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.poweroff_on_fatal= [ACPI] 194 {0 | 1} 195 Causes the system to poweroff when the ACPI bytecode signals 196 a fatal error. The default value of this setting is 1. 197 Overriding this value should only be done for diagnosing 198 ACPI firmware problems, as the system might behave erratically 199 after having encountered a fatal ACPI error. 200 201 acpi_enforce_resources= [ACPI] 202 { strict | lax | no } 203 Check for resource conflicts between native drivers 204 and ACPI OperationRegions (SystemIO and SystemMemory 205 only). IO ports and memory declared in ACPI might be 206 used by the ACPI subsystem in arbitrary AML code and 207 can interfere with legacy drivers. 208 strict (default): access to resources claimed by ACPI 209 is denied; legacy drivers trying to access reserved 210 resources will fail to bind to device using them. 211 lax: access to resources claimed by ACPI is allowed; 212 legacy drivers trying to access reserved resources 213 will bind successfully but a warning message is logged. 214 no: ACPI OperationRegions are not marked as reserved, 215 no further checks are performed. 216 217 acpi_force_table_verification [HW,ACPI,EARLY] 218 Enable table checksum verification during early stage. 219 By default, this is disabled due to x86 early mapping 220 size limitation. 221 222 acpi_irq_balance [HW,ACPI] 223 ACPI will balance active IRQs 224 default in APIC mode 225 226 acpi_irq_nobalance [HW,ACPI] 227 ACPI will not move active IRQs (default) 228 default in PIC mode 229 230 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 231 Format: <irq>,<irq>... 232 233 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 234 use by PCI 235 Format: <irq>,<irq>... 236 237 acpi_mask_gpe= [HW,ACPI] 238 Due to the existence of _Lxx/_Exx, some GPEs triggered 239 by unsupported hardware/firmware features can result in 240 GPE floodings that cannot be automatically disabled by 241 the GPE dispatcher. 242 This facility can be used to prevent such uncontrolled 243 GPE floodings. 244 Format: <byte> or <bitmap-list> 245 246 acpi_no_auto_serialize [HW,ACPI] 247 Disable auto-serialization of AML methods 248 AML control methods that contain the opcodes to create 249 named objects will be marked as "Serialized" by the 250 auto-serialization feature. 251 This feature is enabled by default. 252 This option allows to turn off the feature. 253 254 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 255 kernels. 256 257 acpi_no_static_ssdt [HW,ACPI,EARLY] 258 Disable installation of static SSDTs at early boot time 259 By default, SSDTs contained in the RSDT/XSDT will be 260 installed automatically and they will appear under 261 /sys/firmware/acpi/tables. 262 This option turns off this feature. 263 Note that specifying this option does not affect 264 dynamic table installation which will install SSDT 265 tables to /sys/firmware/acpi/tables/dynamic. 266 267 acpi_no_watchdog [HW,ACPI,WDT] 268 Ignore the ACPI-based watchdog interface (WDAT) and let 269 a native driver control the watchdog device instead. 270 271 acpi_rsdp= [ACPI,EFI,KEXEC,EARLY] 272 Pass the RSDP address to the kernel, mostly used 273 on machines running EFI runtime service to boot the 274 second kernel for kdump. 275 276 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 277 Format: To spoof as Windows 98: ="Microsoft Windows" 278 279 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 280 of 2 which is mandated by ACPI 6) as the supported ACPI 281 specification revision (when using this switch, it may 282 be necessary to carry out a cold reboot _twice_ in a 283 row to make it take effect on the platform firmware). 284 285 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 286 acpi_osi="string1" # add string1 287 acpi_osi="!string2" # remove string2 288 acpi_osi=!* # remove all strings 289 acpi_osi=! # disable all built-in OS vendor 290 strings 291 acpi_osi=!! # enable all built-in OS vendor 292 strings 293 acpi_osi= # disable all strings 294 295 'acpi_osi=!' can be used in combination with single or 296 multiple 'acpi_osi="string1"' to support specific OS 297 vendor string(s). Note that such command can only 298 affect the default state of the OS vendor strings, thus 299 it cannot affect the default state of the feature group 300 strings and the current state of the OS vendor strings, 301 specifying it multiple times through kernel command line 302 is meaningless. This command is useful when one do not 303 care about the state of the feature group strings which 304 should be controlled by the OSPM. 305 Examples: 306 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 307 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 308 can make '_OSI("Windows 2000")' TRUE. 309 310 'acpi_osi=' cannot be used in combination with other 311 'acpi_osi=' command lines, the _OSI method will not 312 exist in the ACPI namespace. NOTE that such command can 313 only affect the _OSI support state, thus specifying it 314 multiple times through kernel command line is also 315 meaningless. 316 Examples: 317 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 318 FALSE. 319 320 'acpi_osi=!*' can be used in combination with single or 321 multiple 'acpi_osi="string1"' to support specific 322 string(s). Note that such command can affect the 323 current state of both the OS vendor strings and the 324 feature group strings, thus specifying it multiple times 325 through kernel command line is meaningful. But it may 326 still not able to affect the final state of a string if 327 there are quirks related to this string. This command 328 is useful when one want to control the state of the 329 feature group strings to debug BIOS issues related to 330 the OSPM features. 331 Examples: 332 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 333 '_OSI("Module Device")' FALSE. 334 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 335 '_OSI("Module Device")' TRUE. 336 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 337 equivalent to 338 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 339 and 340 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 341 they all will make '_OSI("Windows 2000")' TRUE. 342 343 acpi_pm_good [X86] 344 Override the pmtimer bug detection: force the kernel 345 to assume that this machine's pmtimer latches its value 346 and always returns good values. 347 348 acpi_sci= [HW,ACPI,EARLY] ACPI System Control Interrupt trigger mode 349 Format: { level | edge | high | low } 350 351 acpi_skip_timer_override [HW,ACPI,EARLY] 352 Recognize and ignore IRQ0/pin2 Interrupt Override. 353 For broken nForce2 BIOS resulting in XT-PIC timer. 354 355 acpi_sleep= [HW,ACPI] Sleep options 356 Format: { s3_bios, s3_mode, s3_beep, s4_hwsig, 357 s4_nohwsig, old_ordering, nonvs, 358 sci_force_enable, nobl } 359 See Documentation/power/video.rst for information on 360 s3_bios and s3_mode. 361 s3_beep is for debugging; it makes the PC's speaker beep 362 as soon as the kernel's real-mode entry point is called. 363 s4_hwsig causes the kernel to check the ACPI hardware 364 signature during resume from hibernation, and gracefully 365 refuse to resume if it has changed. This complies with 366 the ACPI specification but not with reality, since 367 Windows does not do this and many laptops do change it 368 on docking. So the default behaviour is to allow resume 369 and simply warn when the signature changes, unless the 370 s4_hwsig option is enabled. 371 s4_nohwsig prevents ACPI hardware signature from being 372 used (or even warned about) during resume. 373 old_ordering causes the ACPI 1.0 ordering of the _PTS 374 control method, with respect to putting devices into 375 low power states, to be enforced (the ACPI 2.0 ordering 376 of _PTS is used by default). 377 nonvs prevents the kernel from saving/restoring the 378 ACPI NVS memory during suspend/hibernation and resume. 379 sci_force_enable causes the kernel to set SCI_EN directly 380 on resume from S1/S3 (which is against the ACPI spec, 381 but some broken systems don't work without it). 382 nobl causes the internal blacklist of systems known to 383 behave incorrectly in some ways with respect to system 384 suspend and resume to be ignored (use wisely). 385 386 acpi_use_timer_override [HW,ACPI,EARLY] 387 Use timer override. For some broken Nvidia NF5 boards 388 that require a timer override, but don't have HPET 389 390 add_efi_memmap [EFI,X86,EARLY] Include EFI memory map in 391 kernel's map of available physical RAM. 392 393 agp= [AGP] 394 { off | try_unsupported } 395 off: disable AGP support 396 try_unsupported: try to drive unsupported chipsets 397 (may crash computer or cause data corruption) 398 399 ALSA [HW,ALSA] 400 See Documentation/sound/alsa-configuration.rst 401 402 alignment= [KNL,ARM] 403 Allow the default userspace alignment fault handler 404 behaviour to be specified. Bit 0 enables warnings, 405 bit 1 enables fixups, and bit 2 sends a segfault. 406 407 align_va_addr= [X86-64] 408 Align virtual addresses by clearing slice [14:12] when 409 allocating a VMA at process creation time. This option 410 gives you up to 3% performance improvement on AMD F15h 411 machines (where it is enabled by default) for a 412 CPU-intensive style benchmark, and it can vary highly in 413 a microbenchmark depending on workload and compiler. 414 415 32: only for 32-bit processes 416 64: only for 64-bit processes 417 on: enable for both 32- and 64-bit processes 418 off: disable for both 32- and 64-bit processes 419 420 alloc_snapshot [FTRACE] 421 Allocate the ftrace snapshot buffer on boot up when the 422 main buffer is allocated. This is handy if debugging 423 and you need to use tracing_snapshot() on boot up, and 424 do not want to use tracing_snapshot_alloc() as it needs 425 to be done where GFP_KERNEL allocations are allowed. 426 427 allow_mismatched_32bit_el0 [ARM64,EARLY] 428 Allow execve() of 32-bit applications and setting of the 429 PER_LINUX32 personality on systems where only a strict 430 subset of the CPUs support 32-bit EL0. When this 431 parameter is present, the set of CPUs supporting 32-bit 432 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0 433 and hot-unplug operations may be restricted. 434 435 See Documentation/arch/arm64/asymmetric-32bit.rst for more 436 information. 437 438 amd_iommu= [HW,X86-64] 439 Pass parameters to the AMD IOMMU driver in the system. 440 Possible values are: 441 fullflush - Deprecated, equivalent to iommu.strict=1 442 off - do not initialize any AMD IOMMU found in 443 the system 444 force_isolation - Force device isolation for all 445 devices. The IOMMU driver is not 446 allowed anymore to lift isolation 447 requirements as needed. This option 448 does not override iommu=pt 449 force_enable - Force enable the IOMMU on platforms known 450 to be buggy with IOMMU enabled. Use this 451 option with care. 452 pgtbl_v1 - Use v1 page table for DMA-API (Default). 453 pgtbl_v2 - Use v2 page table for DMA-API. 454 irtcachedis - Disable Interrupt Remapping Table (IRT) caching. 455 nohugepages - Limit page-sizes used for v1 page-tables 456 to 4 KiB. 457 v2_pgsizes_only - Limit page-sizes used for v1 page-tables 458 to 4KiB/2Mib/1GiB. 459 460 461 amd_iommu_dump= [HW,X86-64] 462 Enable AMD IOMMU driver option to dump the ACPI table 463 for AMD IOMMU. With this option enabled, AMD IOMMU 464 driver will print ACPI tables for AMD IOMMU during 465 IOMMU initialization. 466 467 amd_iommu_intr= [HW,X86-64] 468 Specifies one of the following AMD IOMMU interrupt 469 remapping modes: 470 legacy - Use legacy interrupt remapping mode. 471 vapic - Use virtual APIC mode, which allows IOMMU 472 to inject interrupts directly into guest. 473 This mode requires kvm-amd.avic=1. 474 (Default when IOMMU HW support is present.) 475 476 amd_pstate= [X86,EARLY] 477 disable 478 Do not enable amd_pstate as the default 479 scaling driver for the supported processors 480 passive 481 Use amd_pstate with passive mode as a scaling driver. 482 In this mode autonomous selection is disabled. 483 Driver requests a desired performance level and platform 484 tries to match the same performance level if it is 485 satisfied by guaranteed performance level. 486 active 487 Use amd_pstate_epp driver instance as the scaling driver, 488 driver provides a hint to the hardware if software wants 489 to bias toward performance (0x0) or energy efficiency (0xff) 490 to the CPPC firmware. then CPPC power algorithm will 491 calculate the runtime workload and adjust the realtime cores 492 frequency. 493 guided 494 Activate guided autonomous mode. Driver requests minimum and 495 maximum performance level and the platform autonomously 496 selects a performance level in this range and appropriate 497 to the current workload. 498 499 amd_prefcore= 500 [X86] 501 disable 502 Disable amd-pstate preferred core. 503 504 amd_dynamic_epp= 505 [X86] 506 disable 507 Disable amd-pstate dynamic EPP. 508 enable 509 Enable amd-pstate dynamic EPP. 510 511 amijoy.map= [HW,JOY] Amiga joystick support 512 Map of devices attached to JOY0DAT and JOY1DAT 513 Format: <a>,<b> 514 See also Documentation/input/joydev/joystick.rst 515 516 analog.map= [HW,JOY] Analog joystick and gamepad support 517 Specifies type or capabilities of an analog joystick 518 connected to one of 16 gameports 519 Format: <type1>,<type2>,..<type16> 520 521 apc= [HW,SPARC] 522 Power management functions (SPARCstation-4/5 + deriv.) 523 Format: noidle 524 Disable APC CPU standby support. SPARCstation-Fox does 525 not play well with APC CPU idle - disable it if you have 526 APC and your system crashes randomly. 527 528 apic [APIC,X86-64] Use IO-APIC. Default. 529 530 apic= [APIC,X86,EARLY] Advanced Programmable Interrupt Controller 531 Change the output verbosity while booting 532 Format: { quiet (default) | verbose | debug } 533 Change the amount of debugging information output 534 when initialising the APIC and IO-APIC components. 535 536 apic_extnmi= [APIC,X86,EARLY] External NMI delivery setting 537 Format: { bsp (default) | all | none } 538 bsp: External NMI is delivered only to CPU 0 539 all: External NMIs are broadcast to all CPUs as a 540 backup of CPU 0 541 none: External NMI is masked for all CPUs. This is 542 useful so that a dump capture kernel won't be 543 shot down by NMI 544 545 apicpmtimer Do APIC timer calibration using the pmtimer. Implies 546 apicmaintimer. Useful when your PIT timer is totally 547 broken. 548 549 autoconf= [IPV6] 550 See Documentation/networking/ipv6.rst. 551 552 apm= [APM] Advanced Power Management 553 See header of arch/x86/kernel/apm_32.c. 554 555 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 556 Format: { "0" | "1" } 557 See security/apparmor/Kconfig help text 558 0 -- disable. 559 1 -- enable. 560 Default value is set via kernel config option. 561 562 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards 563 Format: <io>,<irq>,<nodeID> 564 565 arm64.no32bit_el0 [ARM64] Unconditionally disable the execution of 566 32 bit applications. 567 568 arm64.nobti [ARM64] Unconditionally disable Branch Target 569 Identification support 570 571 arm64.nogcs [ARM64] Unconditionally disable Guarded Control Stack 572 support 573 574 arm64.nomops [ARM64] Unconditionally disable Memory Copy and Memory 575 Set instructions support 576 577 arm64.nompam [ARM64] Unconditionally disable Memory Partitioning And 578 Monitoring support 579 580 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension 581 support 582 583 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 584 support 585 586 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix 587 Extension support 588 589 arm64.nosve [ARM64] Unconditionally disable Scalable Vector 590 Extension support 591 592 ataflop= [HW,M68k] 593 594 atarimouse= [HW,MOUSE] Atari Mouse 595 596 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 597 EzKey and similar keyboards 598 599 atkbd.reset= [HW] Reset keyboard during initialization 600 601 atkbd.set= [HW] Select keyboard code set 602 Format: <int> (2 = AT (default), 3 = PS/2) 603 604 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 605 keyboards 606 607 atkbd.softraw= [HW] Choose between synthetic and real raw mode 608 Format: <bool> (0 = real, 1 = synthetic (default)) 609 610 atkbd.softrepeat= [HW] 611 Use software keyboard repeat 612 613 audit= [KNL] Enable the audit sub-system 614 Format: { "0" | "1" | "off" | "on" } 615 0 | off - kernel audit is disabled and can not be 616 enabled until the next reboot 617 unset - kernel audit is initialized but disabled and 618 will be fully enabled by the userspace auditd. 619 1 | on - kernel audit is initialized and partially 620 enabled, storing at most audit_backlog_limit 621 messages in RAM until it is fully enabled by the 622 userspace auditd. 623 Default: unset 624 625 audit_backlog_limit= [KNL] Set the audit queue size limit. 626 Format: <int> (must be >=0) 627 Default: 64 628 629 bau= [X86_UV] Enable the BAU on SGI UV. The default 630 behavior is to disable the BAU (i.e. bau=0). 631 Format: { "0" | "1" } 632 0 - Disable the BAU. 633 1 - Enable the BAU. 634 unset - Disable the BAU. 635 636 baycom_epp= [HW,AX25] 637 Format: <io>,<mode> 638 639 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem 640 Format: <io>,<mode> 641 See header of drivers/net/hamradio/baycom_par.c. 642 643 baycom_ser_fdx= [HW,AX25] 644 BayCom Serial Port AX.25 Modem (Full Duplex Mode) 645 Format: <io>,<irq>,<mode>[,<baud>] 646 See header of drivers/net/hamradio/baycom_ser_fdx.c. 647 648 baycom_ser_hdx= [HW,AX25] 649 BayCom Serial Port AX.25 Modem (Half Duplex Mode) 650 Format: <io>,<irq>,<mode> 651 See header of drivers/net/hamradio/baycom_ser_hdx.c. 652 653 bdev_allow_write_mounted= 654 Format: <bool> 655 Control the ability to open a mounted block device 656 for writing, i.e., allow / disallow writes that bypass 657 the FS. This was implemented as a means to prevent 658 fuzzers from crashing the kernel by overwriting the 659 metadata underneath a mounted FS without its awareness. 660 This also prevents destructive formatting of mounted 661 filesystems by naive storage tooling that don't use 662 O_EXCL. Default is Y and can be changed through the 663 Kconfig option CONFIG_BLK_DEV_WRITE_MOUNTED. 664 665 bert_disable [ACPI] 666 Disable BERT OS support on buggy BIOSes. 667 668 bgrt_disable [ACPI,X86,EARLY] 669 Disable BGRT to avoid flickering OEM logo. 670 671 blkdevparts= Manual partition parsing of block device(s) for 672 embedded devices based on command line input. 673 See Documentation/block/cmdline-partition.rst 674 675 boot_delay= [KNL,EARLY] 676 Milliseconds to delay each printk during boot. 677 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled, 678 and you may also have to specify "lpj=". Boot_delay 679 values larger than 10 seconds (10000) are assumed 680 erroneous and ignored. 681 Format: integer 682 683 bootconfig [KNL,EARLY] 684 Extended command line options can be added to an initrd 685 and this will cause the kernel to look for it. 686 687 See Documentation/admin-guide/bootconfig.rst 688 689 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 690 bttv.radio= Most important insmod options are available as 691 kernel args too. 692 bttv.pll= See Documentation/admin-guide/media/bttv.rst 693 bttv.tuner= 694 695 bulk_remove=off [PPC] This parameter disables the use of the pSeries 696 firmware feature for flushing multiple hpte entries 697 at a time. 698 699 c101= [NET] Moxa C101 synchronous serial card 700 701 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 702 Sometimes CPU hardware bugs make them report the cache 703 size incorrectly. The kernel will attempt work arounds 704 to fix known problems, but for some CPUs it is not 705 possible to determine what the correct size should be. 706 This option provides an override for these situations. 707 708 carrier_timeout= 709 [NET] Specifies amount of time (in seconds) that 710 the kernel should wait for a network carrier. By default 711 it waits 120 seconds. 712 713 ca_keys= [KEYS] This parameter identifies a specific key(s) on 714 the system trusted keyring to be used for certificate 715 trust validation. 716 format: { id:<keyid> | builtin } 717 718 cca= [MIPS,EARLY] Override the kernel pages' cache coherency 719 algorithm. Accepted values range from 0 to 7 720 inclusive. See arch/mips/include/asm/pgtable-bits.h 721 for platform specific values (SB1, Loongson3 and 722 others). 723 724 ccw_timeout_log [S390] 725 See Documentation/arch/s390/common_io.rst for details. 726 727 cfi= [X86-64] Set Control Flow Integrity checking features 728 when CONFIG_FINEIBT is enabled. 729 Format: feature[,feature...] 730 Default: auto 731 732 auto: Use FineIBT if IBT available, otherwise kCFI. 733 Under FineIBT, enable "paranoid" mode when 734 FRED is not available. 735 off: Turn off CFI checking. 736 kcfi: Use kCFI (disable FineIBT). 737 fineibt: Use FineIBT (even if IBT not available). 738 norand: Do not re-randomize CFI hashes. 739 paranoid: Add caller hash checking under FineIBT. 740 bhi: Enable register poisoning to stop speculation 741 across FineIBT. (Disabled by default.) 742 warn: Do not enforce CFI checking: warn only. 743 debug: Report CFI initialization details. 744 745 cgroup_disable= [KNL] Disable a particular controller or optional feature 746 Format: {name of the controller(s) or feature(s) to disable} 747 The effects of cgroup_disable=foo are: 748 - foo isn't auto-mounted if you mount all cgroups in 749 a single hierarchy 750 - foo isn't visible as an individually mountable 751 subsystem 752 - if foo is an optional feature then the feature is 753 disabled and corresponding cgroup files are not 754 created 755 {Currently only "memory" controller deal with this and 756 cut the overhead, others just disable the usage. So 757 only cgroup_disable=memory is actually worthy} 758 Specifying "pressure" disables per-cgroup pressure 759 stall information accounting feature 760 761 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 762 Format: { { controller | "all" | "named" } 763 [,{ controller | "all" | "named" }...] } 764 Like cgroup_disable, but only applies to cgroup v1; 765 the blacklisted controllers remain available in cgroup2. 766 "all" blacklists all controllers and "named" disables 767 named mounts. Specifying both "all" and "named" disables 768 all v1 hierarchies. 769 770 cgroup_v1_proc= [KNL] Show also missing controllers in /proc/cgroups 771 Format: { "true" | "false" } 772 /proc/cgroups lists only v1 controllers by default. 773 This compatibility option enables listing also v2 774 controllers (whose v1 code is not compiled!), so that 775 semi-legacy software can check this file to decide 776 about usage of v2 (sic) controllers. 777 778 cgroup_favordynmods= [KNL] Enable or Disable favordynmods. 779 Format: { "true" | "false" } 780 Defaults to the value of CONFIG_CGROUP_FAVOR_DYNMODS. 781 782 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 783 Format: <string> 784 nosocket -- Disable socket memory accounting. 785 nokmem -- Disable kernel memory accounting. 786 nobpf -- Disable BPF memory accounting. 787 788 check_pages= [MM,EARLY] Enable sanity checking of pages after 789 allocations / before freeing. This adds checks to catch 790 double-frees, use-after-frees, and other sources of 791 page corruption by inspecting page internals (flags, 792 mapcount/refcount, memcg_data, etc.). 793 Format: { "0" | "1" } 794 Default: 0 (1 if CONFIG_DEBUG_VM is set) 795 796 checkreqprot= [SELINUX] Set initial checkreqprot flag value. 797 Format: { "0" | "1" } 798 See security/selinux/Kconfig help text. 799 0 -- check protection applied by kernel (includes 800 any implied execute protection). 801 1 -- check protection requested by application. 802 Default value is set via a kernel config option. 803 Value can be changed at runtime via 804 /sys/fs/selinux/checkreqprot. 805 Setting checkreqprot to 1 is deprecated. 806 807 cio_ignore= [S390] 808 See Documentation/arch/s390/common_io.rst for details. 809 810 clearcpuid=X[,X...] [X86] 811 Disable CPUID feature X for the kernel. See 812 arch/x86/include/asm/cpufeatures.h for the valid bit 813 numbers X. Note the Linux-specific bits are not necessarily 814 stable over kernel options, but the vendor-specific 815 ones should be. 816 X can also be a string as appearing in the flags: line 817 in /proc/cpuinfo which does not have the above 818 instability issue. However, not all features have names 819 in /proc/cpuinfo. 820 Note that using this option will taint your kernel. 821 Also note that user programs calling CPUID directly 822 or using the feature without checking anything 823 will still see it. This just prevents it from 824 being used by the kernel or shown in /proc/cpuinfo. 825 Also note the kernel might malfunction if you disable 826 some critical bits. 827 828 clk_ignore_unused 829 [CLK] 830 Prevents the clock framework from automatically gating 831 clocks that have not been explicitly enabled by a Linux 832 device driver but are enabled in hardware at reset or 833 by the bootloader/firmware. Note that this does not 834 force such clocks to be always-on nor does it reserve 835 those clocks in any way. This parameter is useful for 836 debug and development, but should not be needed on a 837 platform with proper driver support. For more 838 information, see Documentation/driver-api/clk.rst. 839 840 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 841 [Deprecated] 842 Forces specified clocksource (if available) to be used 843 when calculating gettimeofday(). If specified 844 clocksource is not available, it defaults to PIT. 845 Format: { pit | tsc | cyclone | pmtmr } 846 847 clocksource= Override the default clocksource 848 Format: <string> 849 Override the default clocksource and use the clocksource 850 with the name specified. 851 Some clocksource names to choose from, depending on 852 the platform: 853 [all] jiffies (this is the base, fallback clocksource) 854 [ACPI] acpi_pm 855 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 856 pxa_timer,timer3,32k_counter,timer0_1 857 [X86-32] pit,hpet,tsc; 858 scx200_hrt on Geode; cyclone on IBM x440 859 [MIPS] MIPS 860 [PARISC] cr16 861 [S390] tod 862 [SH] SuperH 863 [SPARC64] tick 864 [X86-64] hpet,tsc 865 866 clocksource.arm_arch_timer.evtstrm= 867 [ARM,ARM64,EARLY] 868 Format: <bool> 869 Enable/disable the eventstream feature of the ARM 870 architected timer so that code using WFE-based polling 871 loops can be debugged more effectively on production 872 systems. 873 874 clocksource.verify_n_cpus= [KNL] 875 Limit the number of CPUs checked for clocksources 876 marked with CLOCK_SOURCE_VERIFY_PERCPU that 877 are marked unstable due to excessive skew. 878 A negative value says to check all CPUs, while 879 zero says not to check any. Values larger than 880 nr_cpu_ids are silently truncated to nr_cpu_ids. 881 The actual CPUs are chosen randomly, with 882 no replacement if the same CPU is chosen twice. 883 884 clocksource-wdtest.holdoff= [KNL] 885 Set the time in seconds that the clocksource 886 watchdog test waits before commencing its tests. 887 Defaults to zero when built as a module and to 888 10 seconds when built into the kernel. 889 890 cma=nn[MG]@[start[MG][-end[MG]]] 891 [KNL,CMA,EARLY] 892 Sets the size of kernel global memory area for 893 contiguous memory allocations and optionally the 894 placement constraint by the physical address range of 895 memory allocations. A value of 0 disables CMA 896 altogether. For more information, see 897 kernel/dma/contiguous.c 898 899 cma_pernuma=nn[MG] 900 [KNL,CMA,EARLY] 901 Sets the size of kernel per-numa memory area for 902 contiguous memory allocations. A value of 0 disables 903 per-numa CMA altogether. And If this option is not 904 specified, the default value is 0. 905 With per-numa CMA enabled, DMA users on node nid will 906 first try to allocate buffer from the pernuma area 907 which is located in node nid, if the allocation fails, 908 they will fallback to the global default memory area. 909 910 numa_cma=<node>:nn[MG][,<node>:nn[MG]] 911 [KNL,CMA,EARLY] 912 Sets the size of kernel numa memory area for 913 contiguous memory allocations. It will reserve CMA 914 area for the specified node. 915 916 With numa CMA enabled, DMA users on node nid will 917 first try to allocate buffer from the numa area 918 which is located in node nid, if the allocation fails, 919 they will fallback to the global default memory area. 920 921 cmo_free_hint= [PPC] Format: { yes | no } 922 Specify whether pages are marked as being inactive 923 when they are freed. This is used in CMO environments 924 to determine OS memory pressure for page stealing by 925 a hypervisor. 926 Default: yes 927 928 coherent_pool=nn[KMG] [ARM,KNL,EARLY] 929 Sets the size of memory pool for coherent, atomic dma 930 allocations, by default set to 256K. 931 932 com20020= [HW,NET] ARCnet - COM20020 chipset 933 Format: 934 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 935 936 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 937 Format: <io>[,<irq>] 938 939 com90xx= [HW,NET] 940 ARCnet - COM90xx chipset (memory-mapped buffers) 941 Format: <io>[,<irq>[,<memstart>]] 942 943 condev= [HW,S390] console device 944 conmode= 945 946 con3215_drop= [S390,EARLY] 3215 console drop mode. 947 Format: y|n|Y|N|1|0 948 When set to true, drop data on the 3215 console when 949 the console buffer is full. In this case the 950 operator using a 3270 terminal emulator (for example 951 x3270) does not have to enter the clear key for the 952 console output to advance and the kernel to continue. 953 This leads to a much faster boot time when a 3270 954 terminal emulator is active. If no 3270 terminal 955 emulator is used, this parameter has no effect. 956 957 console= [KNL] Output console device and options. 958 959 tty<n> Use the virtual console device <n>. 960 961 ttyS<n>[,options] 962 ttyUSB0[,options] 963 Use the specified serial port. The options are of 964 the form "bbbbpnf", where "bbbb" is the baud rate, 965 "p" is parity ("n", "o", or "e"), "n" is number of 966 bits, and "f" is flow control ("r" for RTS or 967 omit it). Default is "9600n8". 968 969 See Documentation/admin-guide/serial-console.rst for more 970 information. See 971 Documentation/networking/netconsole.rst for an 972 alternative. 973 974 <DEVNAME>:<n>.<n>[,options] 975 Use the specified serial port on the serial core bus. 976 The addressing uses DEVNAME of the physical serial port 977 device, followed by the serial core controller instance, 978 and the serial port instance. The options are the same 979 as documented for the ttyS addressing above. 980 981 The mapping of the serial ports to the tty instances 982 can be viewed with: 983 984 $ ls -d /sys/bus/serial-base/devices/*:*.*/tty/* 985 /sys/bus/serial-base/devices/00:04:0.0/tty/ttyS0 986 987 In the above example, the console can be addressed with 988 console=00:04:0.0. Note that a console addressed this 989 way will only get added when the related device driver 990 is ready. The use of an earlycon parameter in addition to 991 the console may be desired for console output early on. 992 993 uart[8250],io,<addr>[,options] 994 uart[8250],mmio,<addr>[,options] 995 uart[8250],mmio16,<addr>[,options] 996 uart[8250],mmio32,<addr>[,options] 997 uart[8250],0x<addr>[,options] 998 Start an early, polled-mode console on the 8250/16550 999 UART at the specified I/O port or MMIO address, 1000 switching to the matching ttyS device later. 1001 MMIO inter-register address stride is either 8-bit 1002 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 1003 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 1004 to be equivalent to 'mmio'. 'options' are specified in 1005 the same format described for ttyS above; if unspecified, 1006 the h/w is not re-initialized. 1007 1008 hvc<n> Use the hypervisor console device <n>. This is for 1009 both Xen and PowerPC hypervisors. 1010 1011 { null | "" } 1012 Use to disable console output, i.e., to have kernel 1013 console messages discarded. 1014 This must be the only console= parameter used on the 1015 kernel command line. 1016 1017 If the device connected to the port is not a TTY but a braille 1018 device, prepend "brl," before the device type, for instance 1019 console=brl,ttyS0 1020 For now, only VisioBraille is supported. 1021 1022 console_msg_format= 1023 [KNL] Change console messages format 1024 default 1025 By default we print messages on consoles in 1026 "[time stamp] text\n" format (time stamp may not be 1027 printed, depending on CONFIG_PRINTK_TIME or 1028 `printk_time' param). 1029 syslog 1030 Switch to syslog format: "<%u>[time stamp] text\n" 1031 IOW, each message will have a facility and loglevel 1032 prefix. The format is similar to one used by syslog() 1033 syscall, or to executing "dmesg -S --raw" or to reading 1034 from /proc/kmsg. 1035 1036 consoleblank= [KNL] The console blank (screen saver) timeout in 1037 seconds. A value of 0 disables the blank timer. 1038 Defaults to 0. 1039 1040 coredump_filter= 1041 [KNL] Change the default value for 1042 /proc/<pid>/coredump_filter. 1043 See also Documentation/filesystems/proc.rst. 1044 1045 coresight_cpu_debug.enable 1046 [ARM,ARM64] 1047 Format: <bool> 1048 Enable/disable the CPU sampling based debugging. 1049 0: default value, disable debugging 1050 1: enable debugging at boot time 1051 1052 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 1053 Format: 1054 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 1055 1056 cpuidle.off=1 [CPU_IDLE] 1057 disable the cpuidle sub-system 1058 1059 cpuidle.governor= 1060 [CPU_IDLE] Name of the cpuidle governor to use. 1061 1062 cpufreq.off=1 [CPU_FREQ] 1063 disable the cpufreq sub-system 1064 1065 cpufreq.default_governor= 1066 [CPU_FREQ] Name of the default cpufreq governor or 1067 policy to use. This governor must be registered in the 1068 kernel before the cpufreq driver probes. 1069 1070 cpu_init_udelay=N 1071 [X86,EARLY] Delay for N microsec between assert and de-assert 1072 of APIC INIT to start processors. This delay occurs 1073 on every CPU online, such as boot, and resume from suspend. 1074 Default: 10000 1075 1076 cpuhp.parallel= 1077 [SMP] Enable/disable parallel bringup of secondary CPUs 1078 Format: <bool> 1079 Default is enabled if CONFIG_HOTPLUG_PARALLEL=y. Otherwise 1080 the parameter has no effect. 1081 1082 crash_kexec_post_notifiers 1083 Only jump to kdump kernel after running the panic 1084 notifiers and dumping kmsg. This option increases 1085 the risks of a kdump failure, since some panic 1086 notifiers can make the crashed kernel more unstable. 1087 In configurations where kdump may not be reliable, 1088 running the panic notifiers could allow collecting 1089 more data on dmesg, like stack traces from other CPUS 1090 or extra data dumped by panic_print. Note that some 1091 configurations enable this option unconditionally, 1092 like Hyper-V, PowerPC (fadump) and AMD SEV-SNP. 1093 1094 crashkernel=size[KMG][@offset[KMG]] 1095 [KNL,EARLY] Using kexec, Linux can switch to a 'crash kernel' 1096 upon panic. This parameter reserves the physical 1097 memory region [offset, offset + size] for that kernel 1098 image. If '@offset' is omitted, then a suitable offset 1099 is selected automatically. 1100 [KNL, X86-64, ARM64, RISCV, LoongArch] Select a region 1101 under 4G first, and fall back to reserve region above 1102 4G when '@offset' hasn't been specified. 1103 See Documentation/admin-guide/kdump/kdump.rst for further details. 1104 1105 crashkernel=range1:size1[,range2:size2,...][@offset] 1106 [KNL] Same as above, but depends on the memory 1107 in the running system. The syntax of range is 1108 start-[end] where start and end are both 1109 a memory unit (amount[KMG]). See also 1110 Documentation/admin-guide/kdump/kdump.rst for an example. 1111 1112 crashkernel=size[KMG],high 1113 [KNL, X86-64, ARM64, RISCV, LoongArch] range could be 1114 above 4G. 1115 Allow kernel to allocate physical memory region from top, 1116 so could be above 4G if system have more than 4G ram 1117 installed. Otherwise memory region will be allocated 1118 below 4G, if available. 1119 It will be ignored if crashkernel=X is specified. 1120 crashkernel=size[KMG],low 1121 [KNL, X86-64, ARM64, RISCV, LoongArch] range under 4G. 1122 When crashkernel=X,high is passed, kernel could allocate 1123 physical memory region above 4G, that cause second kernel 1124 crash on system that require some amount of low memory, 1125 e.g. swiotlb requires at least 64M+32K low memory, also 1126 enough extra low memory is needed to make sure DMA buffers 1127 for 32-bit devices won't run out. Kernel would try to allocate 1128 default size of memory below 4G automatically. The default 1129 size is platform dependent. 1130 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB) 1131 --> arm64: 128MiB 1132 --> riscv: 128MiB 1133 --> loongarch: 128MiB 1134 This one lets the user specify own low range under 4G 1135 for second kernel instead. 1136 0: to disable low allocation. 1137 It will be ignored when crashkernel=X,high is not used 1138 or memory reserved is below 4G. 1139 crashkernel=size[KMG],cma 1140 [KNL, X86, ppc] Reserve additional crash kernel memory from 1141 CMA. This reservation is usable by the first system's 1142 userspace memory and kernel movable allocations (memory 1143 balloon, zswap). Pages allocated from this memory range 1144 will not be included in the vmcore so this should not 1145 be used if dumping of userspace memory is intended and 1146 it has to be expected that some movable kernel pages 1147 may be missing from the dump. 1148 1149 A standard crashkernel reservation, as described above, 1150 is still needed to hold the crash kernel and initrd. 1151 1152 This option increases the risk of a kdump failure: DMA 1153 transfers configured by the first kernel may end up 1154 corrupting the second kernel's memory. 1155 1156 This reservation method is intended for systems that 1157 can't afford to sacrifice enough memory for standard 1158 crashkernel reservation and where less reliable and 1159 possibly incomplete kdump is preferable to no kdump at 1160 all. 1161 1162 cryptomgr.notests 1163 [KNL] Disable crypto self-tests 1164 1165 cs89x0_dma= [HW,NET] 1166 Format: <dma> 1167 1168 cs89x0_media= [HW,NET] 1169 Format: { rj45 | aui | bnc } 1170 1171 csdlock_debug= [KNL] Enable or disable debug add-ons of cross-CPU 1172 function call handling. When switched on, 1173 additional debug data is printed to the console 1174 in case a hanging CPU is detected, and that 1175 CPU is pinged again in order to try to resolve 1176 the hang situation. The default value of this 1177 option depends on the CSD_LOCK_WAIT_DEBUG_DEFAULT 1178 Kconfig option. 1179 1180 dasd= [HW,NET] 1181 See header of drivers/s390/block/dasd_devmap.c. 1182 1183 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 1184 (one device per port) 1185 Format: <port#>,<type> 1186 See also Documentation/input/devices/joystick-parport.rst 1187 1188 debug [KNL,EARLY] Enable kernel debugging (events log level). 1189 1190 debug_boot_weak_hash 1191 [KNL,EARLY] Enable printing [hashed] pointers early in the 1192 boot sequence. If enabled, we use a weak hash instead 1193 of siphash to hash pointers. Use this option if you are 1194 seeing instances of '(___ptrval___)') and need to see a 1195 value (hashed pointer) instead. Cryptographically 1196 insecure, please do not use on production kernels. 1197 1198 debug_locks_verbose= 1199 [KNL] verbose locking self-tests 1200 Format: <int> 1201 Print debugging info while doing the locking API 1202 self-tests. 1203 Bitmask for the various LOCKTYPE_ tests. Defaults to 0 1204 (no extra messages), setting it to -1 (all bits set) 1205 will print _a_lot_ more information - normally only 1206 useful to lockdep developers. 1207 1208 debug_objects [KNL,EARLY] Enable object debugging 1209 1210 debug_guardpage_minorder= 1211 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this 1212 parameter allows control of the order of pages that will 1213 be intentionally kept free (and hence protected) by the 1214 buddy allocator. Bigger value increase the probability 1215 of catching random memory corruption, but reduce the 1216 amount of memory for normal system use. The maximum 1217 possible value is MAX_PAGE_ORDER/2. Setting this 1218 parameter to 1 or 2 should be enough to identify most 1219 random memory corruption problems caused by bugs in 1220 kernel or driver code when a CPU writes to (or reads 1221 from) a random memory location. Note that there exists 1222 a class of memory corruptions problems caused by buggy 1223 H/W or F/W or by drivers badly programming DMA 1224 (basically when memory is written at bus level and the 1225 CPU MMU is bypassed) which are not detectable by 1226 CONFIG_DEBUG_PAGEALLOC, hence this option will not 1227 help tracking down these problems. 1228 1229 debug_pagealloc= 1230 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 1231 enables the feature at boot time. By default, it is 1232 disabled and the system will work mostly the same as a 1233 kernel built without CONFIG_DEBUG_PAGEALLOC. 