1 ACPI ACPI support is enabled. 2 AGP AGP (Accelerated Graphics Port) is enabled. 3 ALSA ALSA sound support is enabled. 4 APIC APIC support is enabled. 5 APM Advanced Power Management support is enabled. 6 APPARMOR AppArmor support is enabled. 7 ARM ARM architecture is enabled. 8 ARM64 ARM64 architecture is enabled. 9 CLK Common clock infrastructure is enabled. 10 CMA Contiguous Memory Area support is enabled. 11 DRM Direct Rendering Management support is enabled. 12 DYNAMIC_DEBUG Build in debug messages and enable them at runtime 13 EARLY Parameter processed too early to be embedded in initrd. 14 EDD BIOS Enhanced Disk Drive Services (EDD) is enabled 15 EFI EFI Partitioning (GPT) is enabled 16 EVM Extended Verification Module 17 FB The frame buffer device is enabled. 18 FTRACE Function tracing enabled. 19 GCOV GCOV profiling is enabled. 20 HIBERNATION HIBERNATION is enabled. 21 HW Appropriate hardware is enabled. 22 HYPER_V HYPERV support is enabled. 23 IMA Integrity measurement architecture is enabled. 24 IP_PNP IP DHCP, BOOTP, or RARP is enabled. 25 IPV6 IPv6 support is enabled. 26 ISAPNP ISA PnP code is enabled. 27 ISOL CPU Isolation is enabled. 28 JOY Appropriate joystick support is enabled. 29 KGDB Kernel debugger support is enabled. 30 KVM Kernel Virtual Machine support is enabled. 31 LIBATA Libata driver is enabled 32 LOONGARCH LoongArch architecture is enabled. 33 LOOP Loopback device support is enabled. 34 LP Printer support is enabled. 35 M68k M68k architecture is enabled. 36 These options have more detailed description inside of 37 Documentation/arch/m68k/kernel-options.rst. 38 MIPS MIPS architecture is enabled. 39 MOUSE Appropriate mouse support is enabled. 40 MSI Message Signaled Interrupts (PCI). 41 MTD MTD (Memory Technology Device) support is enabled. 42 NET Appropriate network support is enabled. 43 NFS Appropriate NFS support is enabled. 44 NUMA NUMA support is enabled. 45 OF Devicetree is enabled. 46 PARISC The PA-RISC architecture is enabled. 47 PCI PCI bus support is enabled. 48 PCIE PCI Express support is enabled. 49 PCMCIA The PCMCIA subsystem is enabled. 50 PM Power Management support is enabled. 51 PNP Plug & Play support is enabled. 52 PPC PowerPC architecture is enabled. 53 PPT Parallel port support is enabled. 54 PS2 Appropriate PS/2 support is enabled. 55 PV_OPS A paravirtualized kernel is enabled. 56 RAM RAM disk support is enabled. 57 RDT Intel Resource Director Technology. 58 RISCV RISCV architecture is enabled. 59 S390 S390 architecture is enabled. 60 SCSI Appropriate SCSI support is enabled. 61 A lot of drivers have their options described inside 62 the Documentation/scsi/ sub-directory. 63 SDW SoundWire support is enabled. 64 SECURITY Different security models are enabled. 65 SELINUX SELinux support is enabled. 66 SERIAL Serial support is enabled. 67 SH SuperH architecture is enabled. 68 SMP The kernel is an SMP kernel. 69 SPARC Sparc architecture is enabled. 70 SUSPEND System suspend states are enabled. 71 SWSUSP Software suspend (hibernation) is enabled. 72 TPM TPM drivers are enabled. 73 UMS USB Mass Storage support is enabled. 74 USB USB support is enabled. 75 NVME NVMe support is enabled 76 USBHID USB Human Interface Device support is enabled. 77 V4L Video For Linux support is enabled. 78 VGA The VGA console has been enabled. 79 VMMIO Driver for memory mapped virtio devices is enabled. 80 VT Virtual terminal support is enabled. 81 WDT Watchdog support is enabled. 82 X86-32 X86-32, aka i386 architecture is enabled. 83 X86-64 X86-64 architecture is enabled. 84 X86 Either 32-bit or 64-bit x86 (same as X86-32+X86-64) 85 X86_UV SGI UV support is enabled. 86 XEN Xen support is enabled 87 XTENSA xtensa architecture is enabled. 88 89In addition, the following text indicates that the option 90 91 BOOT Is a boot loader parameter. 92 BUGS= Relates to possible processor bugs on the said processor. 93 KNL Is a kernel start-up parameter. 94 95 96Kernel parameters 97 98 accept_memory= [MM] 99 Format: { eager | lazy } 100 default: lazy 101 By default, unaccepted memory is accepted lazily to 102 avoid prolonged boot times. The lazy option will add 103 some runtime overhead until all memory is eventually 104 accepted. In most cases the overhead is negligible. 105 For some workloads or for debugging purposes 106 accept_memory=eager can be used to accept all memory 107 at once during boot. 108 109 acpi= [HW,ACPI,X86,ARM64,RISCV64,EARLY] 110 Advanced Configuration and Power Interface 111 Format: { force | on | off | strict | noirq | rsdt | 112 copy_dsdt | nospcr } 113 force -- enable ACPI if default was off 114 on -- enable ACPI but allow fallback to DT [arm64,riscv64] 115 off -- disable ACPI if default was on 116 noirq -- do not use ACPI for IRQ routing 117 strict -- Be less tolerant of platforms that are not 118 strictly ACPI specification compliant. 119 rsdt -- prefer RSDT over (default) XSDT 120 copy_dsdt -- copy DSDT to memory 121 nocmcff -- Disable firmware first mode for corrected 122 errors. This disables parsing the HEST CMC error 123 source to check if firmware has set the FF flag. This 124 may result in duplicate corrected error reports. 125 nospcr -- disable console in ACPI SPCR table as 126 default _serial_ console on ARM64 127 spcr -- enable console in ACPI SPCR table as 128 default _serial_ console on x86 129 For ARM64, ONLY "acpi=off", "acpi=on", "acpi=force" or 130 "acpi=nospcr" are available 131 For RISCV64, ONLY "acpi=off", "acpi=on" or "acpi=force" 132 are available 133 134 See also Documentation/power/runtime_pm.rst, pci=noacpi 135 136 acpi_apic_instance= [ACPI,IOAPIC,EARLY] 137 Format: <int> 138 2: use 2nd APIC table, if available 139 1,0: use 1st APIC table 140 default: 0 141 142 acpi_backlight= [HW,ACPI] 143 { vendor | video | native | none } 144 If set to vendor, prefer vendor-specific driver 145 (e.g. thinkpad_acpi, sony_acpi, etc.) instead 146 of the ACPI video.ko driver. 147 If set to video, use the ACPI video.ko driver. 148 If set to native, use the device's native backlight mode. 149 If set to none, disable the ACPI backlight interface. 150 151 acpi_force_32bit_fadt_addr [ACPI,EARLY] 152 force FADT to use 32 bit addresses rather than the 153 64 bit X_* addresses. Some firmware have broken 64 154 bit addresses for force ACPI ignore these and use 155 the older legacy 32 bit addresses. 156 157 acpica_no_return_repair [HW, ACPI] 158 Disable AML predefined validation mechanism 159 This mechanism can repair the evaluation result to make 160 the return objects more ACPI specification compliant. 161 This option is useful for developers to identify the 162 root cause of an AML interpreter issue when the issue 163 has something to do with the repair mechanism. 164 165 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG] 166 acpi.debug_level= [HW,ACPI,ACPI_DEBUG] 167 Format: <int> 168 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI 169 debug output. Bits in debug_layer correspond to a 170 _COMPONENT in an ACPI source file, e.g., 171 #define _COMPONENT ACPI_EVENTS 172 Bits in debug_level correspond to a level in 173 ACPI_DEBUG_PRINT statements, e.g., 174 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ... 175 The debug_level mask defaults to "info". See 176 Documentation/firmware-guide/acpi/debug.rst for more information about 177 debug layers and levels. 178 179 Enable processor driver info messages: 180 acpi.debug_layer=0x20000000 181 Enable AML "Debug" output, i.e., stores to the Debug 182 object while interpreting AML: 183 acpi.debug_layer=0xffffffff acpi.debug_level=0x2 184 Enable all messages related to ACPI hardware: 185 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff 186 187 Some values produce so much output that the system is 188 unusable. The "log_buf_len" parameter may be useful 189 if you need to capture more output. 190 191 acpi.poweroff_on_fatal= [ACPI] 192 {0 | 1} 193 Causes the system to poweroff when the ACPI bytecode signals 194 a fatal error. The default value of this setting is 1. 195 Overriding this value should only be done for diagnosing 196 ACPI firmware problems, as the system might behave erratically 197 after having encountered a fatal ACPI error. 198 199 acpi_enforce_resources= [ACPI] 200 { strict | lax | no } 201 Check for resource conflicts between native drivers 202 and ACPI OperationRegions (SystemIO and SystemMemory 203 only). IO ports and memory declared in ACPI might be 204 used by the ACPI subsystem in arbitrary AML code and 205 can interfere with legacy drivers. 206 strict (default): access to resources claimed by ACPI 207 is denied; legacy drivers trying to access reserved 208 resources will fail to bind to device using them. 209 lax: access to resources claimed by ACPI is allowed; 210 legacy drivers trying to access reserved resources 211 will bind successfully but a warning message is logged. 212 no: ACPI OperationRegions are not marked as reserved, 213 no further checks are performed. 214 215 acpi_force_table_verification [HW,ACPI,EARLY] 216 Enable table checksum verification during early stage. 217 By default, this is disabled due to x86 early mapping 218 size limitation. 219 220 acpi_irq_balance [HW,ACPI] 221 ACPI will balance active IRQs 222 default in APIC mode 223 224 acpi_irq_nobalance [HW,ACPI] 225 ACPI will not move active IRQs (default) 226 default in PIC mode 227 228 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 229 Format: <irq>,<irq>... 230 231 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 232 use by PCI 233 Format: <irq>,<irq>... 234 235 acpi_mask_gpe= [HW,ACPI] 236 Due to the existence of _Lxx/_Exx, some GPEs triggered 237 by unsupported hardware/firmware features can result in 238 GPE floodings that cannot be automatically disabled by 239 the GPE dispatcher. 240 This facility can be used to prevent such uncontrolled 241 GPE floodings. 242 Format: <byte> or <bitmap-list> 243 244 acpi_no_auto_serialize [HW,ACPI] 245 Disable auto-serialization of AML methods 246 AML control methods that contain the opcodes to create 247 named objects will be marked as "Serialized" by the 248 auto-serialization feature. 249 This feature is enabled by default. 250 This option allows to turn off the feature. 251 252 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 253 kernels. 254 255 acpi_no_static_ssdt [HW,ACPI,EARLY] 256 Disable installation of static SSDTs at early boot time 257 By default, SSDTs contained in the RSDT/XSDT will be 258 installed automatically and they will appear under 259 /sys/firmware/acpi/tables. 260 This option turns off this feature. 261 Note that specifying this option does not affect 262 dynamic table installation which will install SSDT 263 tables to /sys/firmware/acpi/tables/dynamic. 