_tis.interrupts= [HW,TPM] 7672 Enable interrupts for the MMIO based physical layer 7673 for the FIFO interface. By default it is set to false 7674 (0). For more information about TPM hardware interfaces 7675 defined by Trusted Computing Group (TCG) see 7676 https://trustedcomputinggroup.org/resource/pc-client-platform-tpm-profile-ptp-specification/ 7677 7678 tp_printk [FTRACE] 7679 Have the tracepoints sent to printk as well as the 7680 tracing ring buffer. This is useful for early boot up 7681 where the system hangs or reboots and does not give the 7682 option for reading the tracing buffer or performing a 7683 ftrace_dump_on_oops. 7684 7685 To turn off having tracepoints sent to printk, 7686 echo 0 > /proc/sys/kernel/tracepoint_printk 7687 Note, echoing 1 into this file without the 7688 tp_printk kernel cmdline option has no effect. 7689 7690 The tp_printk_stop_on_boot (see below) can also be used 7691 to stop the printing of events to console at 7692 late_initcall_sync. 7693 7694 ** CAUTION ** 7695 7696 Having tracepoints sent to printk() and activating high 7697 frequency tracepoints such as irq or sched, can cause 7698 the system to live lock. 7699 7700 tp_printk_stop_on_boot [FTRACE] 7701 When tp_printk (above) is set, it can cause a lot of noise 7702 on the console. It may be useful to only include the 7703 printing of events during boot up, as user space may 7704 make the system inoperable. 7705 7706 This command line option will stop the printing of events 7707 to console at the late_initcall_sync() time frame. 7708 7709 trace_buf_size=nn[KMG] 7710 [FTRACE] will set tracing buffer size on each cpu. 7711 7712 trace_clock= [FTRACE] Set the clock used for tracing events 7713 at boot up. 7714 local - Use the per CPU time stamp counter 7715 (converted into nanoseconds). Fast, but 7716 depending on the architecture, may not be 7717 in sync between CPUs. 7718 global - Event time stamps are synchronized across 7719 CPUs. May be slower than the local clock, 7720 but better for some race conditions. 7721 counter - Simple counting of events (1, 2, ..) 7722 note, some counts may be skipped due to the 7723 infrastructure grabbing the clock more than 7724 once per event. 7725 uptime - Use jiffies as the time stamp. 7726 perf - Use the same clock that perf uses. 7727 mono - Use ktime_get_mono_fast_ns() for time stamps. 7728 mono_raw - Use ktime_get_raw_fast_ns() for time 7729 stamps. 7730 boot - Use ktime_get_boot_fast_ns() for time stamps. 7731 Architectures may add more clocks. See 7732 Documentation/trace/ftrace.rst for more details. 7733 7734 trace_event=[event-list] 7735 [FTRACE] Set and start specified trace events in order 7736 to facilitate early boot debugging. The event-list is a 7737 comma-separated list of trace events to enable. See 7738 also Documentation/trace/events.rst 7739 7740 To enable modules, use :mod: keyword: 7741 7742 trace_event=:mod:<module> 7743 7744 The value before :mod: will only enable specific events 7745 that are part of the module. See the above mentioned 7746 document for more information. 7747 7748 trace_instance=[instance-info] 7749 [FTRACE] Create a ring buffer instance early in boot up. 7750 This will be listed in: 7751 7752 /sys/kernel/tracing/instances 7753 7754 Events can be enabled at the time the instance is created 7755 via: 7756 7757 trace_instance=<name>,<system1>:<event1>,<system2>:<event2> 7758 7759 Note, the "<system*>:" portion is optional if the event is 7760 unique. 7761 7762 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall 7763 7764 will enable the "sched_switch" event (note, the "sched:" is optional, and 7765 the same thing would happen if it was left off). The irq_handler_entry 7766 event, and all events under the "initcall" system. 7767 7768 Flags can be added to the instance to modify its behavior when it is 7769 created. The flags are separated by '^'. 7770 7771 The available flags are: 7772 7773 traceoff - Have the tracing instance tracing disabled after it is created. 7774 traceprintk - Have trace_printk() write into this trace instance 7775 (note, "printk" and "trace_printk" can also be used) 7776 7777 trace_instance=foo^traceoff^traceprintk,sched,irq 7778 7779 The flags must come before the defined events. 7780 7781 If memory has been reserved (see memmap for x86), the instance 7782 can use that memory: 7783 7784 memmap=12M$0x284500000 trace_instance=boot_map@0x284500000:12M 7785 7786 The above will create a "boot_map" instance that uses the physical 7787 memory at 0x284500000 that is 12Megs. The per CPU buffers of that 7788 instance will be split up accordingly. 7789 7790 Alternatively, the memory can be reserved by the reserve_mem option: 7791 7792 reserve_mem=12M:4096:trace trace_instance=boot_map@trace 7793 7794 This will reserve 12 megabytes at boot up with a 4096 byte alignment 7795 and place the ring buffer in this memory. Note that due to KASLR, the 7796 memory may not be the same location each time, which will not preserve 7797 the buffer content. 