1234 Note: to get most of debug_pagealloc error reports, it's 1235 useful to also enable the page_owner functionality. 1236 on: enable the feature 1237 1238 debugfs= [KNL,EARLY] This parameter enables what is exposed to 1239 userspace and debugfs internal clients. 1240 Format: { on, off } 1241 on: All functions are enabled. 1242 off: Filesystem is not registered and clients 1243 get a -EPERM as result when trying to register files 1244 or directories within debugfs. 1245 This is equivalent of the runtime functionality if 1246 debugfs was not enabled in the kernel at all. 1247 Default value is set in build-time with a kernel configuration. 1248 1249 debugpat [X86] Enable PAT debugging 1250 1251 default_hugepagesz= 1252 [HW] The size of the default HugeTLB page. This is 1253 the size represented by the legacy /proc/ hugepages 1254 APIs. In addition, this is the default hugetlb size 1255 used for shmget(), mmap() and mounting hugetlbfs 1256 filesystems. If not specified, defaults to the 1257 architecture's default huge page size. Huge page 1258 sizes are architecture dependent. See also 1259 Documentation/admin-guide/mm/hugetlbpage.rst. 1260 Format: size[KMG] 1261 1262 deferred_probe_timeout= 1263 [KNL] Debugging option to set a timeout in seconds for 1264 deferred probe to give up waiting on dependencies to 1265 probe. Only specific dependencies (subsystems or 1266 drivers) that have opted in will be ignored. A timeout 1267 of 0 will timeout at the end of initcalls. If the time 1268 out hasn't expired, it'll be restarted by each 1269 successful driver registration. This option will also 1270 dump out devices still on the deferred probe list after 1271 retrying. 1272 1273 delayacct [KNL] Enable per-task delay accounting 1274 1275 dell_smm_hwmon.ignore_dmi= 1276 [HW] Continue probing hardware even if DMI data 1277 indicates that the driver is running on unsupported 1278 hardware. 1279 1280 dell_smm_hwmon.force= 1281 [HW] Activate driver even if SMM BIOS signature does 1282 not match list of supported models and enable otherwise 1283 blacklisted features. 1284 1285 dell_smm_hwmon.power_status= 1286 [HW] Report power status in /proc/i8k 1287 (disabled by default). 1288 1289 dell_smm_hwmon.restricted= 1290 [HW] Allow controlling fans only if SYS_ADMIN 1291 capability is set. 1292 1293 dell_smm_hwmon.fan_mult= 1294 [HW] Factor to multiply fan speed with. 1295 1296 dell_smm_hwmon.fan_max= 1297 [HW] Maximum configurable fan speed. 1298 1299 dfltcc= [HW,S390] 1300 Format: { on | off | def_only | inf_only | always } 1301 on: s390 zlib hardware support for compression on 1302 level 1 and decompression (default) 1303 off: No s390 zlib hardware support 1304 def_only: s390 zlib hardware support for deflate 1305 only (compression on level 1) 1306 inf_only: s390 zlib hardware support for inflate 1307 only (decompression) 1308 always: Same as 'on' but ignores the selected compression 1309 level always using hardware support (used for debugging) 1310 1311 dhash_entries= [KNL] 1312 Set number of hash buckets for dentry cache. 1313 1314 disable_1tb_segments [PPC,EARLY] 1315 Disables the use of 1TB hash page table segments. This 1316 causes the kernel to fall back to 256MB segments which 1317 can be useful when debugging issues that require an SLB 1318 miss to occur. 1319 1320 disable= [IPV6] 1321 See Documentation/networking/ipv6.rst. 1322 1323 disable_radix [PPC,EARLY] 1324 Disable RADIX MMU mode on POWER9 1325 1326 disable_tlbie [PPC] 1327 Disable TLBIE instruction. Currently does not work 1328 with KVM, with HASH MMU, or with coherent accelerators. 1329 1330 disable_ddw [PPC/PSERIES,EARLY] 1331 Disable Dynamic DMA Window support. Use this 1332 to workaround buggy firmware. 1333 1334 disable_ipv6= [IPV6] 1335 See Documentation/networking/ipv6.rst. 1336 1337 disable_mtrr_cleanup [X86,EARLY] 1338 The kernel tries to adjust MTRR layout from continuous 1339 to discrete, to make X server driver able to add WB 1340 entry later. This parameter disables that. 1341 1342 disable_mtrr_trim [X86, Intel and AMD only,EARLY] 1343 By default the kernel will trim any uncacheable 1344 memory out of your available memory pool based on 1345 MTRR settings. This parameter disables that behavior, 1346 possibly causing your machine to run very slowly. 1347 1348 disable_timer_pin_1 [X86,EARLY] 1349 Disable PIN 1 of APIC timer 1350 Can be useful to work around chipset bugs. 1351 1352 dis_ucode_ldr [X86] Disable the microcode loader. 1353 1354 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 1355 this option disables the debugging code at boot. 1356 1357 dma_debug_entries=<number> 1358 This option allows to tune the number of preallocated 1359 entries for DMA-API debugging code. One entry is 1360 required per DMA-API allocation. Use this if the 1361 DMA-API debugging code disables itself because the 1362 architectural default is too low. 1363 1364 dma_debug_driver=<driver_name> 1365 With this option the DMA-API debugging driver 1366 filter feature can be enabled at boot time. Just 1367 pass the driver to filter for as the parameter. 1368 The filter can be disabled or changed to another 1369 driver later using sysfs. 1370 1371 reg_file_data_sampling= 1372 [X86] Controls mitigation for Register File Data 1373 Sampling (RFDS) vulnerability. RFDS is a CPU 1374 vulnerability which may allow userspace to infer 1375 kernel data values previously stored in floating point 1376 registers, vector registers, or integer registers. 1377 RFDS only affects Intel Atom processors. 1378 1379 on: Turns ON the mitigation. 1380 off: Turns OFF the mitigation. 1381 1382 This parameter overrides the compile time default set 1383 by CONFIG_MITIGATION_RFDS. Mitigation cannot be 1384 disabled when other VERW based mitigations (like MDS) 1385 are enabled. In order to disable RFDS mitigation all 1386 VERW based mitigations need to be disabled. 1387 1388 For details see: 1389 Documentation/admin-guide/hw-vuln/reg-file-data-sampling.rst 1390 1391 dm_verity.keyring_unsealed= 1392 [KNL] When set to 1, leave the dm-verity keyring 1393 unsealed after initialization so userspace can 1394 provision keys. Once the keyring is restricted 1395 it becomes active and is searched during signature 1396 verification. 1397 1398 driver_async_probe= [KNL] 1399 List of driver names to be probed asynchronously. * 1400 matches with all driver names. If * is specified, the 1401 rest of the listed driver names are those that will NOT 1402 match the *. 1403 Format: <driver_name1>,<driver_name2>... 1404 1405 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1406 Broken monitors, graphic adapters, KVMs and EDIDless 1407 panels may send no or incorrect EDID data sets. 1408 This parameter allows to specify an EDID data sets 1409 in the /lib/firmware directory that are used instead. 1410 An EDID data set will only be used for a particular 1411 connector, if its name and a colon are prepended to 1412 the EDID name. Each connector may use a unique EDID 1413 data set by separating the files with a comma. An EDID 1414 data set with no connector name will be used for 1415 any connectors not explicitly specified. 1416 1417 dscc4.setup= [NET] 1418 1419 dt_cpu_ftrs= [PPC,EARLY] 1420 Format: {"off" | "known"} 1421 Control how the dt_cpu_ftrs device-tree binding is 1422 used for CPU feature discovery and setup (if it 1423 exists). 1424 off: Do not use it, fall back to legacy cpu table. 1425 known: Do not pass through unknown features to guests 1426 or userspace, only those that the kernel is aware of. 1427 1428 dump_apple_properties [X86] 1429 Dump name and content of EFI device properties on 1430 x86 Macs. Useful for driver authors to determine 1431 what data is available or for reverse-engineering. 1432 1433 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1434 <module>.dyndbg[="val"] 1435 Enable debug messages at boot time. See 1436 Documentation/admin-guide/dynamic-debug-howto.rst 1437 for details. 1438 1439 early_ioremap_debug [KNL,EARLY] 1440 Enable debug messages in early_ioremap support. This 1441 is useful for tracking down temporary early mappings 1442 which are not unmapped. 1443 1444 earlycon= [KNL,EARLY] Output early console device and options. 1445 1446 When used with no options, the early console is 1447 determined by stdout-path property in device tree's 1448 chosen node or the ACPI SPCR table if supported by 1449 the platform. 1450 1451 cdns,<addr>[,options] 1452 Start an early, polled-mode console on a Cadence 1453 (xuartps) serial port at the specified address. Only 1454 supported option is baud rate. If baud rate is not 1455 specified, the serial port must already be setup and 1456 configured. 1457 1458 uart[8250],io,<addr>[,options[,uartclk]] 1459 uart[8250],mmio,<addr>[,options[,uartclk]] 1460 uart[8250],mmio32,<addr>[,options[,uartclk]] 1461 uart[8250],mmio32be,<addr>[,options[,uartclk]] 1462 uart[8250],0x<addr>[,options] 1463 Start an early, polled-mode console on the 8250/16550 1464 UART at the specified I/O port or MMIO address. 1465 MMIO inter-register address stride is either 8-bit 1466 (mmio) or 32-bit (mmio32 or mmio32be). 1467 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1468 to be equivalent to 'mmio'. 'options' are specified 1469 in the same format described for "console=ttyS<n>"; if 1470 unspecified, the h/w is not initialized. 'uartclk' is 1471 the uart clock frequency; if unspecified, it is set 1472 to 'BASE_BAUD' * 16. 1473 1474 pl011,<addr> 1475 pl011,mmio32,<addr> 1476 Start an early, polled-mode console on a pl011 serial 1477 port at the specified address. The pl011 serial port 1478 must already be setup and configured. Options are not 1479 yet supported. If 'mmio32' is specified, then only 1480 the driver will use only 32-bit accessors to read/write 1481 the device registers. 1482 1483 liteuart,<addr> 1484 Start an early console on a litex serial port at the 1485 specified address. The serial port must already be 1486 setup and configured. Options are not yet supported. 1487 1488 meson,<addr> 1489 Start an early, polled-mode console on a meson serial 1490 port at the specified address. The serial port must 1491 already be setup and configured. Options are not yet 1492 supported. 1493 1494 msm_serial,<addr> 1495 Start an early, polled-mode console on an msm serial 1496 port at the specified address. The serial port 1497 must already be setup and configured. Options are not 1498 yet supported. 1499 1500 msm_serial_dm,<addr> 1501 Start an early, polled-mode console on an msm serial 1502 dm port at the specified address. The serial port 1503 must already be setup and configured. Options are not 1504 yet supported. 1505 1506 owl,<addr> 1507 Start an early, polled-mode console on a serial port 1508 of an Actions Semi SoC, such as S500 or S900, at the 1509 specified address. The serial port must already be 1510 setup and configured. Options are not yet supported. 1511 1512 rda,<addr> 1513 Start an early, polled-mode console on a serial port 1514 of an RDA Micro SoC, such as RDA8810PL, at the 1515 specified address. The serial port must already be 1516 setup and configured. Options are not yet supported. 1517 1518 sbi 1519 Use RISC-V SBI (Supervisor Binary Interface) for early 1520 console. 1521 1522 smh Use ARM semihosting calls for early console. 1523 1524 s3c2410,<addr> 1525 s3c2412,<addr> 1526 s3c2440,<addr> 1527 s3c6400,<addr> 1528 s5pv210,<addr> 1529 exynos4210,<addr> 1530 Use early console provided by serial driver available 1531 on Samsung SoCs, requires selecting proper type and 1532 a correct base address of the selected UART port. The 1533 serial port must already be setup and configured. 1534 Options are not yet supported. 1535 1536 lantiq,<addr> 1537 Start an early, polled-mode console on a lantiq serial 1538 (lqasc) port at the specified address. The serial port 1539 must already be setup and configured. Options are not 1540 yet supported. 1541 1542 lpuart,<addr> 1543 lpuart32,<addr> 1544 Use early console provided by Freescale LP UART driver 1545 found on Freescale Vybrid and QorIQ LS1021A processors. 1546 A valid base address must be provided, and the serial 1547 port must already be setup and configured. 1548 1549 ec_imx21,<addr> 1550 ec_imx6q,<addr> 1551 Start an early, polled-mode, output-only console on the 1552 Freescale i.MX UART at the specified address. The UART 1553 must already be setup and configured. 1554 1555 ar3700_uart,<addr> 1556 Start an early, polled-mode console on the 1557 Armada 3700 serial port at the specified 1558 address. The serial port must already be setup 1559 and configured. Options are not yet supported. 1560 1561 qcom_geni,<addr> 1562 Start an early, polled-mode console on a Qualcomm 1563 Generic Interface (GENI) based serial port at the 1564 specified address. The serial port must already be 1565 setup and configured. Options are not yet supported. 1566 1567 efifb,[options] 1568 Start an early, unaccelerated console on the EFI 1569 memory mapped framebuffer (if available). On cache 1570 coherent non-x86 systems that use system memory for 1571 the framebuffer, pass the 'ram' option so that it is 1572 mapped with the correct attributes. 1573 1574 linflex,<addr> 1575 Use early console provided by Freescale LINFlexD UART 1576 serial driver for NXP S32V234 SoCs. A valid base 1577 address must be provided, and the serial port must 1578 already be setup and configured. 1579 1580 earlyprintk= [X86,SH,ARM,M68k,S390,UM,EARLY] 1581 earlyprintk=vga 1582 earlyprintk=sclp 1583 earlyprintk=xen 1584 earlyprintk=serial[,ttySn[,baudrate]] 1585 earlyprintk=serial[,0x...[,baudrate]] 1586 earlyprintk=ttySn[,baudrate] 1587 earlyprintk=dbgp[debugController#] 1588 earlyprintk=mmio32,membase[,{nocfg|baudrate}] 1589 earlyprintk=pciserial[,force],bus:device.function[,{nocfg|baudrate}] 1590 earlyprintk=xdbc[xhciController#] 1591 earlyprintk=bios 1592 1593 earlyprintk is useful when the kernel crashes before 1594 the normal console is initialized. It is not enabled by 1595 default because it has some cosmetic problems. 1596 1597 Use "nocfg" to skip UART configuration, assume 1598 BIOS/firmware has configured UART correctly. 1599 1600 Append ",keep" to not disable it when the real console 1601 takes over. 1602 1603 Only one of vga, serial, or usb debug port can 1604 be used at a time. 1605 1606 Currently only ttyS0 and ttyS1 may be specified by 1607 name. Other I/O ports may be explicitly specified 1608 on some architectures (x86 and arm at least) by 1609 replacing ttySn with an I/O port address, like this: 1610 earlyprintk=serial,0x1008,115200 1611 You can find the port for a given device in 1612 /proc/tty/driver/serial: 1613 2: uart:ST16650V2 port:00001008 irq:18 ... 1614 1615 Interaction with the standard serial driver is not 1616 very good. 1617 1618 The VGA output is eventually overwritten by 1619 the real console. 1620 1621 The xen option can only be used in Xen domains. 1622 1623 The sclp output can only be used on s390. 1624 1625 The bios output can only be used on SuperH. 1626 1627 The optional "force" to "pciserial" enables use of a 1628 PCI device even when its classcode is not of the 1629 UART class. 1630 1631 edac_report= [HW,EDAC] Control how to report EDAC event 1632 Format: {"on" | "off" | "force"} 1633 on: enable EDAC to report H/W event. May be overridden 1634 by other higher priority error reporting module. 1635 off: disable H/W event reporting through EDAC. 1636 force: enforce the use of EDAC to report H/W event. 1637 default: on. 1638 1639 edd= [EDD] 1640 Format: {"off" | "on" | "skip[mbr]"} 1641 1642 efi= [EFI,EARLY] 1643 Format: { "debug", "disable_early_pci_dma", 1644 "nochunk", "noruntime", "nosoftreserve", 1645 "novamap", "no_disable_early_pci_dma" } 1646 debug: enable misc debug output. 1647 disable_early_pci_dma: disable the busmaster bit on all 1648 PCI bridges while in the EFI boot stub. 1649 nochunk: disable reading files in "chunks" in the EFI 1650 boot stub, as chunking can cause problems with some 1651 firmware implementations. 1652 noruntime : disable EFI runtime services support 1653 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1654 attribute may cause the kernel to reserve the 1655 memory range for a memory mapping driver to 1656 claim. Specify efi=nosoftreserve to disable this 1657 reservation and treat the memory by its base type 1658 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1659 novamap: do not call SetVirtualAddressMap(). 1660 no_disable_early_pci_dma: Leave the busmaster bit set 1661 on all PCI bridges while in the EFI boot stub 1662 1663 efi_no_storage_paranoia [EFI,X86,EARLY] 1664 Using this parameter you can use more than 50% of 1665 your efi variable storage. Use this parameter only if 1666 you are really sure that your UEFI does sane gc and 1667 fulfills the spec otherwise your board may brick. 1668 1669 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1670 that is to be dynamically loaded by Linux. If there are 1671 multiple variables with the same name but with different 1672 vendor GUIDs, all of them will be loaded. See 1673 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1674 1675 1676 eisa_irq_edge= [PARISC,HW] 1677 See header of drivers/parisc/eisa.c. 1678 1679 ekgdboc= [X86,KGDB,EARLY] Allow early kernel console debugging 1680 Format: ekgdboc=kbd 1681 1682 This is designed to be used in conjunction with 1683 the boot argument: earlyprintk=vga 1684 1685 This parameter works in place of the kgdboc parameter 1686 but can only be used if the backing tty is available 1687 very early in the boot process. For early debugging 1688 via a serial port see kgdboc_earlycon instead. 1689 1690 elanfreq= [X86-32] 1691 See comment before function elanfreq_setup() in 1692 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1693 1694 elfcorehdr=[size[KMG]@]offset[KMG] [PPC,SH,X86,S390,EARLY] 1695 Specifies physical address of start of kernel core 1696 image elf header and optionally the size. Generally 1697 kexec loader will pass this option to capture kernel. 1698 See Documentation/admin-guide/kdump/kdump.rst for details. 1699 1700 enable_mtrr_cleanup [X86,EARLY] 1701 The kernel tries to adjust MTRR layout from continuous 1702 to discrete, to make X server driver able to add WB 1703 entry later. This parameter enables that. 1704 1705 enable_timer_pin_1 [X86] 1706 Enable PIN 1 of APIC timer 1707 Can be useful to work around chipset bugs 1708 (in particular on some ATI chipsets). 1709 The kernel tries to set a reasonable default. 1710 1711 enforcing= [SELINUX] Set initial enforcing status. 1712 Format: {"0" | "1"} 1713 See security/selinux/Kconfig help text. 1714 0 -- permissive (log only, no denials). 1715 1 -- enforcing (deny and log). 1716 Default value is 0. 1717 Value can be changed at runtime via 1718 /sys/fs/selinux/enforce. 1719 1720 erst_disable [ACPI] 1721 Disable Error Record Serialization Table (ERST) 1722 support. 1723 1724 ether= [HW,NET] Ethernet cards parameters 1725 This option is obsoleted by the "netdev=" option, which 1726 has equivalent usage. See its documentation for details. 1727 1728 evm= [EVM] 1729 Format: { "fix" } 1730 Permit 'security.evm' to be updated regardless of 1731 current integrity status. 1732 1733 early_page_ext [KNL,EARLY] Enforces page_ext initialization to earlier 1734 stages so cover more early boot allocations. 1735 Please note that as side effect some optimizations 1736 might be disabled to achieve that (e.g. parallelized 1737 memory initialization is disabled) so the boot process 1738 might take longer, especially on systems with a lot of 1739 memory. Available with CONFIG_PAGE_EXTENSION=y. 1740 1741 failslab= 1742 fail_usercopy= 1743 fail_page_alloc= 1744 fail_skb_realloc= 1745 fail_make_request=[KNL] 1746 General fault injection mechanism. 1747 Format: <interval>,<probability>,<space>,<times> 1748 See also Documentation/fault-injection/. 1749 1750 fb_tunnels= [NET] 1751 Format: { initns | none } 1752 See Documentation/admin-guide/sysctl/net.rst for 1753 fb_tunnels_only_for_init_ns 1754 1755 floppy= [HW] 1756 See Documentation/admin-guide/blockdev/floppy.rst. 1757 1758 forcepae [X86-32] 1759 Forcefully enable Physical Address Extension (PAE). 1760 Many Pentium M systems disable PAE but may have a 1761 functionally usable PAE implementation. 1762 Warning: use of this parameter will taint the kernel 1763 and may cause unknown problems. 1764 1765 fred= [X86-64] 1766 Enable/disable Flexible Return and Event Delivery. 1767 Format: { on | off } 1768 on: enable FRED when it's present. 1769 off: disable FRED, the default setting. 1770 1771 ftrace=[tracer] 1772 [FTRACE] will set and start the specified tracer 1773 as early as possible in order to facilitate early 1774 boot debugging. 1775 1776 ftrace_boot_snapshot 1777 [FTRACE] On boot up, a snapshot will be taken of the 1778 ftrace ring buffer that can be read at: 1779 /sys/kernel/tracing/snapshot. 1780 This is useful if you need tracing information from kernel 1781 boot up that is likely to be overridden by user space 1782 start up functionality. 1783 1784 Optionally, the snapshot can also be defined for a tracing 1785 instance that was created by the trace_instance= command 1786 line parameter. 1787 1788 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo 1789 1790 The above will cause the "foo" tracing instance to trigger 1791 a snapshot at the end of boot up. 1792 1793 ftrace_dump_on_oops[=2(orig_cpu) | =<instance>][,<instance> | 1794 ,<instance>=2(orig_cpu)] 1795 [FTRACE] will dump the trace buffers on oops. 1796 If no parameter is passed, ftrace will dump global 1797 buffers of all CPUs, if you pass 2 or orig_cpu, it 1798 will dump only the buffer of the CPU that triggered 1799 the oops, or the specific instance will be dumped if 1800 its name is passed. Multiple instance dump is also 1801 supported, and instances are separated by commas. Each 1802 instance supports only dump on CPU that triggered the 1803 oops by passing 2 or orig_cpu to it. 1804 1805 ftrace_dump_on_oops=foo=orig_cpu 1806 1807 The above will dump only the buffer of "foo" instance 1808 on CPU that triggered the oops. 1809 1810 ftrace_dump_on_oops,foo,bar=orig_cpu 1811 1812 The above will dump global buffer on all CPUs, the 1813 buffer of "foo" instance on all CPUs and the buffer 1814 of "bar" instance on CPU that triggered the oops. 1815 1816 ftrace_filter=[function-list] 1817 [FTRACE] Limit the functions traced by the function 1818 tracer at boot up. function-list is a comma-separated 1819 list of functions. This list can be changed at run 1820 time by the set_ftrace_filter file in the debugfs 1821 tracing directory. 1822 1823 ftrace_notrace=[function-list] 1824 [FTRACE] Do not trace the functions specified in 1825 function-list. This list can be changed at run time 1826 by the set_ftrace_notrace file in the debugfs 1827 tracing directory. 1828 1829 ftrace_graph_filter=[function-list] 1830 [FTRACE] Limit the top level callers functions traced 1831 by the function graph tracer at boot up. 1832 function-list is a comma-separated list of functions 1833 that can be changed at run time by the 1834 set_graph_function file in the debugfs tracing directory. 1835 1836 ftrace_graph_notrace=[function-list] 1837 [FTRACE] Do not trace from the functions specified in 1838 function-list. This list is a comma-separated list of 1839 functions that can be changed at run time by the 1840 set_graph_notrace file in the debugfs tracing directory. 1841 1842 ftrace_graph_max_depth=<uint> 1843 [FTRACE] Used with the function graph tracer. This is 1844 the max depth it will trace into a function. This value 1845 can be changed at run time by the max_graph_depth file 1846 in the tracefs tracing directory. default: 0 (no limit) 1847 1848 fw_devlink= [KNL,EARLY] Create device links between consumer and supplier 1849 devices by scanning the firmware to infer the 1850 consumer/supplier relationships. This feature is 1851 especially useful when drivers are loaded as modules as 1852 it ensures proper ordering of tasks like device probing 1853 (suppliers first, then consumers), supplier boot state 1854 clean up (only after all consumers have probed), 1855 suspend/resume & runtime PM (consumers first, then 1856 suppliers). 1857 Format: { off | permissive | on | rpm } 1858 off -- Don't create device links from firmware info. 1859 permissive -- Create device links from firmware info 1860 but use it only for ordering boot state clean 1861 up (sync_state() calls). 1862 on -- Create device links from firmware info and use it 1863 to enforce probe and suspend/resume ordering. 1864 rpm -- Like "on", but also use to order runtime PM. 1865 1866 fw_devlink.strict=<bool> 1867 [KNL,EARLY] Treat all inferred dependencies as mandatory 1868 dependencies. This only applies for fw_devlink=on|rpm. 1869 Format: <bool> 1870 1871 fw_devlink.sync_state = 1872 [KNL,EARLY] When all devices that could probe have finished 1873 probing, this parameter controls what to do with 1874 devices that haven't yet received their sync_state() 1875 calls. 1876 Format: { strict | timeout } 1877 strict -- Default. Continue waiting on consumers to 1878 probe successfully. 1879 timeout -- Give up waiting on consumers and call 1880 sync_state() on any devices that haven't yet 1881 received their sync_state() calls after 1882 deferred_probe_timeout has expired or by 1883 late_initcall() if !CONFIG_MODULES. 1884 1885 gamecon.map[2|3]= 1886 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1887 support via parallel port (up to 5 devices per port) 1888 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1889 See also Documentation/input/devices/joystick-parport.rst 1890 1891 gamma= [HW,DRM] 1892 1893 gart_fix_e820= [X86-64,EARLY] disable the fix e820 for K8 GART 1894 Format: off | on 1895 default: on 1896 1897 gather_data_sampling= 1898 [X86,INTEL,EARLY] Control the Gather Data Sampling (GDS) 1899 mitigation. 1900 1901 Gather Data Sampling is a hardware vulnerability which 1902 allows unprivileged speculative access to data which was 1903 previously stored in vector registers. 1904 1905 This issue is mitigated by default in updated microcode. 1906 The mitigation may have a performance impact but can be 1907 disabled. On systems without the microcode mitigation 1908 disabling AVX serves as a mitigation. 1909 1910 force: Disable AVX to mitigate systems without 1911 microcode mitigation. No effect if the microcode 1912 mitigation is present. Known to cause crashes in 1913 userspace with buggy AVX enumeration. 1914 1915 off: Disable GDS mitigation. 1916 1917 gbpages [X86] Use GB pages for kernel direct mappings. 1918 1919 gcov_persist= [GCOV] When non-zero (default), profiling data for 1920 kernel modules is saved and remains accessible via 1921 debugfs, even when the module is unloaded/reloaded. 1922 When zero, profiling data is discarded and associated 1923 debugfs files are removed at module unload time. 1924 1925 goldfish [X86] Enable the goldfish android emulator platform. 1926 Don't use this when you are not running on the 1927 android emulator 1928 1929 gpio-mockup.gpio_mockup_ranges 1930 [HW] Sets the ranges of gpiochip of for this device. 1931 Format: <start1>,<end1>,<start2>,<end2>... 1932 gpio-mockup.gpio_mockup_named_lines 1933 [HW] Let the driver know GPIO lines should be named. 1934 1935 gpt [EFI] Forces disk with valid GPT signature but 1936 invalid Protective MBR to be treated as GPT. If the 1937 primary GPT is corrupted, it enables the backup/alternate 1938 GPT to be used instead. 1939 1940 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1941 the "Enable 0" bit of the configuration register. 1942 Format: 0 | 1 1943 Default: 0 1944 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1945 the "Enable 0" bit of the configuration register. 1946 Format: 0 | 1 1947 Default: 0 1948 grcan.select= [HW] Select which physical interface to use. 1949 Format: 0 | 1 1950 Default: 0 1951 grcan.txsize= [HW] Sets the size of the tx buffer. 1952 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1953 Default: 1024 1954 grcan.rxsize= [HW] Sets the size of the rx buffer. 1955 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1956 Default: 1024 1957 1958 hardened_usercopy= 1959 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 1960 hardening is enabled for this boot. Hardened 1961 usercopy checking is used to protect the kernel 1962 from reading or writing beyond known memory 1963 allocation boundaries as a proactive defense 1964 against bounds-checking flaws in the kernel's 1965 copy_to_user()/copy_from_user() interface. 1966 The default is determined by 1967 CONFIG_HARDENED_USERCOPY_DEFAULT_ON. 1968 on Perform hardened usercopy checks. 1969 off Disable hardened usercopy checks. 1970 1971 hardlockup_all_cpu_backtrace= 1972 [KNL] Should the hard-lockup detector generate 1973 backtraces on all cpus. 1974 Format: 0 | 1 1975 1976 hash_pointers= 1977 [KNL,EARLY] 1978 By default, when pointers are printed to the console 1979 or buffers via the %p format string, that pointer is 1980 "hashed", i.e. obscured by hashing the pointer value. 1981 This is a security feature that hides actual kernel 1982 addresses from unprivileged users, but it also makes 1983 debugging the kernel more difficult since unequal 1984 pointers can no longer be compared. The choices are: 1985 Format: { auto | always | never } 1986 Default: auto 1987 1988 auto - Hash pointers unless slab_debug is enabled. 1989 always - Always hash pointers (even if slab_debug is 1990 enabled). 1991 never - Never hash pointers. This option should only 1992 be specified when debugging the kernel. Do 1993 not use on production kernels. The boot 1994 param "no_hash_pointers" is an alias for 1995 this mode. 1996 1997 For controlling hashing dynamically at runtime, 1998 use the "kernel.kptr_restrict" sysctl instead. 1999 2000 hashdist= [KNL,NUMA] Large hashes allocated during boot 2001 are distributed across NUMA nodes. Defaults on 2002 for 64-bit NUMA, off otherwise. 2003 Format: 0 | 1 (for off | on) 2004 2005 hd= [EIDE] (E)IDE hard drive subsystem geometry 2006 Format: <cyl>,<head>,<sect> 2007 2008 hest_disable [ACPI] 2009 Disable Hardware Error Source Table (HEST) support; 2010 corresponding firmware-first mode error processing 2011 logic will be disabled. 2012 2013 hibernate= [HIBERNATION] 2014 noresume Don't check if there's a hibernation image 2015 present during boot. 2016 nocompress Don't compress/decompress hibernation images. 2017 no Disable hibernation and resume. 2018 protect_image Turn on image protection during restoration 2019 (that will set all pages holding image data 2020 during restoration read-only). 2021 2022 hibernate.compressor= [HIBERNATION] Compression algorithm to be 2023 used with hibernation. 2024 Format: { lzo | lz4 } 2025 Default: lzo 2026 2027 lzo: Select LZO compression algorithm to 2028 compress/decompress hibernation image. 2029 2030 lz4: Select LZ4 compression algorithm to 2031 compress/decompress hibernation image. 2032 2033 hibernate.pm_test_delay= 2034 [HIBERNATION] 2035 Sets the number of seconds to remain in a hibernation test 2036 mode before resuming the system (see 2037 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 2038 is set. Default value is 5. 2039 2040 hibernate_compression_threads= 2041 [HIBERNATION] 2042 Set the number of threads used for compressing or decompressing 2043 hibernation images. 2044 2045 Format: <integer> 2046 Default: 3 2047 Minimum: 1 2048 Example: hibernate_compression_threads=4 2049 2050 highmem=nn[KMG] [KNL,BOOT,EARLY] forces the highmem zone to have an exact 2051 size of <nn>. This works even on boxes that have no 2052 highmem otherwise. This also works to reduce highmem 2053 size on bigger boxes. 2054 2055 highres= [KNL] Enable/disable high resolution timer mode. 2056 Valid parameters: "on", "off" 2057 Default: "on" 2058 2059 hlt [BUGS=ARM,SH] 2060 2061 hostname= [KNL,EARLY] Set the hostname (aka UTS nodename). 2062 Format: <string> 2063 This allows setting the system's hostname during early 2064 startup. This sets the name returned by gethostname. 2065 Using this parameter to set the hostname makes it 2066 possible to ensure the hostname is correctly set before 2067 any userspace processes run, avoiding the possibility 2068 that a process may call gethostname before the hostname 2069 has been explicitly set, resulting in the calling 2070 process getting an incorrect result. The string must 2071 not exceed the maximum allowed hostname length (usually 2072 64 characters) and will be truncated otherwise. 2073 2074 hpet= [X86-32,HPET] option to control HPET usage 2075 Format: { enable (default) | disable | force | 2076 verbose } 2077 disable: disable HPET and use PIT instead 2078 force: allow force enabled of undocumented chips (ICH4, 2079 VIA, nVidia) 2080 verbose: show contents of HPET registers during setup 2081 2082 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 2083 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 2084 2085 hugepages= [HW,EARLY] Number of HugeTLB pages to allocate at boot. 2086 If this follows hugepagesz (below), it specifies 2087 the number of pages of hugepagesz to be allocated. 2088 If this is the first HugeTLB parameter on the command 2089 line, it specifies the number of pages to allocate for 2090 the default huge page size. If using node format, the 2091 number of pages to allocate per-node can be specified. 2092 See also Documentation/admin-guide/mm/hugetlbpage.rst. 2093 Format: <integer> or (node format) 2094 <node>:<integer>[,<node>:<integer>] 2095 2096 hugepagesz= 2097 [HW,EARLY] The size of the HugeTLB pages. This is 2098 used in conjunction with hugepages (above) to 2099 allocate huge pages of a specific size at boot. The 2100 pair hugepagesz=X hugepages=Y can be specified once 2101 for each supported huge page size. Huge page sizes 2102 are architecture dependent. See also 2103 Documentation/admin-guide/mm/hugetlbpage.rst. 2104 Format: size[KMG] 2105 2106 hugepage_alloc_threads= 2107 [HW] The number of threads that should be used to 2108 allocate hugepages during boot. This option can be 2109 used to improve system bootup time when allocating 2110 a large amount of huge pages. 2111 The default value is 25% of the available hardware threads. 2112 2113 Note that this parameter only applies to non-gigantic huge pages. 2114 2115 hugetlb_cma= [HW,CMA,EARLY] The size of a CMA area used for allocation 2116 of gigantic hugepages. Or using node format, the size 2117 of a CMA area per node can be specified. 2118 Format: nn[KMGTPE] or (node format) 2119 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]] 2120 2121 Reserve a CMA area of given size and allocate gigantic 2122 hugepages using the CMA allocator. If enabled, the 2123 boot-time allocation of gigantic hugepages is skipped. 2124 2125 hugetlb_cma_only= 2126 [HW,CMA,EARLY] When allocating new HugeTLB pages, only 2127 try to allocate from the CMA areas. 2128 2129 This option does nothing if hugetlb_cma= is not also 2130 specified. 2131 2132 hugetlb_free_vmemmap= 2133 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP 2134 enabled. 2135 Control if HugeTLB Vmemmap Optimization (HVO) is enabled. 2136 Allows heavy hugetlb users to free up some more 2137 memory (7 * PAGE_SIZE for each 2MB hugetlb page). 2138 Format: { on | off (default) } 2139 2140 on: enable HVO 2141 off: disable HVO 2142 2143 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y, 2144 the default is on. 2145 2146 Note that the vmemmap pages may be allocated from the added 2147 memory block itself when memory_hotplug.memmap_on_memory is 2148 enabled, those vmemmap pages cannot be optimized even if this 2149 feature is enabled. Other vmemmap pages not allocated from 2150 the added memory block itself do not be affected. 2151 2152 hung_task_panic= 2153 [KNL] Number of hung tasks to trigger kernel panic. 2154 Format: <int> 2155 2156 When set to a non-zero value, a kernel panic will be triggered if 2157 the number of detected hung tasks reaches this value. 2158 2159 0: don't panic 2160 1: panic immediately on first hung task 2161 N: panic after N hung tasks are detected in a single scan 2162 2163 The default value is controlled by the 2164 CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time option. The value 2165 selected by this boot parameter can be changed later by the 2166 kernel.hung_task_panic sysctl. 2167 2168 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 2169 terminal devices. Valid values: 0..8 2170 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 2171 If specified, z/VM IUCV HVC accepts connections 2172 from listed z/VM user IDs only. 2173 2174 hv_nopvspin [X86,HYPER_V,EARLY] 2175 Disables the paravirt spinlock optimizations 2176 which allow the hypervisor to 'idle' the guest 2177 on lock contention. 2178 2179 hw_protection= [HW] 2180 Format: reboot | shutdown 2181 2182 Hardware protection action taken on critical events like 2183 overtemperature or imminent voltage loss. 2184 2185 i2c_bus= [HW] Override the default board specific I2C bus speed 2186 or register an additional I2C bus that is not 2187 registered from board initialization code. 2188 Format: 2189 <bus_id>,<clkrate> 2190 2191 i2c_touchscreen_props= [HW,ACPI,X86] 2192 Set device-properties for ACPI-enumerated I2C-attached 2193 touchscreen, to e.g. fix coordinates of upside-down 2194 mounted touchscreens. If you need this option please 2195 submit a drivers/platform/x86/touchscreen_dmi.c patch 2196 adding a DMI quirk for this. 2197 2198 Format: 2199 <ACPI_HW_ID>:<prop_name>=<val>[:prop_name=val][:...] 2200 Where <val> is one of: 2201 Omit "=<val>" entirely Set a boolean device-property 2202 Unsigned number Set a u32 device-property 2203 Anything else Set a string device-property 2204 2205 Examples (split over multiple lines): 2206 i2c_touchscreen_props=GDIX1001:touchscreen-inverted-x: 2207 touchscreen-inverted-y 2208 2209 i2c_touchscreen_props=MSSL1680:touchscreen-size-x=1920: 2210 touchscreen-size-y=1080:touchscreen-inverted-y: 2211 firmware-name=gsl1680-vendor-model.fw:silead,home-button 2212 2213 i8042.debug [HW] Toggle i8042 debug mode 2214 i8042.unmask_kbd_data 2215 [HW] Enable printing of interrupt data from the KBD port 2216 (disabled by default, and as a pre-condition 2217 requires that i8042.debug=1 be enabled) 2218 i8042.direct [HW] Put keyboard port into non-translated mode 2219 i8042.dumbkbd [HW] Pretend that controller can only read data from 2220 keyboard and cannot control its state 2221 (Don't attempt to blink the leds) 2222 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 2223 i8042.nokbd [HW] Don't check/create keyboard port 2224 i8042.noloop [HW] Disable the AUX Loopback command while probing 2225 for the AUX port 2226 i8042.nomux [HW] Don't check presence of an active multiplexing 2227 controller 2228 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 2229 controllers 2230 i8042.notimeout [HW] Ignore timeout condition signalled by controller 2231 i8042.reset [HW] Reset the controller during init, cleanup and 2232 suspend-to-ram transitions, only during s2r 2233 transitions, or never reset 2234 Format: { 1 | Y | y | 0 | N | n } 2235 1, Y, y: always reset controller 2236 0, N, n: don't ever reset controller 2237 Default: only on s2r transitions on x86; most other 2238 architectures force reset to be always executed 2239 i8042.unlock [HW] Unlock (ignore) the keylock 2240 i8042.kbdreset [HW] Reset device connected to KBD port 2241 i8042.probe_defer 2242 [HW] Allow deferred probing upon i8042 probe errors 2243 2244 i810= [HW,DRM] 2245 2246 i915.invert_brightness= 2247 [DRM] Invert the sense of the variable that is used to 2248 set the brightness of the panel backlight. Normally a 2249 brightness value of 0 indicates backlight switched off, 2250 and the maximum of the brightness value sets the backlight 2251 to maximum brightness. If this parameter is set to 0 2252 (default) and the machine requires it, or this parameter 2253 is set to 1, a brightness value of 0 sets the backlight 2254 to maximum brightness, and the maximum of the brightness 2255 value switches the backlight off. 2256 -1 -- never invert brightness 2257 0 -- machine default 2258 1 -- force brightness inversion 2259 2260 ia32_emulation= [X86-64] 2261 Format: <bool> 2262 When true, allows loading 32-bit programs and executing 32-bit 2263 syscalls, essentially overriding IA32_EMULATION_DEFAULT_DISABLED at 2264 boot time. When false, unconditionally disables IA32 emulation. 2265 2266 icn= [HW,ISDN] 2267 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 2268 2269 2270 idle= [X86,EARLY] 2271 Format: idle=poll, idle=halt, idle=nomwait 2272 2273 idle=poll: Don't do power saving in the idle loop 2274 using HLT, but poll for rescheduling event. This will 2275 make the CPUs eat a lot more power, but may be useful 2276 to get slightly better performance in multiprocessor 2277 benchmarks. It also makes some profiling using 2278 performance counters more accurate. Please note that 2279 on systems with MONITOR/MWAIT support (like Intel 2280 EM64T CPUs) this option has no performance advantage 2281 over the normal idle loop. It may also interact badly 2282 with hyperthreading. 