264 265 acpi_no_watchdog [HW,ACPI,WDT] 266 Ignore the ACPI-based watchdog interface (WDAT) and let 267 a native driver control the watchdog device instead. 268 269 acpi_rsdp= [ACPI,EFI,KEXEC,EARLY] 270 Pass the RSDP address to the kernel, mostly used 271 on machines running EFI runtime service to boot the 272 second kernel for kdump. 273 274 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 275 Format: To spoof as Windows 98: ="Microsoft Windows" 276 277 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 278 of 2 which is mandated by ACPI 6) as the supported ACPI 279 specification revision (when using this switch, it may 280 be necessary to carry out a cold reboot _twice_ in a 281 row to make it take effect on the platform firmware). 282 283 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 284 acpi_osi="string1" # add string1 285 acpi_osi="!string2" # remove string2 286 acpi_osi=!* # remove all strings 287 acpi_osi=! # disable all built-in OS vendor 288 strings 289 acpi_osi=!! # enable all built-in OS vendor 290 strings 291 acpi_osi= # disable all strings 292 293 'acpi_osi=!' can be used in combination with single or 294 multiple 'acpi_osi="string1"' to support specific OS 295 vendor string(s). Note that such command can only 296 affect the default state of the OS vendor strings, thus 297 it cannot affect the default state of the feature group 298 strings and the current state of the OS vendor strings, 299 specifying it multiple times through kernel command line 300 is meaningless. This command is useful when one do not 301 care about the state of the feature group strings which 302 should be controlled by the OSPM. 303 Examples: 304 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 305 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 306 can make '_OSI("Windows 2000")' TRUE. 307 308 'acpi_osi=' cannot be used in combination with other 309 'acpi_osi=' command lines, the _OSI method will not 310 exist in the ACPI namespace. NOTE that such command can 311 only affect the _OSI support state, thus specifying it 312 multiple times through kernel command line is also 313 meaningless. 314 Examples: 315 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 316 FALSE. 317 318 'acpi_osi=!*' can be used in combination with single or 319 multiple 'acpi_osi="string1"' to support specific 320 string(s). Note that such command can affect the 321 current state of both the OS vendor strings and the 322 feature group strings, thus specifying it multiple times 323 through kernel command line is meaningful. But it may 324 still not able to affect the final state of a string if 325 there are quirks related to this string. This command 326 is useful when one want to control the state of the 327 feature group strings to debug BIOS issues related to 328 the OSPM features. 329 Examples: 330 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 331 '_OSI("Module Device")' FALSE. 332 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 333 '_OSI("Module Device")' TRUE. 334 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 335 equivalent to 336 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 337 and 338 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 339 they all will make '_OSI("Windows 2000")' TRUE. 340 341 acpi_pm_good [X86] 342 Override the pmtimer bug detection: force the kernel 343 to assume that this machine's pmtimer latches its value 344 and always returns good values. 345 346 acpi_sci= [HW,ACPI,EARLY] ACPI System Control Interrupt trigger mode 347 Format: { level | edge | high | low } 348 349 acpi_skip_timer_override [HW,ACPI,EARLY] 350 Recognize and ignore IRQ0/pin2 Interrupt Override. 351 For broken nForce2 BIOS resulting in XT-PIC timer. 352 353 acpi_sleep= [HW,ACPI] Sleep options 354 Format: { s3_bios, s3_mode, s3_beep, s4_hwsig, 355 s4_nohwsig, old_ordering, nonvs, 356 sci_force_enable, nobl } 357 See Documentation/power/video.rst for information on 358 s3_bios and s3_mode. 359 s3_beep is for debugging; it makes the PC's speaker beep 360 as soon as the kernel's real-mode entry point is called. 361 s4_hwsig causes the kernel to check the ACPI hardware 362 signature during resume from hibernation, and gracefully 363 refuse to resume if it has changed. This complies with 364 the ACPI specification but not with reality, since 365 Windows does not do this and many laptops do change it 366 on docking. So the default behaviour is to allow resume 367 and simply warn when the signature changes, unless the 368 s4_hwsig option is enabled. 369 s4_nohwsig prevents ACPI hardware signature from being 370 used (or even warned about) during resume. 371 old_ordering causes the ACPI 1.0 ordering of the _PTS 372 control method, with respect to putting devices into 373 low power states, to be enforced (the ACPI 2.0 ordering 374 of _PTS is used by default). 375 nonvs prevents the kernel from saving/restoring the 376 ACPI NVS memory during suspend/hibernation and resume. 377 sci_force_enable causes the kernel to set SCI_EN directly 378 on resume from S1/S3 (which is against the ACPI spec, 379 but some broken systems don't work without it). 380 nobl causes the internal blacklist of systems known to 381 behave incorrectly in some ways with respect to system 382 suspend and resume to be ignored (use wisely). 383 384 acpi_use_timer_override [HW,ACPI,EARLY] 385 Use timer override. For some broken Nvidia NF5 boards 386 that require a timer override, but don't have HPET 387 388 add_efi_memmap [EFI,X86,EARLY] Include EFI memory map in 389 kernel's map of available physical RAM. 390 391 agp= [AGP] 392 { off | try_unsupported } 393 off: disable AGP support 394 try_unsupported: try to drive unsupported chipsets 395 (may crash computer or cause data corruption) 396 397 ALSA [HW,ALSA] 398 See Documentation/sound/alsa-configuration.rst 399 400 alignment= [KNL,ARM] 401 Allow the default userspace alignment fault handler 402 behaviour to be specified. Bit 0 enables warnings, 403 bit 1 enables fixups, and bit 2 sends a segfault. 404 405 align_va_addr= [X86-64] 406 Align virtual addresses by clearing slice [14:12] when 407 allocating a VMA at process creation time. This option 408 gives you up to 3% performance improvement on AMD F15h 409 machines (where it is enabled by default) for a 410 CPU-intensive style benchmark, and it can vary highly in 411 a microbenchmark depending on workload and compiler. 412 413 32: only for 32-bit processes 414 64: only for 64-bit processes 415 on: enable for both 32- and 64-bit processes 416 off: disable for both 32- and 64-bit processes 417 418 alloc_snapshot [FTRACE] 419 Allocate the ftrace snapshot buffer on boot up when the 420 main buffer is allocated. This is handy if debugging 421 and you need to use tracing_snapshot() on boot up, and 422 do not want to use tracing_snapshot_alloc() as it needs 423 to be done where GFP_KERNEL allocations are allowed. 424 425 allow_mismatched_32bit_el0 [ARM64,EARLY] 426 Allow execve() of 32-bit applications and setting of the 427 PER_LINUX32 personality on systems where only a strict 428 subset of the CPUs support 32-bit EL0. When this 429 parameter is present, the set of CPUs supporting 32-bit 430 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0 431 and hot-unplug operations may be restricted. 432 433 See Documentation/arch/arm64/asymmetric-32bit.rst for more 434 information. 435 436 amd_iommu= [HW,X86-64] 437 Pass parameters to the AMD IOMMU driver in the system. 438 Possible values are: 439 fullflush - Deprecated, equivalent to iommu.strict=1 440 off - do not initialize any AMD IOMMU found in 441 the system 442 force_isolation - Force device isolation for all 443 devices. The IOMMU driver is not 444 allowed anymore to lift isolation 445 requirements as needed. This option 446 does not override iommu=pt 447 force_enable - Force enable the IOMMU on platforms known 448 to be buggy with IOMMU enabled. Use this 449 option with care. 450 pgtbl_v1 - Use v1 page table for DMA-API (Default). 451 pgtbl_v2 - Use v2 page table for DMA-API. 452 irtcachedis - Disable Interrupt Remapping Table (IRT) caching. 453 nohugepages - Limit page-sizes used for v1 page-tables 454 to 4 KiB. 455 v2_pgsizes_only - Limit page-sizes used for v1 page-tables 456 to 4KiB/2Mib/1GiB. 457 458 459 amd_iommu_dump= [HW,X86-64] 460 Enable AMD IOMMU driver option to dump the ACPI table 461 for AMD IOMMU. With this option enabled, AMD IOMMU 462 driver will print ACPI tables for AMD IOMMU during 463 IOMMU initialization. 464 465 amd_iommu_intr= [HW,X86-64] 466 Specifies one of the following AMD IOMMU interrupt 467 remapping modes: 468 legacy - Use legacy interrupt remapping mode. 469 vapic - Use virtual APIC mode, which allows IOMMU 470 to inject interrupts directly into guest. 471 This mode requires kvm-amd.avic=1. 472 (Default when IOMMU HW support is present.) 473 474 amd_pstate= [X86,EARLY] 475 disable 476 Do not enable amd_pstate as the default 477 scaling driver for the supported processors 478 passive 479 Use amd_pstate with passive mode as a scaling driver. 480 In this mode autonomous selection is disabled. 481 Driver requests a desired performance level and platform 482 tries to match the same performance level if it is 483 satisfied by guaranteed performance level. 484 active 485 Use amd_pstate_epp driver instance as the scaling driver, 486 driver provides a hint to the hardware if software wants 487 to bias toward performance (0x0) or energy efficiency (0xff) 488 to the CPPC firmware. then CPPC power algorithm will 489 calculate the runtime workload and adjust the realtime cores 490 frequency. 491 guided 492 Activate guided autonomous mode. Driver requests minimum and 493 maximum performance level and the platform autonomously 494 selects a performance level in this range and appropriate 495 to the current workload. 496 497 amd_prefcore= 498 [X86] 499 disable 500 Disable amd-pstate preferred core. 501 502 amd_dynamic_epp= 503 [X86] 504 disable 505 Disable amd-pstate dynamic EPP. 506 enable 507 Enable amd-pstate dynamic EPP. 508 509 amijoy.map= [HW,JOY] Amiga joystick support 510 Map of devices attached to JOY0DAT and JOY1DAT 511 Format: <a>,<b> 512 See also Documentation/input/joydev/joystick.rst 513 514 analog.map= [HW,JOY] Analog joystick and gamepad support 515 Specifies type or capabilities of an analog joystick 516 connected to one of 16 gameports 517 Format: <type1>,<type2>,..<type16> 518 519 apc= [HW,SPARC] 520 Power management functions (SPARCstation-4/5 + deriv.) 521 Format: noidle 522 Disable APC CPU standby support. SPARCstation-Fox does 523 not play well with APC CPU idle - disable it if you have 524 APC and your system crashes randomly. 