7798 7799 Also note that the layout of the ring buffer data may change between 7800 kernel versions where the validator will fail and reset the ring buffer 7801 if the layout is not the same as the previous kernel. 7802 7803 If the ring buffer is used for persistent bootups and has events enabled, 7804 it is recommend to disable tracing so that events from a previous boot do not 7805 mix with events of the current boot (unless you are debugging a random crash 7806 at boot up). 7807 7808 reserve_mem=12M:4096:trace trace_instance=boot_map^traceoff^traceprintk@trace,sched,irq 7809 7810 Note, saving the trace buffer across reboots does require that the system 7811 is set up to not wipe memory. For instance, CONFIG_RESET_ATTACK_MITIGATION 7812 can force a memory reset on boot which will clear any trace that was stored. 7813 This is just one of many ways that can clear memory. Make sure your system 7814 keeps the content of memory across reboots before relying on this option. 7815 7816 NB: Both the mapped address and size must be page aligned for the architecture. 7817 7818 See also Documentation/trace/debugging.rst 7819 7820 7821 trace_options=[option-list] 7822 [FTRACE] Enable or disable tracer options at boot. 7823 The option-list is a comma delimited list of options 7824 that can be enabled or disabled just as if you were 7825 to echo the option name into 7826 7827 /sys/kernel/tracing/trace_options 7828 7829 For example, to enable stacktrace option (to dump the 7830 stack trace of each event), add to the command line: 7831 7832 trace_options=stacktrace 7833 7834 See also Documentation/trace/ftrace.rst "trace options" 7835 section. 7836 7837 trace_trigger=[trigger-list] 7838 [FTRACE] Add an event trigger on specific events. 7839 Set a trigger on top of a specific event, with an optional 7840 filter. 7841 7842 The format is "trace_trigger=<event>.<trigger>[ if <filter>],..." 7843 Where more than one trigger may be specified that are comma delimited. 7844 7845 For example: 7846 7847 trace_trigger="sched_switch.stacktrace if prev_state == 2" 7848 7849 The above will enable the "stacktrace" trigger on the "sched_switch" 7850 event but only trigger it if the "prev_state" of the "sched_switch" 7851 event is "2" (TASK_UNINTERRUPTIBLE). 7852 7853 See also "Event triggers" in Documentation/trace/events.rst 7854 7855 7856 traceoff_after_boot 7857 [FTRACE] Sometimes tracing is used to debug issues 7858 during the boot process. Since the trace buffer has a 7859 limited amount of storage, it may be prudent to 7860 disable tracing after the boot is finished, otherwise 7861 the critical information may be overwritten. With this 7862 option, the main tracing buffer will be turned off at 7863 the end of the boot process. 7864 7865 traceoff_on_warning 7866 [FTRACE] enable this option to disable tracing when a 7867 warning is hit. This turns off "tracing_on". Tracing can 7868 be enabled again by echoing '1' into the "tracing_on" 7869 file located in /sys/kernel/tracing/ 7870 7871 This option is useful, as it disables the trace before 7872 the WARNING dump is called, which prevents the trace to 7873 be filled with content caused by the warning output. 7874 7875 This option can also be set at run time via the sysctl 7876 option: kernel/traceoff_on_warning 7877 7878 transparent_hugepage= 7879 [KNL] 7880 Format: [always|madvise|never] 7881 Can be used to control the default behavior of the system 7882 with respect to transparent hugepages. 7883 See Documentation/admin-guide/mm/transhuge.rst 7884 for more details. 7885 7886 transparent_hugepage_shmem= [KNL] 7887 Format: [always|within_size|advise|never|deny|force] 7888 Can be used to control the hugepage allocation policy for 7889 the internal shmem mount. 7890 See Documentation/admin-guide/mm/transhuge.rst 7891 for more details. 7892 7893 transparent_hugepage_tmpfs= [KNL] 7894 Format: [always|within_size|advise|never] 7895 Can be used to control the default hugepage allocation policy 7896 for the tmpfs mount. 7897 See Documentation/admin-guide/mm/transhuge.rst 7898 for more details. 7899 7900 trusted.source= [KEYS] 7901 Format: <string> 7902 This parameter identifies the trust source as a backend 7903 for trusted keys implementation. Supported trust 7904 sources: 7905 - "tpm" 7906 - "tee" 7907 - "caam" 7908 - "dcp" 7909 - "pkwm" 7910 If not specified then it defaults to iterating through 7911 the trust source list starting with TPM and assigns the 7912 first trust source as a backend which is initialized 7913 successfully during iteration. 7914 7915 trusted.rng= [KEYS] 7916 Format: <string> 7917 The RNG used to generate key material for trusted keys. 7918 Can be one of: 7919 - "kernel" 7920 - the same value as trusted.source: "tpm" or "tee" 7921 - "default" 7922 If not specified, "default" is used. In this case, 7923 the RNG's choice is left to each individual trust source. 7924 7925 trusted.dcp_use_otp_key 7926 This is intended to be used in combination with 7927 trusted.source=dcp and will select the DCP OTP key 7928 instead of the DCP UNIQUE key blob encryption. 