2283 2284 idle=halt: Halt is forced to be used for CPU idle. 2285 In such case C2/C3 won't be used again. 2286 2287 idle=nomwait: Disable mwait for CPU C-states 2288 2289 idxd.sva= [HW] 2290 Format: <bool> 2291 Allow force disabling of Shared Virtual Memory (SVA) 2292 support for the idxd driver. By default it is set to 2293 true (1). 2294 2295 idxd.tc_override= [HW] 2296 Format: <bool> 2297 Allow override of default traffic class configuration 2298 for the device. By default it is set to false (0). 2299 2300 ieee754= [MIPS] Select IEEE Std 754 conformance mode 2301 Format: { strict | legacy | 2008 | relaxed | emulated } 2302 Default: strict 2303 2304 Choose which programs will be accepted for execution 2305 based on the IEEE 754 NaN encoding(s) supported by 2306 the FPU and the NaN encoding requested with the value 2307 of an ELF file header flag individually set by each 2308 binary. Hardware implementations are permitted to 2309 support either or both of the legacy and the 2008 NaN 2310 encoding mode. 2311 2312 Available settings are as follows: 2313 strict accept binaries that request a NaN encoding 2314 supported by the FPU 2315 legacy only accept legacy-NaN binaries, if supported 2316 by the FPU 2317 2008 only accept 2008-NaN binaries, if supported 2318 by the FPU 2319 relaxed accept any binaries regardless of whether 2320 supported by the FPU 2321 emulated accept any binaries but enable FPU emulator 2322 if binary mode is unsupported by the FPU. 2323 2324 The FPU emulator is always able to support both NaN 2325 encodings, so if no FPU hardware is present or it has 2326 been disabled with 'nofpu', then the settings of 2327 'legacy' and '2008' strap the emulator accordingly, 2328 'relaxed' straps the emulator for both legacy-NaN and 2329 2008-NaN, whereas 'strict' enables legacy-NaN only on 2330 legacy processors and both NaN encodings on MIPS32 or 2331 MIPS64 CPUs. 2332 2333 The setting for ABS.fmt/NEG.fmt instruction execution 2334 mode generally follows that for the NaN encoding, 2335 except where unsupported by hardware. 2336 2337 ignore_loglevel [KNL,EARLY] 2338 Ignore loglevel setting - this will print /all/ 2339 kernel messages to the console. Useful for debugging. 2340 We also add it as printk module parameter, so users 2341 could change it dynamically, usually by 2342 /sys/module/printk/parameters/ignore_loglevel. 2343 2344 ignore_rlimit_data 2345 Ignore RLIMIT_DATA setting for data mappings, 2346 print warning at first misuse. Can be changed via 2347 /sys/module/kernel/parameters/ignore_rlimit_data. 2348 2349 ihash_entries= [KNL] 2350 Set number of hash buckets for inode cache. 2351 2352 ima_appraise= [IMA] appraise integrity measurements 2353 Format: { "off" | "enforce" | "fix" | "log" } 2354 default: "enforce" 2355 2356 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 2357 The builtin appraise policy appraises all files 2358 owned by uid=0. 2359 2360 ima_canonical_fmt [IMA] 2361 Use the canonical format for the binary runtime 2362 measurements, instead of host native format. 2363 2364 ima_hash= [IMA] 2365 Format: { md5 | sha1 | rmd160 | sha256 | sha384 2366 | sha512 | ... } 2367 default: "sha1" 2368 2369 The list of supported hash algorithms is defined 2370 in crypto/hash_info.h. 2371 2372 ima_policy= [IMA] 2373 The builtin policies to load during IMA setup. 2374 Format: "tcb | appraise_tcb | secure_boot | 2375 fail_securely | critical_data" 2376 2377 The "tcb" policy measures all programs exec'd, files 2378 mmap'd for exec, and all files opened with the read 2379 mode bit set by either the effective uid (euid=0) or 2380 uid=0. 2381 2382 The "appraise_tcb" policy appraises the integrity of 2383 all files owned by root. 2384 2385 The "secure_boot" policy appraises the integrity 2386 of files (eg. kexec kernel image, kernel modules, 2387 firmware, policy, etc) based on file signatures. 2388 2389 The "fail_securely" policy forces file signature 2390 verification failure also on privileged mounted 2391 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 2392 flag. 2393 2394 The "critical_data" policy measures kernel integrity 2395 critical data. 2396 2397 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 2398 Load a policy which meets the needs of the Trusted 2399 Computing Base. This means IMA will measure all 2400 programs exec'd, files mmap'd for exec, and all files 2401 opened for read by uid=0. 2402 2403 ima_template= [IMA] 2404 Select one of defined IMA measurements template formats. 2405 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" | 2406 "ima-sigv2" } 2407 Default: "ima-ng" 2408 2409 ima_template_fmt= 2410 [IMA] Define a custom template format. 2411 Format: { "field1|...|fieldN" } 2412 2413 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 2414 Format: <min_file_size> 2415 Set the minimal file size for using asynchronous hash. 2416 If left unspecified, ahash usage is disabled. 2417 2418 ahash performance varies for different data sizes on 2419 different crypto accelerators. This option can be used 2420 to achieve the best performance for a particular HW. 2421 2422 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size 2423 Format: <bufsize> 2424 Set hashing buffer size. Default: 4k. 2425 2426 ahash performance varies for different chunk sizes on 2427 different crypto accelerators. This option can be used 2428 to achieve best performance for particular HW. 2429 2430 ima= [IMA] Enable or disable IMA 2431 Format: { "off" | "on" } 2432 Default: "on" 2433 Note that disabling IMA is limited to kdump kernel. 2434 2435 indirect_target_selection= [X86,Intel] Mitigation control for Indirect 2436 Target Selection(ITS) bug in Intel CPUs. Updated 2437 microcode is also required for a fix in IBPB. 2438 2439 on: Enable mitigation (default). 2440 off: Disable mitigation. 2441 force: Force the ITS bug and deploy default 2442 mitigation. 2443 vmexit: Only deploy mitigation if CPU is affected by 2444 guest/host isolation part of ITS. 2445 stuff: Deploy RSB-fill mitigation when retpoline is 2446 also deployed. Otherwise, deploy the default 2447 mitigation. 2448 2449 For details see: 2450 Documentation/admin-guide/hw-vuln/indirect-target-selection.rst 2451 2452 init= [KNL] 2453 Format: <full_path> 2454 Run specified binary instead of /sbin/init as init 2455 process. 2456 2457 initcall_debug [KNL] Trace initcalls as they are executed. Useful 2458 for working out where the kernel is dying during 2459 startup. 2460 2461 initcall_blacklist= [KNL] Do not execute a comma-separated list of 2462 initcall functions. Useful for debugging built-in 2463 modules and initcalls. 2464 2465 initramfs_async= [KNL] 2466 Format: <bool> 2467 Default: 1 2468 This parameter controls whether the initramfs 2469 image is unpacked asynchronously, concurrently 2470 with devices being probed and 2471 initialized. This should normally just work, 2472 but as a debugging aid, one can get the 2473 historical behaviour of the initramfs 2474 unpacking being completed before device_ and 2475 late_ initcalls. 2476 2477 initrd= [BOOT,EARLY] Specify the location of the initial ramdisk 2478 2479 initrdmem= [KNL,EARLY] Specify a physical address and size from which to 2480 load the initrd. If an initrd is compiled in or 2481 specified in the bootparams, it takes priority over this 2482 setting. 2483 Format: ss[KMG],nn[KMG] 2484 Default is 0, 0 2485 2486 init_on_alloc= [MM,EARLY] Fill newly allocated pages and heap objects with 2487 zeroes. 2488 Format: 0 | 1 2489 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 2490 2491 init_on_free= [MM,EARLY] Fill freed pages and heap objects with zeroes. 2492 Format: 0 | 1 2493 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 2494 2495 init_pkru= [X86] Specify the default memory protection keys rights 2496 register contents for all processes. 0x55555554 by 2497 default (disallow access to all but pkey 0). Can 2498 override in debugfs after boot. 2499 2500 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 2501 Format: <irq> 2502 2503 int_pln_enable [X86] Enable power limit notification interrupt 2504 2505 integrity_audit=[IMA] 2506 Format: { "0" | "1" } 2507 0 -- basic integrity auditing messages. (Default) 2508 1 -- additional integrity auditing messages. 2509 2510 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 2511 on 2512 Enable intel iommu driver. 2513 off 2514 Disable intel iommu driver. 2515 igfx_off [Default Off] 2516 By default, gfx is mapped as normal device. If a gfx 2517 device has a dedicated DMAR unit, the DMAR unit is 2518 bypassed by not enabling DMAR with this option. In 2519 this case, gfx device will use physical address for 2520 DMA. 2521 strict [Default Off] 2522 Deprecated, equivalent to iommu.strict=1. 2523 sp_off [Default Off] 2524 By default, super page will be supported if Intel IOMMU 2525 has the capability. With this option, super page will 2526 not be supported. 2527 sm_on 2528 Enable the Intel IOMMU scalable mode if the hardware 2529 advertises that it has support for the scalable mode 2530 translation. 2531 sm_off 2532 Disallow use of the Intel IOMMU scalable mode. 2533 tboot_noforce [Default Off] 2534 Do not force the Intel IOMMU enabled under tboot. 2535 By default, tboot will force Intel IOMMU on, which 2536 could harm performance of some high-throughput 2537 devices like 40GBit network cards, even if identity 2538 mapping is enabled. 2539 Note that using this option lowers the security 2540 provided by tboot because it makes the system 2541 vulnerable to DMA attacks. 2542 2543 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 2544 0 disables intel_idle and fall back on acpi_idle. 2545 1 to 9 specify maximum depth of C-state. 2546 2547 intel_pstate= [X86,EARLY] 2548 disable 2549 Do not enable intel_pstate as the default 2550 scaling driver for the supported processors 2551 active 2552 Use intel_pstate driver to bypass the scaling 2553 governors layer of cpufreq and provides it own 2554 algorithms for p-state selection. There are two 2555 P-state selection algorithms provided by 2556 intel_pstate in the active mode: powersave and 2557 performance. The way they both operate depends 2558 on whether or not the hardware managed P-states 2559 (HWP) feature has been enabled in the processor 2560 and possibly on the processor model. 2561 passive 2562 Use intel_pstate as a scaling driver, but configure it 2563 to work with generic cpufreq governors (instead of 2564 enabling its internal governor). This mode cannot be 2565 used along with the hardware-managed P-states (HWP) 2566 feature. 2567 force 2568 Enable intel_pstate on systems that prohibit it by default 2569 in favor of acpi-cpufreq. Forcing the intel_pstate driver 2570 instead of acpi-cpufreq may disable platform features, such 2571 as thermal controls and power capping, that rely on ACPI 2572 P-States information being indicated to OSPM and therefore 2573 should be used with caution. This option does not work with 2574 processors that aren't supported by the intel_pstate driver 2575 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 2576 no_hwp 2577 Do not enable hardware P state control (HWP) 2578 if available. 2579 hwp_only 2580 Only load intel_pstate on systems which support 2581 hardware P state control (HWP) if available. 2582 support_acpi_ppc 2583 Enforce ACPI _PPC performance limits. If the Fixed ACPI 2584 Description Table, specifies preferred power management 2585 profile as "Enterprise Server" or "Performance Server", 2586 then this feature is turned on by default. 2587 per_cpu_perf_limits 2588 Allow per-logical-CPU P-State performance control limits using 2589 cpufreq sysfs interface 2590 no_cas 2591 Do not enable capacity-aware scheduling (CAS) on 2592 hybrid systems 2593 2594 intremap= [X86-64,Intel-IOMMU,EARLY] 2595 on enable Interrupt Remapping (default) 2596 off disable Interrupt Remapping 2597 nosid disable Source ID checking 2598 no_x2apic_optout 2599 BIOS x2APIC opt-out request will be ignored 2600 nopost disable Interrupt Posting 2601 posted_msi 2602 enable MSIs delivered as posted interrupts 2603 2604 iomem= Disable strict checking of access to MMIO memory 2605 strict regions from userspace. 2606 relaxed 2607 2608 iommu= [X86,EARLY] 2609 2610 off 2611 Don't initialize and use any kind of IOMMU. 2612 2613 force 2614 Force the use of the hardware IOMMU even when 2615 it is not actually needed (e.g. because < 3 GB 2616 memory). 2617 2618 noforce 2619 Don't force hardware IOMMU usage when it is not 2620 needed. (default). 2621 2622 biomerge 2623 panic 2624 nopanic 2625 merge 2626 nomerge 2627 2628 soft 2629 Use software bounce buffering (SWIOTLB) (default for 2630 Intel machines). This can be used to prevent the usage 2631 of an available hardware IOMMU. 2632 2633 [X86] 2634 pt 2635 [X86] 2636 nopt 2637 [PPC/POWERNV] 2638 nobypass 2639 Disable IOMMU bypass, using IOMMU for PCI devices. 2640 2641 [X86] 2642 AMD Gart HW IOMMU-specific options: 2643 2644 <size> 2645 Set the size of the remapping area in bytes. 2646 2647 allowed 2648 Overwrite iommu off workarounds for specific chipsets 2649 2650 fullflush 2651 Flush IOMMU on each allocation (default). 2652 2653 nofullflush 2654 Don't use IOMMU fullflush. 2655 2656 memaper[=<order>] 2657 Allocate an own aperture over RAM with size 2658 32MB<<order. (default: order=1, i.e. 64MB) 2659 2660 merge 2661 Do scatter-gather (SG) merging. Implies "force" 2662 (experimental). 2663 2664 nomerge 2665 Don't do scatter-gather (SG) merging. 2666 2667 noaperture 2668 Ask the IOMMU not to touch the aperture for AGP. 2669 2670 noagp 2671 Don't initialize the AGP driver and use full aperture. 2672 2673 panic 2674 Always panic when IOMMU overflows. 2675 2676 iommu.forcedac= [ARM64,X86,EARLY] Control IOVA allocation for PCI devices. 2677 Format: { "0" | "1" } 2678 0 - Try to allocate a 32-bit DMA address first, before 2679 falling back to the full range if needed. 2680 1 - Allocate directly from the full usable range, 2681 forcing Dual Address Cycle for PCI cards supporting 2682 greater than 32-bit addressing. 2683 2684 iommu.strict= [ARM64,X86,S390,EARLY] Configure TLB invalidation behaviour 2685 Format: { "0" | "1" } 2686 0 - Lazy mode. 2687 Request that DMA unmap operations use deferred 2688 invalidation of hardware TLBs, for increased 2689 throughput at the cost of reduced device isolation. 2690 Will fall back to strict mode if not supported by 2691 the relevant IOMMU driver. 2692 1 - Strict mode. 2693 DMA unmap operations invalidate IOMMU hardware TLBs 2694 synchronously. 2695 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}. 2696 Note: on x86, strict mode specified via one of the 2697 legacy driver-specific options takes precedence. 2698 2699 iommu.passthrough= 2700 [ARM64,X86,EARLY] Configure DMA to bypass the IOMMU by default. 2701 Format: { "0" | "1" } 2702 0 - Use IOMMU translation for DMA. 2703 1 - Bypass the IOMMU for DMA. 2704 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2705 2706 iommu.debug_pagealloc= 2707 [KNL,EARLY] When CONFIG_IOMMU_DEBUG_PAGEALLOC is set, this 2708 parameter enables the feature at boot time. By default, it 2709 is disabled and the system behaves the same way as a kernel 2710 built without CONFIG_IOMMU_DEBUG_PAGEALLOC. 2711 Format: { "0" | "1" } 2712 0 - Sanitizer disabled. 2713 1 - Sanitizer enabled, expect runtime overhead. 2714 2715 io7= [HW] IO7 for Marvel-based Alpha systems 2716 See comment before marvel_specify_io7 in 2717 arch/alpha/kernel/core_marvel.c. 2718 2719 io_delay= [X86,EARLY] I/O delay method 2720 0x80 2721 Standard port 0x80 based delay 2722 0xed 2723 Alternate port 0xed based delay (needed on some systems) 2724 udelay 2725 Simple two microseconds delay 2726 none 2727 No delay 2728 2729 ip= [IP_PNP] 2730 See Documentation/admin-guide/nfs/nfsroot.rst. 2731 2732 ipcmni_extend [KNL,EARLY] Extend the maximum number of unique System V 2733 IPC identifiers from 32,768 to 16,777,216. 2734 2735 ipe.enforce= [IPE] 2736 Format: <bool> 2737 Determine whether IPE starts in permissive (0) or 2738 enforce (1) mode. The default is enforce. 2739 2740 ipe.success_audit= 2741 [IPE] 2742 Format: <bool> 2743 Start IPE with success auditing enabled, emitting 2744 an audit event when a binary is allowed. The default 2745 is 0. 2746 2747 irqaffinity= [SMP] Set the default irq affinity mask 2748 The argument is a cpu list, as described above. 2749 2750 irqchip.gicv2_force_probe= 2751 [ARM,ARM64,EARLY] 2752 Format: <bool> 2753 Force the kernel to look for the second 4kB page 2754 of a GICv2 controller even if the memory range 2755 exposed by the device tree is too small. 2756 2757 irqchip.gicv3_nolpi= 2758 [ARM,ARM64,EARLY] 2759 Force the kernel to ignore the availability of 2760 LPIs (and by consequence ITSs). Intended for system 2761 that use the kernel as a bootloader, and thus want 2762 to let secondary kernels in charge of setting up 2763 LPIs. 2764 2765 irqchip.gicv3_pseudo_nmi= [ARM64,EARLY] 2766 Enables support for pseudo-NMIs in the kernel. This 2767 requires the kernel to be built with 2768 CONFIG_ARM64_PSEUDO_NMI. 2769 2770 irqchip.riscv_imsic_noipi 2771 [RISC-V,EARLY] 2772 Force the kernel to not use IMSIC software injected MSIs 2773 as IPIs. Intended for system where IMSIC is trap-n-emulated, 2774 and thus want to reduce MMIO traps when triggering IPIs 2775 to multiple harts. 2776 2777 irqfixup [HW] 2778 When an interrupt is not handled search all handlers 2779 for it. Intended to get systems with badly broken 2780 firmware running. 2781 2782 irqhandler.duration_warn_us= [KNL] 2783 Warn if an IRQ handler exceeds the specified duration 2784 threshold in microseconds. Useful for identifying 2785 long-running IRQs in the system. 2786 2787 irqpoll [HW] 2788 When an interrupt is not handled search all handlers 2789 for it. Also check all handlers each timer 2790 interrupt. Intended to get systems with badly broken 2791 firmware running. 2792 2793 isapnp= [ISAPNP] 2794 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2795 2796 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2797 [Deprecated - use cpusets instead] 2798 Format: [flag-list,]<cpu-list> 2799 2800 Specify one or more CPUs to isolate from disturbances 2801 specified in the flag list (default: domain): 2802 2803 nohz 2804 Disable the tick when a single task runs as well as 2805 disabling other kernel noises like having RCU callbacks 2806 offloaded. This is equivalent to the nohz_full parameter. 2807 2808 A residual 1Hz tick is offloaded to workqueues, which you 2809 need to affine to housekeeping through the global 2810 workqueue's affinity configured via the 2811 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2812 by using the 'domain' flag described below. 2813 2814 NOTE: by default the global workqueue runs on all CPUs, 2815 so to protect individual CPUs the 'cpumask' file has to 2816 be configured manually after bootup. 2817 2818 domain 2819 Isolate from the general SMP balancing and scheduling 2820 algorithms. Note that performing domain isolation this way 2821 is irreversible: it's not possible to bring back a CPU to 2822 the domains once isolated through isolcpus. It's strongly 2823 advised to use cpusets instead to disable scheduler load 2824 balancing through the "cpuset.sched_load_balance" file. 2825 It offers a much more flexible interface where CPUs can 2826 move in and out of an isolated set anytime. 2827 2828 You can move a process onto or off an "isolated" CPU via 2829 the CPU affinity syscalls or cpuset. 2830 <cpu number> begins at 0 and the maximum value is 2831 "number of CPUs in system - 1". 2832 2833 managed_irq 2834 2835 Isolate from being targeted by managed interrupts 2836 which have an interrupt mask containing isolated 2837 CPUs. The affinity of managed interrupts is 2838 handled by the kernel and cannot be changed via 2839 the /proc/irq/* interfaces. 2840 2841 This isolation is best effort and only effective 2842 if the automatically assigned interrupt mask of a 2843 device queue contains isolated and housekeeping 2844 CPUs. If housekeeping CPUs are online then such 2845 interrupts are directed to the housekeeping CPU 2846 so that IO submitted on the housekeeping CPU 2847 cannot disturb the isolated CPU. 2848 2849 If a queue's affinity mask contains only isolated 2850 CPUs then this parameter has no effect on the 2851 interrupt routing decision, though interrupts are 2852 only delivered when tasks running on those 2853 isolated CPUs submit IO. IO submitted on 2854 housekeeping CPUs has no influence on those 2855 queues. 2856 2857 The format of <cpu-list> is described above. 2858 2859 iucv= [HW,NET] 2860 2861 ivrs_ioapic [HW,X86-64] 2862 Provide an override to the IOAPIC-ID<->DEVICE-ID 2863 mapping provided in the IVRS ACPI table. 2864 By default, PCI segment is 0, and can be omitted. 2865 2866 For example, to map IOAPIC-ID decimal 10 to 2867 PCI segment 0x1 and PCI device 00:14.0, 2868 write the parameter as: 2869 ivrs_ioapic=10@0001:00:14.0 2870 2871 Deprecated formats: 2872 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0 2873 write the parameter as: 2874 ivrs_ioapic[10]=00:14.0 2875 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and 2876 PCI device 00:14.0 write the parameter as: 2877 ivrs_ioapic[10]=0001:00:14.0 2878 2879 ivrs_hpet [HW,X86-64] 2880 Provide an override to the HPET-ID<->DEVICE-ID 2881 mapping provided in the IVRS ACPI table. 2882 By default, PCI segment is 0, and can be omitted. 2883 2884 For example, to map HPET-ID decimal 10 to 2885 PCI segment 0x1 and PCI device 00:14.0, 2886 write the parameter as: 2887 ivrs_hpet=10@0001:00:14.0 2888 2889 Deprecated formats: 2890 * To map HPET-ID decimal 0 to PCI device 00:14.0 2891 write the parameter as: 2892 ivrs_hpet[0]=00:14.0 2893 * To map HPET-ID decimal 10 to PCI segment 0x1 and 2894 PCI device 00:14.0 write the parameter as: 2895 ivrs_ioapic[10]=0001:00:14.0 2896 2897 ivrs_acpihid [HW,X86-64] 2898 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2899 mapping provided in the IVRS ACPI table. 2900 By default, PCI segment is 0, and can be omitted. 2901 2902 For example, to map UART-HID:UID AMD0020:0 to 2903 PCI segment 0x1 and PCI device ID 00:14.5, 2904 write the parameter as: 2905 ivrs_acpihid=AMD0020:0@0001:00:14.5 2906 2907 Deprecated formats: 2908 * To map UART-HID:UID AMD0020:0 to PCI segment is 0, 2909 PCI device ID 00:14.5, write the parameter as: 2910 ivrs_acpihid[00:14.5]=AMD0020:0 2911 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and 2912 PCI device ID 00:14.5, write the parameter as: 2913 ivrs_acpihid[0001:00:14.5]=AMD0020:0 2914 2915 js= [HW,JOY] Analog joystick 2916 See Documentation/input/joydev/joystick.rst. 2917 2918 kasan_multi_shot 2919 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2920 report on every invalid memory access. Without this 2921 parameter KASAN will print report only for the first 2922 invalid access. 2923 2924 keep_bootcon [KNL,EARLY] 2925 Do not unregister boot console at start. This is only 2926 useful for debugging when something happens in the window 2927 between unregistering the boot console and initializing 2928 the real console. 2929 2930 keepinitrd [HW,ARM] See retain_initrd. 2931 2932 kernelcore= [KNL,X86,PPC,EARLY] 2933 Format: nn[KMGTPE] | nn% | "mirror" 2934 This parameter specifies the amount of memory usable by 2935 the kernel for non-movable allocations. The requested 2936 amount is spread evenly throughout all nodes in the 2937 system as ZONE_NORMAL. The remaining memory is used for 2938 movable memory in its own zone, ZONE_MOVABLE. In the 2939 event, a node is too small to have both ZONE_NORMAL and 2940 ZONE_MOVABLE, kernelcore memory will take priority and 2941 other nodes will have a larger ZONE_MOVABLE. 2942 2943 ZONE_MOVABLE is used for the allocation of pages that 2944 may be reclaimed or moved by the page migration 2945 subsystem. Note that allocations like PTEs-from-HighMem 2946 still use the HighMem zone if it exists, and the Normal 2947 zone if it does not. 2948 2949 It is possible to specify the exact amount of memory in 2950 the form of "nn[KMGTPE]", a percentage of total system 2951 memory in the form of "nn%", or "mirror". If "mirror" 2952 option is specified, mirrored (reliable) memory is used 2953 for non-movable allocations and remaining memory is used 2954 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2955 are exclusive, so you cannot specify multiple forms. 2956 2957 kfence.burst= [MM,KFENCE] The number of additional successive 2958 allocations to be attempted through KFENCE for each 2959 sample interval. 2960 Format: <unsigned integer> 2961 Default: 0 2962 2963 kfence.check_on_panic= 2964 [MM,KFENCE] Whether to check all KFENCE-managed objects' 2965 canaries on panic. 2966 Format: <bool> 2967 Default: false 2968 2969 kfence.deferrable= 2970 [MM,KFENCE] Whether to use a deferrable timer to trigger 2971 allocations. This avoids forcing CPU wake-ups if the 2972 system is idle, at the risk of a less predictable 2973 sample interval. 2974 Format: <bool> 2975 Default: CONFIG_KFENCE_DEFERRABLE 2976 2977 kfence.sample_interval= 2978 [MM,KFENCE] KFENCE's sample interval in milliseconds. 2979 Format: <unsigned integer> 2980 0 - Disable KFENCE. 2981 >0 - Enabled KFENCE with given sample interval. 2982 Default: CONFIG_KFENCE_SAMPLE_INTERVAL 2983 2984 kfence.skip_covered_thresh= 2985 [MM,KFENCE] If pool utilization reaches this threshold 2986 (pool usage%), KFENCE limits currently covered 2987 allocations of the same source from further filling 2988 up the pool. 2989 Format: <unsigned integer> 2990 Default: 75 2991 2992 kgdbdbgp= [KGDB,HW,EARLY] kgdb over EHCI usb debug port. 2993 Format: <Controller#>[,poll interval] 2994 The controller # is the number of the ehci usb debug 2995 port as it is probed via PCI. The poll interval is 2996 optional and is the number seconds in between 2997 each poll cycle to the debug port in case you need 2998 the functionality for interrupting the kernel with 2999 gdb or control-c on the dbgp connection. When 3000 not using this parameter you use sysrq-g to break into 3001 the kernel debugger. 3002 3003 kgdboc= [KGDB,HW] kgdb over consoles. 3004 Requires a tty driver that supports console polling, 3005 or a supported polling keyboard driver (non-usb). 3006 Serial only format: <serial_device>[,baud] 3007 keyboard only format: kbd 3008 keyboard and serial format: kbd,<serial_device>[,baud] 3009 Optional Kernel mode setting: 3010 kms, kbd format: kms,kbd 3011 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 3012 3013 kgdboc_earlycon= [KGDB,HW,EARLY] 3014 If the boot console provides the ability to read 3015 characters and can work in polling mode, you can use 3016 this parameter to tell kgdb to use it as a backend 3017 until the normal console is registered. Intended to 3018 be used together with the kgdboc parameter which 3019 specifies the normal console to transition to. 3020 3021 The name of the early console should be specified 3022 as the value of this parameter. Note that the name of 3023 the early console might be different than the tty 3024 name passed to kgdboc. It's OK to leave the value 3025 blank and the first boot console that implements 3026 read() will be picked. 3027 3028 kgdbwait [KGDB,EARLY] Stop kernel execution and enter the 3029 kernel debugger at the earliest opportunity. 3030 3031 kho= [KEXEC,EARLY] 3032 Format: { "0" | "1" | "off" | "on" | "y" | "n" } 3033 Enables or disables Kexec HandOver. 3034 "0" | "off" | "n" - kexec handover is disabled 3035 "1" | "on" | "y" - kexec handover is enabled 3036 3037 kho_scratch= [KEXEC,EARLY] 3038 Format: ll[KMG],mm[KMG],nn[KMG] | nn% 3039 Defines the size of the KHO scratch region. The KHO 3040 scratch regions are physically contiguous memory 3041 ranges that can only be used for non-kernel 3042 allocations. That way, even when memory is heavily 3043 fragmented with handed over memory, the kexeced 3044 kernel will always have enough contiguous ranges to 3045 bootstrap itself. 3046 3047 It is possible to specify the exact amount of 3048 memory in the form of "ll[KMG],mm[KMG],nn[KMG]" 3049 where the first parameter defines the size of a low 3050 memory scratch area, the second parameter defines 3051 the size of a global scratch area and the third 3052 parameter defines the size of additional per-node 3053 scratch areas. The form "nn%" defines scale factor 3054 (in percents) of memory that was used during boot. 3055 3056 kmac= [MIPS] Korina ethernet MAC address. 3057 Configure the RouterBoard 532 series on-chip 3058 Ethernet adapter MAC address. 3059 3060 kmemleak= [KNL,EARLY] Boot-time kmemleak enable/disable 3061 Valid arguments: on, off 3062 Default: on 3063 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 3064 the default is off. 3065 3066 kprobe_event=[probe-list] 3067 [FTRACE] Add kprobe events and enable at boot time. 3068 The probe-list is a semicolon delimited list of probe 3069 definitions. Each definition is same as kprobe_events 3070 interface, but the parameters are comma delimited. 3071 For example, to add a kprobe event on vfs_read with 3072 arg1 and arg2, add to the command line; 3073 3074 kprobe_event=p,vfs_read,$arg1,$arg2 3075 3076 See also Documentation/trace/kprobetrace.rst "Kernel 3077 Boot Parameter" section. 3078 3079 kpti= [ARM64,EARLY] Control page table isolation of 3080 user and kernel address spaces. 3081 Default: enabled on cores which need mitigation. 3082 0: force disabled 3083 1: force enabled 3084 3085 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires 3086 CONFIG_KUNIT to be set to be fully enabled. The 3087 default value can be overridden via 3088 KUNIT_DEFAULT_ENABLED. 3089 Default is 1 (enabled) 3090 3091 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 3092 Default is 0 (don't ignore, but inject #GP) 3093 3094 kvm.eager_page_split= 3095 [KVM,X86] Controls whether or not KVM will try to 3096 proactively split all huge pages during dirty logging. 3097 Eager page splitting reduces interruptions to vCPU 3098 execution by eliminating the write-protection faults 3099 and MMU lock contention that would otherwise be 3100 required to split huge pages lazily. 3101 3102 VM workloads that rarely perform writes or that write 3103 only to a small region of VM memory may benefit from 3104 disabling eager page splitting to allow huge pages to 3105 still be used for reads. 3106 3107 The behavior of eager page splitting depends on whether 3108 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If 3109 disabled, all huge pages in a memslot will be eagerly 3110 split when dirty logging is enabled on that memslot. If 3111 enabled, eager page splitting will be performed during 3112 the KVM_CLEAR_DIRTY ioctl, and only for the pages being 3113 cleared. 3114 3115 Eager page splitting is only supported when kvm.tdp_mmu=Y. 3116 3117 Default is Y (on). 3118 3119 kvm.enable_pmu=[KVM,X86] 3120 If enabled, KVM will virtualize PMU functionality based 3121 on the virtual CPU model defined by userspace. This 3122 can be overridden on a per-VM basis via 3123 KVM_CAP_PMU_CAPABILITY. 3124 3125 If disabled, KVM will not virtualize PMU functionality, 3126 e.g. MSRs, PMCs, PMIs, etc., even if userspace defines 3127 a virtual CPU model that contains PMU assets. 3128 3129 Note, KVM's vPMU support implicitly requires running 3130 with an in-kernel local APIC, e.g. to deliver PMIs to 3131 the guest. Running without an in-kernel local APIC is 3132 not supported, though KVM will allow such a combination 3133 (with severely degraded functionality). 3134 3135 See also enable_mediated_pmu. 3136 3137 Default is Y (on). 3138 3139 kvm.enable_virt_at_load=[KVM,ARM64,LOONGARCH,MIPS,RISCV,X86] 3140 If enabled, KVM will enable virtualization in hardware 3141 when KVM is loaded, and disable virtualization when KVM 3142 is unloaded (if KVM is built as a module). 3143 3144 If disabled, KVM will dynamically enable and disable 3145 virtualization on-demand when creating and destroying 3146 VMs, i.e. on the 0=>1 and 1=>0 transitions of the 3147 number of VMs. 3148 3149 Enabling virtualization at module load avoids potential 3150 latency for creation of the 0=>1 VM, as KVM serializes 3151 virtualization enabling across all online CPUs. The 3152 "cost" of enabling virtualization when KVM is loaded, 3153 is that doing so may interfere with using out-of-tree 3154 hypervisors that want to "own" virtualization hardware. 3155 3156 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 3157 Default is false (don't support). 3158 3159 kvm.nx_huge_pages= 3160 [KVM] Controls the software workaround for the 3161 X86_BUG_ITLB_MULTIHIT bug. 3162 force : Always deploy workaround. 3163 off : Never deploy workaround. 3164 auto : Deploy workaround based on the presence of 3165 X86_BUG_ITLB_MULTIHIT. 3166 3167 Default is 'auto'. 3168 3169 If the software workaround is enabled for the host, 3170 guests do need not to enable it for nested guests. 3171 3172 kvm.nx_huge_pages_recovery_ratio= 3173 [KVM] Controls how many 4KiB pages are periodically zapped 3174 back to huge pages. 0 disables the recovery, otherwise if 3175 the value is N KVM will zap 1/Nth of the 4KiB pages every 3176 period (see below). The default is 60. 3177 3178 kvm.nx_huge_pages_recovery_period_ms= 3179 [KVM] Controls the time period at which KVM zaps 4KiB pages 3180 back to huge pages. If the value is a non-zero N, KVM will 3181 zap a portion (see ratio above) of the pages every N msecs. 3182 If the value is 0 (the default), KVM will pick a period based 3183 on the ratio, such that a page is zapped after 1 hour on average. 3184 3185 kvm-{amd,intel}.enable_mediated_pmu=[KVM,AMD,INTEL] 3186 If enabled, KVM will provide a mediated virtual PMU, 3187 instead of the default perf-based virtual PMU (if 3188 kvm.enable_pmu is true and PMU is enumerated via the 3189 virtual CPU model). 3190 3191 With a perf-based vPMU, KVM operates as a user of perf, 3192 i.e. emulates guest PMU counters using perf events. 3193 KVM-created perf events are managed by perf as regular 3194 (guest-only) events, e.g. are scheduled in/out, contend 3195 for hardware resources, etc. Using a perf-based vPMU 3196 allows guest and host usage of the PMU to co-exist, but 3197 incurs non-trivial overhead and can result in silently 3198 dropped guest events (due to resource contention). 3199 3200 With a mediated vPMU, hardware PMU state is context 3201 switched around the world switch to/from the guest. 3202 KVM mediates which events the guest can utilize, but 3203 gives the guest direct access to all other PMU assets 3204 when possible (KVM may intercept some accesses if the 3205 virtual CPU model provides a subset of hardware PMU 3206 functionality). Using a mediated vPMU significantly 3207 reduces PMU virtualization overhead and eliminates lost 3208 guest events, but is mutually exclusive with using perf 3209 to profile KVM guests and adds latency to most VM-Exits 3210 (to context switch PMU state). 3211 3212 Default is N (off). 3213 3214 kvm-amd.nested= [KVM,AMD] Control nested virtualization feature in 3215 KVM/SVM. Default is 1 (enabled). 3216 3217 kvm-amd.npt= [KVM,AMD] Control KVM's use of Nested Page Tables, 3218 a.k.a. Two-Dimensional Page Tables. Default is 1 3219 (enabled). Disable by KVM if hardware lacks support 3220 for NPT. 3221 3222 kvm-amd.ciphertext_hiding_asids= 3223 [KVM,AMD] Ciphertext hiding prevents disallowed accesses 3224 to SNP private memory from reading ciphertext. Instead, 3225 reads will see constant default values (0xff). 3226 3227 If ciphertext hiding is enabled, the joint SEV-ES and 3228 SEV-SNP ASID space is partitioned into separate SEV-ES 3229 and SEV-SNP ASID ranges, with the SEV-SNP range being 3230 [1..max_snp_asid] and the SEV-ES range being 3231 (max_snp_asid..min_sev_asid), where min_sev_asid is 3232 enumerated by CPUID.0x.8000_001F[EDX]. 3233 3234 A non-zero value enables SEV-SNP ciphertext hiding and 3235 adjusts the ASID ranges for SEV-ES and SEV-SNP guests. 3236 KVM caps the number of SEV-SNP ASIDs at the maximum 3237 possible value, e.g. specifying -1u will assign all 3238 joint SEV-ES and SEV-SNP ASIDs to SEV-SNP. Note, 3239 assigning all joint ASIDs to SEV-SNP, i.e. configuring 3240 max_snp_asid == min_sev_asid-1, will effectively make 3241 SEV-ES unusable. 3242 3243 kvm-arm.mode= 3244 [KVM,ARM,EARLY] Select one of KVM/arm64's modes of 3245 operation. 3246 3247 none: Forcefully disable KVM. 3248 3249 nvhe: Standard nVHE-based mode, without support for 3250 protected guests. 3251 3252 protected: Mode with support for guests whose state is 3253 kept private from the host, using VHE or 3254 nVHE depending on HW support. 3255 3256 nested: VHE-based mode with support for nested 3257 virtualization. Requires at least ARMv8.4 3258 hardware (with FEAT_NV2). 3259 3260 Defaults to VHE/nVHE based on hardware support. Setting 3261 mode to "protected" will disable kexec and hibernation 3262 for the host. To force nVHE on VHE hardware, add 3263 "arm64_sw.hvhe=0 id_aa64mmfr1.vh=0" to the 3264 command-line. 3265 "nested" is experimental and should be used with 3266 extreme caution. 3267 3268 kvm-arm.vgic_v3_group0_trap= 3269 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-0 3270 system registers 3271 3272 kvm-arm.vgic_v3_group1_trap= 3273 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-1 3274 system registers 3275 3276 kvm-arm.vgic_v3_common_trap= 3277 [KVM,ARM,EARLY] Trap guest accesses to GICv3 common 3278 system registers 3279 3280 kvm-arm.vgic_v4_enable= 3281 [KVM,ARM,EARLY] Allow use of GICv4 for direct 3282 injection of LPIs. 3283 3284 kvm-arm.wfe_trap_policy= 3285 [KVM,ARM] Control when to set WFE instruction trap for 3286 KVM VMs. Traps are allowed but not guaranteed by the 3287 CPU architecture. 3288 3289 trap: set WFE instruction trap 3290 3291 notrap: clear WFE instruction trap 3292 3293 kvm-arm.wfi_trap_policy= 3294 [KVM,ARM] Control when to set WFI instruction trap for 3295 KVM VMs. Traps are allowed but not guaranteed by the 3296 CPU architecture. 3297 3298 trap: set WFI instruction trap 3299 3300 notrap: clear WFI instruction trap 3301 3302 kvm_cma_resv_ratio=n [PPC,EARLY] 3303 Reserves given percentage from system memory area for 3304 contiguous memory allocation for KVM hash pagetable 3305 allocation. 3306 By default it reserves 5% of total system memory. 3307 Format: <integer> 3308 Default: 5 3309 3310 kvm-intel.ept= [KVM,Intel] Control KVM's use of Extended Page Tables, 3311 a.k.a. Two-Dimensional Page Tables. Default is 1 3312 (enabled). Disable by KVM if hardware lacks support 3313 for EPT. 3314 3315 kvm-intel.emulate_invalid_guest_state= 3316 [KVM,Intel] Control whether to emulate invalid guest 3317 state. Ignored if kvm-intel.enable_unrestricted_guest=1, 3318 as guest state is never invalid for unrestricted 3319 guests. This param doesn't apply to nested guests (L2), 3320 as KVM never emulates invalid L2 guest state. 3321 Default is 1 (enabled). 3322 3323 kvm-intel.flexpriority= 3324 [KVM,Intel] Control KVM's use of FlexPriority feature 3325 (TPR shadow). Default is 1 (enabled). Disable by KVM if 3326 hardware lacks support for it. 3327 3328 kvm-intel.nested= 3329 [KVM,Intel] Control nested virtualization feature in 3330 KVM/VMX. Default is 1 (enabled). 3331 3332 kvm-intel.unrestricted_guest= 3333 [KVM,Intel] Control KVM's use of unrestricted guest 3334 feature (virtualized real and unpaged mode). Default 3335 is 1 (enabled). Disable by KVM if EPT is disabled or 3336 hardware lacks support for it. 3337 3338 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 3339 CVE-2018-3620. 3340 3341 Valid arguments: never, cond, always 3342 3343 always: L1D cache flush on every VMENTER. 3344 cond: Flush L1D on VMENTER only when the code between 3345 VMEXIT and VMENTER can leak host memory. 3346 never: Disables the mitigation 3347 3348 Default is cond (do L1 cache flush in specific instances) 3349 3350 kvm-intel.vpid= [KVM,Intel] Control KVM's use of Virtual Processor 3351 Identification feature (tagged TLBs). Default is 1 3352 (enabled). Disable by KVM if hardware lacks support 3353 for it. 3354 3355 l1d_flush= [X86,INTEL,EARLY] 3356 Control mitigation for L1D based snooping vulnerability. 3357 3358 Certain CPUs are vulnerable to an exploit against CPU 3359 internal buffers which can forward information to a 3360 disclosure gadget under certain conditions. 3361 3362 In vulnerable processors, the speculatively 3363 forwarded data can be used in a cache side channel 3364 attack, to access data to which the attacker does 3365 not have direct access. 3366 3367 This parameter controls the mitigation. The 3368 options are: 3369 3370 on - enable the interface for the mitigation 3371 3372 l1tf= [X86,EARLY] Control mitigation of the L1TF vulnerability on 3373 affected CPUs 3374 3375 The kernel PTE inversion protection is unconditionally 3376 enabled and cannot be disabled. 3377 3378 full 3379 Provides all available mitigations for the 3380 L1TF vulnerability. Disables SMT and 3381 enables all mitigations in the 3382 hypervisors, i.e. unconditional L1D flush. 