525 526 apic [APIC,X86-64] Use IO-APIC. Default. 527 528 apic= [APIC,X86,EARLY] Advanced Programmable Interrupt Controller 529 Change the output verbosity while booting 530 Format: { quiet (default) | verbose | debug } 531 Change the amount of debugging information output 532 when initialising the APIC and IO-APIC components. 533 534 apic_extnmi= [APIC,X86,EARLY] External NMI delivery setting 535 Format: { bsp (default) | all | none } 536 bsp: External NMI is delivered only to CPU 0 537 all: External NMIs are broadcast to all CPUs as a 538 backup of CPU 0 539 none: External NMI is masked for all CPUs. This is 540 useful so that a dump capture kernel won't be 541 shot down by NMI 542 543 apicpmtimer Do APIC timer calibration using the pmtimer. Implies 544 apicmaintimer. Useful when your PIT timer is totally 545 broken. 546 547 autoconf= [IPV6] 548 See Documentation/networking/ipv6.rst. 549 550 apm= [APM] Advanced Power Management 551 See header of arch/x86/kernel/apm_32.c. 552 553 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 554 Format: { "0" | "1" } 555 See security/apparmor/Kconfig help text 556 0 -- disable. 557 1 -- enable. 558 Default value is set via kernel config option. 559 560 arm64.no32bit_el0 [ARM64] Unconditionally disable the execution of 561 32 bit applications. 562 563 arm64.nobti [ARM64] Unconditionally disable Branch Target 564 Identification support 565 566 arm64.nogcs [ARM64] Unconditionally disable Guarded Control Stack 567 support 568 569 arm64.nohaft [ARM64] Unconditionally disable Hardware managed Access 570 Flag for Table descriptors support 571 572 arm64.nomops [ARM64] Unconditionally disable Memory Copy and Memory 573 Set instructions support 574 575 arm64.nompam [ARM64] Unconditionally disable Memory Partitioning And 576 Monitoring support 577 578 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension 579 support 580 581 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 582 support 583 584 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix 585 Extension support 586 587 arm64.nosve [ARM64] Unconditionally disable Scalable Vector 588 Extension support 589 590 ataflop= [HW,M68k] 591 592 atarimouse= [HW,MOUSE] Atari Mouse 593 594 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 595 EzKey and similar keyboards 596 597 atkbd.reset= [HW] Reset keyboard during initialization 598 599 atkbd.set= [HW] Select keyboard code set 600 Format: <int> (2 = AT (default), 3 = PS/2) 601 602 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 603 keyboards 604 605 atkbd.softraw= [HW] Choose between synthetic and real raw mode 606 Format: <bool> (0 = real, 1 = synthetic (default)) 607 608 atkbd.softrepeat= [HW] 609 Use software keyboard repeat 610 611 audit= [KNL] Enable the audit sub-system 612 Format: { "0" | "1" | "off" | "on" } 613 0 | off - kernel audit is disabled and can not be 614 enabled until the next reboot 615 unset - kernel audit is initialized but disabled and 616 will be fully enabled by the userspace auditd. 617 1 | on - kernel audit is initialized and partially 618 enabled, storing at most audit_backlog_limit 619 messages in RAM until it is fully enabled by the 620 userspace auditd. 621 Default: unset 622 623 audit_backlog_limit= [KNL] Set the audit queue size limit. 624 Format: <int> (must be >=0) 625 Default: 64 626 627 bau= [X86_UV] Enable the BAU on SGI UV. The default 628 behavior is to disable the BAU (i.e. bau=0). 629 Format: { "0" | "1" } 630 0 - Disable the BAU. 631 1 - Enable the BAU. 632 unset - Disable the BAU. 633 634 bdev_allow_write_mounted= 635 Format: <bool> 636 Control the ability to open a mounted block device 637 for writing, i.e., allow / disallow writes that bypass 638 the FS. This was implemented as a means to prevent 639 fuzzers from crashing the kernel by overwriting the 640 metadata underneath a mounted FS without its awareness. 641 This also prevents destructive formatting of mounted 642 filesystems by naive storage tooling that don't use 643 O_EXCL. Default is Y and can be changed through the 644 Kconfig option CONFIG_BLK_DEV_WRITE_MOUNTED. 645 646 bert_disable [ACPI] 647 Disable BERT OS support on buggy BIOSes. 648 649 bgrt_disable [ACPI,X86,EARLY] 650 Disable BGRT to avoid flickering OEM logo. 651 652 blkdevparts= Manual partition parsing of block device(s) for 653 embedded devices based on command line input. 654 See Documentation/block/cmdline-partition.rst 655 656 boot_delay= [KNL,EARLY] 657 Milliseconds to delay each printk during boot. 658 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled, 659 and you may also have to specify "lpj=". Boot_delay 660 values larger than 10 seconds (10000) are assumed 661 erroneous and ignored. 662 Format: integer 663 664 bootconfig [KNL,EARLY] 665 Extended command line options can be added to an initrd 666 and this will cause the kernel to look for it. 667 668 See Documentation/admin-guide/bootconfig.rst 669 670 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 671 bttv.radio= Most important insmod options are available as 672 kernel args too. 673 bttv.pll= See Documentation/admin-guide/media/bttv.rst 674 bttv.tuner= 675 676 bulk_remove=off [PPC] This parameter disables the use of the pSeries 677 firmware feature for flushing multiple hpte entries 678 at a time. 679 680 c101= [NET] Moxa C101 synchronous serial card 681 682 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 683 Sometimes CPU hardware bugs make them report the cache 684 size incorrectly. The kernel will attempt work arounds 685 to fix known problems, but for some CPUs it is not 686 possible to determine what the correct size should be. 687 This option provides an override for these situations. 688 689 carrier_timeout= 690 [NET] Specifies amount of time (in seconds) that 691 the kernel should wait for a network carrier. By default 692 it waits 120 seconds. 693 694 ca_keys= [KEYS] This parameter identifies a specific key(s) on 695 the system trusted keyring to be used for certificate 696 trust validation. 697 format: { id:<keyid> | builtin } 698 699 cca= [MIPS,EARLY] Override the kernel pages' cache coherency 700 algorithm. Accepted values range from 0 to 7 701 inclusive. See arch/mips/include/asm/pgtable-bits.h 702 for platform specific values (SB1, Loongson3 and 703 others). 704 705 ccw_timeout_log [S390] 706 See Documentation/arch/s390/common_io.rst for details. 707 708 cfi= [X86-64] Set Control Flow Integrity checking features 709 when CONFIG_FINEIBT is enabled. 710 Format: feature[,feature...] 711 Default: auto 712 713 auto: Use FineIBT if IBT available, otherwise kCFI. 714 Under FineIBT, enable "paranoid" mode when 715 FRED is not available. 716 off: Turn off CFI checking. 717 kcfi: Use kCFI (disable FineIBT). 718 fineibt: Use FineIBT (even if IBT not available). 719 norand: Do not re-randomize CFI hashes. 720 paranoid: Add caller hash checking under FineIBT. 721 bhi: Enable register poisoning to stop speculation 722 across FineIBT. (Disabled by default.) 723 warn: Do not enforce CFI checking: warn only. 724 debug: Report CFI initialization details. 725 726 cgroup_disable= [KNL] Disable a particular controller or optional feature 727 Format: {name of the controller(s) or feature(s) to disable} 728 The effects of cgroup_disable=foo are: 729 - foo isn't auto-mounted if you mount all cgroups in 730 a single hierarchy 731 - foo isn't visible as an individually mountable 732 subsystem 733 - if foo is an optional feature then the feature is 734 disabled and corresponding cgroup files are not 735 created 736 {Currently only "memory" controller deal with this and 737 cut the overhead, others just disable the usage. So 738 only cgroup_disable=memory is actually worthy} 739 Specifying "pressure" disables per-cgroup pressure 740 stall information accounting feature 741 742 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 743 Format: { { controller | "all" | "named" } 744 [,{ controller | "all" | "named" }...] } 745 Like cgroup_disable, but only applies to cgroup v1; 746 the blacklisted controllers remain available in cgroup2. 747 "all" blacklists all controllers and "named" disables 748 named mounts. Specifying both "all" and "named" disables 749 all v1 hierarchies. 750 751 cgroup_v1_proc= [KNL] Show also missing controllers in /proc/cgroups 752 Format: { "true" | "false" } 753 /proc/cgroups lists only v1 controllers by default. 754 This compatibility option enables listing also v2 755 controllers (whose v1 code is not compiled!), so that 756 semi-legacy software can check this file to decide 757 about usage of v2 (sic) controllers. 758 759 cgroup_favordynmods= [KNL] Enable or Disable favordynmods. 760 Format: { "true" | "false" } 761 Defaults to the value of CONFIG_CGROUP_FAVOR_DYNMODS. 762 763 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 764 Format: <string> 765 nosocket -- Disable socket memory accounting. 766 nokmem -- Disable kernel memory accounting. 767 nobpf -- Disable BPF memory accounting. 768 769 check_pages= [MM,EARLY] Enable sanity checking of pages after 770 allocations / before freeing. This adds checks to catch 771 double-frees, use-after-frees, and other sources of 772 page corruption by inspecting page internals (flags, 773 mapcount/refcount, memcg_data, etc.). 774 Format: { "0" | "1" } 775 Default: 0 (1 if CONFIG_DEBUG_VM is set) 776 777 checkreqprot= [SELINUX] Set initial checkreqprot flag value. 778 Format: { "0" | "1" } 779 See security/selinux/Kconfig help text. 780 0 -- check protection applied by kernel (includes 781 any implied execute protection). 782 1 -- check protection requested by application. 783 Default value is set via a kernel config option. 784 Value can be changed at runtime via 785 /sys/fs/selinux/checkreqprot. 786 Setting checkreqprot to 1 is deprecated. 787 788 cio_ignore= [S390] 789 See Documentation/arch/s390/common_io.rst for details. 790 791 clk_ignore_unused 792 [CLK] 793 Prevents the clock framework from automatically gating 794 clocks that have not been explicitly enabled by a Linux 795 device driver but are enabled in hardware at reset or 796 by the bootloader/firmware. Note that this does not 797 force such clocks to be always-on nor does it reserve 798 those clocks in any way. This parameter is useful for 799 debug and development, but should not be needed on a 800 platform with proper driver support. For more 801 information, see Documentation/driver-api/clk.rst. 802 803 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 804 [Deprecated] 805 Forces specified clocksource (if available) to be used 806 when calculating gettimeofday(). If specified 807 clocksource is not available, it defaults to PIT. 808 Format: { pit | tsc | cyclone | pmtmr } 809 810 clocksource= Override the default clocksource 811 Format: <string> 812 Override the default clocksource and use the clocksource 813 with the name specified. 