7929 7930 trusted.dcp_skip_zk_test 7931 This is intended to be used in combination with 7932 trusted.source=dcp and will disable the check if the 7933 blob key is all zeros. This is helpful for situations where 7934 having this key zero'ed is acceptable. E.g. in testing 7935 scenarios. 7936 7937 tsa= [X86] Control mitigation for Transient Scheduler 7938 Attacks on AMD CPUs. Search the following in your 7939 favourite search engine for more details: 7940 7941 "Technical guidance for mitigating transient scheduler 7942 attacks". 7943 7944 off - disable the mitigation 7945 on - enable the mitigation (default) 7946 user - mitigate only user/kernel transitions 7947 vm - mitigate only guest/host transitions 7948 7949 7950 tsc= Disable clocksource stability checks for TSC. 7951 Format: <string> 7952 [x86] reliable: mark tsc clocksource as reliable, this 7953 disables clocksource verification at runtime, as well 7954 as the stability checks done at bootup. Used to enable 7955 high-resolution timer mode on older hardware, and in 7956 virtualized environment. 7957 [x86] noirqtime: Do not use TSC to do irq accounting. 7958 Used to run time disable IRQ_TIME_ACCOUNTING on any 7959 platforms where RDTSC is slow and this accounting 7960 can add overhead. 7961 [x86] unstable: mark the TSC clocksource as unstable, this 7962 marks the TSC unconditionally unstable at bootup and 7963 avoids any further wobbles once the TSC watchdog notices. 7964 [x86] nowatchdog: disable clocksource watchdog. Used 7965 in situations with strict latency requirements (where 7966 interruptions from clocksource watchdog are not 7967 acceptable). 7968 [x86] recalibrate: force recalibration against a HW timer 7969 (HPET or PM timer) on systems whose TSC frequency was 7970 obtained from HW or FW using either an MSR or CPUID(0x15). 7971 Warn if the difference is more than 500 ppm. 7972 [x86] watchdog: Use TSC as the watchdog clocksource with 7973 which to check other HW timers (HPET or PM timer), but 7974 only on systems where TSC has been deemed trustworthy. 7975 This will be suppressed by an earlier tsc=nowatchdog and 7976 can be overridden by a later tsc=nowatchdog. A console 7977 message will flag any such suppression or overriding. 7978 7979 tsc_early_khz= [X86,EARLY] Skip early TSC calibration and use the given 7980 value instead. Useful when the early TSC frequency discovery 7981 procedure is not reliable, such as on overclocked systems 7982 with CPUID.16h support and partial CPUID.15h support. 7983 Format: <unsigned int> 7984 7985 tsx= [X86] Control Transactional Synchronization 7986 Extensions (TSX) feature in Intel processors that 7987 support TSX control. 7988 7989 This parameter controls the TSX feature. The options are: 7990 7991 on - Enable TSX on the system. Although there are 7992 mitigations for all known security vulnerabilities, 7993 TSX has been known to be an accelerator for 7994 several previous speculation-related CVEs, and 7995 so there may be unknown security risks associated 7996 with leaving it enabled. 7997 7998 off - Disable TSX on the system. (Note that this 7999 option takes effect only on newer CPUs which are 8000 not vulnerable to MDS, i.e., have 8001 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 8002 the new IA32_TSX_CTRL MSR through a microcode 8003 update. This new MSR allows for the reliable 8004 deactivation of the TSX functionality.) 8005 8006 auto - Disable TSX if X86_BUG_TAA is present, 8007 otherwise enable TSX on the system. 8008 8009 Not specifying this option is equivalent to tsx=off. 8010 8011 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 8012 for more details. 8013 8014 tsx_async_abort= [X86,INTEL,EARLY] Control mitigation for the TSX Async 8015 Abort (TAA) vulnerability. 8016 8017 Similar to Micro-architectural Data Sampling (MDS) 8018 certain CPUs that support Transactional 8019 Synchronization Extensions (TSX) are vulnerable to an 8020 exploit against CPU internal buffers which can forward 8021 information to a disclosure gadget under certain 8022 conditions. 8023 8024 In vulnerable processors, the speculatively forwarded 8025 data can be used in a cache side channel attack, to 8026 access data to which the attacker does not have direct 8027 access. 8028 8029 This parameter controls the TAA mitigation. The 8030 options are: 8031 8032 full - Enable TAA mitigation on vulnerable CPUs 8033 if TSX is enabled. 8034 8035 full,nosmt - Enable TAA mitigation and disable SMT on 8036 vulnerable CPUs. If TSX is disabled, SMT 8037 is not disabled because CPU is not 8038 vulnerable to cross-thread TAA attacks. 8039 off - Unconditionally disable TAA mitigation 8040 8041 On MDS-affected machines, tsx_async_abort=off can be 8042 prevented by an active MDS mitigation as both vulnerabilities 8043 are mitigated with the same mechanism so in order to disable 8044 this mitigation, you need to specify mds=off too. 8045 8046 Not specifying this option is equivalent to 8047 tsx_async_abort=full. On CPUs which are MDS affected 8048 and deploy MDS mitigation, TAA mitigation is not 8049 required and doesn't provide any additional 8050 mitigation. 