3383 3384 SMT control and L1D flush control via the 3385 sysfs interface is still possible after 3386 boot. Hypervisors will issue a warning 3387 when the first VM is started in a 3388 potentially insecure configuration, 3389 i.e. SMT enabled or L1D flush disabled. 3390 3391 full,force 3392 Same as 'full', but disables SMT and L1D 3393 flush runtime control. Implies the 3394 'nosmt=force' command line option. 3395 (i.e. sysfs control of SMT is disabled.) 3396 3397 flush 3398 Leaves SMT enabled and enables the default 3399 hypervisor mitigation, i.e. conditional 3400 L1D flush. 3401 3402 SMT control and L1D flush control via the 3403 sysfs interface is still possible after 3404 boot. Hypervisors will issue a warning 3405 when the first VM is started in a 3406 potentially insecure configuration, 3407 i.e. SMT enabled or L1D flush disabled. 3408 3409 flush,nosmt 3410 3411 Disables SMT and enables the default 3412 hypervisor mitigation. 3413 3414 SMT control and L1D flush control via the 3415 sysfs interface is still possible after 3416 boot. Hypervisors will issue a warning 3417 when the first VM is started in a 3418 potentially insecure configuration, 3419 i.e. SMT enabled or L1D flush disabled. 3420 3421 flush,nowarn 3422 Same as 'flush', but hypervisors will not 3423 warn when a VM is started in a potentially 3424 insecure configuration. 3425 3426 off 3427 Disables hypervisor mitigations and doesn't 3428 emit any warnings. 3429 It also drops the swap size and available 3430 RAM limit restriction on both hypervisor and 3431 bare metal. 3432 3433 Default is 'flush'. 3434 3435 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 3436 3437 l2cr= [PPC] 3438 3439 l3cr= [PPC] 3440 3441 lapic [X86-32,APIC,EARLY] Enable the local APIC even if BIOS 3442 disabled it. 3443 3444 lapic= [X86,APIC] Do not use TSC deadline 3445 value for LAPIC timer one-shot implementation. Default 3446 back to the programmable timer unit in the LAPIC. 3447 Format: notscdeadline 3448 3449 lapic_timer_c2_ok [X86,APIC,EARLY] trust the local apic timer 3450 in C2 power state. 3451 3452 libata.dma= [LIBATA] DMA control 3453 libata.dma=0 Disable all PATA and SATA DMA 3454 libata.dma=1 PATA and SATA Disk DMA only 3455 libata.dma=2 ATAPI (CDROM) DMA only 3456 libata.dma=4 Compact Flash DMA only 3457 Combinations also work, so libata.dma=3 enables DMA 3458 for disks and CDROMs, but not CFs. 3459 3460 libata.ignore_hpa= [LIBATA] Ignore HPA limit 3461 libata.ignore_hpa=0 keep BIOS limits (default) 3462 libata.ignore_hpa=1 ignore limits, using full disk 3463 3464 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 3465 when set. 3466 Format: <int> 3467 3468 libata.force= [LIBATA] Force configurations. The format is a comma- 3469 separated list of "[ID:]VAL" where ID is PORT[.DEVICE]. 3470 PORT and DEVICE are decimal numbers matching port, link 3471 or device. Basically, it matches the ATA ID string 3472 printed on console by libata. If the whole ID part is 3473 omitted, the last PORT and DEVICE values are used. If 3474 ID hasn't been specified yet, the configuration applies 3475 to all ports, links and devices. 3476 3477 If only DEVICE is omitted, the parameter applies to 3478 the port and all links and devices behind it. DEVICE 3479 number of 0 either selects the first device or the 3480 first fan-out link behind PMP device. It does not 3481 select the host link. DEVICE number of 15 selects the 3482 host link and device attached to it. 3483 3484 The VAL specifies the configuration to force. As long 3485 as there is no ambiguity, shortcut notation is allowed. 3486 For example, both 1.5 and 1.5G would work for 1.5Gbps. 3487 The following configurations can be forced. 3488 3489 * Cable type: 40c, 80c, short40c, unk, ign or sata. 3490 Any ID with matching PORT is used. 3491 3492 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 3493 3494 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 3495 udma[/][16,25,33,44,66,100,133] notation is also 3496 allowed. 3497 3498 * nohrst, nosrst, norst: suppress hard, soft and both 3499 resets. 3500 3501 * rstonce: only attempt one reset during hot-unplug 3502 link recovery. 3503 3504 * [no]dbdelay: Enable or disable the extra 200ms delay 3505 before debouncing a link PHY and device presence 3506 detection. 3507 3508 * [no]ncq: Turn on or off NCQ. 3509 3510 * [no]ncqtrim: Enable or disable queued DSM TRIM. 3511 3512 * [no]ncqati: Enable or disable NCQ trim on ATI chipset. 3513 3514 * [no]trim: Enable or disable (unqueued) TRIM. 3515 3516 * trim_zero: Indicate that TRIM command zeroes data. 3517 3518 * max_trim_128m: Set 128M maximum trim size limit. 3519 3520 * [no]dma: Turn on or off DMA transfers. 3521 3522 * atapi_dmadir: Enable ATAPI DMADIR bridge support. 3523 3524 * atapi_mod16_dma: Enable the use of ATAPI DMA for 3525 commands that are not a multiple of 16 bytes. 3526 3527 * [no]dmalog: Enable or disable the use of the 3528 READ LOG DMA EXT command to access logs. 3529 3530 * [no]iddevlog: Enable or disable access to the 3531 identify device data log. 3532 3533 * [no]logdir: Enable or disable access to the general 3534 purpose log directory. 3535 3536 * max_sec=<sectors>: Set the transfer size limit, in 3537 number of 512-byte sectors, to the value specified in 3538 <sectors>. The value specified in <sectors> has to be 3539 a non-zero positive integer. 3540 3541 * max_sec_128: Set transfer size limit to 128 sectors. 3542 3543 * max_sec_1024: Set or clear transfer size limit to 3544 1024 sectors. 3545 3546 * max_sec_lba48: Set or clear transfer size limit to 3547 65535 sectors. 3548 3549 * external: Mark port as external (hotplug-capable). 3550 3551 * [no]lpm: Enable or disable link power management. 3552 3553 * [no]setxfer: Indicate if transfer speed mode setting 3554 should be skipped. 3555 3556 * [no]fua: Disable or enable FUA (Force Unit Access) 3557 support for devices supporting this feature. 3558 3559 * dump_id: Dump IDENTIFY data. 3560 3561 * disable: Disable this device. 3562 3563 If there are multiple matching configurations changing 3564 the same attribute, the last one is used. 3565 3566 liveupdate= [KNL,EARLY] 3567 Format: <bool> 3568 Enable Live Update Orchestrator (LUO). 3569 Default: off. 3570 3571 lockd.nlm_grace_period=P [NFS] Assign grace period. 3572 Format: <integer> 3573 3574 lockd.nlm_tcpport=N [NFS] Assign TCP port. 3575 Format: <integer> 3576 3577 lockd.nlm_timeout=T [NFS] Assign timeout value. 3578 Format: <integer> 3579 3580 lockd.nlm_udpport=M [NFS] Assign UDP port. 3581 Format: <integer> 3582 3583 lockdown= [SECURITY,EARLY] 3584 { integrity | confidentiality } 3585 Enable the kernel lockdown feature. If set to 3586 integrity, kernel features that allow userland to 3587 modify the running kernel are disabled. If set to 3588 confidentiality, kernel features that allow userland 3589 to extract confidential information from the kernel 3590 are also disabled. 3591 3592 locktorture.acq_writer_lim= [KNL] 3593 Set the time limit in jiffies for a lock 3594 acquisition. Acquisitions exceeding this limit 3595 will result in a splat once they do complete. 3596 3597 locktorture.bind_readers= [KNL] 3598 Specify the list of CPUs to which the readers are 3599 to be bound. 3600 3601 locktorture.bind_writers= [KNL] 3602 Specify the list of CPUs to which the writers are 3603 to be bound. 3604 3605 locktorture.call_rcu_chains= [KNL] 3606 Specify the number of self-propagating call_rcu() 3607 chains to set up. These are used to ensure that 3608 there is a high probability of an RCU grace period 3609 in progress at any given time. Defaults to 0, 3610 which disables these call_rcu() chains. 3611 3612 locktorture.long_hold= [KNL] 3613 Specify the duration in milliseconds for the 3614 occasional long-duration lock hold time. Defaults 3615 to 100 milliseconds. Select 0 to disable. 3616 3617 locktorture.nested_locks= [KNL] 3618 Specify the maximum lock nesting depth that 3619 locktorture is to exercise, up to a limit of 8 3620 (MAX_NESTED_LOCKS). Specify zero to disable. 3621 Note that this parameter is ineffective on types 3622 of locks that do not support nested acquisition. 3623 3624 locktorture.nreaders_stress= [KNL] 3625 Set the number of locking read-acquisition kthreads. 3626 Defaults to being automatically set based on the 3627 number of online CPUs. 3628 3629 locktorture.nwriters_stress= [KNL] 3630 Set the number of locking write-acquisition kthreads. 3631 3632 locktorture.onoff_holdoff= [KNL] 3633 Set time (s) after boot for CPU-hotplug testing. 3634 3635 locktorture.onoff_interval= [KNL] 3636 Set time (s) between CPU-hotplug operations, or 3637 zero to disable CPU-hotplug testing. 3638 3639 locktorture.rt_boost= [KNL] 3640 Do periodic testing of real-time lock priority 3641 boosting. Select 0 to disable, 1 to boost 3642 only rt_mutex, and 2 to boost unconditionally. 3643 Defaults to 2, which might seem to be an 3644 odd choice, but which should be harmless for 3645 non-real-time spinlocks, due to their disabling 3646 of preemption. Note that non-realtime mutexes 3647 disable boosting. 3648 3649 locktorture.rt_boost_factor= [KNL] 3650 Number that determines how often and for how 3651 long priority boosting is exercised. This is 3652 scaled down by the number of writers, so that the 3653 number of boosts per unit time remains roughly 3654 constant as the number of writers increases. 3655 On the other hand, the duration of each boost 3656 increases with the number of writers. 3657 3658 locktorture.shuffle_interval= [KNL] 3659 Set task-shuffle interval (jiffies). Shuffling 3660 tasks allows some CPUs to go into dyntick-idle 3661 mode during the locktorture test. 3662 3663 locktorture.shutdown_secs= [KNL] 3664 Set time (s) after boot system shutdown. This 3665 is useful for hands-off automated testing. 3666 3667 locktorture.stat_interval= [KNL] 3668 Time (s) between statistics printk()s. 3669 3670 locktorture.stutter= [KNL] 3671 Time (s) to stutter testing, for example, 3672 specifying five seconds causes the test to run for 3673 five seconds, wait for five seconds, and so on. 3674 This tests the locking primitive's ability to 3675 transition abruptly to and from idle. 3676 3677 locktorture.torture_type= [KNL] 3678 Specify the locking implementation to test. 3679 3680 locktorture.verbose= [KNL] 3681 Enable additional printk() statements. 3682 3683 locktorture.writer_fifo= [KNL] 3684 Run the write-side locktorture kthreads at 3685 sched_set_fifo() real-time priority. 3686 3687 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 3688 Format: <irq> 3689 3690 loglevel= [KNL,EARLY] 3691 All Kernel Messages with a loglevel smaller than the 3692 console loglevel will be printed to the console. It can 3693 also be changed with klogd or other programs. The 3694 loglevels are defined as follows: 3695 3696 0 (KERN_EMERG) system is unusable 3697 1 (KERN_ALERT) action must be taken immediately 3698 2 (KERN_CRIT) critical conditions 3699 3 (KERN_ERR) error conditions 3700 4 (KERN_WARNING) warning conditions 3701 5 (KERN_NOTICE) normal but significant condition 3702 6 (KERN_INFO) informational 3703 7 (KERN_DEBUG) debug-level messages 3704 3705 log_buf_len=n[KMG] [KNL,EARLY] 3706 Sets the size of the printk ring buffer, in bytes. 3707 n must be a power of two and greater than the 3708 minimal size. The minimal size is defined by 3709 LOG_BUF_SHIFT kernel config parameter. There 3710 is also CONFIG_LOG_CPU_MAX_BUF_SHIFT config 3711 parameter that allows to increase the default size 3712 depending on the number of CPUs. See init/Kconfig 3713 for more details. 3714 3715 logo.nologo [FB] Disables display of the built-in Linux logo. 3716 This may be used to provide more screen space for 3717 kernel log messages and is useful when debugging 3718 kernel boot problems. 3719 3720 lp=0 [LP] Specify parallel ports to use, e.g, 3721 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 3722 lp=reset first parallel port). 'lp=0' disables the 3723 lp=auto printer driver. 'lp=reset' (which can be 3724 specified in addition to the ports) causes 3725 attached printers to be reset. Using 3726 lp=port1,port2,... specifies the parallel ports 3727 to associate lp devices with, starting with 3728 lp0. A port specification may be 'none' to skip 3729 that lp device, or a parport name such as 3730 'parport0'. Specifying 'lp=auto' instead of a 3731 port specification list means that device IDs 3732 from each port should be examined, to see if 3733 an IEEE 1284-compliant printer is attached; if 3734 so, the driver will manage that printer. 3735 See also header of drivers/char/lp.c. 3736 3737 lpj=n [KNL] 3738 Sets loops_per_jiffy to given constant, thus avoiding 3739 time-consuming boot-time autodetection (up to 250 ms per 3740 CPU). 0 enables autodetection (default). To determine 3741 the correct value for your kernel, boot with normal 3742 autodetection and see what value is printed. Note that 3743 on SMP systems the preset will be applied to all CPUs, 3744 which is likely to cause problems if your CPUs need 3745 significantly divergent settings. An incorrect value 3746 will cause delays in the kernel to be wrong, leading to 3747 unpredictable I/O errors and other breakage. Although 3748 unlikely, in the extreme case this might damage your 3749 hardware. 3750 3751 lsm.debug [SECURITY] Enable LSM initialization debugging output. 3752 3753 lsm=lsm1,...,lsmN 3754 [SECURITY] Choose order of LSM initialization. This 3755 overrides CONFIG_LSM, and the "security=" parameter. 3756 3757 machtype= [Loongson] Share the same kernel image file between 3758 different yeeloong laptops. 3759 Example: machtype=lemote-yeeloong-2f-7inch 3760 3761 maxcpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 3762 will bring up during bootup. maxcpus=n : n >= 0 limits 3763 the kernel to bring up 'n' processors. Surely after 3764 bootup you can bring up the other plugged cpu by executing 3765 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 3766 only takes effect during system bootup. 3767 While n=0 is a special case, it is equivalent to "nosmp", 3768 which also disables the IO APIC. 3769 3770 max_loop= [LOOP] The number of loop block devices that get 3771 (loop.max_loop) unconditionally pre-created at init time. The default 3772 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 3773 of statically allocating a predefined number, loop 3774 devices can be requested on-demand with the 3775 /dev/loop-control interface. 3776 3777 mce= [X86-{32,64}] 3778 3779 Please see Documentation/arch/x86/x86_64/machinecheck.rst for sysfs runtime tunables. 3780 3781 off 3782 disable machine check 3783 3784 no_cmci 3785 disable CMCI(Corrected Machine Check Interrupt) that 3786 Intel processor supports. Usually this disablement is 3787 not recommended, but it might be handy if your 3788 hardware is misbehaving. 3789 3790 Note that you'll get more problems without CMCI than 3791 with due to the shared banks, i.e. you might get 3792 duplicated error logs. 3793 3794 dont_log_ce 3795 don't make logs for corrected errors. All events 3796 reported as corrected are silently cleared by OS. This 3797 option will be useful if you have no interest in any 3798 of corrected errors. 3799 3800 ignore_ce 3801 disable features for corrected errors, e.g. 3802 polling timer and CMCI. All events reported as 3803 corrected are not cleared by OS and remained in its 3804 error banks. 3805 3806 Usually this disablement is not recommended, however 3807 if there is an agent checking/clearing corrected 3808 errors (e.g. BIOS or hardware monitoring 3809 applications), conflicting with OS's error handling, 3810 and you cannot deactivate the agent, then this option 3811 will be a help. 3812 3813 no_lmce 3814 do not opt-in to Local MCE delivery. Use legacy method 3815 to broadcast MCEs. 3816 3817 bootlog 3818 enable logging of machine checks left over from 3819 booting. Disabled by default on AMD Fam10h and older 3820 because some BIOS leave bogus ones. 3821 3822 If your BIOS doesn't do that it's a good idea to 3823 enable though to make sure you log even machine check 3824 events that result in a reboot. On Intel systems it is 3825 enabled by default. 3826 3827 nobootlog 3828 disable boot machine check logging. 3829 3830 monarchtimeout (number) 3831 sets the time in us to wait for other CPUs on machine 3832 checks. 0 to disable. 3833 3834 bios_cmci_threshold 3835 don't overwrite the bios-set CMCI threshold. This boot 3836 option prevents Linux from overwriting the CMCI 3837 threshold set by the bios. Without this option, Linux 3838 always sets the CMCI threshold to 1. Enabling this may 3839 make memory predictive failure analysis less effective 3840 if the bios sets thresholds for memory errors since we 3841 will not see details for all errors. 3842 3843 recovery 3844 force-enable recoverable machine check code paths 3845 3846 Everything else is in sysfs now. 3847 3848 3849 md= [HW] RAID subsystems devices and level 3850 See Documentation/admin-guide/md.rst. 3851 3852 mdacon= [MDA] 3853 Format: <first>,<last> 3854 Specifies range of consoles to be captured by the MDA. 3855 3856 mds= [X86,INTEL,EARLY] 3857 Control mitigation for the Micro-architectural Data 3858 Sampling (MDS) vulnerability. 3859 3860 Certain CPUs are vulnerable to an exploit against CPU 3861 internal buffers which can forward information to a 3862 disclosure gadget under certain conditions. 3863 3864 In vulnerable processors, the speculatively 3865 forwarded data can be used in a cache side channel 3866 attack, to access data to which the attacker does 3867 not have direct access. 3868 3869 This parameter controls the MDS mitigation. The 3870 options are: 3871 3872 full - Enable MDS mitigation on vulnerable CPUs 3873 full,nosmt - Enable MDS mitigation and disable 3874 SMT on vulnerable CPUs 3875 off - Unconditionally disable MDS mitigation 3876 3877 On TAA-affected machines, mds=off can be prevented by 3878 an active TAA mitigation as both vulnerabilities are 3879 mitigated with the same mechanism so in order to disable 3880 this mitigation, you need to specify tsx_async_abort=off 3881 too. 3882 3883 Not specifying this option is equivalent to 3884 mds=full. 3885 3886 For details see: Documentation/admin-guide/hw-vuln/mds.rst 3887 3888 mem=nn[KMG] [HEXAGON,EARLY] Set the memory size. 3889 Must be specified, otherwise memory size will be 0. 3890 3891 mem=nn[KMG] [KNL,BOOT,EARLY] Force usage of a specific amount 3892 of memory Amount of memory to be used in cases 3893 as follows: 3894 3895 1 for test; 3896 2 when the kernel is not able to see the whole system memory; 3897 3 memory that lies after 'mem=' boundary is excluded from 3898 the hypervisor, then assigned to KVM guests. 3899 4 to limit the memory available for kdump kernel. 3900 3901 [ARC,MICROBLAZE] - the limit applies only to low memory, 3902 high memory is not affected. 3903 3904 [ARM64] - only limits memory covered by the linear 3905 mapping. The NOMAP regions are not affected. 3906 3907 [X86] Work as limiting max address. Use together 3908 with memmap= to avoid physical address space collisions. 3909 Without memmap= PCI devices could be placed at addresses 3910 belonging to unused RAM. 3911 3912 Note that this only takes effects during boot time since 3913 in above case 3, memory may need be hot added after boot 3914 if system memory of hypervisor is not sufficient. 3915 3916 mem=nn[KMG]@ss[KMG] 3917 [ARM,MIPS,EARLY] - override the memory layout 3918 reported by firmware. 3919 Define a memory region of size nn[KMG] starting at 3920 ss[KMG]. 3921 Multiple different regions can be specified with 3922 multiple mem= parameters on the command line. 3923 3924 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 3925 memory. 3926 3927 memblock=debug [KNL,EARLY] Enable memblock debug messages. 3928 3929 memchunk=nn[KMG] 3930 [KNL,SH] Allow user to override the default size for 3931 per-device physically contiguous DMA buffers. 3932 3933 memhp_default_state=online/offline/online_kernel/online_movable 3934 [KNL] Set the initial state for the memory hotplug 3935 onlining policy. If not specified, the default value is 3936 set according to the 3937 CONFIG_MHP_DEFAULT_ONLINE_TYPE kernel config 3938 options. 3939 See Documentation/admin-guide/mm/memory-hotplug.rst. 3940 3941 memmap=exactmap [KNL,X86,EARLY] Enable setting of an exact 3942 E820 memory map, as specified by the user. 3943 Such memmap=exactmap lines can be constructed based on 3944 BIOS output or other requirements. See the memmap=nn@ss 3945 option description. 3946 3947 memmap=nn[KMG]@ss[KMG] 3948 [KNL, X86,MIPS,XTENSA,EARLY] Force usage of a specific region of memory. 3949 Region of memory to be used is from ss to ss+nn. 3950 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 3951 which limits max address to nn[KMG]. 3952 Multiple different regions can be specified, 3953 comma delimited. 3954 Example: 3955 memmap=100M@2G,100M#3G,1G!1024G 3956 3957 memmap=nn[KMG]#ss[KMG] 3958 [KNL,ACPI,EARLY] Mark specific memory as ACPI data. 3959 Region of memory to be marked is from ss to ss+nn. 3960 3961 memmap=nn[KMG]$ss[KMG] 3962 [KNL,ACPI,EARLY] Mark specific memory as reserved. 3963 Region of memory to be reserved is from ss to ss+nn. 3964 Example: Exclude memory from 0x18690000-0x1869ffff 3965 memmap=64K$0x18690000 3966 or 3967 memmap=0x10000$0x18690000 3968 Some bootloaders may need an escape character before '$', 3969 like Grub2, otherwise '$' and the following number 3970 will be eaten. 3971 3972 memmap=nn[KMG]!ss[KMG,EARLY] 3973 [KNL,X86] Mark specific memory as protected. 3974 Region of memory to be used, from ss to ss+nn. 3975 The memory region may be marked as e820 type 12 (0xc) 3976 and is NVDIMM or ADR memory. 3977 3978 memmap=<size>%<offset>-<oldtype>+<newtype> 3979 [KNL,ACPI,EARLY] Convert memory within the specified region 3980 from <oldtype> to <newtype>. If "-<oldtype>" is left 3981 out, the whole region will be marked as <newtype>, 3982 even if previously unavailable. If "+<newtype>" is left 3983 out, matching memory will be removed. Types are 3984 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 3985 3 = ACPI, 12 = PRAM. 3986 3987 memory_corruption_check=0/1 [X86,EARLY] 3988 Some BIOSes seem to corrupt the first 64k of 3989 memory when doing things like suspend/resume. 3990 Setting this option will scan the memory 3991 looking for corruption. Enabling this will 3992 both detect corruption and prevent the kernel 3993 from using the memory being corrupted. 3994 However, it's intended as a diagnostic tool; if 3995 repeatable BIOS-originated corruption always 3996 affects the same memory, you can use memmap= 3997 to prevent the kernel from using that memory. 3998 3999 memory_corruption_check_size=size [X86,EARLY] 4000 By default it checks for corruption in the low 4001 64k, making this memory unavailable for normal 4002 use. Use this parameter to scan for 4003 corruption in more or less memory. 4004 4005 memory_corruption_check_period=seconds [X86,EARLY] 4006 By default it checks for corruption every 60 4007 seconds. Use this parameter to check at some 4008 other rate. 0 disables periodic checking. 4009 4010 memory_hotplug.memmap_on_memory 4011 [KNL,X86,ARM] Boolean flag to enable this feature. 4012 Format: {on | off (default)} 4013 When enabled, runtime hotplugged memory will 4014 allocate its internal metadata (struct pages, 4015 those vmemmap pages cannot be optimized even 4016 if hugetlb_free_vmemmap is enabled) from the 4017 hotadded memory which will allow to hotadd a 4018 lot of memory without requiring additional 4019 memory to do so. 4020 This feature is disabled by default because it 4021 has some implication on large (e.g. GB) 4022 allocations in some configurations (e.g. small 4023 memory blocks). 4024 The state of the flag can be read in 4025 /sys/module/memory_hotplug/parameters/memmap_on_memory. 4026 Note that even when enabled, there are a few cases where 4027 the feature is not effective. 4028 4029 memtest= [KNL,X86,ARM,M68K,PPC,RISCV,EARLY] Enable memtest 4030 Format: <integer> 4031 default : 0 <disable> 4032 Specifies the number of memtest passes to be 4033 performed. Each pass selects another test 4034 pattern from a given set of patterns. Memtest 4035 fills the memory with this pattern, validates 4036 memory contents and reserves bad memory 4037 regions that are detected. 4038 4039 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 4040 Valid arguments: on, off 4041 Default: off 4042 mem_encrypt=on: Activate SME 4043 mem_encrypt=off: Do not activate SME 4044 4045 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst 4046 for details on when memory encryption can be activated. 4047 4048 mem_sleep_default= [SUSPEND] Default system suspend mode: 4049 s2idle - Suspend-To-Idle 4050 shallow - Power-On Suspend or equivalent (if supported) 4051 deep - Suspend-To-RAM or equivalent (if supported) 4052 See Documentation/admin-guide/pm/sleep-states.rst. 4053 4054 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 4055 the BIOS has incorrectly applied a workaround. TinyBIOS 4056 version 0.98 is known to be affected, 0.99 fixes the 4057 problem by letting the user disable the workaround. 4058 4059 mga= [HW,DRM] 4060 4061 microcode= [X86] Control the behavior of the microcode loader. 4062 Available options, comma separated: 4063 4064 base_rev=X - with <X> with format: <u32> 4065 Set the base microcode revision of each thread when in 4066 debug mode. 4067 4068 dis_ucode_ldr: disable the microcode loader 4069 4070 force_minrev: 4071 Enable or disable the microcode minimal revision 4072 enforcement for the runtime microcode loader. 4073 4074 mini2440= [ARM,HW,KNL] 4075 Format:[0..2][b][c][t] 4076 Default: "0tb" 4077 MINI2440 configuration specification: 4078 0 - The attached screen is the 3.5" TFT 4079 1 - The attached screen is the 7" TFT 4080 2 - The VGA Shield is attached (1024x768) 4081 Leaving out the screen size parameter will not load 4082 the TFT driver, and the framebuffer will be left 4083 unconfigured. 4084 b - Enable backlight. The TFT backlight pin will be 4085 linked to the kernel VESA blanking code and a GPIO 4086 LED. This parameter is not necessary when using the 4087 VGA shield. 4088 c - Enable the s3c camera interface. 4089 t - Reserved for enabling touchscreen support. The 4090 touchscreen support is not enabled in the mainstream 4091 kernel as of 2.6.30, a preliminary port can be found 4092 in the "bleeding edge" mini2440 support kernel at 4093 https://repo.or.cz/w/linux-2.6/mini2440.git 4094 4095 mitigations= 4096 [X86,PPC,S390,ARM64,EARLY] Control optional mitigations for 4097 CPU vulnerabilities. This is a set of curated, 4098 arch-independent options, each of which is an 4099 aggregation of existing arch-specific options. 4100 4101 Note, "mitigations" is supported if and only if the 4102 kernel was built with CPU_MITIGATIONS=y. 4103 4104 off 4105 Disable all optional CPU mitigations. This 4106 improves system performance, but it may also 4107 expose users to several CPU vulnerabilities. 4108 Equivalent to: if nokaslr then kpti=0 [ARM64] 4109 gather_data_sampling=off [X86] 4110 indirect_target_selection=off [X86] 4111 kvm.nx_huge_pages=off [X86] 4112 l1tf=off [X86] 4113 mds=off [X86] 4114 mmio_stale_data=off [X86] 4115 no_entry_flush [PPC] 4116 no_uaccess_flush [PPC] 4117 nobp=0 [S390] 4118 nopti [X86,PPC] 4119 nospectre_bhb [ARM64] 4120 nospectre_v1 [X86,PPC] 4121 nospectre_v2 [X86,PPC,S390,ARM64] 4122 reg_file_data_sampling=off [X86] 4123 retbleed=off [X86] 4124 spec_rstack_overflow=off [X86] 4125 spec_store_bypass_disable=off [X86,PPC] 4126 spectre_bhi=off [X86] 4127 spectre_v2_user=off [X86] 4128 srbds=off [X86,INTEL] 4129 ssbd=force-off [ARM64] 4130 tsa=off [X86,AMD] 4131 tsx_async_abort=off [X86] 4132 vmscape=off [X86] 4133 4134 Exceptions: 4135 This does not have any effect on 4136 kvm.nx_huge_pages when 4137 kvm.nx_huge_pages=force. 4138 4139 auto (default) 4140 Mitigate all CPU vulnerabilities, but leave SMT 4141 enabled, even if it's vulnerable. This is for 4142 users who don't want to be surprised by SMT 4143 getting disabled across kernel upgrades, or who 4144 have other ways of avoiding SMT-based attacks. 4145 Equivalent to: (default behavior) 4146 4147 auto,nosmt 4148 Mitigate all CPU vulnerabilities, disabling SMT 4149 if needed. This is for users who always want to 4150 be fully mitigated, even if it means losing SMT. 4151 Equivalent to: l1tf=flush,nosmt [X86] 4152 mds=full,nosmt [X86] 4153 tsx_async_abort=full,nosmt [X86] 4154 mmio_stale_data=full,nosmt [X86] 4155 retbleed=auto,nosmt [X86] 4156 4157 [X86] After one of the above options, additionally 4158 supports attack-vector based controls as documented in 4159 Documentation/admin-guide/hw-vuln/attack_vector_controls.rst 4160 4161 mminit_loglevel= 4162 [KNL,EARLY] When CONFIG_DEBUG_MEMORY_INIT is set, this 4163 parameter allows control of the logging verbosity for 4164 the additional memory initialisation checks. A value 4165 of 0 disables mminit logging and a level of 4 will 4166 log everything. Information is printed at KERN_DEBUG 4167 so loglevel=8 may also need to be specified. 4168 4169 mmio_stale_data= 4170 [X86,INTEL,EARLY] Control mitigation for the Processor 4171 MMIO Stale Data vulnerabilities. 4172 4173 Processor MMIO Stale Data is a class of 4174 vulnerabilities that may expose data after an MMIO 4175 operation. Exposed data could originate or end in 4176 the same CPU buffers as affected by MDS and TAA. 4177 Therefore, similar to MDS and TAA, the mitigation 4178 is to clear the affected CPU buffers. 4179 4180 This parameter controls the mitigation. The 4181 options are: 4182 4183 full - Enable mitigation on vulnerable CPUs 4184 4185 full,nosmt - Enable mitigation and disable SMT on 4186 vulnerable CPUs. 4187 4188 off - Unconditionally disable mitigation 4189 4190 On MDS or TAA affected machines, 4191 mmio_stale_data=off can be prevented by an active 4192 MDS or TAA mitigation as these vulnerabilities are 4193 mitigated with the same mechanism so in order to 4194 disable this mitigation, you need to specify 4195 mds=off and tsx_async_abort=off too. 4196 4197 Not specifying this option is equivalent to 4198 mmio_stale_data=full. 4199 4200 For details see: 4201 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst 4202 4203 <module>.async_probe[=<bool>] [KNL] 4204 If no <bool> value is specified or if the value 4205 specified is not a valid <bool>, enable asynchronous 4206 probe on this module. Otherwise, enable/disable 4207 asynchronous probe on this module as indicated by the 4208 <bool> value. See also: module.async_probe 4209 4210 module.async_probe=<bool> 4211 [KNL] When set to true, modules will use async probing 4212 by default. To enable/disable async probing for a 4213 specific module, use the module specific control that 4214 is documented under <module>.async_probe. When both 4215 module.async_probe and <module>.async_probe are 4216 specified, <module>.async_probe takes precedence for 4217 the specific module. 4218 4219 module.enable_dups_trace 4220 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set, 4221 this means that duplicate request_module() calls will 4222 trigger a WARN_ON() instead of a pr_warn(). Note that 4223 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s 4224 will always be issued and this option does nothing. 4225 module.sig_enforce 4226 [KNL] When CONFIG_MODULE_SIG is set, this means that 4227 modules without (valid) signatures will fail to load. 4228 Note that if CONFIG_MODULE_SIG_FORCE is set, that 4229 is always true, so this option does nothing. 4230 4231 module_blacklist= [KNL] Do not load a comma-separated list of 4232 modules. Useful for debugging problem modules. 4233 4234 mousedev.tap_time= 4235 [MOUSE] Maximum time between finger touching and 4236 leaving touchpad surface for touch to be considered 4237 a tap and be reported as a left button click (for 4238 touchpads working in absolute mode only). 4239 Format: <msecs> 4240 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 4241 reporting absolute coordinates, such as tablets 4242 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 4243 reporting absolute coordinates, such as tablets 4244 4245 movablecore= [KNL,X86,PPC,EARLY] 4246 Format: nn[KMGTPE] | nn% 4247 This parameter is the complement to kernelcore=, it 4248 specifies the amount of memory used for migratable 4249 allocations. If both kernelcore and movablecore is 4250 specified, then kernelcore will be at *least* the 4251 specified value but may be more. If movablecore on its 4252 own is specified, the administrator must be careful 4253 that the amount of memory usable for all allocations 4254 is not too small. 4255 4256 movable_node [KNL,EARLY] Boot-time switch to make hotplugable memory 4257 NUMA nodes to be movable. This means that the memory 4258 of such nodes will be usable only for movable 4259 allocations which rules out almost all kernel 4260 allocations. Use with caution! 4261 4262 MTD_Partition= [MTD] 4263 Format: <name>,<region-number>,<size>,<offset> 4264 4265 MTD_Region= [MTD] Format: 4266 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 4267 4268 mtdparts= [MTD] 4269 See drivers/mtd/parsers/cmdlinepart.c 4270 4271 mtouchusb.raw_coordinates= 4272 [HW] Make the MicroTouch USB driver use raw coordinates 4273 ('y', default) or cooked coordinates ('n') 4274 4275 mtrr=debug [X86,EARLY] 4276 Enable printing debug information related to MTRR 4277 registers at boot time. 4278 4279 mtrr_chunk_size=nn[KMG,X86,EARLY] 4280 used for mtrr cleanup. It is largest continuous chunk 4281 that could hold holes aka. UC entries. 4282 4283 mtrr_gran_size=nn[KMG,X86,EARLY] 4284 Used for mtrr cleanup. It is granularity of mtrr block. 4285 Default is 1. 4286 Large value could prevent small alignment from 4287 using up MTRRs. 4288 4289 mtrr_spare_reg_nr=n [X86,EARLY] 4290 Format: <integer> 4291 Range: 0,7 : spare reg number 4292 Default : 1 4293 Used for mtrr cleanup. It is spare mtrr entries number. 4294 Set to 2 or more if your graphical card needs more. 4295 4296 multitce=off [PPC] This parameter disables the use of the pSeries 4297 firmware feature for updating multiple TCE entries 4298 at a time. 4299 4300 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 4301 4302 netdev= [NET] Network devices parameters 4303 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 4304 Note that mem_start is often overloaded to mean 4305 something different and driver-specific. 4306 This usage is only documented in each driver source 4307 file if at all. 4308 4309 netpoll.carrier_timeout= 4310 [NET] Specifies amount of time (in seconds) that 4311 netpoll should wait for a carrier. By default netpoll 4312 waits 4 seconds. 4313 4314 nf_conntrack.acct= 4315 [NETFILTER] Enable connection tracking flow accounting 4316 0 to disable accounting 4317 1 to enable accounting 4318 Default value is 0. 4319 4320 nfs.cache_getent= 4321 [NFS] sets the pathname to the program which is used 4322 to update the NFS client cache entries. 4323 4324 nfs.cache_getent_timeout= 4325 [NFS] sets the timeout after which an attempt to 4326 update a cache entry is deemed to have failed. 4327 4328 nfs.callback_nr_threads= 4329 [NFSv4] set the total number of threads that the 4330 NFS client will assign to service NFSv4 callback 4331 requests. 4332 4333 nfs.callback_tcpport= 4334 [NFS] set the TCP port on which the NFSv4 callback 4335 channel should listen. 4336 4337 nfs.delay_retrans= 4338 [NFS] specifies the number of times the NFSv4 client 4339 retries the request before returning an EAGAIN error, 4340 after a reply of NFS4ERR_DELAY from the server. 4341 Only applies if the softerr mount option is enabled, 4342 and the specified value is >= 0. 4343 4344 nfs.enable_ino64= 4345 [NFS] enable 64-bit inode numbers. 4346 If zero, the NFS client will fake up a 32-bit inode 4347 number for the readdir() and stat() syscalls instead 4348 of returning the full 64-bit number. 4349 The default is to return 64-bit inode numbers. 4350 4351 nfs.idmap_cache_timeout= 4352 [NFS] set the maximum lifetime for idmapper cache 4353 entries. 4354 4355 nfs.max_session_cb_slots= 4356 [NFSv4.1] Sets the maximum number of session 4357 slots the client will assign to the callback 4358 channel. This determines the maximum number of 4359 callbacks the client will process in parallel for 4360 a particular server. 4361 4362 nfs.max_session_slots= 4363 [NFSv4.1] Sets the maximum number of session slots 4364 the client will attempt to negotiate with the server. 4365 This limits the number of simultaneous RPC requests 4366 that the client can send to the NFSv4.1 server. 4367 Note that there is little point in setting this 4368 value higher than the max_tcp_slot_table_limit. 4369 4370 nfs.nfs4_disable_idmapping= 4371 [NFSv4] When set to the default of '1', this option 4372 ensures that both the RPC level authentication 4373 scheme and the NFS level operations agree to use 4374 numeric uids/gids if the mount is using the 4375 'sec=sys' security flavour. In effect it is 4376 disabling idmapping, which can make migration from 4377 legacy NFSv2/v3 systems to NFSv4 easier. 4378 Servers that do not support this mode of operation 4379 will be autodetected by the client, and it will fall 4380 back to using the idmapper. 4381 To turn off this behaviour, set the value to '0'. 4382 4383 nfs.nfs4_unique_id= 4384 [NFS4] Specify an additional fixed unique ident- 4385 ification string that NFSv4 clients can insert into 4386 their nfs_client_id4 string. This is typically a 4387 UUID that is generated at system install time. 4388 4389 nfs.recover_lost_locks= 4390 [NFSv4] Attempt to recover locks that were lost due 4391 to a lease timeout on the server. Please note that 4392 doing this risks data corruption, since there are 4393 no guarantees that the file will remain unchanged 4394 after the locks are lost. 4395 If you want to enable the kernel legacy behaviour of 4396 attempting to recover these locks, then set this 4397 parameter to '1'. 4398 The default parameter value of '0' causes the kernel 4399 not to attempt recovery of lost locks. 4400 4401 nfs.send_implementation_id= 4402 [NFSv4.1] Send client implementation identification 4403 information in exchange_id requests. 4404 If zero, no implementation identification information 4405 will be sent. 4406 The default is to send the implementation identification 4407 information. 4408 4409 nfs4.layoutstats_timer= 4410 [NFSv4.2] Change the rate at which the kernel sends 4411 layoutstats to the pNFS metadata server. 4412 4413 Setting this to value to 0 causes the kernel to use 4414 whatever value is the default set by the layout 4415 driver. A non-zero value sets the minimum interval 4416 in seconds between layoutstats transmissions. 4417 4418 nfsd.inter_copy_offload_enable= 4419 [NFSv4.2] When set to 1, the server will support 4420 server-to-server copies for which this server is 4421 the destination of the copy. 4422 4423 nfsd.nfs4_disable_idmapping= 4424 [NFSv4] When set to the default of '1', the NFSv4 4425 server will return only numeric uids and gids to 4426 clients using auth_sys, and will accept numeric uids 4427 and gids from such clients. This is intended to ease 4428 migration from NFSv2/v3. 4429 4430 nfsd.nfsd4_ssc_umount_timeout= 4431 [NFSv4.2] When used as the destination of a 4432 server-to-server copy, knfsd temporarily mounts 4433 the source server. It caches the mount in case 4434 it will be needed again, and discards it if not 4435 used for the number of milliseconds specified by 4436 this parameter. 4437 4438 nfsaddrs= [NFS] Deprecated. Use ip= instead. 4439 See Documentation/admin-guide/nfs/nfsroot.rst. 4440 4441 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 4442 See Documentation/admin-guide/nfs/nfsroot.rst. 4443 4444 nfsrootdebug [NFS] enable nfsroot debugging messages. 4445 See Documentation/admin-guide/nfs/nfsroot.rst. 4446 4447 nmi_backtrace.backtrace_idle [KNL] 4448 Dump stacks even of idle CPUs in response to an 4449 NMI stack-backtrace request. 4450 4451 nmi_debug= [KNL,SH] Specify one or more actions to take 4452 when a NMI is triggered. 4453 Format: [state][,regs][,debounce][,die] 4454 4455 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 4456 Format: [panic,][nopanic,][rNNN,][num] 4457 Valid num: 0 or 1 4458 0 - turn hardlockup detector in nmi_watchdog off 4459 1 - turn hardlockup detector in nmi_watchdog on 4460 rNNN - configure the watchdog with raw perf event 0xNNN 4461 4462 When panic is specified, panic when an NMI watchdog 4463 timeout occurs (or 'nopanic' to not panic on an NMI 4464 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 4465 To disable both hard and soft lockup detectors, 4466 please see 'nowatchdog'. 4467 This is useful when you use a panic=... timeout and 4468 need the box quickly up again. 4469 4470 These settings can be accessed at runtime via 4471 the nmi_watchdog and hardlockup_panic sysctls. 4472 4473 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 4474 emulation library even if a 387 maths coprocessor 4475 is present. 4476 4477 no4lvl [RISCV,EARLY] Disable 4-level and 5-level paging modes. 