814 Some clocksource names to choose from, depending on 815 the platform: 816 [all] jiffies (this is the base, fallback clocksource) 817 [ACPI] acpi_pm 818 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 819 pxa_timer,timer3,32k_counter,timer0_1 820 [X86-32] pit,hpet,tsc; 821 scx200_hrt on Geode; cyclone on IBM x440 822 [MIPS] MIPS 823 [PARISC] cr16 824 [S390] tod 825 [SH] SuperH 826 [SPARC64] tick 827 [X86-64] hpet,tsc 828 829 clocksource.arm_arch_timer.evtstrm= 830 [ARM,ARM64,EARLY] 831 Format: <bool> 832 Enable/disable the eventstream feature of the ARM 833 architected timer so that code using WFE-based polling 834 loops can be debugged more effectively on production 835 systems. 836 837 clocksource.verify_n_cpus= [KNL] 838 Limit the number of CPUs checked for clocksources 839 marked with CLOCK_SOURCE_VERIFY_PERCPU that 840 are marked unstable due to excessive skew. 841 A negative value says to check all CPUs, while 842 zero says not to check any. Values larger than 843 nr_cpu_ids are silently truncated to nr_cpu_ids. 844 The actual CPUs are chosen randomly, with 845 no replacement if the same CPU is chosen twice. 846 847 clocksource-wdtest.holdoff= [KNL] 848 Set the time in seconds that the clocksource 849 watchdog test waits before commencing its tests. 850 Defaults to zero when built as a module and to 851 10 seconds when built into the kernel. 852 853 cma=nn[MG]@[start[MG][-end[MG]]] 854 [KNL,CMA,EARLY] 855 Sets the size of kernel global memory area for 856 contiguous memory allocations and optionally the 857 placement constraint by the physical address range of 858 memory allocations. A value of 0 disables CMA 859 altogether. For more information, see 860 kernel/dma/contiguous.c 861 862 cma_pernuma=nn[MG] 863 [KNL,CMA,EARLY] 864 Sets the size of kernel per-numa memory area for 865 contiguous memory allocations. A value of 0 disables 866 per-numa CMA altogether. And If this option is not 867 specified, the default value is 0. 868 With per-numa CMA enabled, DMA users on node nid will 869 first try to allocate buffer from the pernuma area 870 which is located in node nid, if the allocation fails, 871 they will fallback to the global default memory area. 872 873 numa_cma=<node>:nn[MG][,<node>:nn[MG]] 874 [KNL,CMA,EARLY] 875 Sets the size of kernel numa memory area for 876 contiguous memory allocations. It will reserve CMA 877 area for the specified node. 878 879 If it is setup together with upper 'cmd_pernuma=' 880 (unlikely), its size setting takes priority for the 881 specified numa nodes. 882 883 With numa CMA enabled, DMA users on node nid will 884 first try to allocate buffer from the numa area 885 which is located in node nid, if the allocation fails, 886 they will fallback to the global default memory area. 887 888 cmo_free_hint= [PPC] Format: { yes | no } 889 Specify whether pages are marked as being inactive 890 when they are freed. This is used in CMO environments 891 to determine OS memory pressure for page stealing by 892 a hypervisor. 893 Default: yes 894 895 coherent_pool=nn[KMG] [ARM,KNL,EARLY] 896 Sets the size of memory pool for coherent, atomic dma 897 allocations, by default set to 256K. 898 899 condev= [HW,S390] console device 900 conmode= 901 902 con3215_drop= [S390,EARLY] 3215 console drop mode. 903 Format: y|n|Y|N|1|0 904 When set to true, drop data on the 3215 console when 905 the console buffer is full. In this case the 906 operator using a 3270 terminal emulator (for example 907 x3270) does not have to enter the clear key for the 908 console output to advance and the kernel to continue. 909 This leads to a much faster boot time when a 3270 910 terminal emulator is active. If no 3270 terminal 911 emulator is used, this parameter has no effect. 912 913 console= [KNL] Output console device and options. 914 915 tty<n> Use the virtual console device <n>. 916 917 ttyS<n>[,options] 918 ttyUSB0[,options] 919 Use the specified serial port. The options are of 920 the form "bbbbpnf", where "bbbb" is the baud rate, 921 "p" is parity ("n", "o", or "e"), "n" is number of 922 bits, and "f" is flow control ("r" for RTS or 923 omit it). Default is "9600n8". 924 925 See Documentation/admin-guide/serial-console.rst for more 926 information. See 927 Documentation/networking/netconsole.rst for an 928 alternative. 929 930 <DEVNAME>:<n>.<n>[,options] 931 Use the specified serial port on the serial core bus. 932 The addressing uses DEVNAME of the physical serial port 933 device, followed by the serial core controller instance, 934 and the serial port instance. The options are the same 935 as documented for the ttyS addressing above. 936 937 The mapping of the serial ports to the tty instances 938 can be viewed with: 939 940 $ ls -d /sys/bus/serial-base/devices/*:*.*/tty/* 941 /sys/bus/serial-base/devices/00:04:0.0/tty/ttyS0 942 943 In the above example, the console can be addressed with 944 console=00:04:0.0. Note that a console addressed this 945 way will only get added when the related device driver 946 is ready. The use of an earlycon parameter in addition to 947 the console may be desired for console output early on. 948 949 uart[8250],io,<addr>[,options] 950 uart[8250],mmio,<addr>[,options] 951 uart[8250],mmio16,<addr>[,options] 952 uart[8250],mmio32,<addr>[,options] 953 uart[8250],0x<addr>[,options] 954 Start an early, polled-mode console on the 8250/16550 955 UART at the specified I/O port or MMIO address, 956 switching to the matching ttyS device later. 957 MMIO inter-register address stride is either 8-bit 958 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 959 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 960 to be equivalent to 'mmio'. 'options' are specified in 961 the same format described for ttyS above; if unspecified, 962 the h/w is not re-initialized. 963 964 hvc<n> Use the hypervisor console device <n>. This is for 965 both Xen and PowerPC hypervisors. 966 967 { null | "" } 968 Use to disable console output, i.e., to have kernel 969 console messages discarded. 970 This must be the only console= parameter used on the 971 kernel command line. 972 973 If the device connected to the port is not a TTY but a braille 974 device, prepend "brl," before the device type, for instance 975 console=brl,ttyS0 976 For now, only VisioBraille is supported. 977 978 console_msg_format= 979 [KNL] Change console messages format 980 default 981 By default we print messages on consoles in 982 "[time stamp] text\n" format (time stamp may not be 983 printed, depending on CONFIG_PRINTK_TIME or 984 `printk_time' param). 985 syslog 986 Switch to syslog format: "<%u>[time stamp] text\n" 987 IOW, each message will have a facility and loglevel 988 prefix. The format is similar to one used by syslog() 989 syscall, or to executing "dmesg -S --raw" or to reading 990 from /proc/kmsg. 991 992 consoleblank= [KNL] The console blank (screen saver) timeout in 993 seconds. A value of 0 disables the blank timer. 994 Defaults to 0. 995 996 coredump_filter= 997 [KNL] Change the default value for 998 /proc/<pid>/coredump_filter. 999 See also Documentation/filesystems/proc.rst. 1000 1001 coresight_cpu_debug.enable 1002 [ARM,ARM64] 1003 Format: <bool> 1004 Enable/disable the CPU sampling based debugging. 1005 0: default value, disable debugging 1006 1: enable debugging at boot time 1007 1008 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 1009 Format: 1010 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 1011 1012 cpuidle.off=1 [CPU_IDLE] 1013 disable the cpuidle sub-system 1014 1015 cpuidle.governor= 1016 [CPU_IDLE] Name of the cpuidle governor to use. 1017 1018 cpufreq.off=1 [CPU_FREQ] 1019 disable the cpufreq sub-system 1020 1021 cpufreq.default_governor= 1022 [CPU_FREQ] Name of the default cpufreq governor or 1023 policy to use. This governor must be registered in the 1024 kernel before the cpufreq driver probes. 1025 1026 cpu_init_udelay=N 1027 [X86,EARLY] Delay for N microsec between assert and de-assert 1028 of APIC INIT to start processors. This delay occurs 1029 on every CPU online, such as boot, and resume from suspend. 1030 Default: 10000 1031 1032 cpuhp.parallel= 1033 [SMP] Enable/disable parallel bringup of secondary CPUs 1034 Format: <bool> 1035 Default is enabled if CONFIG_HOTPLUG_PARALLEL=y. Otherwise 1036 the parameter has no effect. 1037 1038 crash_kexec_post_notifiers 1039 Only jump to kdump kernel after running the panic 1040 notifiers and dumping kmsg. This option increases 1041 the risks of a kdump failure, since some panic 1042 notifiers can make the crashed kernel more unstable. 1043 In configurations where kdump may not be reliable, 1044 running the panic notifiers could allow collecting 1045 more data on dmesg, like stack traces from other CPUS 1046 or extra data dumped by panic_print. Note that some 1047 configurations enable this option unconditionally, 1048 like Hyper-V, PowerPC (fadump) and AMD SEV-SNP. 1049 1050 crashkernel=size[KMG][@offset[KMG]] 1051 [KNL,EARLY] Using kexec, Linux can switch to a 'crash kernel' 1052 upon panic. This parameter reserves the physical 1053 memory region [offset, offset + size] for that kernel 1054 image. If '@offset' is omitted, then a suitable offset 1055 is selected automatically. 1056 [KNL, X86-64, ARM64, RISCV, LoongArch] Select a region 1057 under 4G first, and fall back to reserve region above 1058 4G when '@offset' hasn't been specified. 1059 See Documentation/admin-guide/kdump/kdump.rst for further details. 1060 1061 crashkernel=range1:size1[,range2:size2,...][@offset] 1062 [KNL] Same as above, but depends on the memory 1063 in the running system. The syntax of range is 1064 start-[end] where start and end are both 1065 a memory unit (amount[KMG]). See also 1066 Documentation/admin-guide/kdump/kdump.rst for an example. 1067 1068 crashkernel=size[KMG],high 1069 [KNL, X86-64, ARM64, RISCV, LoongArch] range could be 1070 above 4G. 1071 Allow kernel to allocate physical memory region from top, 1072 so could be above 4G if system have more than 4G ram 1073 installed. Otherwise memory region will be allocated 1074 below 4G, if available. 1075 It will be ignored if crashkernel=X is specified. 