8051 8052 For details see: 8053 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 8054 8055 turbografx.map[2|3]= [HW,JOY] 8056 TurboGraFX parallel port interface 8057 Format: 8058 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 8059 See also Documentation/input/devices/joystick-parport.rst 8060 8061 udbg-immortal [PPC] When debugging early kernel crashes that 8062 happen after console_init() and before a proper 8063 console driver takes over, this boot options might 8064 help "seeing" what's going on. 8065 8066 uhash_entries= [KNL,NET] 8067 Set number of hash buckets for UDP/UDP-Lite connections 8068 8069 uhci-hcd.ignore_oc= 8070 [USB] Ignore overcurrent events (default N). 8071 Some badly-designed motherboards generate lots of 8072 bogus events, for ports that aren't wired to 8073 anything. Set this parameter to avoid log spamming. 8074 Note that genuine overcurrent events won't be 8075 reported either. 8076 8077 unaligned_scalar_speed= 8078 [RISCV] 8079 Format: {slow | fast | unsupported} 8080 Allow skipping scalar unaligned access speed tests. This 8081 is useful for testing alternative code paths and to skip 8082 the tests in environments where they run too slowly. All 8083 CPUs must have the same scalar unaligned access speed. 8084 8085 unaligned_vector_speed= 8086 [RISCV] 8087 Format: {slow | fast | unsupported} 8088 Allow skipping vector unaligned access speed tests. This 8089 is useful for testing alternative code paths and to skip 8090 the tests in environments where they run too slowly. All 8091 CPUs must have the same vector unaligned access speed. 8092 8093 unknown_nmi_panic 8094 [X86] Cause panic on unknown NMI. 8095 8096 unwind_debug [X86-64,EARLY] 8097 Enable unwinder debug output. This can be 8098 useful for debugging certain unwinder error 8099 conditions, including corrupt stacks and 8100 bad/missing unwinder metadata. 8101 8102 usbcore.authorized_default= 8103 [USB] Default USB device authorization: 8104 (default -1 = authorized (same as 1), 8105 0 = not authorized, 1 = authorized, 2 = authorized 8106 if device connected to internal port) 8107 8108 usbcore.autosuspend= 8109 [USB] The autosuspend time delay (in seconds) used 8110 for newly-detected USB devices (default 2). This 8111 is the time required before an idle device will be 8112 autosuspended. Devices for which the delay is set 8113 to a negative value won't be autosuspended at all. 8114 8115 usbcore.usbfs_snoop= 8116 [USB] Set to log all usbfs traffic (default 0 = off). 8117 8118 usbcore.usbfs_snoop_max= 8119 [USB] Maximum number of bytes to snoop in each URB 8120 (default = 65536). 8121 8122 usbcore.blinkenlights= 8123 [USB] Set to cycle leds on hubs (default 0 = off). 8124 8125 usbcore.old_scheme_first= 8126 [USB] Start with the old device initialization 8127 scheme (default 0 = off). 8128 8129 usbcore.usbfs_memory_mb= 8130 [USB] Memory limit (in MB) for buffers allocated by 8131 usbfs (default = 16, 0 = max = 2047). 8132 8133 usbcore.use_both_schemes= 8134 [USB] Try the other device initialization scheme 8135 if the first one fails (default 1 = enabled). 8136 8137 usbcore.initial_descriptor_timeout= 8138 [USB] Specifies timeout for the initial 64-byte 8139 USB_REQ_GET_DESCRIPTOR request in milliseconds 8140 (default 5000 = 5.0 seconds). 8141 8142 usbcore.nousb [USB] Disable the USB subsystem 8143 8144 usbcore.quirks= 8145 [USB] A list of quirk entries to augment the built-in 8146 usb core quirk list. List entries are separated by 8147 commas. Each entry has the form 8148 VendorID:ProductID:Flags. The IDs are 4-digit hex 8149 numbers and Flags is a set of letters. Each letter 8150 will change the built-in quirk; setting it if it is 8151 clear and clearing it if it is set. The letters have 8152 the following meanings: 8153 a = USB_QUIRK_STRING_FETCH_255 (string 8154 descriptors must not be fetched using 8155 a 255-byte read); 8156 b = USB_QUIRK_RESET_RESUME (device can't resume 8157 correctly so reset it instead); 8158 c = USB_QUIRK_NO_SET_INTF (device can't handle 8159 Set-Interface requests); 8160 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 8161 handle its Configuration or Interface 8162 strings); 8163 e = USB_QUIRK_RESET (device can't be reset 8164 (e.g morph devices), don't use reset); 8165 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 8166 more interface descriptions than the 8167 bNumInterfaces count, and can't handle 8168 talking to these interfaces); 8169 g = USB_QUIRK_DELAY_INIT (device needs a pause 8170 during initialization, after we read 8171 the device descriptor); 8172 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 8173 high speed and super speed interrupt 8174 endpoints, the USB 2.0 and USB 3.0 spec 8175 require the interval in microframes (1 8176 microframe = 125 microseconds) to be 8177 calculated as interval = 2 ^ 8178 (bInterval-1). 