4478 Forces kernel to use 3-level paging instead. 4479 4480 no5lvl [X86-64,RISCV,EARLY] Disable 5-level paging mode. Forces 4481 kernel to use 4-level paging instead. 4482 4483 noalign [KNL,ARM] 4484 4485 noapic [SMP,APIC,EARLY] Tells the kernel to not make use of any 4486 IOAPICs that may be present in the system. 4487 4488 noapictimer [APIC,X86] Don't set up the APIC timer 4489 4490 noautogroup Disable scheduler automatic task group creation. 4491 4492 nocache [ARM,EARLY] 4493 4494 no_console_suspend 4495 [HW] Never suspend the console 4496 Disable suspending of consoles during suspend and 4497 hibernate operations. Once disabled, debugging 4498 messages can reach various consoles while the rest 4499 of the system is being put to sleep (ie, while 4500 debugging driver suspend/resume hooks). This may 4501 not work reliably with all consoles, but is known 4502 to work with serial and VGA consoles. 4503 To facilitate more flexible debugging, we also add 4504 console_suspend, a printk module parameter to control 4505 it. Users could use console_suspend (usually 4506 /sys/module/printk/parameters/console_suspend) to 4507 turn on/off it dynamically. 4508 4509 no_debug_objects 4510 [KNL,EARLY] Disable object debugging 4511 4512 nodsp [SH] Disable hardware DSP at boot time. 4513 4514 noefi [EFI,EARLY] Disable EFI runtime services support. 4515 4516 no_entry_flush [PPC,EARLY] Don't flush the L1-D cache when entering the kernel. 4517 4518 noexec32 [X86-64] 4519 This affects only 32-bit executables. 4520 noexec32=on: enable non-executable mappings (default) 4521 read doesn't imply executable mappings 4522 noexec32=off: disable non-executable mappings 4523 read implies executable mappings 4524 4525 no_file_caps Tells the kernel not to honor file capabilities. The 4526 only way then for a file to be executed with privilege 4527 is to be setuid root or executed by root. 4528 4529 nofpu [MIPS,SH] Disable hardware FPU at boot time. 4530 4531 nofsgsbase [X86] Disables FSGSBASE instructions. 4532 4533 nofxsr [BUGS=X86-32] Disables x86 floating point extended 4534 register save and restore. The kernel will only save 4535 legacy floating-point registers on task switch. 4536 4537 nogbpages [X86] Do not use GB pages for kernel direct mappings. 4538 4539 no_hash_pointers 4540 [KNL,EARLY] 4541 Alias for "hash_pointers=never". 4542 4543 nohibernate [HIBERNATION] Disable hibernation and resume. 4544 4545 nohlt [ARM,ARM64,MICROBLAZE,MIPS,PPC,RISCV,SH] Forces the kernel to 4546 busy wait in do_idle() and not use the arch_cpu_idle() 4547 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP 4548 to be effective. This is useful on platforms where the 4549 sleep(SH) or wfi(ARM,ARM64) instructions do not work 4550 correctly or when doing power measurements to evaluate 4551 the impact of the sleep instructions. This is also 4552 useful when using JTAG debugger. 4553 4554 nohpet [X86] Don't use the HPET timer. 4555 4556 nohugeiomap [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge I/O mappings. 4557 4558 nohugevmalloc [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge vmalloc mappings. 4559 4560 nohz= [KNL] Boottime enable/disable dynamic ticks 4561 Valid arguments: on, off 4562 Default: on 4563 4564 nohz_full= [KNL,BOOT,SMP,ISOL] 4565 The argument is a cpu list, as described above. 4566 In kernels built with CONFIG_NO_HZ_FULL=y, set 4567 the specified list of CPUs whose tick will be stopped 4568 whenever possible. The boot CPU will be forced outside 4569 the range to maintain the timekeeping. Any CPUs 4570 in this list will have their RCU callbacks offloaded, 4571 just as if they had also been called out in the 4572 rcu_nocbs= boot parameter. 4573 4574 Note that this argument takes precedence over 4575 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 4576 4577 noinitrd [Deprecated,RAM] Tells the kernel not to load any configured 4578 initial RAM disk. Currently this parameter applies to 4579 initrd only, not to initramfs. But it applies to both 4580 in EFI mode. 4581 4582 nointremap [X86-64,Intel-IOMMU,EARLY] Do not enable interrupt 4583 remapping. 4584 [Deprecated - use intremap=off] 4585 4586 noinvpcid [X86,EARLY] Disable the INVPCID cpu feature. 4587 4588 noiotrap [SH] Disables trapped I/O port accesses. 4589 4590 noirqdebug [X86-32] Disables the code which attempts to detect and 4591 disable unhandled interrupt sources. 4592 4593 noisapnp [ISAPNP] Disables ISA PnP code. 4594 4595 nokaslr [KNL,EARLY] 4596 When CONFIG_RANDOMIZE_BASE is set, this disables 4597 kernel and module base offset ASLR (Address Space 4598 Layout Randomization). 4599 4600 no-kvmapf [X86,KVM,EARLY] Disable paravirtualized asynchronous page 4601 fault handling. 4602 4603 no-kvmclock [X86,KVM,EARLY] Disable paravirtualized KVM clock driver 4604 4605 nolapic [X86-32,APIC,EARLY] Do not enable or use the local APIC. 4606 4607 nolapic_timer [X86-32,APIC,EARLY] Do not use the local APIC timer. 4608 4609 nomce [X86-32] Disable Machine Check Exception 4610 4611 nomfgpt [X86-32] Disable Multi-Function General Purpose 4612 Timer usage (for AMD Geode machines). 4613 4614 nomodeset Disable kernel modesetting. Most systems' firmware 4615 sets up a display mode and provides framebuffer memory 4616 for output. With nomodeset, DRM and fbdev drivers will 4617 not load if they could possibly displace the pre- 4618 initialized output. Only the system framebuffer will 4619 be available for use. The respective drivers will not 4620 perform display-mode changes or accelerated rendering. 4621 4622 Useful as error fallback, or for testing and debugging. 4623 4624 nomodule Disable module load 4625 4626 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 4627 shutdown the other cpus. Instead use the REBOOT_VECTOR 4628 irq. 4629 4630 nopat [X86,EARLY] Disable PAT (page attribute table extension of 4631 pagetables) support. 4632 4633 nopcid [X86-64,EARLY] Disable the PCID cpu feature. 4634 4635 nopku [X86] Disable Memory Protection Keys CPU feature found 4636 in some Intel CPUs. 4637 4638 nopti [X86-64,EARLY] 4639 Equivalent to pti=off 4640 4641 nopv= [X86,XEN,KVM,HYPER_V,VMWARE,EARLY] 4642 Disables the PV optimizations forcing the guest to run 4643 as generic guest with no PV drivers. Currently support 4644 XEN HVM, KVM, HYPER_V and VMWARE guest. 4645 4646 nopvspin [X86,XEN,KVM,EARLY] 4647 Disables the qspinlock slow path using PV optimizations 4648 which allow the hypervisor to 'idle' the guest on lock 4649 contention. 4650 4651 norandmaps Don't use address space randomization. Equivalent to 4652 echo 0 > /proc/sys/kernel/randomize_va_space 4653 4654 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 4655 with UP alternatives 4656 4657 noresume [SWSUSP] Disables resume and restores original swap 4658 space. 4659 4660 no-scroll [VGA] Disables scrollback. 4661 This is required for the Braillex ib80-piezo Braille 4662 reader made by F.H. Papenmeier (Germany). 4663 4664 nosgx [X86-64,SGX,EARLY] Disables Intel SGX kernel support. 4665 4666 nosmap [PPC,EARLY] 4667 Disable SMAP (Supervisor Mode Access Prevention) 4668 even if it is supported by processor. 4669 4670 nosmep [PPC64s,EARLY] 4671 Disable SMEP (Supervisor Mode Execution Prevention) 4672 even if it is supported by processor. 4673 4674 nosmp [SMP,EARLY] Tells an SMP kernel to act as a UP kernel, 4675 and disable the IO APIC. legacy for "maxcpus=0". 4676 4677 nosmt [KNL,MIPS,PPC,EARLY] Disable symmetric multithreading (SMT). 4678 Equivalent to smt=1. 4679 4680 [KNL,LOONGARCH,X86,PPC,S390] Disable symmetric multithreading (SMT). 4681 nosmt=force: Force disable SMT, cannot be undone 4682 via the sysfs control file. 4683 4684 nosoftlockup [KNL] Disable the soft-lockup detector. 4685 4686 nospec_store_bypass_disable 4687 [HW,EARLY] Disable all mitigations for the Speculative 4688 Store Bypass vulnerability 4689 4690 nospectre_bhb [ARM64,EARLY] Disable all mitigations for Spectre-BHB (branch 4691 history injection) vulnerability. System may allow data leaks 4692 with this option. 4693 4694 nospectre_v1 [X86,PPC,EARLY] Disable mitigations for Spectre Variant 1 4695 (bounds check bypass). With this option data leaks are 4696 possible in the system. 4697 4698 nospectre_v2 [X86,PPC_E500,ARM64,EARLY] Disable all mitigations 4699 for the Spectre variant 2 (indirect branch 4700 prediction) vulnerability. System may allow data 4701 leaks with this option. 4702 4703 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES,RISCV,LOONGARCH,EARLY] 4704 Disable paravirtualized steal time accounting. steal time 4705 is computed, but won't influence scheduler behaviour 4706 4707 nosync [HW,M68K] Disables sync negotiation for all devices. 4708 4709 no_timer_check [X86,APIC] Disables the code which tests for broken 4710 timer IRQ sources, i.e., the IO-APIC timer. This can 4711 work around problems with incorrect timer 4712 initialization on some boards. 4713 4714 no_uaccess_flush 4715 [PPC,EARLY] Don't flush the L1-D cache after accessing user data. 4716 4717 novmcoredd [KNL,KDUMP] 4718 Disable device dump. Device dump allows drivers to 4719 append dump data to vmcore so you can collect driver 4720 specified debug info. Drivers can append the data 4721 without any limit and this data is stored in memory, 4722 so this may cause significant memory stress. Disabling 4723 device dump can help save memory but the driver debug 4724 data will be no longer available. This parameter 4725 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 4726 is set. 4727 4728 no-vmw-sched-clock 4729 [X86,PV_OPS,EARLY] Disable paravirtualized VMware 4730 scheduler clock and use the default one. 4731 4732 nowatchdog [KNL] Disable both lockup detectors, i.e. 4733 soft-lockup and NMI watchdog (hard-lockup). 4734 4735 nowb [ARM,EARLY] 4736 4737 nox2apic [X86-64,APIC,EARLY] Do not enable x2APIC mode. 4738 4739 NOTE: this parameter will be ignored on systems with the 4740 LEGACY_XAPIC_DISABLED bit set in the 4741 IA32_XAPIC_DISABLE_STATUS MSR. 4742 4743 noxsave [BUGS=X86] Disables x86 extended register state save 4744 and restore using xsave. The kernel will fallback to 4745 enabling legacy floating-point and sse state. 4746 4747 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 4748 register states. The kernel will fall back to use 4749 xsave to save the states. By using this parameter, 4750 performance of saving the states is degraded because 4751 xsave doesn't support modified optimization while 4752 xsaveopt supports it on xsaveopt enabled systems. 4753 4754 noxsaves [X86] Disables xsaves and xrstors used in saving and 4755 restoring x86 extended register state in compacted 4756 form of xsave area. The kernel will fall back to use 4757 xsaveopt and xrstor to save and restore the states 4758 in standard form of xsave area. By using this 4759 parameter, xsave area per process might occupy more 4760 memory on xsaves enabled systems. 4761 4762 nr_cpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 4763 could support. nr_cpus=n : n >= 1 limits the kernel to 4764 support 'n' processors. It could be larger than the 4765 number of already plugged CPU during bootup, later in 4766 runtime you can physically add extra cpu until it reaches 4767 n. So during boot up some boot time memory for per-cpu 4768 variables need be pre-allocated for later physical cpu 4769 hot plugging. 4770 4771 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 4772 4773 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86, EARLY] 4774 Disable NUMA, Only set up a single NUMA node 4775 spanning all memory. 4776 4777 numa=fake=<size>[MG] 4778 [KNL, ARM64, RISCV, X86, EARLY] 4779 If given as a memory unit, fills all system RAM with 4780 nodes of size interleaved over physical nodes. 4781 4782 numa=fake=<N> 4783 [KNL, ARM64, RISCV, X86, EARLY] 4784 If given as an integer, fills all system RAM with N 4785 fake nodes interleaved over physical nodes. 4786 4787 numa=fake=<N>U 4788 [KNL, ARM64, RISCV, X86, EARLY] 4789 If given as an integer followed by 'U', it will 4790 divide each physical node into N emulated nodes. 4791 4792 numa=noacpi [X86] Don't parse the SRAT table for NUMA setup 4793 4794 numa=nohmat [X86] Don't parse the HMAT table for NUMA setup, or 4795 soft-reserved memory partitioning. 4796 4797 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic 4798 NUMA balancing. 4799 Allowed values are enable and disable 4800 4801 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 4802 'node', 'default' can be specified 4803 This can be set from sysctl after boot. 4804 See Documentation/admin-guide/sysctl/vm.rst for details. 4805 4806 nvme.quirks= [NVME] A list of quirk entries to augment the built-in 4807 nvme quirk list. List entries are separated by a 4808 '-' character. 4809 Each entry has the form VendorID:ProductID:quirk_names. 4810 The IDs are 4-digits hex numbers and quirk_names is a 4811 list of quirk names separated by commas. A quirk name 4812 can be prefixed by '^', meaning that the specified 4813 quirk must be disabled. 4814 4815 Example: 4816 nvme.quirks=7710:2267:bogus_nid,^identify_cns-9900:7711:broken_msi 4817 4818 ohci1394_dma=early [HW,EARLY] enable debugging via the ohci1394 driver. 4819 See Documentation/core-api/debugging-via-ohci1394.rst for more 4820 info. 4821 4822 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 4823 Rather than timing out after 20 ms if an EC 4824 command is not properly ACKed, override the length 4825 of the timeout. We have interrupts disabled while 4826 waiting for the ACK, so if this is set too high 4827 interrupts *may* be lost! 4828 4829 omap_mux= [OMAP] Override bootloader pin multiplexing. 4830 Format: <mux_mode0.mode_name=value>... 4831 For example, to override I2C bus2: 4832 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 4833 4834 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 4835 4836 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 4837 4838 boundary - index of last SLC block on Flex-OneNAND. 4839 The remaining blocks are configured as MLC blocks. 4840 lock - Configure if Flex-OneNAND boundary should be locked. 4841 Once locked, the boundary cannot be changed. 4842 1 indicates lock status, 0 indicates unlock status. 4843 4844 oops=panic [KNL,EARLY] 4845 Always panic on oopses. Default is to just kill the 4846 process, but there is a small probability of 4847 deadlocking the machine. 4848 This will also cause panics on machine check exceptions. 4849 Useful together with panic=30 to trigger a reboot. 4850 4851 page_alloc.shuffle= 4852 [KNL] Boolean flag to control whether the page allocator 4853 should randomize its free lists. This parameter can be 4854 used to enable/disable page randomization. The state of 4855 the flag can be read from sysfs at: 4856 /sys/module/page_alloc/parameters/shuffle. 4857 This parameter is only available if CONFIG_SHUFFLE_PAGE_ALLOCATOR=y. 4858 4859 page_owner= [KNL,EARLY] Boot-time page_owner enabling option. 4860 Storage of the information about who allocated 4861 each page is disabled in default. With this switch, 4862 we can turn it on. 4863 on: enable the feature 4864 4865 page_poison= [KNL,EARLY] Boot-time parameter changing the state of 4866 poisoning on the buddy allocator, available with 4867 CONFIG_PAGE_POISONING=y. 4868 off: turn off poisoning (default) 4869 on: turn on poisoning 4870 4871 page_reporting.page_reporting_order= 4872 [KNL] Minimal page reporting order 4873 Format: <integer> 4874 Adjust the minimal page reporting order. The page 4875 reporting is disabled when it exceeds MAX_PAGE_ORDER. 4876 4877 panic= [KNL] Kernel behaviour on panic: delay <timeout> 4878 timeout > 0: seconds before rebooting 4879 timeout = 0: wait forever 4880 timeout < 0: reboot immediately 4881 Format: <timeout> 4882 4883 panic_on_taint= [KNL,EARLY] 4884 Bitmask for conditionally calling panic() in add_taint() 4885 Format: <hex>[,nousertaint] 4886 Hexadecimal bitmask representing the set of TAINT flags 4887 that will cause the kernel to panic when add_taint() is 4888 called with any of the flags in this set. 4889 The optional switch "nousertaint" can be utilized to 4890 prevent userspace forced crashes by writing to sysctl 4891 /proc/sys/kernel/tainted any flagset matching with the 4892 bitmask set on panic_on_taint. 4893 See Documentation/admin-guide/tainted-kernels.rst for 4894 extra details on the taint flags that users can pick 4895 to compose the bitmask to assign to panic_on_taint. 4896 4897 panic_on_warn=1 panic() instead of WARN(). Useful to cause kdump 4898 on a WARN(). 4899 4900 panic_force_cpu= 4901 [KNL,SMP] Force panic handling to execute on a specific CPU. 4902 Format: <cpu number> 4903 Some platforms require panic handling to occur on a 4904 specific CPU for the crash kernel to function correctly. 4905 This can be due to firmware limitations, interrupt routing 4906 constraints, or platform-specific requirements where only 4907 a particular CPU can safely enter the crash kernel. 4908 When set, panic() will redirect execution to the specified 4909 CPU before proceeding with the normal panic and kexec flow. 4910 If the target CPU is offline or unavailable, panic proceeds 4911 on the current CPU. 4912 This option should only be used for systems with the above 4913 constraints as it might cause the panic operation to be less reliable. 4914 4915 panic_print= Bitmask for printing system info when panic happens. 4916 User can chose combination of the following bits: 4917 bit 0: print all tasks info 4918 bit 1: print system memory info 4919 bit 2: print timer info 4920 bit 3: print locks info if CONFIG_LOCKDEP is on 4921 bit 4: print ftrace buffer 4922 bit 5: replay all kernel messages on consoles at the end of panic 4923 bit 6: print all CPUs backtrace (if available in the arch) 4924 bit 7: print only tasks in uninterruptible (blocked) state 4925 *Be aware* that this option may print a _lot_ of lines, 4926 so there are risks of losing older messages in the log. 4927 Use this option carefully, maybe worth to setup a 4928 bigger log buffer with "log_buf_len" along with this. 4929 4930 panic_sys_info= A comma separated list of extra information to be dumped 4931 on panic. 4932 Format: val[,val...] 4933 Where @val can be any of the following: 4934 4935 tasks: print all tasks info 4936 mem: print system memory info 4937 timers: print timers info 4938 locks: print locks info if CONFIG_LOCKDEP is on 4939 ftrace: print ftrace buffer 4940 all_bt: print all CPUs backtrace (if available in the arch) 4941 blocked_tasks: print only tasks in uninterruptible (blocked) state 4942 4943 This is a human readable alternative to the 'panic_print' option. 4944 4945 panic_console_replay 4946 When panic happens, replay all kernel messages on 4947 consoles at the end of panic. 4948 4949 parkbd.port= [HW] Parallel port number the keyboard adapter is 4950 connected to, default is 0. 4951 Format: <parport#> 4952 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 4953 0 for XT, 1 for AT (default is AT). 4954 Format: <mode> 4955 4956 parport= [HW,PPT] Specify parallel ports. 0 disables. 4957 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 4958 Use 'auto' to force the driver to use any 4959 IRQ/DMA settings detected (the default is to 4960 ignore detected IRQ/DMA settings because of 4961 possible conflicts). You can specify the base 4962 address, IRQ, and DMA settings; IRQ and DMA 4963 should be numbers, or 'auto' (for using detected 4964 settings on that particular port), or 'nofifo' 4965 (to avoid using a FIFO even if it is detected). 4966 Parallel ports are assigned in the order they 4967 are specified on the command line, starting 4968 with parport0. 4969 4970 parport_init_mode= [HW,PPT] 4971 Configure VIA parallel port to operate in 4972 a specific mode. This is necessary on Pegasos 4973 computer where firmware has no options for setting 4974 up parallel port mode and sets it to spp. 4975 Currently this function knows 686a and 8231 chips. 4976 Format: [spp|ps2|epp|ecp|ecpepp] 4977 4978 pata_legacy.all= [HW,LIBATA] 4979 Format: <int> 4980 Set to non-zero to probe primary and secondary ISA 4981 port ranges on PCI systems where no PCI PATA device 4982 has been found at either range. Disabled by default. 4983 4984 pata_legacy.autospeed= [HW,LIBATA] 4985 Format: <int> 4986 Set to non-zero if a chip is present that snoops speed 4987 changes. Disabled by default. 4988 4989 pata_legacy.ht6560a= [HW,LIBATA] 4990 Format: <int> 4991 Set to 1, 2, or 3 for HT 6560A on the primary channel, 4992 the secondary channel, or both channels respectively. 4993 Disabled by default. 4994 4995 pata_legacy.ht6560b= [HW,LIBATA] 4996 Format: <int> 4997 Set to 1, 2, or 3 for HT 6560B on the primary channel, 4998 the secondary channel, or both channels respectively. 4999 Disabled by default. 5000 5001 pata_legacy.iordy_mask= [HW,LIBATA] 5002 Format: <int> 5003 IORDY enable mask. Set individual bits to allow IORDY 5004 for the respective channel. Bit 0 is for the first 5005 legacy channel handled by this driver, bit 1 is for 5006 the second channel, and so on. The sequence will often 5007 correspond to the primary legacy channel, the secondary 5008 legacy channel, and so on, but the handling of a PCI 5009 bus and the use of other driver options may interfere 5010 with the sequence. By default IORDY is allowed across 5011 all channels. 5012 5013 pata_legacy.opti82c46x= [HW,LIBATA] 5014 Format: <int> 5015 Set to 1, 2, or 3 for Opti 82c611A on the primary 5016 channel, the secondary channel, or both channels 5017 respectively. Disabled by default. 5018 5019 pata_legacy.opti82c611a= [HW,LIBATA] 5020 Format: <int> 5021 Set to 1, 2, or 3 for Opti 82c465MV on the primary 5022 channel, the secondary channel, or both channels 5023 respectively. Disabled by default. 5024 5025 pata_legacy.pio_mask= [HW,LIBATA] 5026 Format: <int> 5027 PIO mode mask for autospeed devices. Set individual 5028 bits to allow the use of the respective PIO modes. 5029 Bit 0 is for mode 0, bit 1 is for mode 1, and so on. 5030 All modes allowed by default. 5031 5032 pata_legacy.probe_all= [HW,LIBATA] 5033 Format: <int> 5034 Set to non-zero to probe tertiary and further ISA 5035 port ranges on PCI systems. Disabled by default. 5036 5037 pata_legacy.probe_mask= [HW,LIBATA] 5038 Format: <int> 5039 Probe mask for legacy ISA PATA ports. Depending on 5040 platform configuration and the use of other driver 5041 options up to 6 legacy ports are supported: 0x1f0, 5042 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing 5043 of individual ports can be disabled by setting the 5044 corresponding bits in the mask to 1. Bit 0 is for 5045 the first port in the list above (0x1f0), and so on. 5046 By default all supported ports are probed. 5047 5048 pata_legacy.qdi= [HW,LIBATA] 5049 Format: <int> 5050 Set to non-zero to probe QDI controllers. By default 5051 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise. 5052 5053 pata_legacy.winbond= [HW,LIBATA] 5054 Format: <int> 5055 Set to non-zero to probe Winbond controllers. Use 5056 the standard I/O port (0x130) if 1, otherwise the 5057 value given is the I/O port to use (typically 0x1b0). 5058 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE, 5059 0 otherwise. 5060 5061 pata_platform.pio_mask= [HW,LIBATA] 5062 Format: <int> 5063 Supported PIO mode mask. Set individual bits to allow 5064 the use of the respective PIO modes. Bit 0 is for 5065 mode 0, bit 1 is for mode 1, and so on. Mode 0 only 5066 allowed by default. 5067 5068 pause_on_oops=<int> 5069 Halt all CPUs after the first oops has been printed for 5070 the specified number of seconds. This is to be used if 5071 your oopses keep scrolling off the screen. 5072 5073 pcbit= [HW,ISDN] 5074 5075 pci=option[,option...] [PCI,EARLY] various PCI subsystem options. 5076 5077 Some options herein operate on a specific device 5078 or a set of devices (<pci_dev>). These are 5079 specified in one of the following formats: 5080 5081 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 5082 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 5083 5084 Note: the first format specifies a PCI 5085 bus/device/function address which may change 5086 if new hardware is inserted, if motherboard 5087 firmware changes, or due to changes caused 5088 by other kernel parameters. If the 5089 domain is left unspecified, it is 5090 taken to be zero. Optionally, a path 5091 to a device through multiple device/function 5092 addresses can be specified after the base 5093 address (this is more robust against 5094 renumbering issues). The second format 5095 selects devices using IDs from the 5096 configuration space which may match multiple 5097 devices in the system. 5098 5099 earlydump dump PCI config space before the kernel 5100 changes anything 5101 off [X86] don't probe for the PCI bus 5102 bios [X86-32] force use of PCI BIOS, don't access 5103 the hardware directly. Use this if your machine 5104 has a non-standard PCI host bridge. 5105 nobios [X86-32] disallow use of PCI BIOS, only direct 5106 hardware access methods are allowed. Use this 5107 if you experience crashes upon bootup and you 5108 suspect they are caused by the BIOS. 5109 conf1 [X86] Force use of PCI Configuration Access 5110 Mechanism 1 (config address in IO port 0xCF8, 5111 data in IO port 0xCFC, both 32-bit). 5112 conf2 [X86] Force use of PCI Configuration Access 5113 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 5114 the function, IO port 0xCFA, also 8-bit, sets 5115 bus number. The config space is then accessed 5116 through ports 0xC000-0xCFFF). 5117 See http://wiki.osdev.org/PCI for more info 5118 on the configuration access mechanisms. 5119 noaer [PCIE] If the PCIEAER kernel config parameter is 5120 enabled, this kernel boot option can be used to 5121 disable the use of PCIE advanced error reporting. 5122 nodomains [PCI] Disable support for multiple PCI 5123 root domains (aka PCI segments, in ACPI-speak). 5124 nommconf [X86] Disable use of MMCONFIG for PCI 5125 Configuration 5126 check_enable_amd_mmconf [X86] check for and enable 5127 properly configured MMIO access to PCI 5128 config space on AMD family 10h CPU 5129 nomsi [MSI] If the PCI_MSI kernel config parameter is 5130 enabled, this kernel boot option can be used to 5131 disable the use of MSI interrupts system-wide. 5132 noioapicquirk [APIC] Disable all boot interrupt quirks. 5133 Safety option to keep boot IRQs enabled. This 5134 should never be necessary. 5135 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 5136 primary IO-APIC for bridges that cannot disable 5137 boot IRQs. This fixes a source of spurious IRQs 5138 when the system masks IRQs. 5139 noioapicreroute [APIC] Disable workaround that uses the 5140 boot IRQ equivalent of an IRQ that connects to 5141 a chipset where boot IRQs cannot be disabled. 5142 The opposite of ioapicreroute. 5143 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 5144 routing table. These calls are known to be buggy 5145 on several machines and they hang the machine 5146 when used, but on other computers it's the only 5147 way to get the interrupt routing table. Try 5148 this option if the kernel is unable to allocate 5149 IRQs or discover secondary PCI buses on your 5150 motherboard. 5151 rom [X86] Assign address space to expansion ROMs. 5152 Use with caution as certain devices share 5153 address decoders between ROMs and other 5154 resources. 5155 norom [X86] Do not assign address space to 5156 expansion ROMs that do not already have 5157 BIOS assigned address ranges. 5158 nobar [X86] Do not assign address space to the 5159 BARs that weren't assigned by the BIOS. 5160 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 5161 assigned automatically to PCI devices. You can 5162 make the kernel exclude IRQs of your ISA cards 5163 this way. 5164 pirqaddr=0xAAAAA [X86] Specify the physical address 5165 of the PIRQ table (normally generated 5166 by the BIOS) if it is outside the 5167 F0000h-100000h range. 5168 lastbus=N [X86] Scan all buses thru bus #N. Can be 5169 useful if the kernel is unable to find your 5170 secondary buses and you want to tell it 5171 explicitly which ones they are. 5172 assign-busses [X86] Always assign all PCI bus 5173 numbers ourselves, overriding 5174 whatever the firmware may have done. 5175 usepirqmask [X86] Honor the possible IRQ mask stored 5176 in the BIOS $PIR table. This is needed on 5177 some systems with broken BIOSes, notably 5178 some HP Pavilion N5400 and Omnibook XE3 5179 notebooks. This will have no effect if ACPI 5180 IRQ routing is enabled. 5181 noacpi [X86] Do not use ACPI for IRQ routing 5182 or for PCI scanning. 5183 use_crs [X86] Use PCI host bridge window information 5184 from ACPI. On BIOSes from 2008 or later, this 5185 is enabled by default. If you need to use this, 5186 please report a bug. 5187 nocrs [X86] Ignore PCI host bridge windows from ACPI. 5188 If you need to use this, please report a bug. 5189 use_e820 [X86] Use E820 reservations to exclude parts of 5190 PCI host bridge windows. This is a workaround 5191 for BIOS defects in host bridge _CRS methods. 5192 If you need to use this, please report a bug to 5193 <linux-pci@vger.kernel.org>. 5194 no_e820 [X86] Ignore E820 reservations for PCI host 5195 bridge windows. This is the default on modern 5196 hardware. If you need to use this, please report 5197 a bug to <linux-pci@vger.kernel.org>. 5198 routeirq Do IRQ routing for all PCI devices. 5199 This is normally done in pci_enable_device(), 5200 so this option is a temporary workaround 5201 for broken drivers that don't call it. 5202 skip_isa_align [X86] do not align io start addr, so can 5203 handle more pci cards 5204 noearly [X86] Don't do any early type 1 scanning. 5205 This might help on some broken boards which 5206 machine check when some devices' config space 5207 is read. But various workarounds are disabled 5208 and some IOMMU drivers will not work. 5209 bfsort Sort PCI devices into breadth-first order. 5210 This sorting is done to get a device 5211 order compatible with older (<= 2.4) kernels. 5212 nobfsort Don't sort PCI devices into breadth-first order. 5213 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 5214 tuning and use the BIOS-configured MPS defaults. 5215 pcie_bus_safe Set every device's MPS to the largest value 5216 supported by all devices below the root complex. 5217 pcie_bus_perf Set device MPS to the largest allowable MPS 5218 based on its parent bus. Also set MRRS (Max 5219 Read Request Size) to the largest supported 5220 value (no larger than the MPS that the device 5221 or bus can support) for best performance. 5222 pcie_bus_peer2peer Set every device's MPS to 128B, which 5223 every device is guaranteed to support. This 5224 configuration allows peer-to-peer DMA between 5225 any pair of devices, possibly at the cost of 5226 reduced performance. This also guarantees 5227 that hot-added devices will work. 5228 cbiosize=nn[KMG] The fixed amount of bus space which is 5229 reserved for the CardBus bridge's IO window. 5230 The default value is 256 bytes. 5231 cbmemsize=nn[KMG] The fixed amount of bus space which is 5232 reserved for the CardBus bridge's memory 5233 window. The default value is 64 megabytes. 5234 resource_alignment= 5235 Format: 5236 [<order of align>@]<pci_dev>[; ...] 5237 Specifies alignment and device to reassign 5238 aligned memory resources. How to 5239 specify the device is described above. 5240 If <order of align> is not specified, 5241 PAGE_SIZE is used as alignment. 5242 A PCI-PCI bridge can be specified if resource 5243 windows need to be expanded. 5244 To specify the alignment for several 5245 instances of a device, the PCI vendor, 5246 device, subvendor, and subdevice may be 5247 specified, e.g., 12@pci:8086:9c22:103c:198f 5248 for 4096-byte alignment. 5249 ecrc= Enable/disable PCIe ECRC (transaction layer 5250 end-to-end CRC checking). Only effective if 5251 OS has native AER control (either granted by 5252 ACPI _OSC or forced via "pcie_ports=native") 5253 bios: Use BIOS/firmware settings. This is the 5254 the default. 5255 off: Turn ECRC off 5256 on: Turn ECRC on. 5257 hpiosize=nn[KMG] The fixed amount of bus space which is 5258 reserved for hotplug bridge's IO window. 5259 Default size is 256 bytes. 5260 hpmmiosize=nn[KMG] The fixed amount of bus space which is 5261 reserved for hotplug bridge's MMIO window. 5262 Default size is 2 megabytes. 5263 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 5264 reserved for hotplug bridge's MMIO_PREF window. 5265 Default size is 2 megabytes. 5266 hpmemsize=nn[KMG] The fixed amount of bus space which is 5267 reserved for hotplug bridge's MMIO and 5268 MMIO_PREF window. 5269 Default size is 2 megabytes. 5270 hpbussize=nn The minimum amount of additional bus numbers 5271 reserved for buses below a hotplug bridge. 5272 Default is 1. 5273 realloc= Enable/disable reallocating PCI bridge resources 5274 if allocations done by BIOS are too small to 5275 accommodate resources required by all child 5276 devices. 5277 off: Turn realloc off 5278 on: Turn realloc on 5279 realloc same as realloc=on 5280 noari do not use PCIe ARI. 5281 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 5282 do not use PCIe ATS (and IOMMU device IOTLB). 5283 pcie_scan_all Scan all possible PCIe devices. Otherwise we 5284 only look for one device below a PCIe downstream 5285 port. 5286 big_root_window Try to add a big 64bit memory window to the PCIe 5287 root complex on AMD CPUs. Some GFX hardware 5288 can resize a BAR to allow access to all VRAM. 5289 Adding the window is slightly risky (it may 5290 conflict with unreported devices), so this 5291 taints the kernel. 5292 disable_acs_redir=<pci_dev>[; ...] 5293 Specify one or more PCI devices (in the format 5294 specified above) separated by semicolons. 5295 Each device specified will have the PCI ACS 5296 redirect capabilities forced off which will 5297 allow P2P traffic between devices through 5298 bridges without forcing it upstream. Note: 5299 this removes isolation between devices and 5300 may put more devices in an IOMMU group. 5301 config_acs= 5302 Format: 5303 <ACS flags>@<pci_dev>[; ...] 5304 Specify one or more PCI devices (in the format 5305 specified above) optionally prepended with flags 5306 and separated by semicolons. The respective 5307 capabilities will be enabled, disabled or 5308 unchanged based on what is specified in 5309 flags. 5310 5311 ACS Flags is defined as follows: 5312 bit-0 : ACS Source Validation 5313 bit-1 : ACS Translation Blocking 5314 bit-2 : ACS P2P Request Redirect 5315 bit-3 : ACS P2P Completion Redirect 5316 bit-4 : ACS Upstream Forwarding 5317 bit-5 : ACS P2P Egress Control 5318 bit-6 : ACS Direct Translated P2P 5319 Each bit can be marked as: 5320 '0' – force disabled 5321 '1' – force enabled 5322 'x' – unchanged 5323 For example, 5324 pci=config_acs=10x@pci:0:0 5325 would configure all devices that support 5326 ACS to enable P2P Request Redirect, disable 5327 Translation Blocking, and leave Source 5328 Validation unchanged from whatever power-up 5329 or firmware set it to. 5330 5331 Note: this may remove isolation between devices 5332 and may put more devices in an IOMMU group. 5333 force_floating [S390] Force usage of floating interrupts. 5334 nomio [S390] Do not use MIO instructions. 5335 norid [S390] ignore the RID field and force use of 5336 one PCI domain per PCI function 5337 notph [PCIE] If the PCIE_TPH kernel config parameter 5338 is enabled, this kernel boot option can be used 5339 to disable PCIe TLP Processing Hints support 5340 system-wide. 5341 5342 pcie_aspm= [PCIE] Forcibly enable or ignore PCIe Active State Power 5343 Management. 5344 off Don't touch ASPM configuration at all. Leave any 5345 configuration done by firmware unchanged. 5346 force Enable ASPM even on devices that claim not to support it. 5347 WARNING: Forcing ASPM on may cause system lockups. 5348 5349 pcie_ports= [PCIE] PCIe port services handling: 5350 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 5351 even if the platform doesn't give the OS permission to 5352 use them. This may cause conflicts if the platform 5353 also tries to use these services. 5354 dpc-native Use native PCIe service for DPC only. May 5355 cause conflicts if firmware uses AER or DPC. 5356 compat Disable native PCIe services (PME, AER, DPC, PCIe 5357 hotplug). 5358 5359 pcie_port_pm= [PCIE] PCIe port power management handling: 5360 off Disable power management of all PCIe ports 5361 force Forcibly enable power management of all PCIe ports 5362 5363 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 5364 nomsi Do not use MSI for native PCIe PME signaling (this makes 5365 all PCIe root ports use INTx for all services). 5366 5367 pcmv= [HW,PCMCIA] BadgePAD 4 5368 5369 pd_ignore_unused 5370 [PM] 5371 Keep all power-domains already enabled by bootloader on, 5372 even if no driver has claimed them. This is useful 5373 for debug and development, but should not be 5374 needed on a platform with proper driver support. 5375 5376 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 5377 boot time. 5378 Format: { 0 | 1 } 5379 See arch/parisc/kernel/pdc_chassis.c 5380 5381 percpu_alloc= [MM,EARLY] 5382 Select which percpu first chunk allocator to use. 5383 Currently supported values are "embed" and "page". 5384 Archs may support subset or none of the selections. 5385 See comments in mm/percpu.c for details on each 5386 allocator. This parameter is primarily for debugging 5387 and performance comparison. 5388 5389 pirq= [SMP,APIC] Manual mp-table setup 5390 See Documentation/arch/x86/i386/IO-APIC.rst. 5391 5392 plip= [PPT,NET] Parallel port network link 5393 Format: { parport<nr> | timid | 0 } 5394 See also Documentation/admin-guide/parport.rst. 5395 5396 pmtmr= [X86] Manual setup of pmtmr I/O Port. 5397 Override pmtimer IOPort with a hex value. 5398 e.g. pmtmr=0x508 5399 5400 pmu_override= [PPC] Override the PMU. 5401 This option takes over the PMU facility, so it is no 5402 longer usable by perf. Setting this option starts the 5403 PMU counters by setting MMCR0 to 0 (the FC bit is 5404 cleared). If a number is given, then MMCR1 is set to 5405 that number, otherwise (e.g., 'pmu_override=on'), MMCR1 5406 remains 0. 5407 5408 pm_async= [PM] 5409 Format: off 5410 This parameter sets the initial value of the 5411 /sys/power/pm_async sysfs knob at boot time. 5412 If set to "off", disables asynchronous suspend and 5413 resume of devices during system-wide power transitions. 5414 This can be useful on platforms where device 5415 dependencies are not well-defined, or for debugging 5416 power management issues. Asynchronous operations are 5417 enabled by default. 5418 5419 5420 pm_debug_messages [SUSPEND,KNL] 5421 Enable suspend/resume debug messages during boot up. 5422 5423 pnp.debug=1 [PNP] 5424 Enable PNP debug messages (depends on the 5425 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 5426 via /sys/module/pnp/parameters/debug. We always show 5427 current resource usage; turning this on also shows 5428 possible settings and some assignment information. 5429 5430 pnpacpi= [ACPI] 5431 { off } 5432 5433 pnpbios= [ISAPNP] 5434 { on | off | curr | res | no-curr | no-res } 5435 5436 pnp_reserve_irq= 5437 [ISAPNP] Exclude IRQs for the autoconfiguration 5438 5439 pnp_reserve_dma= 5440 [ISAPNP] Exclude DMAs for the autoconfiguration 5441 5442 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 5443 Ranges are in pairs (I/O port base and size). 5444 5445 pnp_reserve_mem= 5446 [ISAPNP] Exclude memory regions for the 5447 autoconfiguration. 5448 Ranges are in pairs (memory base and size). 5449 5450 ports= [IP_VS_FTP] IPVS ftp helper module 5451 Default is 21. 5452 Up to 8 (IP_VS_APP_MAX_PORTS) ports 5453 may be specified. 5454 Format: <port>,<port>.... 5455 5456 possible_cpus= [SMP,S390,X86] 5457 Format: <unsigned int> 5458 Set the number of possible CPUs, overriding the 5459 regular discovery mechanisms (such as ACPI/FW, etc). 5460 5461 powersave=off [PPC] This option disables power saving features. 5462 It specifically disables cpuidle and sets the 5463 platform machine description specific power_save 5464 function to NULL. On Idle the CPU just reduces 5465 execution priority. 5466 5467 ppc_strict_facility_enable 5468 [PPC,ENABLE] This option catches any kernel floating point, 5469 Altivec, VSX and SPE outside of regions specifically 5470 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 5471 There is some performance impact when enabling this. 5472 5473 ppc_tm= [PPC,EARLY] 5474 Format: {"off"} 5475 Disable Hardware Transactional Memory 5476 5477 preempt= [KNL] 5478 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC 5479 none - Limited to cond_resched() calls 5480 voluntary - Limited to cond_resched() and might_sleep() calls 5481 full - Any section that isn't explicitly preempt disabled 5482 can be preempted anytime. Tasks will also yield 5483 contended spinlocks (if the critical section isn't 5484 explicitly preempt disabled beyond the lock itself). 