1076 crashkernel=size[KMG],low 1077 [KNL, X86-64, ARM64, RISCV, LoongArch] range under 4G. 1078 When crashkernel=X,high is passed, kernel could allocate 1079 physical memory region above 4G, that cause second kernel 1080 crash on system that require some amount of low memory, 1081 e.g. swiotlb requires at least 64M+32K low memory, also 1082 enough extra low memory is needed to make sure DMA buffers 1083 for 32-bit devices won't run out. Kernel would try to allocate 1084 default size of memory below 4G automatically. The default 1085 size is platform dependent. 1086 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB) 1087 --> arm64: 128MiB 1088 --> riscv: 128MiB 1089 --> loongarch: 128MiB 1090 This one lets the user specify own low range under 4G 1091 for second kernel instead. 1092 0: to disable low allocation. 1093 It will be ignored when crashkernel=X,high is not used 1094 or memory reserved is below 4G. 1095 crashkernel=size[KMG],cma 1096 [KNL, X86, ARM64, RISCV, PPC] Reserve additional crash 1097 kernel memory from CMA. This reservation is usable by 1098 the first system's userspace memory and kernel movable 1099 allocations (memory balloon, zswap). Pages allocated 1100 from this memory range will not be included in the vmcore 1101 so this should not be used if dumping of userspace memory 1102 is intended and it has to be expected that some movable 1103 kernel pages may be missing from the dump. 1104 1105 A standard crashkernel reservation, as described above, 1106 is still needed to hold the crash kernel and initrd. 1107 1108 This option increases the risk of a kdump failure: DMA 1109 transfers configured by the first kernel may end up 1110 corrupting the second kernel's memory. 1111 1112 This reservation method is intended for systems that 1113 can't afford to sacrifice enough memory for standard 1114 crashkernel reservation and where less reliable and 1115 possibly incomplete kdump is preferable to no kdump at 1116 all. 1117 1118 cryptomgr.notests 1119 [KNL] Disable crypto self-tests 1120 1121 cs89x0_dma= [HW,NET] 1122 Format: <dma> 1123 1124 cs89x0_media= [HW,NET] 1125 Format: { rj45 | aui | bnc } 1126 1127 csdlock_debug= [KNL] Enable or disable debug add-ons of cross-CPU 1128 function call handling. When switched on, 1129 additional debug data is printed to the console 1130 in case a hanging CPU is detected, and that 1131 CPU is pinged again in order to try to resolve 1132 the hang situation. The default value of this 1133 option depends on the CSD_LOCK_WAIT_DEBUG_DEFAULT 1134 Kconfig option. 1135 1136 dasd= [HW,NET] 1137 See header of drivers/s390/block/dasd_devmap.c. 1138 1139 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 1140 (one device per port) 1141 Format: <port#>,<type> 1142 See also Documentation/input/devices/joystick-parport.rst 1143 1144 debug [KNL,EARLY] Enable kernel debugging (events log level). 1145 1146 debug_boot_weak_hash 1147 [KNL,EARLY] Enable printing [hashed] pointers early in the 1148 boot sequence. If enabled, we use a weak hash instead 1149 of siphash to hash pointers. Use this option if you are 1150 seeing instances of '(___ptrval___)') and need to see a 1151 value (hashed pointer) instead. Cryptographically 1152 insecure, please do not use on production kernels. 1153 1154 debug_locks_verbose= 1155 [KNL] verbose locking self-tests 1156 Format: <int> 1157 Print debugging info while doing the locking API 1158 self-tests. 1159 Bitmask for the various LOCKTYPE_ tests. Defaults to 0 1160 (no extra messages), setting it to -1 (all bits set) 1161 will print _a_lot_ more information - normally only 1162 useful to lockdep developers. 1163 1164 debug_objects [KNL,EARLY] Enable object debugging 1165 1166 debug_guardpage_minorder= 1167 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this 1168 parameter allows control of the order of pages that will 1169 be intentionally kept free (and hence protected) by the 1170 buddy allocator. Bigger value increase the probability 1171 of catching random memory corruption, but reduce the 1172 amount of memory for normal system use. The maximum 1173 possible value is MAX_PAGE_ORDER/2. Setting this 1174 parameter to 1 or 2 should be enough to identify most 1175 random memory corruption problems caused by bugs in 1176 kernel or driver code when a CPU writes to (or reads 1177 from) a random memory location. Note that there exists 1178 a class of memory corruptions problems caused by buggy 1179 H/W or F/W or by drivers badly programming DMA 1180 (basically when memory is written at bus level and the 1181 CPU MMU is bypassed) which are not detectable by 1182 CONFIG_DEBUG_PAGEALLOC, hence this option will not 1183 help tracking down these problems. 1184 1185 debug_pagealloc= 1186 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 1187 enables the feature at boot time. By default, it is 1188 disabled and the system will work mostly the same as a 1189 kernel built without CONFIG_DEBUG_PAGEALLOC. 1190 Note: to get most of debug_pagealloc error reports, it's 1191 useful to also enable the page_owner functionality. 1192 on: enable the feature 1193 1194 debugfs= [KNL,EARLY] This parameter enables what is exposed to 1195 userspace and debugfs internal clients. 1196 Format: { on, off } 1197 on: All functions are enabled. 1198 off: Filesystem is not registered and clients 1199 get a -EPERM as result when trying to register files 1200 or directories within debugfs. 1201 This is equivalent of the runtime functionality if 1202 debugfs was not enabled in the kernel at all. 1203 Default value is set in build-time with a kernel configuration. 1204 1205 debugpat [X86] Enable PAT debugging 1206 1207 default_hugepagesz= 1208 [HW] The size of the default HugeTLB page. This is 1209 the size represented by the legacy /proc/ hugepages 1210 APIs. In addition, this is the default hugetlb size 1211 used for shmget(), mmap() and mounting hugetlbfs 1212 filesystems. If not specified, defaults to the 1213 architecture's default huge page size. Huge page 1214 sizes are architecture dependent. See also 1215 Documentation/admin-guide/mm/hugetlbpage.rst. 1216 Format: size[KMG] 1217 1218 deferred_probe_timeout= 1219 [KNL] Debugging option to set a timeout in seconds for 1220 deferred probe to give up waiting on dependencies to 1221 probe. Only specific dependencies (subsystems or 1222 drivers) that have opted in will be ignored. A timeout 1223 of 0 will timeout at the end of initcalls; a negative 1224 value is treated as an infinite timeout value. If the 1225 timeout hasn't expired, it'll be restarted by each 1226 successful driver registration. This option will also 1227 dump out devices still on the deferred probe list after 1228 retrying. 1229 1230 delayacct [KNL] Enable per-task delay accounting 1231 1232 dell_smm_hwmon.ignore_dmi= 1233 [HW] Continue probing hardware even if DMI data 1234 indicates that the driver is running on unsupported 1235 hardware. 1236 1237 dell_smm_hwmon.force= 1238 [HW] Activate driver even if SMM BIOS signature does 1239 not match list of supported models and enable otherwise 1240 blacklisted features. 1241 1242 dell_smm_hwmon.power_status= 1243 [HW] Report power status in /proc/i8k 1244 (disabled by default). 1245 1246 dell_smm_hwmon.restricted= 1247 [HW] Allow controlling fans only if SYS_ADMIN 1248 capability is set. 1249 1250 dell_smm_hwmon.fan_mult= 1251 [HW] Factor to multiply fan speed with. 1252 1253 dell_smm_hwmon.fan_max= 1254 [HW] Maximum configurable fan speed. 1255 1256 dfltcc= [HW,S390] 1257 Format: { on | off | def_only | inf_only | always } 1258 on: s390 zlib hardware support for compression on 1259 level 1 and decompression (default) 1260 off: No s390 zlib hardware support 1261 def_only: s390 zlib hardware support for deflate 1262 only (compression on level 1) 1263 inf_only: s390 zlib hardware support for inflate 1264 only (decompression) 1265 always: Same as 'on' but ignores the selected compression 1266 level always using hardware support (used for debugging) 1267 1268 dhash_entries= [KNL] 1269 Set number of hash buckets for dentry cache. 1270 1271 disable_1tb_segments [PPC,EARLY] 1272 Disables the use of 1TB hash page table segments. This 1273 causes the kernel to fall back to 256MB segments which 1274 can be useful when debugging issues that require an SLB 1275 miss to occur. 1276 1277 disable= [IPV6] 1278 See Documentation/networking/ipv6.rst. 1279 1280 disable_radix [PPC,EARLY] 1281 Disable RADIX MMU mode on POWER9 1282 1283 disable_tlbie [PPC] 1284 Disable TLBIE instruction. Currently does not work 1285 with KVM, with HASH MMU, or with coherent accelerators. 1286 1287 disable_ddw [PPC/PSERIES,EARLY] 1288 Disable Dynamic DMA Window support. Use this 1289 to workaround buggy firmware. 1290 1291 disable_ipv6= [IPV6] 1292 See Documentation/networking/ipv6.rst. 1293 1294 disable_mtrr_cleanup [X86,EARLY] 1295 The kernel tries to adjust MTRR layout from continuous 1296 to discrete, to make X server driver able to add WB 1297 entry later. This parameter disables that. 1298 1299 disable_mtrr_trim [X86, Intel and AMD only,EARLY] 1300 By default the kernel will trim any uncacheable 1301 memory out of your available memory pool based on 1302 MTRR settings. This parameter disables that behavior, 1303 possibly causing your machine to run very slowly. 1304 1305 disable_timer_pin_1 [X86,EARLY] 1306 Disable PIN 1 of APIC timer 1307 Can be useful to work around chipset bugs. 1308 1309 dis_ucode_ldr [X86] Disable the microcode loader. 1310 1311 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 1312 this option disables the debugging code at boot. 1313 1314 dma_debug_entries=<number> 1315 This option allows to tune the number of preallocated 1316 entries for DMA-API debugging code. One entry is 1317 required per DMA-API allocation. Use this if the 1318 DMA-API debugging code disables itself because the 1319 architectural default is too low. 1320 1321 dma_debug_driver=<driver_name> 1322 With this option the DMA-API debugging driver 1323 filter feature can be enabled at boot time. Just 1324 pass the driver to filter for as the parameter. 1325 The filter can be disabled or changed to another 1326 driver later using sysfs. 1327 1328 reg_file_data_sampling= 1329 [X86] Controls mitigation for Register File Data 1330 Sampling (RFDS) vulnerability. RFDS is a CPU 1331 vulnerability which may allow userspace to infer 1332 kernel data values previously stored in floating point 1333 registers, vector registers, or integer registers. 1334 RFDS only affects Intel Atom processors. 1335 1336 on: Turns ON the mitigation. 1337 off: Turns OFF the mitigation. 1338 1339 This parameter overrides the compile time default set 1340 by CONFIG_MITIGATION_RFDS. Mitigation cannot be 1341 disabled when other VERW based mitigations (like MDS) 1342 are enabled. In order to disable RFDS mitigation all 1343 VERW based mitigations need to be disabled. 