8179 Devices with this quirk report their 8180 bInterval as the result of this 8181 calculation instead of the exponent 8182 variable used in the calculation); 8183 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 8184 handle device_qualifier descriptor 8185 requests); 8186 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 8187 generates spurious wakeup, ignore 8188 remote wakeup capability); 8189 k = USB_QUIRK_NO_LPM (device can't handle Link 8190 Power Management); 8191 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 8192 (Device reports its bInterval as linear 8193 frames instead of the USB 2.0 8194 calculation); 8195 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 8196 to be disconnected before suspend to 8197 prevent spurious wakeup); 8198 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 8199 pause after every control message); 8200 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 8201 delay after resetting its port); 8202 p = USB_QUIRK_SHORT_SET_ADDRESS_REQ_TIMEOUT 8203 (Reduce timeout of the SET_ADDRESS 8204 request from 5000 ms to 500 ms); 8205 q = USB_QUIRK_FORCE_ONE_CONFIG (Device 8206 claims zero configurations, 8207 forcing to 1); 8208 Example: quirks=0781:5580:bk,0a5c:5834:gij 8209 8210 usbhid.mousepoll= 8211 [USBHID] The interval which mice are to be polled at. 8212 8213 usbhid.jspoll= 8214 [USBHID] The interval which joysticks are to be polled at. 8215 8216 usbhid.kbpoll= 8217 [USBHID] The interval which keyboards are to be polled at. 8218 8219 usb-storage.delay_use= 8220 [UMS] The delay in seconds before a new device is 8221 scanned for Logical Units (default 1). 8222 Optionally the delay in milliseconds if the value has 8223 suffix with "ms". 8224 Example: delay_use=2567ms 8225 8226 usb-storage.quirks= 8227 [UMS] A list of quirks entries to supplement or 8228 override the built-in unusual_devs list. List 8229 entries are separated by commas. Each entry has 8230 the form VID:PID:Flags where VID and PID are Vendor 8231 and Product ID values (4-digit hex numbers) and 8232 Flags is a set of characters, each corresponding 8233 to a common usb-storage quirk flag as follows: 8234 a = SANE_SENSE (collect more than 18 bytes 8235 of sense data, not on uas); 8236 b = BAD_SENSE (don't collect more than 18 8237 bytes of sense data, not on uas); 8238 c = FIX_CAPACITY (decrease the reported 8239 device capacity by one sector); 8240 d = NO_READ_DISC_INFO (don't use 8241 READ_DISC_INFO command, not on uas); 8242 e = NO_READ_CAPACITY_16 (don't use 8243 READ_CAPACITY_16 command); 8244 f = NO_REPORT_OPCODES (don't use report opcodes 8245 command, uas only); 8246 g = MAX_SECTORS_240 (don't transfer more than 8247 240 sectors at a time, uas only); 8248 h = CAPACITY_HEURISTICS (decrease the 8249 reported device capacity by one 8250 sector if the number is odd); 8251 i = IGNORE_DEVICE (don't bind to this 8252 device); 8253 j = NO_REPORT_LUNS (don't use report luns 8254 command, uas only); 8255 k = NO_SAME (do not use WRITE_SAME, uas only) 8256 l = NOT_LOCKABLE (don't try to lock and 8257 unlock ejectable media, not on uas); 8258 m = MAX_SECTORS_64 (don't transfer more 8259 than 64 sectors = 32 KB at a time, 8260 not on uas); 8261 n = INITIAL_READ10 (force a retry of the 8262 initial READ(10) command, not on uas); 8263 o = CAPACITY_OK (accept the capacity 8264 reported by the device, not on uas); 8265 p = WRITE_CACHE (the device cache is ON 8266 by default, not on uas); 8267 r = IGNORE_RESIDUE (the device reports 8268 bogus residue values, not on uas); 8269 s = SINGLE_LUN (the device has only one 8270 Logical Unit); 8271 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 8272 commands, uas only); 8273 u = IGNORE_UAS (don't bind to the uas driver); 8274 w = NO_WP_DETECT (don't test whether the 8275 medium is write-protected). 8276 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 8277 even if the device claims no cache, 8278 not on uas) 8279 Example: quirks=0419:aaf5:rl,0421:0433:rc 8280 8281 user_debug= [KNL,ARM] 8282 Format: <int> 8283 See arch/arm/Kconfig.debug help text. 8284 1 - undefined instruction events 8285 2 - system calls 8286 4 - invalid data aborts 8287 8 - SIGSEGV faults 8288 16 - SIGBUS faults 8289 Example: user_debug=31 8290 8291 vdso= [X86,SH,SPARC] 8292 On X86_32, this is an alias for vdso32=. Otherwise: 8293 8294 vdso=1: enable VDSO (the default) 8295 vdso=0: disable VDSO mapping 8296 8297 vdso32= [X86] Control the 32-bit vDSO 8298 vdso32=1: enable 32-bit VDSO 8299 vdso32=0 or vdso32=2: disable 32-bit VDSO 8300 8301 See the help text for CONFIG_COMPAT_VDSO for more 8302 details. If CONFIG_COMPAT_VDSO is set, the default is 8303 vdso32=0; otherwise, the default is vdso32=1. 8304 8305 For compatibility with older kernels, vdso32=2 is an 8306 alias for vdso32=0. 