5485 lazy - Scheduler controlled. Similar to full but instead 5486 of preempting the task immediately, the task gets 5487 one HZ tick time to yield itself before the 5488 preemption will be forced. One preemption is when the 5489 task returns to user space. 5490 5491 print-fatal-signals= 5492 [KNL] debug: print fatal signals 5493 5494 If enabled, warn about various signal handling 5495 related application anomalies: too many signals, 5496 too many POSIX.1 timers, fatal signals causing a 5497 coredump - etc. 5498 5499 If you hit the warning due to signal overflow, 5500 you might want to try "ulimit -i unlimited". 5501 5502 default: off. 5503 5504 printk.always_kmsg_dump= 5505 Trigger kmsg_dump for cases other than kernel oops or 5506 panics 5507 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5508 default: disabled 5509 5510 printk.console_no_auto_verbose= 5511 Disable console loglevel raise on oops, panic 5512 or lockdep-detected issues (only if lock debug is on). 5513 With an exception to setups with low baudrate on 5514 serial console, keeping this 0 is a good choice 5515 in order to provide more debug information. 5516 Format: <bool> 5517 default: 0 (auto_verbose is enabled) 5518 5519 printk.debug_non_panic_cpus= 5520 Allows storing messages from non-panic CPUs into 5521 the printk log buffer during panic(). They are 5522 flushed to consoles by the panic-CPU on 5523 a best-effort basis. 5524 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5525 Default: disabled 5526 5527 printk.devkmsg={on,off,ratelimit} 5528 Control writing to /dev/kmsg. 5529 on - unlimited logging to /dev/kmsg from userspace 5530 off - logging to /dev/kmsg disabled 5531 ratelimit - ratelimit the logging 5532 Default: ratelimit 5533 5534 printk.time= Show timing data prefixed to each printk message line 5535 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5536 5537 proc_mem.force_override= [KNL] 5538 Format: {always | ptrace | never} 5539 Traditionally /proc/pid/mem allows memory permissions to be 5540 overridden without restrictions. This option may be set to 5541 restrict that. Can be one of: 5542 - 'always': traditional behavior always allows mem overrides. 5543 - 'ptrace': only allow mem overrides for active ptracers. 5544 - 'never': never allow mem overrides. 5545 If not specified, default is the CONFIG_PROC_MEM_* choice. 5546 5547 processor.max_cstate= [HW,ACPI] 5548 Limit processor to maximum C-state 5549 max_cstate=9 overrides any DMI blacklist limit. 5550 5551 processor.nocst [HW,ACPI] 5552 Ignore the _CST method to determine C-states, 5553 instead using the legacy FADT method 5554 5555 profile= [KNL] Enable kernel profiling via /proc/profile 5556 Format: [<profiletype>,]<number> 5557 Param: <profiletype>: "schedule" or "kvm" 5558 [defaults to kernel profiling] 5559 Param: "schedule" - profile schedule points. 5560 Param: "kvm" - profile VM exits. 5561 Param: <number> - step/bucket size as a power of 2 for 5562 statistical time based profiling. 5563 5564 prot_virt= [S390] enable hosting protected virtual machines 5565 isolated from the hypervisor (if hardware supports 5566 that). If enabled, the default kernel base address 5567 might be overridden even when Kernel Address Space 5568 Layout Randomization is disabled. 5569 Format: <bool> 5570 5571 psi= [KNL] Enable or disable pressure stall information 5572 tracking. 5573 Format: <bool> 5574 5575 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 5576 probe for; one of (bare|imps|exps|lifebook|any). 5577 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 5578 per second. 5579 psmouse.resetafter= [HW,MOUSE] 5580 Try to reset the device after so many bad packets 5581 (0 = never). 5582 psmouse.resolution= 5583 [HW,MOUSE] Set desired mouse resolution, in dpi. 5584 psmouse.smartscroll= 5585 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 5586 0 = disabled, 1 = enabled (default). 5587 5588 pstore.backend= Specify the name of the pstore backend to use 5589 5590 pti= [X86-64] Control Page Table Isolation of user and 5591 kernel address spaces. Disabling this feature 5592 removes hardening, but improves performance of 5593 system calls and interrupts. 5594 5595 on - unconditionally enable 5596 off - unconditionally disable 5597 auto - kernel detects whether your CPU model is 5598 vulnerable to issues that PTI mitigates 5599 5600 Not specifying this option is equivalent to pti=auto. 5601 5602 pty.legacy_count= 5603 [KNL] Number of legacy pty's. Overwrites compiled-in 5604 default number. 5605 5606 quiet [KNL,EARLY] Disable most log messages 5607 5608 r128= [HW,DRM] 5609 5610 radix_hcall_invalidate=on [PPC/PSERIES] 5611 Disable RADIX GTSE feature and use hcall for TLB 5612 invalidate. 5613 5614 raid= [HW,RAID] 5615 See Documentation/admin-guide/md.rst. 5616 5617 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 5618 See Documentation/admin-guide/blockdev/ramdisk.rst. 5619 5620 ramdisk_start= [Deprecated,RAM] RAM disk image start address 5621 5622 random.trust_cpu=off 5623 [KNL,EARLY] Disable trusting the use of the CPU's 5624 random number generator (if available) to 5625 initialize the kernel's RNG. 5626 5627 random.trust_bootloader=off 5628 [KNL,EARLY] Disable trusting the use of the a seed 5629 passed by the bootloader (if available) to 5630 initialize the kernel's RNG. 5631 5632 randomize_kstack_offset= 5633 [KNL,EARLY] Enable or disable kernel stack offset 5634 randomization, which provides roughly 5 bits of 5635 entropy, frustrating memory corruption attacks 5636 that depend on stack address determinism or 5637 cross-syscall address exposures. This is only 5638 available on architectures that have defined 5639 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET. 5640 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5641 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT. 5642 5643 ras=option[,option,...] [KNL] RAS-specific options 5644 5645 cec_disable [X86] 5646 Disable the Correctable Errors Collector, 5647 see CONFIG_RAS_CEC help text. 5648 5649 rcu_nocbs[=cpu-list] 5650 [KNL] The optional argument is a cpu list, 5651 as described above. 5652 5653 In kernels built with CONFIG_RCU_NOCB_CPU=y, 5654 enable the no-callback CPU mode, which prevents 5655 such CPUs' callbacks from being invoked in 5656 softirq context. Invocation of such CPUs' RCU 5657 callbacks will instead be offloaded to "rcuox/N" 5658 kthreads created for that purpose, where "x" is 5659 "p" for RCU-preempt, "s" for RCU-sched, and "g" 5660 for the kthreads that mediate grace periods; and 5661 "N" is the CPU number. This reduces OS jitter on 5662 the offloaded CPUs, which can be useful for HPC 5663 and real-time workloads. It can also improve 5664 energy efficiency for asymmetric multiprocessors. 5665 5666 If a cpulist is passed as an argument, the specified 5667 list of CPUs is set to no-callback mode from boot. 5668 5669 Otherwise, if the '=' sign and the cpulist 5670 arguments are omitted, no CPU will be set to 5671 no-callback mode from boot but the mode may be 5672 toggled at runtime via cpusets. 5673 5674 Note that this argument takes precedence over 5675 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 5676 5677 rcu_nocb_poll [KNL] 5678 Rather than requiring that offloaded CPUs 5679 (specified by rcu_nocbs= above) explicitly 5680 awaken the corresponding "rcuoN" kthreads, 5681 make these kthreads poll for callbacks. 5682 This improves the real-time response for the 5683 offloaded CPUs by relieving them of the need to 5684 wake up the corresponding kthread, but degrades 5685 energy efficiency by requiring that the kthreads 5686 periodically wake up to do the polling. 5687 5688 rcutree.blimit= [KNL] 5689 Set maximum number of finished RCU callbacks to 5690 process in one batch. 5691 5692 rcutree.csd_lock_suppress_rcu_stall= [KNL] 5693 Do only a one-line RCU CPU stall warning when 5694 there is an ongoing too-long CSD-lock wait. 5695 5696 rcutree.do_rcu_barrier= [KNL] 5697 Request a call to rcu_barrier(). This is 5698 throttled so that userspace tests can safely 5699 hammer on the sysfs variable if they so choose. 5700 If triggered before the RCU grace-period machinery 5701 is fully active, this will error out with EAGAIN. 5702 5703 rcutree.dump_tree= [KNL] 5704 Dump the structure of the rcu_node combining tree 5705 out at early boot. This is used for diagnostic 5706 purposes, to verify correct tree setup. 5707 5708 rcutree.gp_cleanup_delay= [KNL] 5709 Set the number of jiffies to delay each step of 5710 RCU grace-period cleanup. 5711 5712 rcutree.gp_init_delay= [KNL] 5713 Set the number of jiffies to delay each step of 5714 RCU grace-period initialization. 5715 5716 rcutree.gp_preinit_delay= [KNL] 5717 Set the number of jiffies to delay each step of 5718 RCU grace-period pre-initialization, that is, 5719 the propagation of recent CPU-hotplug changes up 5720 the rcu_node combining tree. 5721 5722 rcutree.jiffies_till_first_fqs= [KNL] 5723 Set delay from grace-period initialization to 5724 first attempt to force quiescent states. 5725 Units are jiffies, minimum value is zero, 5726 and maximum value is HZ. 5727 5728 rcutree.jiffies_till_next_fqs= [KNL] 5729 Set delay between subsequent attempts to force 5730 quiescent states. Units are jiffies, minimum 5731 value is one, and maximum value is HZ. 5732 5733 rcutree.jiffies_till_sched_qs= [KNL] 5734 Set required age in jiffies for a 5735 given grace period before RCU starts 5736 soliciting quiescent-state help from 5737 rcu_note_context_switch() and cond_resched(). 5738 If not specified, the kernel will calculate 5739 a value based on the most recent settings 5740 of rcutree.jiffies_till_first_fqs 5741 and rcutree.jiffies_till_next_fqs. 5742 This calculated value may be viewed in 5743 rcutree.jiffies_to_sched_qs. Any attempt to set 5744 rcutree.jiffies_to_sched_qs will be cheerfully 5745 overwritten. 5746 5747 rcutree.kthread_prio= [KNL,BOOT] 5748 Set the SCHED_FIFO priority of the RCU per-CPU 5749 kthreads (rcuc/N). This value is also used for 5750 the priority of the RCU boost threads (rcub/N) 5751 and for the RCU grace-period kthreads (rcu_bh, 5752 rcu_preempt, and rcu_sched). If RCU_BOOST is 5753 set, valid values are 1-99 and the default is 1 5754 (the least-favored priority). Otherwise, when 5755 RCU_BOOST is not set, valid values are 0-99 and 5756 the default is zero (non-realtime operation). 5757 When RCU_NOCB_CPU is set, also adjust the 5758 priority of NOCB callback kthreads. 5759 5760 rcutree.nocb_nobypass_lim_per_jiffy= [KNL] 5761 On callback-offloaded (rcu_nocbs) CPUs, 5762 RCU reduces the lock contention that would 5763 otherwise be caused by callback floods through 5764 use of the ->nocb_bypass list. However, in the 5765 common non-flooded case, RCU queues directly to 5766 the main ->cblist in order to avoid the extra 5767 overhead of the ->nocb_bypass list and its lock. 5768 But if there are too many callbacks queued during 5769 a single jiffy, RCU pre-queues the callbacks into 5770 the ->nocb_bypass queue. The definition of "too 5771 many" is supplied by this kernel boot parameter. 5772 5773 rcutree.nohz_full_patience_delay= [KNL] 5774 On callback-offloaded (rcu_nocbs) CPUs, avoid 5775 disturbing RCU unless the grace period has 5776 reached the specified age in milliseconds. 5777 Defaults to zero. Large values will be capped 5778 at five seconds. All values will be rounded down 5779 to the nearest value representable by jiffies. 5780 5781 rcutree.qhimark= [KNL] 5782 Set threshold of queued RCU callbacks beyond which 5783 batch limiting is disabled. 5784 5785 rcutree.qlowmark= [KNL] 5786 Set threshold of queued RCU callbacks below which 5787 batch limiting is re-enabled. 5788 5789 rcutree.qovld= [KNL] 5790 Set threshold of queued RCU callbacks beyond which 5791 RCU's force-quiescent-state scan will aggressively 5792 enlist help from cond_resched() and sched IPIs to 5793 help CPUs more quickly reach quiescent states. 5794 Set to less than zero to make this be set based 5795 on rcutree.qhimark at boot time and to zero to 5796 disable more aggressive help enlistment. 5797 5798 rcutree.rcu_delay_page_cache_fill_msec= [KNL] 5799 Set the page-cache refill delay (in milliseconds) 5800 in response to low-memory conditions. The range 5801 of permitted values is in the range 0:100000. 5802 5803 rcutree.rcu_divisor= [KNL] 5804 Set the shift-right count to use to compute 5805 the callback-invocation batch limit bl from 5806 the number of callbacks queued on this CPU. 5807 The result will be bounded below by the value of 5808 the rcutree.blimit kernel parameter. Every bl 5809 callbacks, the softirq handler will exit in 5810 order to allow the CPU to do other work. 5811 5812 Please note that this callback-invocation batch 5813 limit applies only to non-offloaded callback 5814 invocation. Offloaded callbacks are instead 5815 invoked in the context of an rcuoc kthread, which 5816 scheduler will preempt as it does any other task. 5817 5818 rcutree.rcu_fanout_exact= [KNL] 5819 Disable autobalancing of the rcu_node combining 5820 tree. This is used by rcutorture, and might 5821 possibly be useful for architectures having high 5822 cache-to-cache transfer latencies. 5823 5824 rcutree.rcu_fanout_leaf= [KNL] 5825 Change the number of CPUs assigned to each 5826 leaf rcu_node structure. Useful for very 5827 large systems, which will choose the value 64, 5828 and for NUMA systems with large remote-access 5829 latencies, which will choose a value aligned 5830 with the appropriate hardware boundaries. 5831 5832 rcutree.rcu_min_cached_objs= [KNL] 5833 Minimum number of objects which are cached and 5834 maintained per one CPU. Object size is equal 5835 to PAGE_SIZE. The cache allows to reduce the 5836 pressure to page allocator, also it makes the 5837 whole algorithm to behave better in low memory 5838 condition. 5839 5840 rcutree.rcu_nocb_gp_stride= [KNL] 5841 Set the number of NOCB callback kthreads in 5842 each group, which defaults to the square root 5843 of the number of CPUs. Larger numbers reduce 5844 the wakeup overhead on the global grace-period 5845 kthread, but increases that same overhead on 5846 each group's NOCB grace-period kthread. 5847 5848 rcutree.rcu_kick_kthreads= [KNL] 5849 Cause the grace-period kthread to get an extra 5850 wake_up() if it sleeps three times longer than 5851 it should at force-quiescent-state time. 5852 This wake_up() will be accompanied by a 5853 WARN_ONCE() splat and an ftrace_dump(). 5854 5855 rcutree.rcu_resched_ns= [KNL] 5856 Limit the time spend invoking a batch of RCU 5857 callbacks to the specified number of nanoseconds. 5858 By default, this limit is checked only once 5859 every 32 callbacks in order to limit the pain 5860 inflicted by local_clock() overhead. 5861 5862 rcutree.rcu_unlock_delay= [KNL] 5863 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 5864 this specifies an rcu_read_unlock()-time delay 5865 in microseconds. This defaults to zero. 5866 Larger delays increase the probability of 5867 catching RCU pointer leaks, that is, buggy use 5868 of RCU-protected pointers after the relevant 5869 rcu_read_unlock() has completed. 5870 5871 rcutree.sysrq_rcu= [KNL] 5872 Commandeer a sysrq key to dump out Tree RCU's 5873 rcu_node tree with an eye towards determining 5874 why a new grace period has not yet started. 5875 5876 rcutree.use_softirq= [KNL] 5877 If set to zero, move all RCU_SOFTIRQ processing to 5878 per-CPU rcuc kthreads. Defaults to a non-zero 5879 value, meaning that RCU_SOFTIRQ is used by default. 5880 Specify rcutree.use_softirq=0 to use rcuc kthreads. 5881 5882 But note that CONFIG_PREEMPT_RT=y kernels disable 5883 this kernel boot parameter, forcibly setting it 5884 to zero. 5885 5886 rcutree.enable_rcu_lazy= [KNL] 5887 To save power, batch RCU callbacks and flush after 5888 delay, memory pressure or callback list growing too 5889 big. 5890 5891 rcutree.rcu_normal_wake_from_gp= [KNL] 5892 Reduces a latency of synchronize_rcu() call. This approach 5893 maintains its own track of synchronize_rcu() callers, so it 5894 does not interact with regular callbacks because it does not 5895 use a call_rcu[_hurry]() path. Please note, this is for a 5896 normal grace period. 5897 5898 How to enable it: 5899 5900 echo 1 > /sys/module/rcutree/parameters/rcu_normal_wake_from_gp 5901 or pass a boot parameter "rcutree.rcu_normal_wake_from_gp=1" 5902 5903 Default is 1 if num_possible_cpus() <= 16 and it is not explicitly 5904 disabled by the boot parameter passing 0. 5905 5906 rcuscale.gp_async= [KNL] 5907 Measure performance of asynchronous 5908 grace-period primitives such as call_rcu(). 5909 5910 rcuscale.gp_async_max= [KNL] 5911 Specify the maximum number of outstanding 5912 callbacks per writer thread. When a writer 5913 thread exceeds this limit, it invokes the 5914 corresponding flavor of rcu_barrier() to allow 5915 previously posted callbacks to drain. 5916 5917 rcuscale.gp_exp= [KNL] 5918 Measure performance of expedited synchronous 5919 grace-period primitives. 5920 5921 rcuscale.holdoff= [KNL] 5922 Set test-start holdoff period. The purpose of 5923 this parameter is to delay the start of the 5924 test until boot completes in order to avoid 5925 interference. 5926 5927 rcuscale.kfree_by_call_rcu= [KNL] 5928 In kernels built with CONFIG_RCU_LAZY=y, test 5929 call_rcu() instead of kfree_rcu(). 5930 5931 rcuscale.kfree_mult= [KNL] 5932 Instead of allocating an object of size kfree_obj, 5933 allocate one of kfree_mult * sizeof(kfree_obj). 5934 Defaults to 1. 5935 5936 rcuscale.kfree_rcu_test= [KNL] 5937 Set to measure performance of kfree_rcu() flooding. 5938 5939 rcuscale.kfree_rcu_test_double= [KNL] 5940 Test the double-argument variant of kfree_rcu(). 5941 If this parameter has the same value as 5942 rcuscale.kfree_rcu_test_single, both the single- 5943 and double-argument variants are tested. 5944 5945 rcuscale.kfree_rcu_test_single= [KNL] 5946 Test the single-argument variant of kfree_rcu(). 5947 If this parameter has the same value as 5948 rcuscale.kfree_rcu_test_double, both the single- 5949 and double-argument variants are tested. 5950 5951 rcuscale.kfree_nthreads= [KNL] 5952 The number of threads running loops of kfree_rcu(). 5953 5954 rcuscale.kfree_alloc_num= [KNL] 5955 Number of allocations and frees done in an iteration. 5956 5957 rcuscale.kfree_loops= [KNL] 5958 Number of loops doing rcuscale.kfree_alloc_num number 5959 of allocations and frees. 5960 5961 rcuscale.minruntime= [KNL] 5962 Set the minimum test run time in seconds. This 5963 does not affect the data-collection interval, 5964 but instead allows better measurement of things 5965 like CPU consumption. 5966 5967 rcuscale.nreaders= [KNL] 5968 Set number of RCU readers. The value -1 selects 5969 N, where N is the number of CPUs. A value 5970 "n" less than -1 selects N-n+1, where N is again 5971 the number of CPUs. For example, -2 selects N 5972 (the number of CPUs), -3 selects N+1, and so on. 5973 A value of "n" less than or equal to -N selects 5974 a single reader. 5975 5976 rcuscale.nwriters= [KNL] 5977 Set number of RCU writers. The values operate 5978 the same as for rcuscale.nreaders. 5979 N, where N is the number of CPUs 5980 5981 rcuscale.scale_type= [KNL] 5982 Specify the RCU implementation to test. 5983 5984 rcuscale.shutdown= [KNL] 5985 Shut the system down after performance tests 5986 complete. This is useful for hands-off automated 5987 testing. 5988 5989 rcuscale.verbose= [KNL] 5990 Enable additional printk() statements. 5991 5992 rcuscale.writer_holdoff= [KNL] 5993 Write-side holdoff between grace periods, 5994 in microseconds. The default of zero says 5995 no holdoff. 5996 5997 rcuscale.writer_holdoff_jiffies= [KNL] 5998 Additional write-side holdoff between grace 5999 periods, but in jiffies. The default of zero 6000 says no holdoff. 6001 6002 rcutorture.fqs_duration= [KNL] 6003 Set duration of force_quiescent_state bursts 6004 in microseconds. 6005 6006 rcutorture.fqs_holdoff= [KNL] 6007 Set holdoff time within force_quiescent_state bursts 6008 in microseconds. 6009 6010 rcutorture.fqs_stutter= [KNL] 6011 Set wait time between force_quiescent_state bursts 6012 in seconds. 6013 6014 rcutorture.fwd_progress= [KNL] 6015 Specifies the number of kthreads to be used 6016 for RCU grace-period forward-progress testing 6017 for the types of RCU supporting this notion. 6018 Defaults to 1 kthread, values less than zero or 6019 greater than the number of CPUs cause the number 6020 of CPUs to be used. 6021 6022 rcutorture.fwd_progress_div= [KNL] 6023 Specify the fraction of a CPU-stall-warning 6024 period to do tight-loop forward-progress testing. 6025 6026 rcutorture.fwd_progress_holdoff= [KNL] 6027 Number of seconds to wait between successive 6028 forward-progress tests. 6029 6030 rcutorture.fwd_progress_need_resched= [KNL] 6031 Enclose cond_resched() calls within checks for 6032 need_resched() during tight-loop forward-progress 6033 testing. 6034 6035 rcutorture.gp_cond= [KNL] 6036 Use conditional/asynchronous update-side 6037 normal-grace-period primitives, if available. 6038 6039 rcutorture.gp_cond_exp= [KNL] 6040 Use conditional/asynchronous update-side 6041 expedited-grace-period primitives, if available. 6042 6043 rcutorture.gp_cond_full= [KNL] 6044 Use conditional/asynchronous update-side 6045 normal-grace-period primitives that also take 6046 concurrent expedited grace periods into account, 6047 if available. 6048 6049 rcutorture.gp_cond_exp_full= [KNL] 6050 Use conditional/asynchronous update-side 6051 expedited-grace-period primitives that also take 6052 concurrent normal grace periods into account, 6053 if available. 6054 6055 rcutorture.gp_cond_wi= [KNL] 6056 Nominal wait interval for normal conditional 6057 grace periods (specified by rcutorture's 6058 gp_cond and gp_cond_full module parameters), 6059 in microseconds. The actual wait interval will 6060 be randomly selected to nanosecond granularity up 6061 to this wait interval. Defaults to 16 jiffies, 6062 for example, 16,000 microseconds on a system 6063 with HZ=1000. 6064 6065 rcutorture.gp_cond_wi_exp= [KNL] 6066 Nominal wait interval for expedited conditional 6067 grace periods (specified by rcutorture's 6068 gp_cond_exp and gp_cond_exp_full module 6069 parameters), in microseconds. The actual wait 6070 interval will be randomly selected to nanosecond 6071 granularity up to this wait interval. Defaults to 6072 128 microseconds. 6073 6074 rcutorture.gp_exp= [KNL] 6075 Use expedited update-side primitives, if available. 6076 6077 rcutorture.gp_normal= [KNL] 6078 Use normal (non-expedited) asynchronous 6079 update-side primitives, if available. 6080 6081 rcutorture.gp_poll= [KNL] 6082 Use polled update-side normal-grace-period 6083 primitives, if available. 6084 6085 rcutorture.gp_poll_exp= [KNL] 6086 Use polled update-side expedited-grace-period 6087 primitives, if available. 6088 6089 rcutorture.gp_poll_full= [KNL] 6090 Use polled update-side normal-grace-period 6091 primitives that also take concurrent expedited 6092 grace periods into account, if available. 6093 6094 rcutorture.gp_poll_exp_full= [KNL] 6095 Use polled update-side expedited-grace-period 6096 primitives that also take concurrent normal 6097 grace periods into account, if available. 6098 6099 rcutorture.gp_poll_wi= [KNL] 6100 Nominal wait interval for normal conditional 6101 grace periods (specified by rcutorture's 6102 gp_poll and gp_poll_full module parameters), 6103 in microseconds. The actual wait interval will 6104 be randomly selected to nanosecond granularity up 6105 to this wait interval. Defaults to 16 jiffies, 6106 for example, 16,000 microseconds on a system 6107 with HZ=1000. 6108 6109 rcutorture.gp_poll_wi_exp= [KNL] 6110 Nominal wait interval for expedited conditional 6111 grace periods (specified by rcutorture's 6112 gp_poll_exp and gp_poll_exp_full module 6113 parameters), in microseconds. The actual wait 6114 interval will be randomly selected to nanosecond 6115 granularity up to this wait interval. Defaults to 6116 128 microseconds. 6117 6118 rcutorture.gp_sync= [KNL] 6119 Use normal (non-expedited) synchronous 6120 update-side primitives, if available. If all 6121 of rcutorture.gp_cond=, rcutorture.gp_exp=, 6122 rcutorture.gp_normal=, and rcutorture.gp_sync= 6123 are zero, rcutorture acts as if is interpreted 6124 they are all non-zero. 6125 6126 rcutorture.gpwrap_lag= [KNL] 6127 Enable grace-period wrap lag testing. Setting 6128 to false prevents the gpwrap lag test from 6129 running. Default is true. 6130 6131 rcutorture.gpwrap_lag_gps= [KNL] 6132 Set the value for grace-period wrap lag during 6133 active lag testing periods. This controls how many 6134 grace periods differences we tolerate between 6135 rdp and rnp's gp_seq before setting overflow flag. 6136 The default is always set to 8. 6137 6138 rcutorture.gpwrap_lag_cycle_mins= [KNL] 6139 Set the total cycle duration for gpwrap lag 6140 testing in minutes. This is the total time for 6141 one complete cycle of active and inactive 6142 testing periods. Default is 30 minutes. 6143 6144 rcutorture.gpwrap_lag_active_mins= [KNL] 6145 Set the duration for which gpwrap lag is active 6146 within each cycle, in minutes. During this time, 6147 the grace-period wrap lag will be set to the 6148 value specified by gpwrap_lag_gps. Default is 6149 5 minutes. 6150 6151 rcutorture.irqreader= [KNL] 6152 Run RCU readers from irq handlers, or, more 6153 accurately, from a timer handler. Not all RCU 6154 flavors take kindly to this sort of thing. 6155 6156 rcutorture.leakpointer= [KNL] 6157 Leak an RCU-protected pointer out of the reader. 6158 This can of course result in splats, and is 6159 intended to test the ability of things like 6160 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 6161 such leaks. 6162 6163 rcutorture.n_barrier_cbs= [KNL] 6164 Set callbacks/threads for rcu_barrier() testing. 6165 6166 rcutorture.nfakewriters= [KNL] 6167 Set number of concurrent RCU writers. These just 6168 stress RCU, they don't participate in the actual 6169 test, hence the "fake". 6170 6171 rcutorture.nocbs_nthreads= [KNL] 6172 Set number of RCU callback-offload togglers. 6173 Zero (the default) disables toggling. 6174 6175 rcutorture.nocbs_toggle= [KNL] 6176 Set the delay in milliseconds between successive 6177 callback-offload toggling attempts. 6178 6179 rcutorture.nreaders= [KNL] 6180 Set number of RCU readers. The value -1 selects 6181 N-1, where N is the number of CPUs. A value 6182 "n" less than -1 selects N-n-2, where N is again 6183 the number of CPUs. For example, -2 selects N 6184 (the number of CPUs), -3 selects N+1, and so on. 6185 6186 rcutorture.object_debug= [KNL] 6187 Enable debug-object double-call_rcu() testing. 6188 6189 rcutorture.onoff_holdoff= [KNL] 6190 Set time (s) after boot for CPU-hotplug testing. 6191 6192 rcutorture.onoff_interval= [KNL] 6193 Set time (jiffies) between CPU-hotplug operations, 6194 or zero to disable CPU-hotplug testing. 6195 6196 rcutorture.preempt_duration= [KNL] 6197 Set duration (in milliseconds) of preemptions 6198 by a high-priority FIFO real-time task. Set to 6199 zero (the default) to disable. The CPUs to 6200 preempt are selected randomly from the set that 6201 are online at a given point in time. Races with 6202 CPUs going offline are ignored, with that attempt 6203 at preemption skipped. 6204 6205 rcutorture.preempt_interval= [KNL] 6206 Set interval (in milliseconds, defaulting to one 6207 second) between preemptions by a high-priority 6208 FIFO real-time task. This delay is mediated 6209 by an hrtimer and is further fuzzed to avoid 6210 inadvertent synchronizations. 6211 6212 rcutorture.read_exit_burst= [KNL] 6213 The number of times in a given read-then-exit 6214 episode that a set of read-then-exit kthreads 6215 is spawned. 6216 6217 rcutorture.read_exit_delay= [KNL] 6218 The delay, in seconds, between successive 6219 read-then-exit testing episodes. 6220 6221 rcutorture.reader_flavor= [KNL] 6222 A bit mask indicating which readers to use. 6223 If there is more than one bit set, the readers 6224 are entered from low-order bit up, and are 6225 exited in the opposite order. For SRCU, the 6226 0x1 bit is normal readers, 0x2 NMI-safe readers, 6227 and 0x4 light-weight readers. 6228 6229 rcutorture.shuffle_interval= [KNL] 6230 Set task-shuffle interval (s). Shuffling tasks 6231 allows some CPUs to go into dyntick-idle mode 6232 during the rcutorture test. 6233 6234 rcutorture.shutdown_secs= [KNL] 6235 Set time (s) after boot system shutdown. This 6236 is useful for hands-off automated testing. 6237 6238 rcutorture.stall_cpu= [KNL] 6239 Duration of CPU stall (s) to test RCU CPU stall 6240 warnings, zero to disable. 6241 6242 rcutorture.stall_cpu_block= [KNL] 6243 Sleep while stalling if set. This will result 6244 in warnings from preemptible RCU in addition to 6245 any other stall-related activity. Note that 6246 in kernels built with CONFIG_PREEMPTION=n and 6247 CONFIG_PREEMPT_COUNT=y, this parameter will 6248 cause the CPU to pass through a quiescent state. 6249 Given CONFIG_PREEMPTION=n, this will suppress 6250 RCU CPU stall warnings, but will instead result 6251 in scheduling-while-atomic splats. 6252 6253 Use of this module parameter results in splats. 6254 6255 6256 rcutorture.stall_cpu_holdoff= [KNL] 6257 Time to wait (s) after boot before inducing stall. 6258 6259 rcutorture.stall_cpu_irqsoff= [KNL] 6260 Disable interrupts while stalling if set, but only 6261 on the first stall in the set. 6262 6263 rcutorture.stall_cpu_repeat= [KNL] 6264 Number of times to repeat the stall sequence, 6265 so that rcutorture.stall_cpu_repeat=3 will result 6266 in four stall sequences. 6267 6268 rcutorture.stall_gp_kthread= [KNL] 6269 Duration (s) of forced sleep within RCU 6270 grace-period kthread to test RCU CPU stall 6271 warnings, zero to disable. If both stall_cpu 6272 and stall_gp_kthread are specified, the 6273 kthread is starved first, then the CPU. 6274 6275 rcutorture.stat_interval= [KNL] 6276 Time (s) between statistics printk()s. 6277 6278 rcutorture.stutter= [KNL] 6279 Time (s) to stutter testing, for example, specifying 6280 five seconds causes the test to run for five seconds, 6281 wait for five seconds, and so on. This tests RCU's 6282 ability to transition abruptly to and from idle. 6283 6284 rcutorture.test_boost= [KNL] 6285 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 6286 "Maybe" means test if the RCU implementation 6287 under test support RCU priority boosting. 6288 6289 rcutorture.test_boost_duration= [KNL] 6290 Duration (s) of each individual boost test. 6291 6292 rcutorture.test_boost_holdoff= [KNL] 6293 Holdoff time (s) from start of test to the start 6294 of RCU priority-boost testing. Defaults to zero, 6295 that is, no holdoff. 6296 6297 rcutorture.test_boost_interval= [KNL] 6298 Interval (s) between each boost test. 6299 6300 rcutorture.test_no_idle_hz= [KNL] 6301 Test RCU's dyntick-idle handling. See also the 6302 rcutorture.shuffle_interval parameter. 6303 6304 rcutorture.torture_type= [KNL] 6305 Specify the RCU implementation to test. 6306 6307 rcutorture.verbose= [KNL] 6308 Enable additional printk() statements. 6309 6310 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 6311 Dump ftrace buffer after reporting RCU CPU 6312 stall warning. 6313 6314 rcupdate.rcu_cpu_stall_notifiers= [KNL] 6315 Provide RCU CPU stall notifiers, but see the 6316 warnings in the RCU_CPU_STALL_NOTIFIER Kconfig 6317 option's help text. TL;DR: You almost certainly 6318 do not want rcupdate.rcu_cpu_stall_notifiers. 6319 6320 rcupdate.rcu_cpu_stall_suppress= [KNL] 6321 Suppress RCU CPU stall warning messages. 6322 6323 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 6324 Suppress RCU CPU stall warning messages and 6325 rcutorture writer stall warnings that occur 6326 during early boot, that is, during the time 6327 before the init task is spawned. 6328 6329 rcupdate.rcu_cpu_stall_timeout= [KNL] 6330 Set timeout for RCU CPU stall warning messages. 6331 The value is in seconds and the maximum allowed 6332 value is 300 seconds. 6333 6334 rcupdate.rcu_exp_cpu_stall_timeout= [KNL] 6335 Set timeout for expedited RCU CPU stall warning 6336 messages. The value is in milliseconds 6337 and the maximum allowed value is 21000 6338 milliseconds. Please note that this value is 6339 adjusted to an arch timer tick resolution. 6340 Setting this to zero causes the value from 6341 rcupdate.rcu_cpu_stall_timeout to be used (after 6342 conversion from seconds to milliseconds). 6343 6344 rcupdate.rcu_cpu_stall_cputime= [KNL] 6345 Provide statistics on the cputime and count of 6346 interrupts and tasks during the sampling period. For 6347 multiple continuous RCU stalls, all sampling periods 6348 begin at half of the first RCU stall timeout. 6349 6350 rcupdate.rcu_exp_stall_task_details= [KNL] 6351 Print stack dumps of any tasks blocking the 6352 current expedited RCU grace period during an 6353 expedited RCU CPU stall warning. 6354 6355 rcupdate.rcu_expedited= [KNL] 6356 Use expedited grace-period primitives, for 6357 example, synchronize_rcu_expedited() instead 6358 of synchronize_rcu(). This reduces latency, 6359 but can increase CPU utilization, degrade 6360 real-time latency, and degrade energy efficiency. 6361 No effect on CONFIG_TINY_RCU kernels. 6362 6363 rcupdate.rcu_normal= [KNL] 6364 Use only normal grace-period primitives, 6365 for example, synchronize_rcu() instead of 6366 synchronize_rcu_expedited(). This improves 6367 real-time latency, CPU utilization, and 6368 energy efficiency, but can expose users to 6369 increased grace-period latency. This parameter 6370 overrides rcupdate.rcu_expedited. No effect on 6371 CONFIG_TINY_RCU kernels. 6372 6373 rcupdate.rcu_normal_after_boot= [KNL] 6374 Once boot has completed (that is, after 6375 rcu_end_inkernel_boot() has been invoked), use 6376 only normal grace-period primitives. No effect 6377 on CONFIG_TINY_RCU kernels. 6378 6379 But note that CONFIG_PREEMPT_RT=y kernels enables 6380 this kernel boot parameter, forcibly setting 6381 it to the value one, that is, converting any 6382 post-boot attempt at an expedited RCU grace 6383 period to instead use normal non-expedited 6384 grace-period processing. 6385 6386 rcupdate.rcu_task_collapse_lim= [KNL] 6387 Set the maximum number of callbacks present 6388 at the beginning of a grace period that allows 6389 the RCU Tasks flavors to collapse back to using 6390 a single callback queue. This switching only 6391 occurs when rcupdate.rcu_task_enqueue_lim is 6392 set to the default value of -1. 6393 6394 rcupdate.rcu_task_contend_lim= [KNL] 6395 Set the minimum number of callback-queuing-time 6396 lock-contention events per jiffy required to 6397 cause the RCU Tasks flavors to switch to per-CPU 6398 callback queuing. This switching only occurs 6399 when rcupdate.rcu_task_enqueue_lim is set to 6400 the default value of -1. 6401 6402 rcupdate.rcu_task_enqueue_lim= [KNL] 6403 Set the number of callback queues to use for the 6404 RCU Tasks family of RCU flavors. The default 6405 of -1 allows this to be automatically (and 6406 dynamically) adjusted. This parameter is intended 6407 for use in testing. 6408 6409 rcupdate.rcu_task_lazy_lim= [KNL] 6410 Number of callbacks on a given CPU that will 6411 cancel laziness on that CPU. Use -1 to disable 6412 cancellation of laziness, but be advised that 6413 doing so increases the danger of OOM due to 6414 callback flooding. 6415 6416 rcupdate.rcu_task_stall_info= [KNL] 6417 Set initial timeout in jiffies for RCU task stall 6418 informational messages, which give some indication 6419 of the problem for those not patient enough to 6420 wait for ten minutes. Informational messages are 6421 only printed prior to the stall-warning message 6422 for a given grace period. Disable with a value 6423 less than or equal to zero. Defaults to ten 6424 seconds. A change in value does not take effect 6425 until the beginning of the next grace period. 6426 6427 rcupdate.rcu_task_stall_info_mult= [KNL] 6428 Multiplier for time interval between successive 6429 RCU task stall informational messages for a given 6430 RCU tasks grace period. This value is clamped 6431 to one through ten, inclusive. It defaults to 6432 the value three, so that the first informational 6433 message is printed 10 seconds into the grace 6434 period, the second at 40 seconds, the third at 6435 160 seconds, and then the stall warning at 600 6436 seconds would prevent a fourth at 640 seconds. 6437 6438 rcupdate.rcu_task_stall_timeout= [KNL] 6439 Set timeout in jiffies for RCU task stall 6440 warning messages. Disable with a value less 6441 than or equal to zero. Defaults to ten minutes. 6442 A change in value does not take effect until 6443 the beginning of the next grace period. 6444 6445 rcupdate.rcu_tasks_lazy_ms= [KNL] 6446 Set timeout in milliseconds RCU Tasks asynchronous 6447 callback batching for call_rcu_tasks(). 6448 A negative value will take the default. A value 6449 of zero will disable batching. Batching is 6450 always disabled for synchronize_rcu_tasks(). 6451 6452 rcupdate.rcu_self_test= [KNL] 6453 Run the RCU early boot self tests 6454 6455 rdinit= [KNL] 6456 Format: <full_path> 6457 Run specified binary instead of /init from the ramdisk, 6458 used for early userspace startup. See initrd. 6459 6460 rdrand= [X86,EARLY] 6461 force - Override the decision by the kernel to hide the 6462 advertisement of RDRAND support (this affects 6463 certain AMD processors because of buggy BIOS 6464 support, specifically around the suspend/resume 6465 path). 6466 6467 rdt= [HW,X86,RDT] 6468 Turn on/off individual RDT features. List is: 6469 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 6470 mba, smba, bmec, abmc, sdciae, energy[:guid], 6471 perf[:guid]. 6472 E.g. to turn on cmt and turn off mba use: 6473 rdt=cmt,!mba 6474 To turn off all energy telemetry monitoring and ensure that 6475 perf telemetry monitoring associated with guid 0x12345 6476 is enabled use: 6477 rdt=!energy,perf:0x12345 6478 6479 reboot= [KNL] 6480 Format (x86 or x86_64): 6481 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \ 6482 [[,]s[mp]#### \ 6483 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 6484 [[,]f[orce] 6485 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 6486 (prefix with 'panic_' to set mode for panic 6487 reboot only), 6488 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 6489 reboot_force is either force or not specified, 6490 reboot_cpu is s[mp]#### with #### being the processor 6491 to be used for rebooting. 6492 6493 acpi 6494 Use the ACPI RESET_REG in the FADT. If ACPI is not 6495 configured or the ACPI reset does not work, the reboot 6496 path attempts the reset using the keyboard controller. 6497 6498 bios 6499 Use the CPU reboot vector for warm reset 6500 6501 cold 6502 Set the cold reboot flag 6503 6504 default 6505 There are some built-in platform specific "quirks" 6506 - you may see: "reboot: <name> series board detected. 6507 Selecting <type> for reboots." In the case where you 6508 think the quirk is in error (e.g. you have newer BIOS, 6509 or newer board) using this option will ignore the 6510 built-in quirk table, and use the generic default 6511 reboot actions. 6512 6513 efi 6514 Use efi reset_system runtime service. If EFI is not 6515 configured or the EFI reset does not work, the reboot 6516 path attempts the reset using the keyboard controller. 6517 6518 force 6519 Don't stop other CPUs on reboot. This can make reboot 6520 more reliable in some cases. 6521 6522 kbd 6523 Use the keyboard controller. cold reset (default) 6524 6525 pci 6526 Use a write to the PCI config space register 0xcf9 to 6527 trigger reboot. 6528 6529 triple 6530 Force a triple fault (init) 6531 6532 warm 6533 Don't set the cold reboot flag 6534 6535 Using warm reset will be much faster especially on big 6536 memory systems because the BIOS will not go through 6537 the memory check. Disadvantage is that not all 6538 hardware will be completely reinitialized on reboot so 6539 there may be boot problems on some systems. 6540 6541 6542 refscale.holdoff= [KNL] 6543 Set test-start holdoff period. The purpose of 6544 this parameter is to delay the start of the 6545 test until boot completes in order to avoid 6546 interference. 6547 6548 refscale.lookup_instances= [KNL] 6549 Number of data elements to use for the forms of 6550 SLAB_TYPESAFE_BY_RCU testing. A negative number 6551 is negated and multiplied by nr_cpu_ids, while 6552 zero specifies nr_cpu_ids. 6553 6554 refscale.loops= [KNL] 6555 Set the number of loops over the synchronization 6556 primitive under test. Increasing this number 6557 reduces noise due to loop start/end overhead, 6558 but the default has already reduced the per-pass 6559 noise to a handful of picoseconds on ca. 2020 6560 x86 laptops. 6561 6562 refscale.nreaders= [KNL] 6563 Set number of readers. The default value of -1 6564 selects N, where N is roughly 75% of the number 6565 of CPUs. A value of zero is an interesting choice. 6566 6567 refscale.nruns= [KNL] 6568 Set number of runs, each of which is dumped onto 6569 the console log. 6570 6571 refscale.readdelay= [KNL] 6572 Set the read-side critical-section duration, 6573 measured in microseconds. 