1344 1345 For details see: 1346 Documentation/admin-guide/hw-vuln/reg-file-data-sampling.rst 1347 1348 dm_verity.keyring_unsealed= 1349 [KNL] When set to 1, leave the dm-verity keyring 1350 unsealed after initialization so userspace can 1351 provision keys. Once the keyring is restricted 1352 it becomes active and is searched during signature 1353 verification. 1354 1355 driver_async_probe= [KNL] 1356 List of driver names to be probed asynchronously. * 1357 matches with all driver names. If * is specified, the 1358 rest of the listed driver names are those that will NOT 1359 match the *. 1360 Format: <driver_name1>,<driver_name2>... 1361 1362 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1363 Broken monitors, graphic adapters, KVMs and EDIDless 1364 panels may send no or incorrect EDID data sets. 1365 This parameter allows to specify an EDID data sets 1366 in the /lib/firmware directory that are used instead. 1367 An EDID data set will only be used for a particular 1368 connector, if its name and a colon are prepended to 1369 the EDID name. Each connector may use a unique EDID 1370 data set by separating the files with a comma. An EDID 1371 data set with no connector name will be used for 1372 any connectors not explicitly specified. 1373 1374 dscc4.setup= [NET] 1375 1376 dt_cpu_ftrs= [PPC,EARLY] 1377 Format: {"off" | "known"} 1378 Control how the dt_cpu_ftrs device-tree binding is 1379 used for CPU feature discovery and setup (if it 1380 exists). 1381 off: Do not use it, fall back to legacy cpu table. 1382 known: Do not pass through unknown features to guests 1383 or userspace, only those that the kernel is aware of. 1384 1385 dump_apple_properties [X86] 1386 Dump name and content of EFI device properties on 1387 x86 Macs. Useful for driver authors to determine 1388 what data is available or for reverse-engineering. 1389 1390 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1391 <module>.dyndbg[="val"] 1392 Enable debug messages at boot time. See 1393 Documentation/admin-guide/dynamic-debug-howto.rst 1394 for details. 1395 1396 early_ioremap_debug [KNL,EARLY] 1397 Enable debug messages in early_ioremap support. This 1398 is useful for tracking down temporary early mappings 1399 which are not unmapped. 1400 1401 earlycon= [KNL,EARLY] Output early console device and options. 1402 1403 When used with no options, the early console is 1404 determined by stdout-path property in device tree's 1405 chosen node or the ACPI SPCR table if supported by 1406 the platform. 1407 1408 cdns,<addr>[,options] 1409 Start an early, polled-mode console on a Cadence 1410 (xuartps) serial port at the specified address. Only 1411 supported option is baud rate. If baud rate is not 1412 specified, the serial port must already be setup and 1413 configured. 1414 1415 uart[8250],io,<addr>[,options[,uartclk]] 1416 uart[8250],mmio,<addr>[,options[,uartclk]] 1417 uart[8250],mmio32,<addr>[,options[,uartclk]] 1418 uart[8250],mmio32be,<addr>[,options[,uartclk]] 1419 uart[8250],0x<addr>[,options] 1420 Start an early, polled-mode console on the 8250/16550 1421 UART at the specified I/O port or MMIO address. 1422 MMIO inter-register address stride is either 8-bit 1423 (mmio) or 32-bit (mmio32 or mmio32be). 1424 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1425 to be equivalent to 'mmio'. 'options' are specified 1426 in the same format described for "console=ttyS<n>"; if 1427 unspecified, the h/w is not initialized. 'uartclk' is 1428 the uart clock frequency; if unspecified, it is set 1429 to 'BASE_BAUD' * 16. 1430 1431 pl011,<addr> 1432 pl011,mmio32,<addr> 1433 Start an early, polled-mode console on a pl011 serial 1434 port at the specified address. The pl011 serial port 1435 must already be setup and configured. Options are not 1436 yet supported. If 'mmio32' is specified, then only 1437 the driver will use only 32-bit accessors to read/write 1438 the device registers. 1439 1440 liteuart,<addr> 1441 Start an early console on a litex serial port at the 1442 specified address. The serial port must already be 1443 setup and configured. Options are not yet supported. 1444 1445 meson,<addr> 1446 Start an early, polled-mode console on a meson serial 1447 port at the specified address. The serial port must 1448 already be setup and configured. Options are not yet 1449 supported. 1450 1451 msm_serial,<addr> 1452 Start an early, polled-mode console on an msm serial 1453 port at the specified address. The serial port 1454 must already be setup and configured. Options are not 1455 yet supported. 1456 1457 msm_serial_dm,<addr> 1458 Start an early, polled-mode console on an msm serial 1459 dm port at the specified address. The serial port 1460 must already be setup and configured. Options are not 1461 yet supported. 1462 1463 owl,<addr> 1464 Start an early, polled-mode console on a serial port 1465 of an Actions Semi SoC, such as S500 or S900, at the 1466 specified address. The serial port must already be 1467 setup and configured. Options are not yet supported. 1468 1469 rda,<addr> 1470 Start an early, polled-mode console on a serial port 1471 of an RDA Micro SoC, such as RDA8810PL, at the 1472 specified address. The serial port must already be 1473 setup and configured. Options are not yet supported. 1474 1475 sbi 1476 Use RISC-V SBI (Supervisor Binary Interface) for early 1477 console. 1478 1479 smh Use ARM semihosting calls for early console. 1480 1481 s3c2410,<addr> 1482 s3c2412,<addr> 1483 s3c2440,<addr> 1484 s3c6400,<addr> 1485 s5pv210,<addr> 1486 exynos4210,<addr> 1487 Use early console provided by serial driver available 1488 on Samsung SoCs, requires selecting proper type and 1489 a correct base address of the selected UART port. The 1490 serial port must already be setup and configured. 1491 Options are not yet supported. 1492 1493 lantiq,<addr> 1494 Start an early, polled-mode console on a lantiq serial 1495 (lqasc) port at the specified address. The serial port 1496 must already be setup and configured. Options are not 1497 yet supported. 1498 1499 lpuart,<addr> 1500 lpuart32,<addr> 1501 Use early console provided by Freescale LP UART driver 1502 found on Freescale Vybrid and QorIQ LS1021A processors. 1503 A valid base address must be provided, and the serial 1504 port must already be setup and configured. 1505 1506 ec_imx21,<addr> 1507 ec_imx6q,<addr> 1508 Start an early, polled-mode, output-only console on the 1509 Freescale i.MX UART at the specified address. The UART 1510 must already be setup and configured. 1511 1512 ar3700_uart,<addr> 1513 Start an early, polled-mode console on the 1514 Armada 3700 serial port at the specified 1515 address. The serial port must already be setup 1516 and configured. Options are not yet supported. 1517 1518 qcom_geni,<addr> 1519 Start an early, polled-mode console on a Qualcomm 1520 Generic Interface (GENI) based serial port at the 1521 specified address. The serial port must already be 1522 setup and configured. Options are not yet supported. 1523 1524 efifb,[options] 1525 Start an early, unaccelerated console on the EFI 1526 memory mapped framebuffer (if available). On cache 1527 coherent non-x86 systems that use system memory for 1528 the framebuffer, pass the 'ram' option so that it is 1529 mapped with the correct attributes. 1530 1531 linflex,<addr> 1532 Use early console provided by Freescale LINFlexD UART 1533 serial driver for NXP S32V234 SoCs. A valid base 1534 address must be provided, and the serial port must 1535 already be setup and configured. 1536 1537 earlyprintk= [X86,SH,ARM,M68k,S390,UM,EARLY] 1538 earlyprintk=vga 1539 earlyprintk=sclp 1540 earlyprintk=xen 1541 earlyprintk=serial[,ttySn[,baudrate]] 1542 earlyprintk=serial[,0x...[,baudrate]] 1543 earlyprintk=ttySn[,baudrate] 1544 earlyprintk=dbgp[debugController#] 1545 earlyprintk=mmio32,membase[,{nocfg|baudrate}] 1546 earlyprintk=pciserial[,force],bus:device.function[,{nocfg|baudrate}] 1547 earlyprintk=xdbc[xhciController#] 1548 earlyprintk=bios 1549 1550 earlyprintk is useful when the kernel crashes before 1551 the normal console is initialized. It is not enabled by 1552 default because it has some cosmetic problems. 1553 1554 Use "nocfg" to skip UART configuration, assume 1555 BIOS/firmware has configured UART correctly. 1556 1557 Append ",keep" to not disable it when the real console 1558 takes over. 1559 1560 Only one of vga, serial, or usb debug port can 1561 be used at a time. 1562 1563 Currently only ttyS0 and ttyS1 may be specified by 1564 name. Other I/O ports may be explicitly specified 1565 on some architectures (x86 and arm at least) by 1566 replacing ttySn with an I/O port address, like this: 1567 earlyprintk=serial,0x1008,115200 1568 You can find the port for a given device in 1569 /proc/tty/driver/serial: 1570 2: uart:ST16650V2 port:00001008 irq:18 ... 1571 1572 Interaction with the standard serial driver is not 1573 very good. 1574 1575 The VGA output is eventually overwritten by 1576 the real console. 1577 1578 The xen option can only be used in Xen domains. 1579 1580 The sclp output can only be used on s390. 1581 1582 The bios output can only be used on SuperH. 1583 1584 The optional "force" to "pciserial" enables use of a 1585 PCI device even when its classcode is not of the 1586 UART class. 1587 1588 edac_report= [HW,EDAC] Control how to report EDAC event 1589 Format: {"on" | "off" | "force"} 1590 on: enable EDAC to report H/W event. May be overridden 1591 by other higher priority error reporting module. 1592 off: disable H/W event reporting through EDAC. 1593 force: enforce the use of EDAC to report H/W event. 1594 default: on. 1595 1596 edd= [EDD] 1597 Format: {"off" | "on" | "skip[mbr]"} 1598 1599 efi= [EFI,EARLY] 1600 Format: { "debug", "disable_early_pci_dma", 1601 "nochunk", "noruntime", "nosoftreserve", 1602 "novamap", "no_disable_early_pci_dma" } 1603 debug: enable misc debug output. 1604 disable_early_pci_dma: disable the busmaster bit on all 1605 PCI bridges while in the EFI boot stub. 1606 nochunk: disable reading files in "chunks" in the EFI 1607 boot stub, as chunking can cause problems with some 1608 firmware implementations. 1609 noruntime : disable EFI runtime services support 1610 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1611 attribute may cause the kernel to reserve the 1612 memory range for a memory mapping driver to 1613 claim. Specify efi=nosoftreserve to disable this 1614 reservation and treat the memory by its base type 1615 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1616 novamap: do not call SetVirtualAddressMap(). 