8307 8308 Try vdso32=0 if you encounter an error that says: 8309 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 8310 8311 video= [FB,EARLY] Frame buffer configuration 8312 See Documentation/fb/modedb.rst. 8313 8314 video.brightness_switch_enabled= [ACPI] 8315 Format: [0|1] 8316 If set to 1, on receiving an ACPI notify event 8317 generated by hotkey, video driver will adjust brightness 8318 level and then send out the event to user space through 8319 the allocated input device. If set to 0, video driver 8320 will only send out the event without touching backlight 8321 brightness level. 8322 default: 1 8323 8324 virtio_mmio.device= 8325 [VMMIO] Memory mapped virtio (platform) device. 8326 8327 <size>@<baseaddr>:<irq>[:<id>] 8328 where: 8329 <size> := size (can use standard suffixes 8330 like K, M and G) 8331 <baseaddr> := physical base address 8332 <irq> := interrupt number (as passed to 8333 request_irq()) 8334 <id> := (optional) platform device id 8335 example: 8336 virtio_mmio.device=1K@0x100b0000:48:7 8337 8338 Can be used multiple times for multiple devices. 8339 8340 vga= [BOOT,X86-32] Select a particular video mode 8341 See Documentation/arch/x86/boot.rst and 8342 Documentation/admin-guide/svga.rst. 8343 Use vga=ask for menu. 8344 This is actually a boot loader parameter; the value is 8345 passed to the kernel using a special protocol. 8346 8347 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 8348 May slow down system boot speed, especially when 8349 enabled on systems with a large amount of memory. 8350 All options are enabled by default, and this 8351 interface is meant to allow for selectively 8352 enabling or disabling specific virtual memory 8353 debugging features. 8354 8355 Available options are: 8356 P Enable page structure init time poisoning 8357 - Disable all of the above options 8358 8359 vmalloc=nn[KMG] [KNL,BOOT,EARLY] Forces the vmalloc area to have an 8360 exact size of <nn>. This can be used to increase 8361 the minimum size (128MB on x86, arm32 platforms). 8362 It can also be used to decrease the size and leave more room 8363 for directly mapped kernel RAM. Note that this parameter does 8364 not exist on many other platforms (including arm64, alpha, 8365 loongarch, arc, csky, hexagon, microblaze, mips, nios2, openrisc, 8366 parisc, m64k, powerpc, riscv, sh, um, xtensa, s390, sparc). 8367 8368 vmcp_cma=nn[MG] [KNL,S390,EARLY] 8369 Sets the memory size reserved for contiguous memory 8370 allocations for the vmcp device driver. 8371 8372 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 8373 Format: <command> 8374 8375 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 8376 Format: <command> 8377 8378 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 8379 Format: <command> 8380 8381 vmscape= [X86] Controls mitigation for VMscape attacks. 8382 VMscape attacks can leak information from a userspace 8383 hypervisor to a guest via speculative side-channels. 8384 8385 off - disable the mitigation 8386 ibpb - use Indirect Branch Prediction Barrier 8387 (IBPB) mitigation (default) 8388 force - force vulnerability detection even on 8389 unaffected processors 8390 8391 vsyscall= [X86-64,EARLY] 8392 Controls the behavior of vsyscalls (i.e. calls to 8393 fixed addresses of 0xffffffffff600x00 from legacy 8394 code). Most statically-linked binaries and older 8395 versions of glibc use these calls. Because these 8396 functions are at fixed addresses, they make nice 8397 targets for exploits that can control RIP. 8398 8399 emulate Vsyscalls turn into traps and are emulated 8400 reasonably safely. The vsyscall page is 8401 readable. 8402 8403 xonly [default] Vsyscalls turn into traps and are 8404 emulated reasonably safely. The vsyscall 8405 page is not readable. 8406 8407 none Vsyscalls don't work at all. This makes 8408 them quite hard to use for exploits but 8409 might break your system. 8410 8411 vt.color= [VT] Default text color. 8412 Format: 0xYX, X = foreground, Y = background. 8413 Default: 0x07 = light gray on black. 8414 8415 vt.cur_default= [VT] Default cursor shape. 8416 Format: 0xCCBBAA, where AA, BB, and CC are the same as 8417 the parameters of the <Esc>[?A;B;Cc escape sequence; 8418 see vga-softcursor.rst. Default: 2 = underline. 8419 8420 vt.default_blu= [VT] 8421 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 8422 Change the default blue palette of the console. 8423 This is a 16-member array composed of values 8424 ranging from 0-255. 8425 8426 vt.default_grn= [VT] 8427 Format: <green0>,<green1>,<green2>,...,<green15> 8428 Change the default green palette of the console. 8429 This is a 16-member array composed of values 8430 ranging from 0-255. 8431 8432 vt.default_red= [VT] 8433 Format: <red0>,<red1>,<red2>,...,<red15> 8434 Change the default red palette of the console. 8435 This is a 16-member array composed of values 8436 ranging from 0-255. 8437 8438 vt.default_utf8= 8439 [VT] 8440 Format=<0|1> 8441 Set system-wide default UTF-8 mode for all tty's. 8442 Default is 1, i.e. UTF-8 mode is enabled for all 8443 newly opened terminals. 8444 8445 vt.global_cursor_default= 8446 [VT] 8447 Format=<-1|0|1> 8448 Set system-wide default for whether a cursor 8449 is shown on new VTs. Default is -1, 8450 i.e. cursors will be created by default unless 8451 overridden by individual drivers. 