6574 6575 refscale.scale_type= [KNL] 6576 Specify the read-protection implementation to test. 6577 6578 refscale.shutdown= [KNL] 6579 Shut down the system at the end of the performance 6580 test. This defaults to 1 (shut it down) when 6581 refscale is built into the kernel and to 0 (leave 6582 it running) when refscale is built as a module. 6583 6584 refscale.verbose= [KNL] 6585 Enable additional printk() statements. 6586 6587 refscale.verbose_batched= [KNL] 6588 Batch the additional printk() statements. If zero 6589 (the default) or negative, print everything. Otherwise, 6590 print every Nth verbose statement, where N is the value 6591 specified. 6592 6593 regulator_ignore_unused 6594 [REGULATOR] 6595 Prevents regulator framework from disabling regulators 6596 that are unused, due no driver claiming them. This may 6597 be useful for debug and development, but should not be 6598 needed on a platform with proper driver support. 6599 6600 relax_domain_level= 6601 [KNL, SMP] Set scheduler's default relax_domain_level. 6602 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 6603 6604 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 6605 Format: <base1>,<size1>[,<base2>,<size2>,...] 6606 Reserve I/O ports or memory so the kernel won't use 6607 them. If <base> is less than 0x10000, the region 6608 is assumed to be I/O ports; otherwise it is memory. 6609 6610 reserve_mem= [RAM] 6611 Format: nn[KMG]:<align>:<label> 6612 Reserve physical memory and label it with a name that 6613 other subsystems can use to access it. This is typically 6614 used for systems that do not wipe the RAM, and this command 6615 line will try to reserve the same physical memory on 6616 soft reboots. Note, it is not guaranteed to be the same 6617 location. For example, if anything about the system changes 6618 or if booting a different kernel. It can also fail if KASLR 6619 places the kernel at the location of where the RAM reservation 6620 was from a previous boot, the new reservation will be at a 6621 different location. 6622 Any subsystem using this feature must add a way to verify 6623 that the contents of the physical memory is from a previous 6624 boot, as there may be cases where the memory will not be 6625 located at the same location. 6626 6627 The format is size:align:label for example, to request 6628 12 megabytes of 4096 alignment for ramoops: 6629 6630 reserve_mem=12M:4096:oops ramoops.mem_name=oops 6631 6632 reservetop= [X86-32,EARLY] 6633 Format: nn[KMG] 6634 Reserves a hole at the top of the kernel virtual 6635 address space. 6636 6637 reset_devices [KNL] Force drivers to reset the underlying device 6638 during initialization. 6639 6640 resume= [SWSUSP] 6641 Specify the partition device for software suspend 6642 Format: 6643 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 6644 6645 resume_offset= [SWSUSP] 6646 Specify the offset from the beginning of the partition 6647 given by "resume=" at which the swap header is located, 6648 in <PAGE_SIZE> units (needed only for swap files). 6649 See Documentation/power/swsusp-and-swap-files.rst 6650 6651 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 6652 read the resume files 6653 6654 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 6655 Useful for devices that are detected asynchronously 6656 (e.g. USB and MMC devices). 6657 6658 retain_initrd [RAM] Keep initrd memory after extraction. After boot, it will 6659 be accessible via /sys/firmware/initrd. 6660 6661 retbleed= [X86] Control mitigation of RETBleed (Arbitrary 6662 Speculative Code Execution with Return Instructions) 6663 vulnerability. 6664 6665 AMD-based UNRET and IBPB mitigations alone do not stop 6666 sibling threads from influencing the predictions of other 6667 sibling threads. For that reason, STIBP is used on pro- 6668 cessors that support it, and mitigate SMT on processors 6669 that don't. 6670 6671 off - no mitigation 6672 auto - automatically select a mitigation 6673 auto,nosmt - automatically select a mitigation, 6674 disabling SMT if necessary for 6675 the full mitigation (only on Zen1 6676 and older without STIBP). 6677 ibpb - On AMD, mitigate short speculation 6678 windows on basic block boundaries too. 6679 Safe, highest perf impact. It also 6680 enables STIBP if present. Not suitable 6681 on Intel. 6682 ibpb,nosmt - Like "ibpb" above but will disable SMT 6683 when STIBP is not available. This is 6684 the alternative for systems which do not 6685 have STIBP. 6686 unret - Force enable untrained return thunks, 6687 only effective on AMD f15h-f17h based 6688 systems. 6689 unret,nosmt - Like unret, but will disable SMT when STIBP 6690 is not available. This is the alternative for 6691 systems which do not have STIBP. 6692 6693 Selecting 'auto' will choose a mitigation method at run 6694 time according to the CPU. 6695 6696 Not specifying this option is equivalent to retbleed=auto. 6697 6698 rfkill.default_state= 6699 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 6700 etc. communication is blocked by default. 6701 1 Unblocked. 6702 6703 rfkill.master_switch_mode= 6704 0 The "airplane mode" button does nothing. 6705 1 The "airplane mode" button toggles between everything 6706 blocked and the previous configuration. 6707 2 The "airplane mode" button toggles between everything 6708 blocked and everything unblocked. 6709 6710 ring3mwait=disable 6711 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 6712 CPUs. 6713 6714 riscv_isa_fallback [RISCV,EARLY] 6715 When CONFIG_RISCV_ISA_FALLBACK is not enabled, permit 6716 falling back to detecting extension support by parsing 6717 "riscv,isa" property on devicetree systems when the 6718 replacement properties are not found. See the Kconfig 6719 entry for RISCV_ISA_FALLBACK. 6720 6721 riscv_nousercfi= 6722 all Disable user CFI ABI to userspace even if cpu extension 6723 are available. 6724 bcfi Disable user backward CFI ABI to userspace even if 6725 the shadow stack extension is available. 6726 fcfi Disable user forward CFI ABI to userspace even if the 6727 landing pad extension is available. 6728 6729 ro [KNL] Mount root device read-only on boot 6730 6731 rodata= [KNL,EARLY] 6732 on Mark read-only kernel memory as read-only (default). 6733 off Leave read-only kernel memory writable for debugging. 6734 noalias Mark read-only kernel memory as read-only but retain 6735 writable aliases in the direct map for regions outside 6736 of the kernel image. [arm64] 6737 6738 rockchip.usb_uart 6739 [EARLY] 6740 Enable the uart passthrough on the designated usb port 6741 on Rockchip SoCs. When active, the signals of the 6742 debug-uart get routed to the D+ and D- pins of the usb 6743 port and the regular usb controller gets disabled. 6744 6745 root= [KNL] Root filesystem 6746 Usually this is a block device specifier of some kind, 6747 see the early_lookup_bdev comment in 6748 block/early-lookup.c for details. 6749 Alternatively this can be "ram" for the legacy initial 6750 ramdisk, "nfs" and "cifs" for root on a network file 6751 system, or "mtd" and "ubi" for mounting from raw flash. 6752 6753 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 6754 mount the root filesystem 6755 6756 rootflags= [KNL] Set root filesystem mount option string 6757 6758 rseq_slice_ext= [KNL] RSEQ based time slice extension 6759 Format: boolean 6760 Control enablement of RSEQ based time slice extension. 6761 Default is 'on'. 6762 6763 initramfs_options= [KNL] 6764 Specify mount options for for the initramfs mount. 6765 6766 rootfstype= [KNL] Set root filesystem type 6767 6768 rootwait [KNL] Wait (indefinitely) for root device to show up. 6769 Useful for devices that are detected asynchronously 6770 (e.g. USB and MMC devices). 6771 6772 rootwait= [KNL] Maximum time (in seconds) to wait for root device 6773 to show up before attempting to mount the root 6774 filesystem. 6775 6776 rproc_mem=nn[KMG][@address] 6777 [KNL,ARM,CMA] Remoteproc physical memory block. 6778 Memory area to be used by remote processor image, 6779 managed by CMA. 6780 6781 rseq_debug= [KNL] Enable or disable restartable sequence 6782 debug mode. Defaults to CONFIG_RSEQ_DEBUG_DEFAULT_ENABLE. 6783 Format: <bool> 6784 6785 rt_group_sched= [KNL] Enable or disable SCHED_RR/FIFO group scheduling 6786 when CONFIG_RT_GROUP_SCHED=y. Defaults to 6787 !CONFIG_RT_GROUP_SCHED_DEFAULT_DISABLED. 6788 Format: <bool> 6789 6790 rw [KNL] Mount root device read-write on boot 6791 6792 S [KNL] Run init in single mode 6793 6794 s390_iommu= [HW,S390] 6795 Set s390 IOTLB flushing mode 6796 strict 6797 With strict flushing every unmap operation will result 6798 in an IOTLB flush. Default is lazy flushing before 6799 reuse, which is faster. Deprecated, equivalent to 6800 iommu.strict=1. 6801 6802 s390_iommu_aperture= [KNL,S390] 6803 Specifies the size of the per device DMA address space 6804 accessible through the DMA and IOMMU APIs as a decimal 6805 factor of the size of main memory. 6806 The default is 1 meaning that one can concurrently use 6807 as many DMA addresses as physical memory is installed, 6808 if supported by hardware, and thus map all of memory 6809 once. With a value of 2 one can map all of memory twice 6810 and so on. As a special case a factor of 0 imposes no 6811 restrictions other than those given by hardware at the 6812 cost of significant additional memory use for tables. 6813 6814 sa1100ir [NET] 6815 See drivers/net/irda/sa1100_ir.c. 6816 6817 sched_proxy_exec= [KNL] 6818 Enables or disables "proxy execution" style 6819 solution to mutex-based priority inversion. 6820 Format: <bool> 6821 6822 sched_verbose [KNL,EARLY] Enables verbose scheduler debug messages. 6823 6824 schedstats= [KNL,X86] Enable or disable scheduled statistics. 6825 Allowed values are enable and disable. This feature 6826 incurs a small amount of overhead in the scheduler 6827 but is useful for debugging and performance tuning. 6828 6829 sched_thermal_decay_shift= 6830 [Deprecated] 6831 [KNL, SMP] Set a decay shift for scheduler thermal 6832 pressure signal. Thermal pressure signal follows the 6833 default decay period of other scheduler pelt 6834 signals(usually 32 ms but configurable). Setting 6835 sched_thermal_decay_shift will left shift the decay 6836 period for the thermal pressure signal by the shift 6837 value. 6838 i.e. with the default pelt decay period of 32 ms 6839 sched_thermal_decay_shift thermal pressure decay pr 6840 1 64 ms 6841 2 128 ms 6842 and so on. 6843 Format: integer between 0 and 10 6844 Default is 0. 6845 6846 scftorture.holdoff= [KNL] 6847 Number of seconds to hold off before starting 6848 test. Defaults to zero for module insertion and 6849 to 10 seconds for built-in smp_call_function() 6850 tests. 6851 6852 scftorture.longwait= [KNL] 6853 Request ridiculously long waits randomly selected 6854 up to the chosen limit in seconds. Zero (the 6855 default) disables this feature. Please note 6856 that requesting even small non-zero numbers of 6857 seconds can result in RCU CPU stall warnings, 6858 softlockup complaints, and so on. 6859 6860 scftorture.nthreads= [KNL] 6861 Number of kthreads to spawn to invoke the 6862 smp_call_function() family of functions. 6863 The default of -1 specifies a number of kthreads 6864 equal to the number of CPUs. 6865 6866 scftorture.onoff_holdoff= [KNL] 6867 Number seconds to wait after the start of the 6868 test before initiating CPU-hotplug operations. 6869 6870 scftorture.onoff_interval= [KNL] 6871 Number seconds to wait between successive 6872 CPU-hotplug operations. Specifying zero (which 6873 is the default) disables CPU-hotplug operations. 6874 6875 scftorture.shutdown_secs= [KNL] 6876 The number of seconds following the start of the 6877 test after which to shut down the system. The 6878 default of zero avoids shutting down the system. 6879 Non-zero values are useful for automated tests. 6880 6881 scftorture.stat_interval= [KNL] 6882 The number of seconds between outputting the 6883 current test statistics to the console. A value 6884 of zero disables statistics output. 6885 6886 scftorture.stutter_cpus= [KNL] 6887 The number of jiffies to wait between each change 6888 to the set of CPUs under test. 6889 6890 scftorture.use_cpus_read_lock= [KNL] 6891 Use use_cpus_read_lock() instead of the default 6892 preempt_disable() to disable CPU hotplug 6893 while invoking one of the smp_call_function*() 6894 functions. 6895 6896 scftorture.verbose= [KNL] 6897 Enable additional printk() statements. 6898 6899 scftorture.weight_single= [KNL] 6900 The probability weighting to use for the 6901 smp_call_function_single() function with a zero 6902 "wait" parameter. A value of -1 selects the 6903 default if all other weights are -1. However, 6904 if at least one weight has some other value, a 6905 value of -1 will instead select a weight of zero. 6906 6907 scftorture.weight_single_wait= [KNL] 6908 The probability weighting to use for the 6909 smp_call_function_single() function with a 6910 non-zero "wait" parameter. See weight_single. 6911 6912 scftorture.weight_many= [KNL] 6913 The probability weighting to use for the 6914 smp_call_function_many() function with a zero 6915 "wait" parameter. See weight_single. 6916 Note well that setting a high probability for 6917 this weighting can place serious IPI load 6918 on the system. 6919 6920 scftorture.weight_many_wait= [KNL] 6921 The probability weighting to use for the 6922 smp_call_function_many() function with a 6923 non-zero "wait" parameter. See weight_single 6924 and weight_many. 6925 6926 scftorture.weight_all= [KNL] 6927 The probability weighting to use for the 6928 smp_call_function_all() function with a zero 6929 "wait" parameter. See weight_single and 6930 weight_many. 6931 6932 scftorture.weight_all_wait= [KNL] 6933 The probability weighting to use for the 6934 smp_call_function_all() function with a 6935 non-zero "wait" parameter. See weight_single 6936 and weight_many. 6937 6938 sdw_mclk_divider=[SDW] 6939 Specify the MCLK divider for Intel SoundWire buses in 6940 case the BIOS does not provide the clock rate properly. 6941 6942 skew_tick= [KNL,EARLY] Offset the periodic timer tick per cpu to mitigate 6943 xtime_lock contention on larger systems, and/or RCU lock 6944 contention on all systems with CONFIG_MAXSMP set. 6945 Format: { "0" | "1" } 6946 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 6947 1 -- enable. 6948 Note: increases power consumption, thus should only be 6949 enabled if running jitter sensitive (HPC/RT) workloads. 6950 6951 security= [SECURITY] Choose a legacy "major" security module to 6952 enable at boot. This has been deprecated by the 6953 "lsm=" parameter. 6954 6955 selinux= [SELINUX] Disable or enable SELinux at boot time. 6956 Format: { "0" | "1" } 6957 See security/selinux/Kconfig help text. 6958 0 -- disable. 6959 1 -- enable. 6960 Default value is 1. 6961 6962 serialnumber [BUGS=X86-32] 6963 6964 sev=option[,option...] [X86-64] 6965 6966 debug 6967 Enable debug messages. 6968 6969 nosnp 6970 Do not enable SEV-SNP (applies to host/hypervisor 6971 only). Setting 'nosnp' avoids the RMP check overhead 6972 in memory accesses when users do not want to run 6973 SEV-SNP guests. 6974 6975 shapers= [NET] 6976 Maximal number of shapers. 6977 6978 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 6979 Limit apic dumping. The parameter defines the maximal 6980 number of local apics being dumped. Also it is possible 6981 to set it to "all" by meaning -- no limit here. 6982 Format: { 1 (default) | 2 | ... | all }. 6983 The parameter valid if only apic=debug or 6984 apic=verbose is specified. 6985 Example: apic=debug show_lapic=all 6986 6987 slab_debug[=options[,slabs][;[options[,slabs]]...] [MM] 6988 Enabling slab_debug allows one to determine the 6989 culprit if slab objects become corrupted. Enabling 6990 slab_debug can create guard zones around objects and 6991 may poison objects when not in use. Also tracks the 6992 last alloc / free. For more information see 6993 Documentation/admin-guide/mm/slab.rst. 6994 (slub_debug legacy name also accepted for now) 6995 6996 Using this option implies the "no_hash_pointers" 6997 option which can be undone by adding the 6998 "hash_pointers=always" option. 6999 7000 slab_max_order= [MM] 7001 Determines the maximum allowed order for slabs. 7002 A high setting may cause OOMs due to memory 7003 fragmentation. For more information see 7004 Documentation/admin-guide/mm/slab.rst. 7005 (slub_max_order legacy name also accepted for now) 7006 7007 slab_merge [MM] 7008 Enable merging of slabs with similar size when the 7009 kernel is built without CONFIG_SLAB_MERGE_DEFAULT. 7010 (slub_merge legacy name also accepted for now) 7011 7012 slab_min_objects= [MM] 7013 The minimum number of objects per slab. SLUB will 7014 increase the slab order up to slab_max_order to 7015 generate a sufficiently large slab able to contain 7016 the number of objects indicated. The higher the number 7017 of objects the smaller the overhead of tracking slabs 7018 and the less frequently locks need to be acquired. 7019 For more information see 7020 Documentation/admin-guide/mm/slab.rst. 7021 (slub_min_objects legacy name also accepted for now) 7022 7023 slab_min_order= [MM] 7024 Determines the minimum page order for slabs. Must be 7025 lower or equal to slab_max_order. For more information see 7026 Documentation/admin-guide/mm/slab.rst. 7027 (slub_min_order legacy name also accepted for now) 7028 7029 slab_nomerge [MM] 7030 Disable merging of slabs with similar size. May be 7031 necessary if there is some reason to distinguish 7032 allocs to different slabs, especially in hardened 7033 environments where the risk of heap overflows and 7034 layout control by attackers can usually be 7035 frustrated by disabling merging. This will reduce 7036 most of the exposure of a heap attack to a single 7037 cache (risks via metadata attacks are mostly 7038 unchanged). Debug options disable merging on their 7039 own. 7040 For more information see 7041 Documentation/admin-guide/mm/slab.rst. 7042 (slub_nomerge legacy name also accepted for now) 7043 7044 slab_strict_numa [MM] 7045 Support memory policies on a per object level 7046 in the slab allocator. The default is for memory 7047 policies to be applied at the folio level when 7048 a new folio is needed or a partial folio is 7049 retrieved from the lists. Increases overhead 7050 in the slab fastpaths but gains more accurate 7051 NUMA kernel object placement which helps with slow 7052 interconnects in NUMA systems. 7053 7054 slram= [HW,MTD] 7055 7056 smart2= [HW] 7057 Format: <io1>[,<io2>[,...,<io8>]] 7058 7059 smp.csd_lock_timeout= [KNL] 7060 Specify the period of time in milliseconds 7061 that smp_call_function() and friends will wait 7062 for a CPU to release the CSD lock. This is 7063 useful when diagnosing bugs involving CPUs 7064 disabling interrupts for extended periods 7065 of time. Defaults to 5,000 milliseconds, and 7066 setting a value of zero disables this feature. 7067 This feature may be more efficiently disabled 7068 using the csdlock_debug- kernel parameter. 7069 7070 smp.panic_on_ipistall= [KNL] 7071 If a csd_lock_timeout extends for more than 7072 the specified number of milliseconds, panic the 7073 system. By default, let CSD-lock acquisition 7074 take as long as they take. Specifying 300,000 7075 for this value provides a 5-minute timeout. 7076 7077 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 7078 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 7079 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 7080 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 7081 smsc-ircc2.ircc_irq= [HW] IRQ line 7082 smsc-ircc2.ircc_dma= [HW] DMA channel 7083 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 7084 0: Toshiba Satellite 1800 (GP data pin select) 7085 1: Fast pin select (default) 7086 2: ATC IRMode 7087 7088 smt= [KNL,MIPS,S390,EARLY] Set the maximum number of threads 7089 (logical CPUs) to use per physical CPU on systems 7090 capable of symmetric multithreading (SMT). Will 7091 be capped to the actual hardware limit. 7092 Format: <integer> 7093 Default: -1 (no limit) 7094 7095 softlockup_panic= 7096 [KNL] Should the soft-lockup detector generate panics. 7097 Format: <int> 7098 7099 A value of non-zero instructs the soft-lockup detector 7100 to panic the machine when a soft-lockup duration exceeds 7101 N thresholds. It is also controlled by the kernel.softlockup_panic 7102 sysctl and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 7103 respective build-time switch to that functionality. 7104 7105 softlockup_all_cpu_backtrace= 7106 [KNL] Should the soft-lockup detector generate 7107 backtraces on all cpus. 7108 Format: 0 | 1 7109 7110 sonypi.*= [HW] Sony Programmable I/O Control Device driver 7111 See Documentation/admin-guide/laptops/sonypi.rst 7112 7113 spectre_bhi= [X86] Control mitigation of Branch History Injection 7114 (BHI) vulnerability. This setting affects the 7115 deployment of the HW BHI control and the SW BHB 7116 clearing sequence. 7117 7118 on - (default) Enable the HW or SW mitigation as 7119 needed. This protects the kernel from 7120 both syscalls and VMs. 7121 vmexit - On systems which don't have the HW mitigation 7122 available, enable the SW mitigation on vmexit 7123 ONLY. On such systems, the host kernel is 7124 protected from VM-originated BHI attacks, but 7125 may still be vulnerable to syscall attacks. 7126 off - Disable the mitigation. 7127 7128 spectre_v2= [X86,EARLY] Control mitigation of Spectre variant 2 7129 (indirect branch speculation) vulnerability. 7130 The default operation protects the kernel from 7131 user space attacks. 7132 7133 on - unconditionally enable, implies 7134 spectre_v2_user=on 7135 off - unconditionally disable, implies 7136 spectre_v2_user=off 7137 auto - kernel detects whether your CPU model is 7138 vulnerable 7139 7140 Selecting 'on' will, and 'auto' may, choose a 7141 mitigation method at run time according to the 7142 CPU, the available microcode, the setting of the 7143 CONFIG_MITIGATION_RETPOLINE configuration option, 7144 and the compiler with which the kernel was built. 7145 7146 Selecting 'on' will also enable the mitigation 7147 against user space to user space task attacks. 7148 Selecting specific mitigation does not force enable 7149 user mitigations. 7150 7151 Selecting 'off' will disable both the kernel and 7152 the user space protections. 7153 7154 Specific mitigations can also be selected manually: 7155 7156 retpoline - replace indirect branches 7157 retpoline,generic - Retpolines 7158 retpoline,lfence - LFENCE; indirect branch 7159 retpoline,amd - alias for retpoline,lfence 7160 eibrs - Enhanced/Auto IBRS 7161 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines 7162 eibrs,lfence - Enhanced/Auto IBRS + LFENCE 7163 ibrs - use IBRS to protect kernel 7164 7165 Not specifying this option is equivalent to 7166 spectre_v2=auto. 7167 7168 spectre_v2_user= 7169 [X86] Control mitigation of Spectre variant 2 7170 (indirect branch speculation) vulnerability between 7171 user space tasks 7172 7173 on - Unconditionally enable mitigations. Is 7174 enforced by spectre_v2=on 7175 7176 off - Unconditionally disable mitigations. Is 7177 enforced by spectre_v2=off 7178 7179 prctl - Indirect branch speculation is enabled, 7180 but mitigation can be enabled via prctl 7181 per thread. The mitigation control state 7182 is inherited on fork. 7183 7184 prctl,ibpb 7185 - Like "prctl" above, but only STIBP is 7186 controlled per thread. IBPB is issued 7187 always when switching between different user 7188 space processes. 7189 7190 seccomp 7191 - Same as "prctl" above, but all seccomp 7192 threads will enable the mitigation unless 7193 they explicitly opt out. 7194 7195 seccomp,ibpb 7196 - Like "seccomp" above, but only STIBP is 7197 controlled per thread. IBPB is issued 7198 always when switching between different 7199 user space processes. 7200 7201 auto - Kernel selects the mitigation depending on 7202 the available CPU features and vulnerability. 7203 7204 Default mitigation: "prctl" 7205 7206 Not specifying this option is equivalent to 7207 spectre_v2_user=auto. 7208 7209 spec_rstack_overflow= 7210 [X86,EARLY] Control RAS overflow mitigation on AMD Zen CPUs 7211 7212 off - Disable mitigation 7213 microcode - Enable microcode mitigation only 7214 safe-ret - Enable sw-only safe RET mitigation (default) 7215 ibpb - Enable mitigation by issuing IBPB on 7216 kernel entry 7217 ibpb-vmexit - Issue IBPB only on VMEXIT 7218 (cloud-specific mitigation) 7219 7220 spec_store_bypass_disable= 7221 [HW,EARLY] Control Speculative Store Bypass (SSB) Disable mitigation 7222 (Speculative Store Bypass vulnerability) 7223 7224 Certain CPUs are vulnerable to an exploit against a 7225 a common industry wide performance optimization known 7226 as "Speculative Store Bypass" in which recent stores 7227 to the same memory location may not be observed by 7228 later loads during speculative execution. The idea 7229 is that such stores are unlikely and that they can 7230 be detected prior to instruction retirement at the 7231 end of a particular speculation execution window. 7232 7233 In vulnerable processors, the speculatively forwarded 7234 store can be used in a cache side channel attack, for 7235 example to read memory to which the attacker does not 7236 directly have access (e.g. inside sandboxed code). 7237 7238 This parameter controls whether the Speculative Store 7239 Bypass optimization is used. 7240 7241 On x86 the options are: 7242 7243 on - Unconditionally disable Speculative Store Bypass 7244 off - Unconditionally enable Speculative Store Bypass 7245 auto - Kernel detects whether the CPU model contains an 7246 implementation of Speculative Store Bypass and 7247 picks the most appropriate mitigation. If the 7248 CPU is not vulnerable, "off" is selected. If the 7249 CPU is vulnerable the default mitigation is 7250 architecture and Kconfig dependent. See below. 7251 prctl - Control Speculative Store Bypass per thread 7252 via prctl. Speculative Store Bypass is enabled 7253 for a process by default. The state of the control 7254 is inherited on fork. 7255 seccomp - Same as "prctl" above, but all seccomp threads 7256 will disable SSB unless they explicitly opt out. 7257 7258 Default mitigations: 7259 X86: "prctl" 7260 7261 On powerpc the options are: 7262 7263 on,auto - On Power8 and Power9 insert a store-forwarding 7264 barrier on kernel entry and exit. On Power7 7265 perform a software flush on kernel entry and 7266 exit. 7267 off - No action. 7268 7269 Not specifying this option is equivalent to 7270 spec_store_bypass_disable=auto. 7271 7272 split_lock_detect= 7273 [X86] Enable split lock detection or bus lock detection 7274 7275 When enabled (and if hardware support is present), atomic 7276 instructions that access data across cache line 7277 boundaries will result in an alignment check exception 7278 for split lock detection or a debug exception for 7279 bus lock detection. 7280 7281 off - not enabled 7282 7283 warn - the kernel will emit rate-limited warnings 7284 about applications triggering the #AC 7285 exception or the #DB exception. This mode is 7286 the default on CPUs that support split lock 7287 detection or bus lock detection. Default 7288 behavior is by #AC if both features are 7289 enabled in hardware. 7290 7291 fatal - the kernel will send SIGBUS to applications 7292 that trigger the #AC exception or the #DB 7293 exception. Default behavior is by #AC if 7294 both features are enabled in hardware. 7295 7296 ratelimit:N - 7297 Set system wide rate limit to N bus locks 7298 per second for bus lock detection. 7299 0 < N <= 1000. 7300 7301 N/A for split lock detection. 7302 7303 7304 If an #AC exception is hit in the kernel or in 7305 firmware (i.e. not while executing in user mode) 7306 the kernel will oops in either "warn" or "fatal" 7307 mode. 7308 7309 #DB exception for bus lock is triggered only when 7310 CPL > 0. 7311 7312 srbds= [X86,INTEL,EARLY] 7313 Control the Special Register Buffer Data Sampling 7314 (SRBDS) mitigation. 7315 7316 Certain CPUs are vulnerable to an MDS-like 7317 exploit which can leak bits from the random 7318 number generator. 7319 7320 By default, this issue is mitigated by 7321 microcode. However, the microcode fix can cause 7322 the RDRAND and RDSEED instructions to become 7323 much slower. Among other effects, this will 7324 result in reduced throughput from /dev/urandom. 7325 7326 The microcode mitigation can be disabled with 7327 the following option: 7328 7329 off: Disable mitigation and remove 7330 performance impact to RDRAND and RDSEED 7331 7332 srcutree.big_cpu_lim [KNL] 7333 Specifies the number of CPUs constituting a 7334 large system, such that srcu_struct structures 7335 should immediately allocate an srcu_node array. 7336 This kernel-boot parameter defaults to 128, 7337 but takes effect only when the low-order four 7338 bits of srcutree.convert_to_big is equal to 3 7339 (decide at boot). 7340 7341 srcutree.convert_to_big [KNL] 7342 Specifies under what conditions an SRCU tree 7343 srcu_struct structure will be converted to big 7344 form, that is, with an rcu_node tree: 7345 7346 0: Never. 7347 1: At init_srcu_struct() time. 7348 2: When rcutorture decides to. 7349 3: Decide at boot time (default). 7350 0x1X: Above plus if high contention. 7351 7352 Either way, the srcu_node tree will be sized based 7353 on the actual runtime number of CPUs (nr_cpu_ids) 7354 instead of the compile-time CONFIG_NR_CPUS. 7355 7356 srcutree.counter_wrap_check [KNL] 7357 Specifies how frequently to check for 7358 grace-period sequence counter wrap for the 7359 srcu_data structure's ->srcu_gp_seq_needed field. 7360 The greater the number of bits set in this kernel 7361 parameter, the less frequently counter wrap will 7362 be checked for. Note that the bottom two bits 7363 are ignored. 7364 7365 srcutree.exp_holdoff [KNL] 7366 Specifies how many nanoseconds must elapse 7367 since the end of the last SRCU grace period for 7368 a given srcu_struct until the next normal SRCU 7369 grace period will be considered for automatic 7370 expediting. Set to zero to disable automatic 7371 expediting. 7372 7373 srcutree.srcu_max_nodelay [KNL] 7374 Specifies the number of no-delay instances 7375 per jiffy for which the SRCU grace period 7376 worker thread will be rescheduled with zero 7377 delay. Beyond this limit, worker thread will 7378 be rescheduled with a sleep delay of one jiffy. 7379 7380 srcutree.srcu_max_nodelay_phase [KNL] 7381 Specifies the per-grace-period phase, number of 7382 non-sleeping polls of readers. Beyond this limit, 7383 grace period worker thread will be rescheduled 7384 with a sleep delay of one jiffy, between each 7385 rescan of the readers, for a grace period phase. 7386 7387 srcutree.srcu_retry_check_delay [KNL] 7388 Specifies number of microseconds of non-sleeping 7389 delay between each non-sleeping poll of readers. 7390 7391 srcutree.small_contention_lim [KNL] 7392 Specifies the number of update-side contention 7393 events per jiffy will be tolerated before 7394 initiating a conversion of an srcu_struct 7395 structure to big form. Note that the value of 7396 srcutree.convert_to_big must have the 0x10 bit 7397 set for contention-based conversions to occur. 7398 7399 ssbd= [ARM64,HW,EARLY] 7400 Speculative Store Bypass Disable control 7401 7402 On CPUs that are vulnerable to the Speculative 7403 Store Bypass vulnerability and offer a 7404 firmware based mitigation, this parameter 7405 indicates how the mitigation should be used: 7406 7407 force-on: Unconditionally enable mitigation for 7408 for both kernel and userspace 7409 force-off: Unconditionally disable mitigation for 7410 for both kernel and userspace 7411 kernel: Always enable mitigation in the 7412 kernel, and offer a prctl interface 7413 to allow userspace to register its 7414 interest in being mitigated too. 7415 7416 stack_guard_gap= [MM] 7417 override the default stack gap protection. The value 7418 is in page units and it defines how many pages prior 7419 to (for stacks growing down) resp. after (for stacks 7420 growing up) the main stack are reserved for no other 7421 mapping. Default value is 256 pages. 7422 7423 stack_depot_disable= [KNL,EARLY] 7424 Setting this to true through kernel command line will 7425 disable the stack depot thereby saving the static memory 7426 consumed by the stack hash table. By default this is set 7427 to false. 7428 7429 stack_depot_max_pools= [KNL,EARLY] 7430 Specify the maximum number of pools to use for storing 7431 stack traces. Pools are allocated on-demand up to this 7432 limit. Default value is 8191 pools. 7433 7434 stacktrace [FTRACE] 7435 Enable the stack tracer on boot up. 7436 7437 stacktrace_filter=[function-list] 7438 [FTRACE] Limit the functions that the stack tracer 7439 will trace at boot up. function-list is a comma-separated 7440 list of functions. This list can be changed at run 7441 time by the stack_trace_filter file in the debugfs 7442 tracing directory. Note, this enables stack tracing 7443 and the stacktrace above is not needed. 7444 7445 sti= [PARISC,HW] 7446 Format: <num> 7447 Set the STI (builtin display/keyboard on the HP-PARISC 7448 machines) console (graphic card) which should be used 7449 as the initial boot-console. 7450 See also comment in drivers/video/console/sticore.c. 7451 7452 sti_font= [HW] 7453 See comment in drivers/video/console/sticore.c. 7454 7455 stifb= [HW] 7456 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 7457 7458 strict_sas_size= 7459 [X86] 7460 Format: <bool> 7461 Enable or disable strict sigaltstack size checks 7462 against the required signal frame size which 7463 depends on the supported FPU features. This can 7464 be used to filter out binaries which have 7465 not yet been made aware of AT_MINSIGSTKSZ. 7466 7467 stress_hpt [PPC,EARLY] 7468 Limits the number of kernel HPT entries in the hash 7469 page table to increase the rate of hash page table 7470 faults on kernel addresses. 7471 7472 stress_slb [PPC,EARLY] 7473 Limits the number of kernel SLB entries, and flushes 7474 them frequently to increase the rate of SLB faults 7475 on kernel addresses. 7476 7477 no_slb_preload [PPC,EARLY] 7478 Disables slb preloading for userspace. 7479 7480 sunrpc.min_resvport= 7481 sunrpc.max_resvport= 7482 [NFS,SUNRPC] 7483 SunRPC servers often require that client requests 7484 originate from a privileged port (i.e. a port in the 7485 range 0 < portnr < 1024). 7486 An administrator who wishes to reserve some of these 7487 ports for other uses may adjust the range that the 7488 kernel's sunrpc client considers to be privileged 7489 using these two parameters to set the minimum and 7490 maximum port values. 7491 7492 sunrpc.svc_rpc_per_connection_limit= 7493 [NFS,SUNRPC] 7494 Limit the number of requests that the server will 7495 process in parallel from a single connection. 7496 The default value is 0 (no limit). 7497 7498 sunrpc.pool_mode= 7499 [NFS] 7500 Control how the NFS server code allocates CPUs to 7501 service thread pools. Depending on how many NICs 7502 you have and where their interrupts are bound, this 7503 option will affect which CPUs will do NFS serving. 7504 Note: this parameter cannot be changed while the 7505 NFS server is running. 7506 7507 auto the server chooses an appropriate mode 7508 automatically using heuristics 7509 global a single global pool contains all CPUs 7510 percpu one pool for each CPU 7511 pernode one pool for each NUMA node (equivalent 7512 to global on non-NUMA machines) 7513 7514 sunrpc.tcp_slot_table_entries= 7515 sunrpc.udp_slot_table_entries= 7516 [NFS,SUNRPC] 7517 Sets the upper limit on the number of simultaneous 7518 RPC calls that can be sent from the client to a 7519 server. Increasing these values may allow you to 7520 improve throughput, but will also increase the 7521 amount of memory reserved for use by the client. 7522 7523 suspend.pm_test_delay= 7524 [SUSPEND] 7525 Sets the number of seconds to remain in a suspend test 7526 mode before resuming the system (see 7527 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 7528 is set. Default value is 5. 7529 7530 svm= [PPC] 7531 Format: { on | off | y | n | 1 | 0 } 7532 This parameter controls use of the Protected 7533 Execution Facility on pSeries. 7534 7535 swiotlb= [ARM,PPC,MIPS,X86,S390,EARLY] 7536 Format: { <int> [,<int>] | force | noforce } 7537 <int> -- Number of I/O TLB slabs 7538 <int> -- Second integer after comma. Number of swiotlb 7539 areas with their own lock. Will be rounded up 7540 to a power of 2. 7541 force -- force using of bounce buffers even if they 7542 wouldn't be automatically used by the kernel 7543 noforce -- Never use bounce buffers (for debugging) 7544 7545 switches= [HW,M68k,EARLY] 7546 7547 sysctl.*= [KNL] 7548 Set a sysctl parameter, right before loading the init 7549 process, as if the value was written to the respective 7550 /proc/sys/... file. Both '.' and '/' are recognized as 7551 separators. Unrecognized parameters and invalid values 7552 are reported in the kernel log. Sysctls registered 7553 later by a loaded module cannot be set this way. 7554 Example: sysctl.vm.swappiness=40 7555 7556 sysrq_always_enabled 7557 [KNL] 7558 Ignore sysrq setting - this boot parameter will 7559 neutralize any effect of /proc/sys/kernel/sysrq. 7560 Useful for debugging. 7561 7562 tcpmhash_entries= [KNL,NET] 7563 Set the number of tcp_metrics_hash slots. 7564 Default value is 8192 or 16384 depending on total 7565 ram pages. This is used to specify the TCP metrics 7566 cache size. See Documentation/networking/ip-sysctl.rst 7567 "tcp_no_metrics_save" section for more details. 7568 7569 tdfx= [HW,DRM] 7570 7571 test_suspend= [SUSPEND] 7572 Format: { "mem" | "standby" | "freeze" }[,N] 7573 Specify "mem" (for Suspend-to-RAM) or "standby" (for 7574 standby suspend) or "freeze" (for suspend type freeze) 7575 as the system sleep state during system startup with 7576 the optional capability to repeat N number of times. 7577 The system is woken from this state using a 7578 wakeup-capable RTC alarm. 7579 7580 thash_entries= [KNL,NET] 7581 Set number of hash buckets for TCP connection 7582 7583 thermal.act= [HW,ACPI] 7584 -1: disable all active trip points in all thermal zones 7585 <degrees C>: override all lowest active trip points 7586 7587 thermal.crt= [HW,ACPI] 7588 -1: disable all critical trip points in all thermal zones 7589 <degrees C>: override all critical trip points 7590 7591 thermal.off= [HW,ACPI] 7592 1: disable ACPI thermal control 7593 7594 thermal.psv= [HW,ACPI] 7595 -1: disable all passive trip points 7596 <degrees C>: override all passive trip points to this 7597 value 7598 7599 thermal.tzp= [HW,ACPI] 7600 Specify global default ACPI thermal zone polling rate 7601 <deci-seconds>: poll all this frequency 7602 0: no polling (default) 7603 7604 thp_anon= [KNL] 7605 Format: <size>[KMG],<size>[KMG]:<state>;<size>[KMG]-<size>[KMG]:<state> 7606 state is one of "always", "madvise", "never" or "inherit". 7607 Control the default behavior of the system with respect 7608 to anonymous transparent hugepages. 7609 Can be used multiple times for multiple anon THP sizes. 7610 See Documentation/admin-guide/mm/transhuge.rst for more 7611 details. 7612 7613 threadirqs [KNL,EARLY] 7614 Force threading of all interrupt handlers except those 7615 marked explicitly IRQF_NO_THREAD. 7616 7617 thp_shmem= [KNL] 7618 Format: <size>[KMG],<size>[KMG]:<policy>;<size>[KMG]-<size>[KMG]:<policy> 7619 Control the default policy of each hugepage size for the 7620 internal shmem mount. <policy> is one of policies available 7621 for the shmem mount ("always", "inherit", "never", "within_size", 7622 and "advise"). 7623 It can be used multiple times for multiple shmem THP sizes. 7624 See Documentation/admin-guide/mm/transhuge.rst for more 7625 details. 7626 7627 topology= [S390,EARLY] 7628 Format: {off | on} 7629 Specify if the kernel should make use of the cpu 7630 topology information if the hardware supports this. 7631 The scheduler will make use of this information and 7632 e.g. base its process migration decisions on it. 7633 Default is on. 7634 7635 torture.disable_onoff_at_boot= [KNL] 7636 Prevent the CPU-hotplug component of torturing 7637 until after init has spawned. 7638 7639 torture.ftrace_dump_at_shutdown= [KNL] 7640 Dump the ftrace buffer at torture-test shutdown, 7641 even if there were no errors. This can be a 7642 very costly operation when many torture tests 7643 are running concurrently, especially on systems 7644 with rotating-rust storage. 7645 7646 torture.verbose_sleep_frequency= [KNL] 7647 Specifies how many verbose printk()s should be 7648 emitted between each sleep. The default of zero 7649 disables verbose-printk() sleeping. 7650 7651 torture.verbose_sleep_duration= [KNL] 7652 Duration of each verbose-printk() sleep in jiffies. 7653 7654 tpm.disable_pcr_integrity= [HW,TPM] 7655 Do not protect PCR registers from unintended physical 7656 access, or interposers in the bus by the means of 7657 having an integrity protected session wrapped around 7658 TPM2_PCR_Extend command. Consider this in a situation 7659 where TPM is heavily utilized by IMA, thus protection 7660 causing a major performance hit, and the space where 7661 machines are deployed is by other means guarded. 