1617 no_disable_early_pci_dma: Leave the busmaster bit set 1618 on all PCI bridges while in the EFI boot stub 1619 1620 efi_no_storage_paranoia [EFI,X86,EARLY] 1621 Using this parameter you can use more than 50% of 1622 your efi variable storage. Use this parameter only if 1623 you are really sure that your UEFI does sane gc and 1624 fulfills the spec otherwise your board may brick. 1625 1626 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1627 that is to be dynamically loaded by Linux. If there are 1628 multiple variables with the same name but with different 1629 vendor GUIDs, all of them will be loaded. See 1630 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1631 1632 1633 eisa_irq_edge= [PARISC,HW] 1634 See header of drivers/parisc/eisa.c. 1635 1636 ekgdboc= [X86,KGDB,EARLY] Allow early kernel console debugging 1637 Format: ekgdboc=kbd 1638 1639 This is designed to be used in conjunction with 1640 the boot argument: earlyprintk=vga 1641 1642 This parameter works in place of the kgdboc parameter 1643 but can only be used if the backing tty is available 1644 very early in the boot process. For early debugging 1645 via a serial port see kgdboc_earlycon instead. 1646 1647 elfcorehdr=[size[KMG]@]offset[KMG] [PPC,SH,X86,S390,EARLY] 1648 Specifies physical address of start of kernel core 1649 image elf header and optionally the size. Generally 1650 kexec loader will pass this option to capture kernel. 1651 See Documentation/admin-guide/kdump/kdump.rst for details. 1652 1653 enable_mtrr_cleanup [X86,EARLY] 1654 The kernel tries to adjust MTRR layout from continuous 1655 to discrete, to make X server driver able to add WB 1656 entry later. This parameter enables that. 1657 1658 enable_timer_pin_1 [X86] 1659 Enable PIN 1 of APIC timer 1660 Can be useful to work around chipset bugs 1661 (in particular on some ATI chipsets). 1662 The kernel tries to set a reasonable default. 1663 1664 enforcing= [SELINUX] Set initial enforcing status. 1665 Format: {"0" | "1"} 1666 See security/selinux/Kconfig help text. 1667 0 -- permissive (log only, no denials). 1668 1 -- enforcing (deny and log). 1669 Default value is 0. 1670 Value can be changed at runtime via 1671 /sys/fs/selinux/enforce. 1672 1673 erst_disable [ACPI] 1674 Disable Error Record Serialization Table (ERST) 1675 support. 1676 1677 ether= [HW,NET] Ethernet cards parameters 1678 This option is obsoleted by the "netdev=" option, which 1679 has equivalent usage. See its documentation for details. 1680 1681 evm= [EVM] 1682 Format: { "fix" } 1683 Permit 'security.evm' to be updated regardless of 1684 current integrity status. 1685 1686 early_page_ext [KNL,EARLY] Enforces page_ext initialization to earlier 1687 stages so cover more early boot allocations. 1688 Please note that as side effect some optimizations 1689 might be disabled to achieve that (e.g. parallelized 1690 memory initialization is disabled) so the boot process 1691 might take longer, especially on systems with a lot of 1692 memory. Available with CONFIG_PAGE_EXTENSION=y. 1693 1694 failslab= 1695 fail_usercopy= 1696 fail_page_alloc= 1697 fail_skb_realloc= 1698 fail_make_request=[KNL] 1699 General fault injection mechanism. 1700 Format: <interval>,<probability>,<space>,<times> 1701 See also Documentation/fault-injection/. 1702 1703 fb_tunnels= [NET] 1704 Format: { initns | none } 1705 See Documentation/admin-guide/sysctl/net.rst for 1706 fb_tunnels_only_for_init_ns 1707 1708 floppy= [HW] 1709 See Documentation/admin-guide/blockdev/floppy.rst. 1710 1711 forcepae [X86-32] 1712 Forcefully enable Physical Address Extension (PAE). 1713 Many Pentium M systems disable PAE but may have a 1714 functionally usable PAE implementation. 1715 Warning: use of this parameter will taint the kernel 1716 and may cause unknown problems. 1717 1718 fred= [X86-64] 1719 Enable/disable Flexible Return and Event Delivery. 1720 Format: { on | off } 1721 on: enable FRED when it's present, the default setting. 1722 off: disable FRED. 1723 1724 ftrace=[tracer] 1725 [FTRACE] will set and start the specified tracer 1726 as early as possible in order to facilitate early 1727 boot debugging. 1728 1729 ftrace_boot_snapshot 1730 [FTRACE] On boot up, a snapshot will be taken of the 1731 ftrace ring buffer that can be read at: 1732 /sys/kernel/tracing/snapshot. 1733 This is useful if you need tracing information from kernel 1734 boot up that is likely to be overridden by user space 1735 start up functionality. 1736 1737 Optionally, the snapshot can also be defined for a tracing 1738 instance that was created by the trace_instance= command 1739 line parameter. 1740 1741 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo 1742 1743 The above will cause the "foo" tracing instance to trigger 1744 a snapshot at the end of boot up. 1745 1746 ftrace_dump_on_oops[=2(orig_cpu) | =<instance>][,<instance> | 1747 ,<instance>=2(orig_cpu)] 1748 [FTRACE] will dump the trace buffers on oops. 1749 If no parameter is passed, ftrace will dump global 1750 buffers of all CPUs, if you pass 2 or orig_cpu, it 1751 will dump only the buffer of the CPU that triggered 1752 the oops, or the specific instance will be dumped if 1753 its name is passed. Multiple instance dump is also 1754 supported, and instances are separated by commas. Each 1755 instance supports only dump on CPU that triggered the 1756 oops by passing 2 or orig_cpu to it. 1757 1758 ftrace_dump_on_oops=foo=orig_cpu 1759 1760 The above will dump only the buffer of "foo" instance 1761 on CPU that triggered the oops. 1762 1763 ftrace_dump_on_oops,foo,bar=orig_cpu 1764 1765 The above will dump global buffer on all CPUs, the 1766 buffer of "foo" instance on all CPUs and the buffer 1767 of "bar" instance on CPU that triggered the oops. 1768 1769 ftrace_filter=[function-list] 1770 [FTRACE] Limit the functions traced by the function 1771 tracer at boot up. function-list is a comma-separated 1772 list of functions. This list can be changed at run 1773 time by the set_ftrace_filter file in the debugfs 1774 tracing directory. 1775 1776 ftrace_notrace=[function-list] 1777 [FTRACE] Do not trace the functions specified in 1778 function-list. This list can be changed at run time 1779 by the set_ftrace_notrace file in the debugfs 1780 tracing directory. 1781 1782 ftrace_graph_filter=[function-list] 1783 [FTRACE] Limit the top level callers functions traced 1784 by the function graph tracer at boot up. 1785 function-list is a comma-separated list of functions 1786 that can be changed at run time by the 1787 set_graph_function file in the debugfs tracing directory. 1788 1789 ftrace_graph_notrace=[function-list] 1790 [FTRACE] Do not trace from the functions specified in 1791 function-list. This list is a comma-separated list of 1792 functions that can be changed at run time by the 1793 set_graph_notrace file in the debugfs tracing directory. 1794 1795 ftrace_graph_max_depth=<uint> 1796 [FTRACE] Used with the function graph tracer. This is 1797 the max depth it will trace into a function. This value 1798 can be changed at run time by the max_graph_depth file 1799 in the tracefs tracing directory. default: 0 (no limit) 1800 1801 fw_devlink= [KNL,EARLY] Create device links between consumer and supplier 1802 devices by scanning the firmware to infer the 1803 consumer/supplier relationships. This feature is 1804 especially useful when drivers are loaded as modules as 1805 it ensures proper ordering of tasks like device probing 1806 (suppliers first, then consumers), supplier boot state 1807 clean up (only after all consumers have probed), 1808 suspend/resume & runtime PM (consumers first, then 1809 suppliers). 1810 Format: { off | permissive | on | rpm } 1811 off -- Don't create device links from firmware info. 1812 permissive -- Create device links from firmware info 1813 but use it only for ordering boot state clean 1814 up (sync_state() calls). 1815 on -- Create device links from firmware info and use it 1816 to enforce probe and suspend/resume ordering. 1817 rpm -- Like "on", but also use to order runtime PM. 1818 1819 fw_devlink.strict=<bool> 1820 [KNL,EARLY] Treat all inferred dependencies as mandatory 1821 dependencies. This only applies for fw_devlink=on|rpm. 1822 Format: <bool> 1823 1824 fw_devlink.sync_state = 1825 [KNL,EARLY] When all devices that could probe have finished 1826 probing, this parameter controls what to do with 1827 devices that haven't yet received their sync_state() 1828 calls. 1829 Format: { strict | timeout } 1830 strict -- Default. Continue waiting on consumers to 1831 probe successfully. 1832 timeout -- Give up waiting on consumers and call 1833 sync_state() on any devices that haven't yet 1834 received their sync_state() calls after 1835 deferred_probe_timeout has expired or by 1836 late_initcall() if !CONFIG_MODULES. 1837 1838 gamecon.map[2|3]= 1839 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1840 support via parallel port (up to 5 devices per port) 1841 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1842 See also Documentation/input/devices/joystick-parport.rst 1843 1844 gamma= [HW,DRM] 1845 1846 gart_fix_e820= [X86-64,EARLY] disable the fix e820 for K8 GART 1847 Format: off | on 1848 default: on 1849 1850 gather_data_sampling= 1851 [X86,INTEL,EARLY] Control the Gather Data Sampling (GDS) 1852 mitigation. 1853 1854 Gather Data Sampling is a hardware vulnerability which 1855 allows unprivileged speculative access to data which was 1856 previously stored in vector registers. 1857 1858 This issue is mitigated by default in updated microcode. 1859 The mitigation may have a performance impact but can be 1860 disabled. On systems without the microcode mitigation 1861 disabling AVX serves as a mitigation. 1862 1863 force: Disable AVX to mitigate systems without 1864 microcode mitigation. No effect if the microcode 1865 mitigation is present. Known to cause crashes in 1866 userspace with buggy AVX enumeration. 1867 1868 off: Disable GDS mitigation. 1869 1870 gbpages [X86] Use GB pages for kernel direct mappings. 1871 1872 gcov_persist= [GCOV] When non-zero (default), profiling data for 1873 kernel modules is saved and remains accessible via 1874 debugfs, even when the module is unloaded/reloaded. 1875 When zero, profiling data is discarded and associated 1876 debugfs files are removed at module unload time. 1877 1878 goldfish [X86] Enable the goldfish android emulator platform. 1879 Don't use this when you are not running on the 1880 android emulator 1881 1882 gpio-mockup.gpio_mockup_ranges 1883 [HW] Sets the ranges of gpiochip of for this device. 1884 Format: <start1>,<end1>,<start2>,<end2>... 1885 gpio-mockup.gpio_mockup_named_lines 1886 [HW] Let the driver know GPIO lines should be named. 1887 1888 gpt [EFI] Forces disk with valid GPT signature but 1889 invalid Protective MBR to be treated as GPT. If the 1890 primary GPT is corrupted, it enables the backup/alternate 1891 GPT to be used instead. 