0 will hide 8452 cursors, 1 will display them. 8453 8454 vt.italic= [VT] Default color for italic text; 0-15. 8455 Default: 2 = green. 8456 8457 vt.underline= [VT] Default color for underlined text; 0-15. 8458 Default: 3 = cyan. 8459 8460 watchdog timers [HW,WDT] For information on watchdog timers, 8461 see Documentation/watchdog/watchdog-parameters.rst 8462 or other driver-specific files in the 8463 Documentation/watchdog/ directory. 8464 8465 watchdog_thresh= 8466 [KNL] 8467 Set the hard lockup detector stall duration 8468 threshold in seconds. The soft lockup detector 8469 threshold is set to twice the value. A value of 0 8470 disables both lockup detectors. Default is 10 8471 seconds. 8472 8473 workqueue.unbound_cpus= 8474 [KNL,SMP] Specify to constrain one or some CPUs 8475 to use in unbound workqueues. 8476 Format: <cpu-list> 8477 By default, all online CPUs are available for 8478 unbound workqueues. 8479 8480 workqueue.watchdog_thresh= 8481 If CONFIG_WQ_WATCHDOG is configured, workqueue can 8482 warn stall conditions and dump internal state to 8483 help debugging. 0 disables workqueue stall 8484 detection; otherwise, it's the stall threshold 8485 duration in seconds. The default value is 30 and 8486 it can be updated at runtime by writing to the 8487 corresponding sysfs file. 8488 8489 workqueue.panic_on_stall=<uint> 8490 Panic when workqueue stall is detected by 8491 CONFIG_WQ_WATCHDOG. It sets the number times of the 8492 stall to trigger panic. 8493 8494 The default is set by CONFIG_BOOTPARAM_WQ_STALL_PANIC, 8495 which is 0 (disabled) if not configured. 8496 8497 workqueue.panic_on_stall_time=<uint> 8498 Panic when a workqueue stall has been continuous for 8499 the specified number of seconds. Unlike panic_on_stall 8500 which counts accumulated stall events, this triggers 8501 based on the duration of a single continuous stall. 8502 8503 The default is 0, which disables the time-based panic. 8504 8505 workqueue.cpu_intensive_thresh_us= 8506 Per-cpu work items which run for longer than this 8507 threshold are automatically considered CPU intensive 8508 and excluded from concurrency management to prevent 8509 them from noticeably delaying other per-cpu work 8510 items. Default is 10000 (10ms). 8511 8512 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8513 will report the work functions which violate this 8514 threshold repeatedly. They are likely good 8515 candidates for using WQ_UNBOUND workqueues instead. 8516 8517 workqueue.cpu_intensive_warning_thresh=<uint> 8518 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8519 will report the work functions which violate the 8520 intensive_threshold_us repeatedly. In order to prevent 8521 spurious warnings, start printing only after a work 8522 function has violated this threshold number of times. 8523 8524 The default is 4 times. 0 disables the warning. 8525 8526 workqueue.power_efficient 8527 Per-cpu workqueues are generally preferred because 8528 they show better performance thanks to cache 8529 locality; unfortunately, per-cpu workqueues tend to 8530 be more power hungry than unbound workqueues. 8531 8532 Enabling this makes the per-cpu workqueues which 8533 were observed to contribute significantly to power 8534 consumption unbound, leading to measurably lower 8535 power usage at the cost of small performance 8536 overhead. 8537 8538 The default value of this parameter is determined by 8539 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 8540 8541 workqueue.default_affinity_scope= 8542 Select the default affinity scope to use for unbound 8543 workqueues. Can be one of "cpu", "smt", "cache", 8544 "numa" and "system". Default is "cache". For more 8545 information, see the Affinity Scopes section in 8546 Documentation/core-api/workqueue.rst. 8547 8548 This can be changed after boot by writing to the 8549 matching /sys/module/workqueue/parameters file. All 8550 workqueues with the "default" affinity scope will be 8551 updated accordingly. 8552 8553 workqueue.debug_force_rr_cpu 8554 Workqueue used to implicitly guarantee that work 8555 items queued without explicit CPU specified are put 8556 on the local CPU. This guarantee is no longer true 8557 and while local CPU is still preferred work items 8558 may be put on foreign CPUs. This debug option 8559 forces round-robin CPU selection to flush out 8560 usages which depend on the now broken guarantee. 8561 When enabled, memory and cache locality will be 8562 impacted. 8563 8564 writecombine= [LOONGARCH,EARLY] Control the MAT (Memory Access 8565 Type) of ioremap_wc(). 8566 8567 on - Enable writecombine, use WUC for ioremap_wc() 8568 off - Disable writecombine, use SUC for ioremap_wc() 8569 8570 x2apic_phys [X86-64,APIC,EARLY] Use x2apic physical mode instead of 8571 default x2apic cluster mode on platforms 8572 supporting x2apic. 8573 8574 xen_512gb_limit [KNL,X86-64,XEN] 8575 Restricts the kernel running paravirtualized under Xen 8576 to use only up to 512 GB of RAM. The reason to do so is 8577 crash analysis tools and Xen tools for doing domain 8578 save/restore/migration must be enabled to handle larger 8579 domains. 