7662 7663 tpm_crb_ffa.busy_timeout_ms= [ARM64,TPM] 7664 Maximum time in milliseconds to retry sending a message 7665 to the TPM service before giving up. This parameter controls 7666 how long the system will continue retrying when the TPM 7667 service is busy. 7668 Format: <unsigned int> 7669 Default: 2000 (2 seconds) 7670 7671 tpm_suspend_pcr=[HW,TPM] 7672 Format: integer pcr id 7673 Specify that at suspend time, the tpm driver 7674 should extend the specified pcr with zeros, 7675 as a workaround for some chips which fail to 7676 flush the last written pcr on TPM_SaveState. 7677 This will guarantee that all the other pcrs 7678 are saved. 7679 7680 tpm_tis.interrupts= [HW,TPM] 7681 Enable interrupts for the MMIO based physical layer 7682 for the FIFO interface. By default it is set to false 7683 (0). For more information about TPM hardware interfaces 7684 defined by Trusted Computing Group (TCG) see 7685 https://trustedcomputinggroup.org/resource/pc-client-platform-tpm-profile-ptp-specification/ 7686 7687 tp_printk [FTRACE] 7688 Have the tracepoints sent to printk as well as the 7689 tracing ring buffer. This is useful for early boot up 7690 where the system hangs or reboots and does not give the 7691 option for reading the tracing buffer or performing a 7692 ftrace_dump_on_oops. 7693 7694 To turn off having tracepoints sent to printk, 7695 echo 0 > /proc/sys/kernel/tracepoint_printk 7696 Note, echoing 1 into this file without the 7697 tp_printk kernel cmdline option has no effect. 7698 7699 The tp_printk_stop_on_boot (see below) can also be used 7700 to stop the printing of events to console at 7701 late_initcall_sync. 7702 7703 ** CAUTION ** 7704 7705 Having tracepoints sent to printk() and activating high 7706 frequency tracepoints such as irq or sched, can cause 7707 the system to live lock. 7708 7709 tp_printk_stop_on_boot [FTRACE] 7710 When tp_printk (above) is set, it can cause a lot of noise 7711 on the console. It may be useful to only include the 7712 printing of events during boot up, as user space may 7713 make the system inoperable. 7714 7715 This command line option will stop the printing of events 7716 to console at the late_initcall_sync() time frame. 7717 7718 trace_buf_size=nn[KMG] 7719 [FTRACE] will set tracing buffer size on each cpu. 7720 7721 trace_clock= [FTRACE] Set the clock used for tracing events 7722 at boot up. 7723 local - Use the per CPU time stamp counter 7724 (converted into nanoseconds). Fast, but 7725 depending on the architecture, may not be 7726 in sync between CPUs. 7727 global - Event time stamps are synchronized across 7728 CPUs. May be slower than the local clock, 7729 but better for some race conditions. 7730 counter - Simple counting of events (1, 2, ..) 7731 note, some counts may be skipped due to the 7732 infrastructure grabbing the clock more than 7733 once per event. 7734 uptime - Use jiffies as the time stamp. 7735 perf - Use the same clock that perf uses. 7736 mono - Use ktime_get_mono_fast_ns() for time stamps. 7737 mono_raw - Use ktime_get_raw_fast_ns() for time 7738 stamps. 7739 boot - Use ktime_get_boot_fast_ns() for time stamps. 7740 Architectures may add more clocks. See 7741 Documentation/trace/ftrace.rst for more details. 7742 7743 trace_event=[event-list] 7744 [FTRACE] Set and start specified trace events in order 7745 to facilitate early boot debugging. The event-list is a 7746 comma-separated list of trace events to enable. See 7747 also Documentation/trace/events.rst 7748 7749 To enable modules, use :mod: keyword: 7750 7751 trace_event=:mod:<module> 7752 7753 The value before :mod: will only enable specific events 7754 that are part of the module. See the above mentioned 7755 document for more information. 7756 7757 trace_instance=[instance-info] 7758 [FTRACE] Create a ring buffer instance early in boot up. 7759 This will be listed in: 7760 7761 /sys/kernel/tracing/instances 7762 7763 Events can be enabled at the time the instance is created 7764 via: 7765 7766 trace_instance=<name>,<system1>:<event1>,<system2>:<event2> 7767 7768 Note, the "<system*>:" portion is optional if the event is 7769 unique. 7770 7771 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall 7772 7773 will enable the "sched_switch" event (note, the "sched:" is optional, and 7774 the same thing would happen if it was left off). The irq_handler_entry 7775 event, and all events under the "initcall" system. 7776 7777 Flags can be added to the instance to modify its behavior when it is 7778 created. The flags are separated by '^'. 7779 7780 The available flags are: 7781 7782 traceoff - Have the tracing instance tracing disabled after it is created. 7783 traceprintk - Have trace_printk() write into this trace instance 7784 (note, "printk" and "trace_printk" can also be used) 7785 7786 trace_instance=foo^traceoff^traceprintk,sched,irq 7787 7788 The flags must come before the defined events. 7789 7790 If memory has been reserved (see memmap for x86), the instance 7791 can use that memory: 7792 7793 memmap=12M$0x284500000 trace_instance=boot_map@0x284500000:12M 7794 7795 The above will create a "boot_map" instance that uses the physical 7796 memory at 0x284500000 that is 12Megs. The per CPU buffers of that 7797 instance will be split up accordingly. 7798 7799 Alternatively, the memory can be reserved by the reserve_mem option: 7800 7801 reserve_mem=12M:4096:trace trace_instance=boot_map@trace 7802 7803 This will reserve 12 megabytes at boot up with a 4096 byte alignment 7804 and place the ring buffer in this memory. Note that due to KASLR, the 7805 memory may not be the same location each time, which will not preserve 7806 the buffer content. 7807 7808 Also note that the layout of the ring buffer data may change between 7809 kernel versions where the validator will fail and reset the ring buffer 7810 if the layout is not the same as the previous kernel. 7811 7812 If the ring buffer is used for persistent bootups and has events enabled, 7813 it is recommend to disable tracing so that events from a previous boot do not 7814 mix with events of the current boot (unless you are debugging a random crash 7815 at boot up). 7816 7817 reserve_mem=12M:4096:trace trace_instance=boot_map^traceoff^traceprintk@trace,sched,irq 7818 7819 Note, saving the trace buffer across reboots does require that the system 7820 is set up to not wipe memory. For instance, CONFIG_RESET_ATTACK_MITIGATION 7821 can force a memory reset on boot which will clear any trace that was stored. 7822 This is just one of many ways that can clear memory. Make sure your system 7823 keeps the content of memory across reboots before relying on this option. 7824 7825 NB: Both the mapped address and size must be page aligned for the architecture. 7826 7827 See also Documentation/trace/debugging.rst 7828 7829 7830 trace_options=[option-list] 7831 [FTRACE] Enable or disable tracer options at boot. 7832 The option-list is a comma delimited list of options 7833 that can be enabled or disabled just as if you were 7834 to echo the option name into 7835 7836 /sys/kernel/tracing/trace_options 7837 7838 For example, to enable stacktrace option (to dump the 7839 stack trace of each event), add to the command line: 7840 7841 trace_options=stacktrace 7842 7843 See also Documentation/trace/ftrace.rst "trace options" 7844 section. 7845 7846 trace_trigger=[trigger-list] 7847 [FTRACE] Add an event trigger on specific events. 7848 Set a trigger on top of a specific event, with an optional 7849 filter. 7850 7851 The format is "trace_trigger=<event>.<trigger>[ if <filter>],..." 7852 Where more than one trigger may be specified that are comma delimited. 7853 7854 For example: 7855 7856 trace_trigger="sched_switch.stacktrace if prev_state == 2" 7857 7858 The above will enable the "stacktrace" trigger on the "sched_switch" 7859 event but only trigger it if the "prev_state" of the "sched_switch" 7860 event is "2" (TASK_UNINTERRUPTIBLE). 7861 7862 See also "Event triggers" in Documentation/trace/events.rst 7863 7864 7865 traceoff_after_boot 7866 [FTRACE] Sometimes tracing is used to debug issues 7867 during the boot process. Since the trace buffer has a 7868 limited amount of storage, it may be prudent to 7869 disable tracing after the boot is finished, otherwise 7870 the critical information may be overwritten. With this 7871 option, the main tracing buffer will be turned off at 7872 the end of the boot process. 7873 7874 traceoff_on_warning 7875 [FTRACE] enable this option to disable tracing when a 7876 warning is hit. This turns off "tracing_on". Tracing can 7877 be enabled again by echoing '1' into the "tracing_on" 7878 file located in /sys/kernel/tracing/ 7879 7880 This option is useful, as it disables the trace before 7881 the WARNING dump is called, which prevents the trace to 7882 be filled with content caused by the warning output. 7883 7884 This option can also be set at run time via the sysctl 7885 option: kernel/traceoff_on_warning 7886 7887 transparent_hugepage= 7888 [KNL] 7889 Format: [always|madvise|never] 7890 Can be used to control the default behavior of the system 7891 with respect to transparent hugepages. 7892 See Documentation/admin-guide/mm/transhuge.rst 7893 for more details. 7894 7895 transparent_hugepage_shmem= [KNL] 7896 Format: [always|within_size|advise|never|deny|force] 7897 Can be used to control the hugepage allocation policy for 7898 the internal shmem mount. 7899 See Documentation/admin-guide/mm/transhuge.rst 7900 for more details. 7901 7902 transparent_hugepage_tmpfs= [KNL] 7903 Format: [always|within_size|advise|never] 7904 Can be used to control the default hugepage allocation policy 7905 for the tmpfs mount. 7906 See Documentation/admin-guide/mm/transhuge.rst 7907 for more details. 7908 7909 trusted.source= [KEYS] 7910 Format: <string> 7911 This parameter identifies the trust source as a backend 7912 for trusted keys implementation. Supported trust 7913 sources: 7914 - "tpm" 7915 - "tee" 7916 - "caam" 7917 - "dcp" 7918 - "pkwm" 7919 If not specified then it defaults to iterating through 7920 the trust source list starting with TPM and assigns the 7921 first trust source as a backend which is initialized 7922 successfully during iteration. 7923 7924 trusted.rng= [KEYS] 7925 Format: <string> 7926 The RNG used to generate key material for trusted keys. 7927 Can be one of: 7928 - "kernel" 7929 - the same value as trusted.source: "tpm" or "tee" 7930 - "default" 7931 If not specified, "default" is used. In this case, 7932 the RNG's choice is left to each individual trust source. 7933 7934 trusted.dcp_use_otp_key 7935 This is intended to be used in combination with 7936 trusted.source=dcp and will select the DCP OTP key 7937 instead of the DCP UNIQUE key blob encryption. 7938 7939 trusted.dcp_skip_zk_test 7940 This is intended to be used in combination with 7941 trusted.source=dcp and will disable the check if the 7942 blob key is all zeros. This is helpful for situations where 7943 having this key zero'ed is acceptable. E.g. in testing 7944 scenarios. 7945 7946 tsa= [X86] Control mitigation for Transient Scheduler 7947 Attacks on AMD CPUs. Search the following in your 7948 favourite search engine for more details: 7949 7950 "Technical guidance for mitigating transient scheduler 7951 attacks". 7952 7953 off - disable the mitigation 7954 on - enable the mitigation (default) 7955 user - mitigate only user/kernel transitions 7956 vm - mitigate only guest/host transitions 7957 7958 7959 tsc= Disable clocksource stability checks for TSC. 7960 Format: <string> 7961 [x86] reliable: mark tsc clocksource as reliable, this 7962 disables clocksource verification at runtime, as well 7963 as the stability checks done at bootup. Used to enable 7964 high-resolution timer mode on older hardware, and in 7965 virtualized environment. 7966 [x86] noirqtime: Do not use TSC to do irq accounting. 7967 Used to run time disable IRQ_TIME_ACCOUNTING on any 7968 platforms where RDTSC is slow and this accounting 7969 can add overhead. 7970 [x86] unstable: mark the TSC clocksource as unstable, this 7971 marks the TSC unconditionally unstable at bootup and 7972 avoids any further wobbles once the TSC watchdog notices. 7973 [x86] nowatchdog: disable clocksource watchdog. Used 7974 in situations with strict latency requirements (where 7975 interruptions from clocksource watchdog are not 7976 acceptable). 7977 [x86] recalibrate: force recalibration against a HW timer 7978 (HPET or PM timer) on systems whose TSC frequency was 7979 obtained from HW or FW using either an MSR or CPUID(0x15). 7980 Warn if the difference is more than 500 ppm. 7981 [x86] watchdog: Use TSC as the watchdog clocksource with 7982 which to check other HW timers (HPET or PM timer), but 7983 only on systems where TSC has been deemed trustworthy. 7984 This will be suppressed by an earlier tsc=nowatchdog and 7985 can be overridden by a later tsc=nowatchdog. A console 7986 message will flag any such suppression or overriding. 7987 7988 tsc_early_khz= [X86,EARLY] Skip early TSC calibration and use the given 7989 value instead. Useful when the early TSC frequency discovery 7990 procedure is not reliable, such as on overclocked systems 7991 with CPUID.16h support and partial CPUID.15h support. 7992 Format: <unsigned int> 7993 7994 tsx= [X86] Control Transactional Synchronization 7995 Extensions (TSX) feature in Intel processors that 7996 support TSX control. 7997 7998 This parameter controls the TSX feature. The options are: 7999 8000 on - Enable TSX on the system. Although there are 8001 mitigations for all known security vulnerabilities, 8002 TSX has been known to be an accelerator for 8003 several previous speculation-related CVEs, and 8004 so there may be unknown security risks associated 8005 with leaving it enabled. 8006 8007 off - Disable TSX on the system. (Note that this 8008 option takes effect only on newer CPUs which are 8009 not vulnerable to MDS, i.e., have 8010 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 8011 the new IA32_TSX_CTRL MSR through a microcode 8012 update. This new MSR allows for the reliable 8013 deactivation of the TSX functionality.) 8014 8015 auto - Disable TSX if X86_BUG_TAA is present, 8016 otherwise enable TSX on the system. 8017 8018 Not specifying this option is equivalent to tsx=off. 8019 8020 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 8021 for more details. 8022 8023 tsx_async_abort= [X86,INTEL,EARLY] Control mitigation for the TSX Async 8024 Abort (TAA) vulnerability. 8025 8026 Similar to Micro-architectural Data Sampling (MDS) 8027 certain CPUs that support Transactional 8028 Synchronization Extensions (TSX) are vulnerable to an 8029 exploit against CPU internal buffers which can forward 8030 information to a disclosure gadget under certain 8031 conditions. 8032 8033 In vulnerable processors, the speculatively forwarded 8034 data can be used in a cache side channel attack, to 8035 access data to which the attacker does not have direct 8036 access. 8037 8038 This parameter controls the TAA mitigation. The 8039 options are: 8040 8041 full - Enable TAA mitigation on vulnerable CPUs 8042 if TSX is enabled. 8043 8044 full,nosmt - Enable TAA mitigation and disable SMT on 8045 vulnerable CPUs. If TSX is disabled, SMT 8046 is not disabled because CPU is not 8047 vulnerable to cross-thread TAA attacks. 8048 off - Unconditionally disable TAA mitigation 8049 8050 On MDS-affected machines, tsx_async_abort=off can be 8051 prevented by an active MDS mitigation as both vulnerabilities 8052 are mitigated with the same mechanism so in order to disable 8053 this mitigation, you need to specify mds=off too. 8054 8055 Not specifying this option is equivalent to 8056 tsx_async_abort=full. On CPUs which are MDS affected 8057 and deploy MDS mitigation, TAA mitigation is not 8058 required and doesn't provide any additional 8059 mitigation. 8060 8061 For details see: 8062 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 8063 8064 turbografx.map[2|3]= [HW,JOY] 8065 TurboGraFX parallel port interface 8066 Format: 8067 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 8068 See also Documentation/input/devices/joystick-parport.rst 8069 8070 udbg-immortal [PPC] When debugging early kernel crashes that 8071 happen after console_init() and before a proper 8072 console driver takes over, this boot options might 8073 help "seeing" what's going on. 8074 8075 uhash_entries= [KNL,NET] 8076 Set number of hash buckets for UDP/UDP-Lite connections 8077 8078 uhci-hcd.ignore_oc= 8079 [USB] Ignore overcurrent events (default N). 8080 Some badly-designed motherboards generate lots of 8081 bogus events, for ports that aren't wired to 8082 anything. Set this parameter to avoid log spamming. 8083 Note that genuine overcurrent events won't be 8084 reported either. 8085 8086 unaligned_scalar_speed= 8087 [RISCV] 8088 Format: {slow | fast | unsupported} 8089 Allow skipping scalar unaligned access speed tests. This 8090 is useful for testing alternative code paths and to skip 8091 the tests in environments where they run too slowly. All 8092 CPUs must have the same scalar unaligned access speed. 8093 8094 unaligned_vector_speed= 8095 [RISCV] 8096 Format: {slow | fast | unsupported} 8097 Allow skipping vector unaligned access speed tests. This 8098 is useful for testing alternative code paths and to skip 8099 the tests in environments where they run too slowly. All 8100 CPUs must have the same vector unaligned access speed. 8101 8102 unknown_nmi_panic 8103 [X86] Cause panic on unknown NMI. 8104 8105 unwind_debug [X86-64,EARLY] 8106 Enable unwinder debug output. This can be 8107 useful for debugging certain unwinder error 8108 conditions, including corrupt stacks and 8109 bad/missing unwinder metadata. 8110 8111 usbcore.authorized_default= 8112 [USB] Default USB device authorization: 8113 (default -1 = authorized (same as 1), 8114 0 = not authorized, 1 = authorized, 2 = authorized 8115 if device connected to internal port) 8116 8117 usbcore.autosuspend= 8118 [USB] The autosuspend time delay (in seconds) used 8119 for newly-detected USB devices (default 2). This 8120 is the time required before an idle device will be 8121 autosuspended. Devices for which the delay is set 8122 to a negative value won't be autosuspended at all. 8123 8124 usbcore.usbfs_snoop= 8125 [USB] Set to log all usbfs traffic (default 0 = off). 8126 8127 usbcore.usbfs_snoop_max= 8128 [USB] Maximum number of bytes to snoop in each URB 8129 (default = 65536). 8130 8131 usbcore.blinkenlights= 8132 [USB] Set to cycle leds on hubs (default 0 = off). 8133 8134 usbcore.old_scheme_first= 8135 [USB] Start with the old device initialization 8136 scheme (default 0 = off). 8137 8138 usbcore.usbfs_memory_mb= 8139 [USB] Memory limit (in MB) for buffers allocated by 8140 usbfs (default = 16, 0 = max = 2047). 8141 8142 usbcore.use_both_schemes= 8143 [USB] Try the other device initialization scheme 8144 if the first one fails (default 1 = enabled). 8145 8146 usbcore.initial_descriptor_timeout= 8147 [USB] Specifies timeout for the initial 64-byte 8148 USB_REQ_GET_DESCRIPTOR request in milliseconds 8149 (default 5000 = 5.0 seconds). 8150 8151 usbcore.nousb [USB] Disable the USB subsystem 8152 8153 usbcore.quirks= 8154 [USB] A list of quirk entries to augment the built-in 8155 usb core quirk list. List entries are separated by 8156 commas. Each entry has the form 8157 VendorID:ProductID:Flags. The IDs are 4-digit hex 8158 numbers and Flags is a set of letters. Each letter 8159 will change the built-in quirk; setting it if it is 8160 clear and clearing it if it is set. The letters have 8161 the following meanings: 8162 a = USB_QUIRK_STRING_FETCH_255 (string 8163 descriptors must not be fetched using 8164 a 255-byte read); 8165 b = USB_QUIRK_RESET_RESUME (device can't resume 8166 correctly so reset it instead); 8167 c = USB_QUIRK_NO_SET_INTF (device can't handle 8168 Set-Interface requests); 8169 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 8170 handle its Configuration or Interface 8171 strings); 8172 e = USB_QUIRK_RESET (device can't be reset 8173 (e.g morph devices), don't use reset); 8174 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 8175 more interface descriptions than the 8176 bNumInterfaces count, and can't handle 8177 talking to these interfaces); 8178 g = USB_QUIRK_DELAY_INIT (device needs a pause 8179 during initialization, after we read 8180 the device descriptor); 8181 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 8182 high speed and super speed interrupt 8183 endpoints, the USB 2.0 and USB 3.0 spec 8184 require the interval in microframes (1 8185 microframe = 125 microseconds) to be 8186 calculated as interval = 2 ^ 8187 (bInterval-1). 8188 Devices with this quirk report their 8189 bInterval as the result of this 8190 calculation instead of the exponent 8191 variable used in the calculation); 8192 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 8193 handle device_qualifier descriptor 8194 requests); 8195 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 8196 generates spurious wakeup, ignore 8197 remote wakeup capability); 8198 k = USB_QUIRK_NO_LPM (device can't handle Link 8199 Power Management); 8200 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 8201 (Device reports its bInterval as linear 8202 frames instead of the USB 2.0 8203 calculation); 8204 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 8205 to be disconnected before suspend to 8206 prevent spurious wakeup); 8207 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 8208 pause after every control message); 8209 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 8210 delay after resetting its port); 8211 p = USB_QUIRK_SHORT_SET_ADDRESS_REQ_TIMEOUT 8212 (Reduce timeout of the SET_ADDRESS 8213 request from 5000 ms to 500 ms); 8214 q = USB_QUIRK_FORCE_ONE_CONFIG (Device 8215 claims zero configurations, 8216 forcing to 1); 8217 Example: quirks=0781:5580:bk,0a5c:5834:gij 8218 8219 usbhid.mousepoll= 8220 [USBHID] The interval which mice are to be polled at. 8221 8222 usbhid.jspoll= 8223 [USBHID] The interval which joysticks are to be polled at. 8224 8225 usbhid.kbpoll= 8226 [USBHID] The interval which keyboards are to be polled at. 8227 8228 usb-storage.delay_use= 8229 [UMS] The delay in seconds before a new device is 8230 scanned for Logical Units (default 1). 8231 Optionally the delay in milliseconds if the value has 8232 suffix with "ms". 8233 Example: delay_use=2567ms 8234 8235 usb-storage.quirks= 8236 [UMS] A list of quirks entries to supplement or 8237 override the built-in unusual_devs list. List 8238 entries are separated by commas. Each entry has 8239 the form VID:PID:Flags where VID and PID are Vendor 8240 and Product ID values (4-digit hex numbers) and 8241 Flags is a set of characters, each corresponding 8242 to a common usb-storage quirk flag as follows: 8243 a = SANE_SENSE (collect more than 18 bytes 8244 of sense data, not on uas); 8245 b = BAD_SENSE (don't collect more than 18 8246 bytes of sense data, not on uas); 8247 c = FIX_CAPACITY (decrease the reported 8248 device capacity by one sector); 8249 d = NO_READ_DISC_INFO (don't use 8250 READ_DISC_INFO command, not on uas); 8251 e = NO_READ_CAPACITY_16 (don't use 8252 READ_CAPACITY_16 command); 8253 f = NO_REPORT_OPCODES (don't use report opcodes 8254 command, uas only); 8255 g = MAX_SECTORS_240 (don't transfer more than 8256 240 sectors at a time, uas only); 8257 h = CAPACITY_HEURISTICS (decrease the 8258 reported device capacity by one 8259 sector if the number is odd); 8260 i = IGNORE_DEVICE (don't bind to this 8261 device); 8262 j = NO_REPORT_LUNS (don't use report luns 8263 command, uas only); 8264 k = NO_SAME (do not use WRITE_SAME, uas only) 8265 l = NOT_LOCKABLE (don't try to lock and 8266 unlock ejectable media, not on uas); 8267 m = MAX_SECTORS_64 (don't transfer more 8268 than 64 sectors = 32 KB at a time, 8269 not on uas); 8270 n = INITIAL_READ10 (force a retry of the 8271 initial READ(10) command, not on uas); 8272 o = CAPACITY_OK (accept the capacity 8273 reported by the device, not on uas); 8274 p = WRITE_CACHE (the device cache is ON 8275 by default, not on uas); 8276 r = IGNORE_RESIDUE (the device reports 8277 bogus residue values, not on uas); 8278 s = SINGLE_LUN (the device has only one 8279 Logical Unit); 8280 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 8281 commands, uas only); 8282 u = IGNORE_UAS (don't bind to the uas driver); 8283 w = NO_WP_DETECT (don't test whether the 8284 medium is write-protected). 8285 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 8286 even if the device claims no cache, 8287 not on uas) 8288 Example: quirks=0419:aaf5:rl,0421:0433:rc 8289 8290 user_debug= [KNL,ARM] 8291 Format: <int> 8292 See arch/arm/Kconfig.debug help text. 8293 1 - undefined instruction events 8294 2 - system calls 8295 4 - invalid data aborts 8296 8 - SIGSEGV faults 8297 16 - SIGBUS faults 8298 Example: user_debug=31 8299 8300 vdso= [X86,SH,SPARC] 8301 On X86_32, this is an alias for vdso32=. Otherwise: 8302 8303 vdso=1: enable VDSO (the default) 8304 vdso=0: disable VDSO mapping 8305 8306 vdso32= [X86] Control the 32-bit vDSO 8307 vdso32=1: enable 32-bit VDSO 8308 vdso32=0 or vdso32=2: disable 32-bit VDSO 8309 8310 See the help text for CONFIG_COMPAT_VDSO for more 8311 details. If CONFIG_COMPAT_VDSO is set, the default is 8312 vdso32=0; otherwise, the default is vdso32=1. 8313 8314 For compatibility with older kernels, vdso32=2 is an 8315 alias for vdso32=0. 8316 8317 Try vdso32=0 if you encounter an error that says: 8318 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 8319 8320 video= [FB,EARLY] Frame buffer configuration 8321 See Documentation/fb/modedb.rst. 8322 8323 video.brightness_switch_enabled= [ACPI] 8324 Format: [0|1] 8325 If set to 1, on receiving an ACPI notify event 8326 generated by hotkey, video driver will adjust brightness 8327 level and then send out the event to user space through 8328 the allocated input device. If set to 0, video driver 8329 will only send out the event without touching backlight 8330 brightness level. 8331 default: 1 8332 8333 virtio_mmio.device= 8334 [VMMIO] Memory mapped virtio (platform) device. 8335 8336 <size>@<baseaddr>:<irq>[:<id>] 8337 where: 8338 <size> := size (can use standard suffixes 8339 like K, M and G) 8340 <baseaddr> := physical base address 8341 <irq> := interrupt number (as passed to 8342 request_irq()) 8343 <id> := (optional) platform device id 8344 example: 8345 virtio_mmio.device=1K@0x100b0000:48:7 8346 8347 Can be used multiple times for multiple devices. 8348 8349 vga= [BOOT,X86-32] Select a particular video mode 8350 See Documentation/arch/x86/boot.rst and 8351 Documentation/admin-guide/svga.rst. 8352 Use vga=ask for menu. 8353 This is actually a boot loader parameter; the value is 8354 passed to the kernel using a special protocol. 8355 8356 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 8357 May slow down system boot speed, especially when 8358 enabled on systems with a large amount of memory. 8359 All options are enabled by default, and this 8360 interface is meant to allow for selectively 8361 enabling or disabling specific virtual memory 8362 debugging features. 8363 8364 Available options are: 8365 P Enable page structure init time poisoning 8366 - Disable all of the above options 8367 8368 vmalloc=nn[KMG] [KNL,BOOT,EARLY] Forces the vmalloc area to have an 8369 exact size of <nn>. This can be used to increase 8370 the minimum size (128MB on x86, arm32 platforms). 8371 It can also be used to decrease the size and leave more room 8372 for directly mapped kernel RAM. Note that this parameter does 8373 not exist on many other platforms (including arm64, alpha, 8374 loongarch, arc, csky, hexagon, microblaze, mips, nios2, openrisc, 8375 parisc, m64k, powerpc, riscv, sh, um, xtensa, s390, sparc). 8376 8377 vmcp_cma=nn[MG] [KNL,S390,EARLY] 8378 Sets the memory size reserved for contiguous memory 8379 allocations for the vmcp device driver. 8380 8381 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 8382 Format: <command> 8383 8384 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 8385 Format: <command> 8386 8387 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 8388 Format: <command> 8389 8390 vmscape= [X86] Controls mitigation for VMscape attacks. 8391 VMscape attacks can leak information from a userspace 8392 hypervisor to a guest via speculative side-channels. 8393 8394 off - disable the mitigation 8395 ibpb - use Indirect Branch Prediction Barrier 8396 (IBPB) mitigation (default) 8397 force - force vulnerability detection even on 8398 unaffected processors 8399 8400 vsyscall= [X86-64,EARLY] 8401 Controls the behavior of vsyscalls (i.e. calls to 8402 fixed addresses of 0xffffffffff600x00 from legacy 8403 code). Most statically-linked binaries and older 8404 versions of glibc use these calls. Because these 8405 functions are at fixed addresses, they make nice 8406 targets for exploits that can control RIP. 8407 8408 emulate Vsyscalls turn into traps and are emulated 8409 reasonably safely. The vsyscall page is 8410 readable. 8411 8412 xonly [default] Vsyscalls turn into traps and are 8413 emulated reasonably safely. The vsyscall 8414 page is not readable. 8415 8416 none Vsyscalls don't work at all. This makes 8417 them quite hard to use for exploits but 8418 might break your system. 8419 8420 vt.color= [VT] Default text color. 8421 Format: 0xYX, X = foreground, Y = background. 8422 Default: 0x07 = light gray on black. 8423 8424 vt.cur_default= [VT] Default cursor shape. 8425 Format: 0xCCBBAA, where AA, BB, and CC are the same as 8426 the parameters of the <Esc>[?A;B;Cc escape sequence; 8427 see vga-softcursor.rst. Default: 2 = underline. 8428 8429 vt.default_blu= [VT] 8430 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 8431 Change the default blue palette of the console. 8432 This is a 16-member array composed of values 8433 ranging from 0-255. 8434 8435 vt.default_grn= [VT] 8436 Format: <green0>,<green1>,<green2>,...,<green15> 8437 Change the default green palette of the console. 8438 This is a 16-member array composed of values 8439 ranging from 0-255. 8440 8441 vt.default_red= [VT] 8442 Format: <red0>,<red1>,<red2>,...,<red15> 8443 Change the default red palette of the console. 8444 This is a 16-member array composed of values 8445 ranging from 0-255. 8446 8447 vt.default_utf8= 8448 [VT] 8449 Format=<0|1> 8450 Set system-wide default UTF-8 mode for all tty's. 8451 Default is 1, i.e. UTF-8 mode is enabled for all 8452 newly opened terminals. 8453 8454 vt.global_cursor_default= 8455 [VT] 8456 Format=<-1|0|1> 8457 Set system-wide default for whether a cursor 8458 is shown on new VTs. Default is -1, 8459 i.e. cursors will be created by default unless 8460 overridden by individual drivers. 0 will hide 8461 cursors, 1 will display them. 8462 8463 vt.italic= [VT] Default color for italic text; 0-15. 8464 Default: 2 = green. 8465 8466 vt.underline= [VT] Default color for underlined text; 0-15. 8467 Default: 3 = cyan. 8468 8469 watchdog timers [HW,WDT] For information on watchdog timers, 8470 see Documentation/watchdog/watchdog-parameters.rst 8471 or other driver-specific files in the 8472 Documentation/watchdog/ directory. 8473 8474 watchdog_thresh= 8475 [KNL] 8476 Set the hard lockup detector stall duration 8477 threshold in seconds. The soft lockup detector 8478 threshold is set to twice the value. A value of 0 8479 disables both lockup detectors. Default is 10 8480 seconds. 8481 8482 workqueue.unbound_cpus= 8483 [KNL,SMP] Specify to constrain one or some CPUs 8484 to use in unbound workqueues. 8485 Format: <cpu-list> 8486 By default, all online CPUs are available for 8487 unbound workqueues. 8488 8489 workqueue.watchdog_thresh= 8490 If CONFIG_WQ_WATCHDOG is configured, workqueue can 8491 warn stall conditions and dump internal state to 8492 help debugging. 0 disables workqueue stall 8493 detection; otherwise, it's the stall threshold 8494 duration in seconds. The default value is 30 and 8495 it can be updated at runtime by writing to the 8496 corresponding sysfs file. 8497 8498 workqueue.panic_on_stall=<uint> 8499 Panic when workqueue stall is detected by 8500 CONFIG_WQ_WATCHDOG. It sets the number times of the 8501 stall to trigger panic. 8502 8503 The default is set by CONFIG_BOOTPARAM_WQ_STALL_PANIC, 8504 which is 0 (disabled) if not configured. 8505 8506 workqueue.panic_on_stall_time=<uint> 8507 Panic when a workqueue stall has been continuous for 8508 the specified number of seconds. Unlike panic_on_stall 8509 which counts accumulated stall events, this triggers 8510 based on the duration of a single continuous stall. 8511 8512 The default is 0, which disables the time-based panic. 8513 8514 workqueue.cpu_intensive_thresh_us= 8515 Per-cpu work items which run for longer than this 8516 threshold are automatically considered CPU intensive 8517 and excluded from concurrency management to prevent 8518 them from noticeably delaying other per-cpu work 8519 items. Default is 10000 (10ms). 8520 8521 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8522 will report the work functions which violate this 8523 threshold repeatedly. They are likely good 8524 candidates for using WQ_UNBOUND workqueues instead. 8525 8526 workqueue.cpu_intensive_warning_thresh=<uint> 8527 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8528 will report the work functions which violate the 8529 intensive_threshold_us repeatedly. In order to prevent 8530 spurious warnings, start printing only after a work 8531 function has violated this threshold number of times. 8532 8533 The default is 4 times. 0 disables the warning. 8534 8535 workqueue.power_efficient 8536 Per-cpu workqueues are generally preferred because 8537 they show better performance thanks to cache 8538 locality; unfortunately, per-cpu workqueues tend to 8539 be more power hungry than unbound workqueues. 8540 8541 Enabling this makes the per-cpu workqueues which 8542 were observed to contribute significantly to power 8543 consumption unbound, leading to measurably lower 8544 power usage at the cost of small performance 8545 overhead. 8546 8547 The default value of this parameter is determined by 8548 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 8549 8550 workqueue.default_affinity_scope= 8551 Select the default affinity scope to use for unbound 8552 workqueues. Can be one of "cpu", "smt", "cache", 8553 "numa" and "system". Default is "cache". For more 8554 information, see the Affinity Scopes section in 8555 Documentation/core-api/workqueue.rst. 8556 8557 This can be changed after boot by writing to the 8558 matching /sys/module/workqueue/parameters file. All 8559 workqueues with the "default" affinity scope will be 8560 updated accordingly. 8561 8562 workqueue.debug_force_rr_cpu 8563 Workqueue used to implicitly guarantee that work 8564 items queued without explicit CPU specified are put 8565 on the local CPU. This guarantee is no longer true 8566 and while local CPU is still preferred work items 8567 may be put on foreign CPUs. This debug option 8568 forces round-robin CPU selection to flush out 8569 usages which depend on the now broken guarantee. 8570 When enabled, memory and cache locality will be 8571 impacted. 8572 8573 writecombine= [LOONGARCH,EARLY] Control the MAT (Memory Access 8574 Type) of ioremap_wc(). 8575 8576 on - Enable writecombine, use WUC for ioremap_wc() 8577 off - Disable writecombine, use SUC for ioremap_wc() 8578 8579 x2apic_phys [X86-64,APIC,EARLY] Use x2apic physical mode instead of 8580 default x2apic cluster mode on platforms 8581 supporting x2apic. 8582 8583 xen_512gb_limit [KNL,X86-64,XEN] 8584 Restricts the kernel running paravirtualized under Xen 8585 to use only up to 512 GB of RAM. The reason to do so is 8586 crash analysis tools and Xen tools for doing domain 8587 save/restore/migration must be enabled to handle larger 8588 domains. 8589 8590 xen_console_io [XEN,EARLY] 8591 Boolean option to enable/disable the usage of the Xen 8592 console_io hypercalls to read and write to the console. 8593 Mostly useful for debugging and development. 8594 8595 xen_emul_unplug= [HW,X86,XEN,EARLY] 8596 Unplug Xen emulated devices 8597 Format: [unplug0,][unplug1] 8598 ide-disks -- unplug primary master IDE devices 8599 aux-ide-disks -- unplug non-primary-master IDE devices 8600 nics -- unplug network devices 8601 all -- unplug all emulated devices (NICs and IDE disks) 8602 unnecessary -- unplugging emulated devices is 8603 unnecessary even if the host did not respond to 8604 the unplug protocol 8605 never -- do not unplug even if version check succeeds 8606 8607 xen_legacy_crash [X86,XEN,EARLY] 8608 Crash from Xen panic notifier, without executing late 8609 panic() code such as dumping handler. 8610 8611 xen_mc_debug [X86,XEN,EARLY] 8612 Enable multicall debugging when running as a Xen PV guest. 8613 Enabling this feature will reduce performance a little 8614 bit, so it should only be enabled for obtaining extended 8615 debug data in case of multicall errors. 8616 8617 xen_msr_safe= [X86,XEN,EARLY] 8618 Format: <bool> 8619 Select whether to always use non-faulting (safe) MSR 8620 access functions when running as Xen PV guest. The 8621 default value is controlled by CONFIG_XEN_PV_MSR_SAFE. 8622 8623 xen_nopv [X86] 8624 Disables the PV optimizations forcing the HVM guest to 8625 run as generic HVM guest with no PV drivers. 8626 This option is obsoleted by the "nopv" option, which 8627 has equivalent effect for XEN platform. 8628 8629 xen_no_vector_callback 8630 [KNL,X86,XEN,EARLY] Disable the vector callback for Xen 8631 event channel interrupts. 8632 8633 xen_scrub_pages= [XEN] 8634 Boolean option to control scrubbing pages before giving them back 8635 to Xen, for use by other domains. Can be also changed at runtime 8636 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 8637 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 8638 8639 xen_timer_slop= [X86-64,XEN,EARLY] 8640 Set the timer slop (in nanoseconds) for the virtual Xen 8641 timers (default is 100000). This adjusts the minimum 8642 delta of virtualized Xen timers, where lower values 8643 improve timer resolution at the expense of processing 8644 more timer interrupts. 8645 8646 xen.balloon_boot_timeout= [XEN] 8647 The time (in seconds) to wait before giving up to boot 8648 in case initial ballooning fails to free enough memory. 8649 Applies only when running as HVM or PVH guest and 8650 started with less memory configured than allowed at 8651 max. Default is 180. 8652 8653 xen.event_eoi_delay= [XEN] 8654 How long to delay EOI handling in case of event 8655 storms (jiffies). Default is 10. 8656 8657 xen.event_loop_timeout= [XEN] 8658 After which time (jiffies) the event handling loop 8659 should start to delay EOI handling. Default is 2. 8660 8661 xen.fifo_events= [XEN] 8662 Boolean parameter to disable using fifo event handling 8663 even if available. Normally fifo event handling is 8664 preferred over the 2-level event handling, as it is 8665 fairer and the number of possible event channels is 8666 much higher. Default is on (use fifo events). 8667 8668 xirc2ps_cs= [NET,PCMCIA] 8669 Format: 8670 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 8671 8672 xive= [PPC] 8673 By default on POWER9 and above, the kernel will 8674 natively use the XIVE interrupt controller. This option 8675 allows the fallback firmware mode to be used: 8676 8677 off Fallback to firmware control of XIVE interrupt 8678 controller on both pseries and powernv 8679 platforms. Only useful on POWER9 and above. 8680 8681 xive.store-eoi=off [PPC] 8682 By default on POWER10 and above, the kernel will use 8683 stores for EOI handling when the XIVE interrupt mode 8684 is active. This option allows the XIVE driver to use 8685 loads instead, as on POWER9. 8686 8687 xhci-hcd.quirks [USB,KNL] 8688 A hex value specifying bitmask with supplemental xhci 8689 host controller quirks. Meaning of each bit can be 8690 consulted in header drivers/usb/host/xhci.h. 8691 8692 xmon [PPC,EARLY] 8693 Format: { early | on | rw | ro | off } 8694 Controls if xmon debugger is enabled. Default is off. 8695 Passing only "xmon" is equivalent to "xmon=early". 8696 early Call xmon as early as possible on boot; xmon 8697 debugger is called from setup_arch(). 8698 on xmon debugger hooks will be installed so xmon 8699 is only called on a kernel crash. Default mode, 8700 i.e. either "ro" or "rw" mode, is controlled 8701 with CONFIG_XMON_DEFAULT_RO_MODE. 8702 rw xmon debugger hooks will be installed so xmon 8703 is called only on a kernel crash, mode is write, 8704 meaning SPR registers, memory and, other data 8705 can be written using xmon commands. 8706 ro same as "rw" option above but SPR registers, 8707 memory, and other data can't be written using 8708 xmon commands. 8709 off xmon is disabled. 8710