8580 8581 xen_console_io [XEN,EARLY] 8582 Boolean option to enable/disable the usage of the Xen 8583 console_io hypercalls to read and write to the console. 8584 Mostly useful for debugging and development. 8585 8586 xen_emul_unplug= [HW,X86,XEN,EARLY] 8587 Unplug Xen emulated devices 8588 Format: [unplug0,][unplug1] 8589 ide-disks -- unplug primary master IDE devices 8590 aux-ide-disks -- unplug non-primary-master IDE devices 8591 nics -- unplug network devices 8592 all -- unplug all emulated devices (NICs and IDE disks) 8593 unnecessary -- unplugging emulated devices is 8594 unnecessary even if the host did not respond to 8595 the unplug protocol 8596 never -- do not unplug even if version check succeeds 8597 8598 xen_legacy_crash [X86,XEN,EARLY] 8599 Crash from Xen panic notifier, without executing late 8600 panic() code such as dumping handler. 8601 8602 xen_mc_debug [X86,XEN,EARLY] 8603 Enable multicall debugging when running as a Xen PV guest. 8604 Enabling this feature will reduce performance a little 8605 bit, so it should only be enabled for obtaining extended 8606 debug data in case of multicall errors. 8607 8608 xen_msr_safe= [X86,XEN,EARLY] 8609 Format: <bool> 8610 Select whether to always use non-faulting (safe) MSR 8611 access functions when running as Xen PV guest. The 8612 default value is controlled by CONFIG_XEN_PV_MSR_SAFE. 8613 8614 xen_nopv [X86] 8615 Disables the PV optimizations forcing the HVM guest to 8616 run as generic HVM guest with no PV drivers. 8617 This option is obsoleted by the "nopv" option, which 8618 has equivalent effect for XEN platform. 8619 8620 xen_no_vector_callback 8621 [KNL,X86,XEN,EARLY] Disable the vector callback for Xen 8622 event channel interrupts. 8623 8624 xen_scrub_pages= [XEN] 8625 Boolean option to control scrubbing pages before giving them back 8626 to Xen, for use by other domains. Can be also changed at runtime 8627 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 8628 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 8629 8630 xen_timer_slop= [X86-64,XEN,EARLY] 8631 Set the timer slop (in nanoseconds) for the virtual Xen 8632 timers (default is 100000). This adjusts the minimum 8633 delta of virtualized Xen timers, where lower values 8634 improve timer resolution at the expense of processing 8635 more timer interrupts. 8636 8637 xen.balloon_boot_timeout= [XEN] 8638 The time (in seconds) to wait before giving up to boot 8639 in case initial ballooning fails to free enough memory. 8640 Applies only when running as HVM or PVH guest and 8641 started with less memory configured than allowed at 8642 max. Default is 180. 8643 8644 xen.event_eoi_delay= [XEN] 8645 How long to delay EOI handling in case of event 8646 storms (jiffies). Default is 10. 8647 8648 xen.event_loop_timeout= [XEN] 8649 After which time (jiffies) the event handling loop 8650 should start to delay EOI handling. Default is 2. 8651 8652 xen.fifo_events= [XEN] 8653 Boolean parameter to disable using fifo event handling 8654 even if available. Normally fifo event handling is 8655 preferred over the 2-level event handling, as it is 8656 fairer and the number of possible event channels is 8657 much higher. Default is on (use fifo events). 8658 8659 xirc2ps_cs= [NET,PCMCIA] 8660 Format: 8661 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 8662 8663 xive= [PPC] 8664 By default on POWER9 and above, the kernel will 8665 natively use the XIVE interrupt controller. This option 8666 allows the fallback firmware mode to be used: 8667 8668 off Fallback to firmware control of XIVE interrupt 8669 controller on both pseries and powernv 8670 platforms. Only useful on POWER9 and above. 8671 8672 xive.store-eoi=off [PPC] 8673 By default on POWER10 and above, the kernel will use 8674 stores for EOI handling when the XIVE interrupt mode 8675 is active. This option allows the XIVE driver to use 8676 loads instead, as on POWER9. 8677 8678 xhci-hcd.quirks [USB,KNL] 8679 A hex value specifying bitmask with supplemental xhci 8680 host controller quirks. Meaning of each bit can be 8681 consulted in header drivers/usb/host/xhci.h. 8682 8683 xmon [PPC,EARLY] 8684 Format: { early | on | rw | ro | off } 8685 Controls if xmon debugger is enabled. Default is off. 8686 Passing only "xmon" is equivalent to "xmon=early". 8687 early Call xmon as early as possible on boot; xmon 8688 debugger is called from setup_arch(). 8689 on xmon debugger hooks will be installed so xmon 8690 is only called on a kernel crash. Default mode, 8691 i.e. either "ro" or "rw" mode, is controlled 8692 with CONFIG_XMON_DEFAULT_RO_MODE. 8693 rw xmon debugger hooks will be installed so xmon 8694 is called only on a kernel crash, mode is write, 8695 meaning SPR registers, memory and, other data 8696 can be written using xmon commands. 8697 ro same as "rw" option above but SPR registers, 8698 memory, and other data can't be written using 8699 xmon commands. 8700 off xmon is disabled. 8701