1892 1893 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1894 the "Enable 0" bit of the configuration register. 1895 Format: 0 | 1 1896 Default: 0 1897 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1898 the "Enable 0" bit of the configuration register. 1899 Format: 0 | 1 1900 Default: 0 1901 grcan.select= [HW] Select which physical interface to use. 1902 Format: 0 | 1 1903 Default: 0 1904 grcan.txsize= [HW] Sets the size of the tx buffer. 1905 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1906 Default: 1024 1907 grcan.rxsize= [HW] Sets the size of the rx buffer. 1908 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1909 Default: 1024 1910 1911 hardened_usercopy= 1912 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 1913 hardening is enabled for this boot. Hardened 1914 usercopy checking is used to protect the kernel 1915 from reading or writing beyond known memory 1916 allocation boundaries as a proactive defense 1917 against bounds-checking flaws in the kernel's 1918 copy_to_user()/copy_from_user() interface. 1919 The default is determined by 1920 CONFIG_HARDENED_USERCOPY_DEFAULT_ON. 1921 on Perform hardened usercopy checks. 1922 off Disable hardened usercopy checks. 1923 1924 hardlockup_all_cpu_backtrace= 1925 [KNL] Should the hard-lockup detector generate 1926 backtraces on all cpus. 1927 Format: 0 | 1 1928 1929 hash_pointers= 1930 [KNL,EARLY] 1931 By default, when pointers are printed to the console 1932 or buffers via the %p format string, that pointer is 1933 "hashed", i.e. obscured by hashing the pointer value. 1934 This is a security feature that hides actual kernel 1935 addresses from unprivileged users, but it also makes 1936 debugging the kernel more difficult since unequal 1937 pointers can no longer be compared. The choices are: 1938 Format: { auto | always | never } 1939 Default: auto 1940 1941 auto - Hash pointers unless slab_debug is enabled. 1942 always - Always hash pointers (even if slab_debug is 1943 enabled). 1944 never - Never hash pointers. This option should only 1945 be specified when debugging the kernel. Do 1946 not use on production kernels. The boot 1947 param "no_hash_pointers" is an alias for 1948 this mode. 1949 1950 For controlling hashing dynamically at runtime, 1951 use the "kernel.kptr_restrict" sysctl instead. 1952 1953 hashdist= [KNL,NUMA] Large hashes allocated during boot 1954 are distributed across NUMA nodes. Defaults on 1955 for 64-bit NUMA, off otherwise. 1956 Format: 0 | 1 (for off | on) 1957 1958 hd= [EIDE] (E)IDE hard drive subsystem geometry 1959 Format: <cyl>,<head>,<sect> 1960 1961 hest_disable [ACPI] 1962 Disable Hardware Error Source Table (HEST) support; 1963 corresponding firmware-first mode error processing 1964 logic will be disabled. 1965 1966 hibernate= [HIBERNATION] 1967 noresume Don't check if there's a hibernation image 1968 present during boot. 1969 nocompress Don't compress/decompress hibernation images. 1970 no Disable hibernation and resume. 1971 protect_image Turn on image protection during restoration 1972 (that will set all pages holding image data 1973 during restoration read-only). 1974 1975 hibernate.compressor= [HIBERNATION] Compression algorithm to be 1976 used with hibernation. 1977 Format: { lzo | lz4 } 1978 Default: lzo 1979 1980 lzo: Select LZO compression algorithm to 1981 compress/decompress hibernation image. 1982 1983 lz4: Select LZ4 compression algorithm to 1984 compress/decompress hibernation image. 1985 1986 hibernate.pm_test_delay= 1987 [HIBERNATION] 1988 Sets the number of seconds to remain in a hibernation test 1989 mode before resuming the system (see 1990 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 1991 is set. Default value is 5. 1992 1993 hibernate_compression_threads= 1994 [HIBERNATION] 1995 Set the number of threads used for compressing or decompressing 1996 hibernation images. 1997 1998 Format: <integer> 1999 Default: 3 2000 Minimum: 1 2001 Example: hibernate_compression_threads=4 2002 2003 highmem=nn[KMG] [KNL,BOOT,EARLY] forces the highmem zone to have an exact 2004 size of <nn>. This works even on boxes that have no 2005 highmem otherwise. This also works to reduce highmem 2006 size on bigger boxes. 2007 2008 highres= [KNL] Enable/disable high resolution timer mode. 2009 Valid parameters: "on", "off" 2010 Default: "on" 2011 2012 hlt [BUGS=ARM,SH] 2013 2014 hostname= [KNL,EARLY] Set the hostname (aka UTS nodename). 2015 Format: <string> 2016 This allows setting the system's hostname during early 2017 startup. This sets the name returned by gethostname. 2018 Using this parameter to set the hostname makes it 2019 possible to ensure the hostname is correctly set before 2020 any userspace processes run, avoiding the possibility 2021 that a process may call gethostname before the hostname 2022 has been explicitly set, resulting in the calling 2023 process getting an incorrect result. The string must 2024 not exceed the maximum allowed hostname length (usually 2025 64 characters) and will be truncated otherwise. 2026 2027 hpet= [X86-32,HPET] option to control HPET usage 2028 Format: { enable (default) | disable | force | 2029 verbose } 2030 disable: disable HPET and use PIT instead 2031 force: allow force enabled of undocumented chips (ICH4, 2032 VIA, nVidia) 2033 verbose: show contents of HPET registers during setup 2034 2035 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 2036 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 2037 2038 hugepages= [HW,EARLY] Number of HugeTLB pages to allocate at boot. 2039 If this follows hugepagesz (below), it specifies 2040 the number of pages of hugepagesz to be allocated. 2041 If this is the first HugeTLB parameter on the command 2042 line, it specifies the number of pages to allocate for 2043 the default huge page size. If using node format, the 2044 number of pages to allocate per-node can be specified. 2045 See also Documentation/admin-guide/mm/hugetlbpage.rst. 2046 Format: <integer> or (node format) 2047 <node>:<integer>[,<node>:<integer>] 2048 2049 hugepagesz= 2050 [HW,EARLY] The size of the HugeTLB pages. This is 2051 used in conjunction with hugepages (above) to 2052 allocate huge pages of a specific size at boot. The 2053 pair hugepagesz=X hugepages=Y can be specified once 2054 for each supported huge page size. Huge page sizes 2055 are architecture dependent. See also 2056 Documentation/admin-guide/mm/hugetlbpage.rst. 2057 Format: size[KMG] 2058 2059 hugepage_alloc_threads= 2060 [HW] The number of threads that should be used to 2061 allocate hugepages during boot. This option can be 2062 used to improve system bootup time when allocating 2063 a large amount of huge pages. 2064 The default value is 25% of the available hardware threads. 2065 2066 Note that this parameter only applies to non-gigantic huge pages. 2067 2068 hugetlb_cma= [HW,CMA,EARLY] The size of a CMA area used for allocation 2069 of gigantic hugepages. Or using node format, the size 2070 of a CMA area per node can be specified. 2071 Format: nn[KMGTPE] or (node format) 2072 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]] 2073 2074 The size must be a multiple of the gigantic page size. 2075 When using node format, this applies to each per-node size. 2076 Missaligned values are dropped with a warning. 2077 2078 Reserve a CMA area of given size and allocate gigantic 2079 hugepages using the CMA allocator. If enabled, the 2080 boot-time allocation of gigantic hugepages is skipped. 2081 2082 hugetlb_cma_only= 2083 [HW,CMA,EARLY] When allocating new HugeTLB pages, only 2084 try to allocate from the CMA areas. 2085 2086 This option does nothing if hugetlb_cma= is not also 2087 specified. 2088 2089 hugetlb_free_vmemmap= 2090 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP 2091 enabled. 2092 Control if HugeTLB Vmemmap Optimization (HVO) is enabled. 2093 Allows heavy hugetlb users to free up some more 2094 memory (7 * PAGE_SIZE for each 2MB hugetlb page). 2095 Format: { on | off (default) } 2096 2097 on: enable HVO 2098 off: disable HVO 2099 2100 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y, 2101 the default is on. 2102 2103 Note that the vmemmap pages may be allocated from the added 2104 memory block itself when memory_hotplug.memmap_on_memory is 2105 enabled, those vmemmap pages cannot be optimized even if this 2106 feature is enabled. Other vmemmap pages not allocated from 2107 the added memory block itself do not be affected. 2108 2109 hung_task_panic= 2110 [KNL] Number of hung tasks to trigger kernel panic. 2111 Format: <int> 2112 2113 When set to a non-zero value, a kernel panic will be triggered if 2114 the number of detected hung tasks reaches this value. 2115 2116 0: don't panic 2117 1: panic immediately on first hung task 2118 N: panic after N hung tasks are detected in a single scan 2119 2120 The default value is controlled by the 2121 CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time option. The value 2122 selected by this boot parameter can be changed later by the 2123 kernel.hung_task_panic sysctl. 2124 2125 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 2126 terminal devices. Valid values: 0..8 2127 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 2128 If specified, z/VM IUCV HVC accepts connections 2129 from listed z/VM user IDs only. 2130 2131 hv_nopvspin [X86,HYPER_V,EARLY] 2132 Disables the paravirt spinlock optimizations 2133 which allow the hypervisor to 'idle' the guest 2134 on lock contention. 2135 2136 hw_protection= [HW] 2137 Format: reboot | shutdown 2138 2139 Hardware protection action taken on critical events like 2140 overtemperature or imminent voltage loss. 2141 2142 i2c_bus= [HW] Override the default board specific I2C bus speed 2143 or register an additional I2C bus that is not 2144 registered from board initialization code. 2145 Format: 2146 <bus_id>,<clkrate> 2147 2148 i2c_touchscreen_props= [HW,ACPI,X86] 2149 Set device-properties for ACPI-enumerated I2C-attached 2150 touchscreen, to e.g. fix coordinates of upside-down 2151 mounted touchscreens. If you need this option please 2152 submit a drivers/platform/x86/touchscreen_dmi.c patch 2153 adding a DMI quirk for this. 2154 2155 Format: 2156 <ACPI_HW_ID>:<prop_name>=<val>[:prop_name=val][:...] 2157 Where <val> is one of: 2158 Omit "=<val>" entirely Set a boolean device-property 2159 Unsigned number Set a u32 device-property 2160 Anything else Set a string device-property 2161 2162 Examples (split over multiple lines): 2163 i2c_touchscreen_props=GDIX1001:touchscreen-inverted-x: 2164 touchscreen-inverted-y 2165 2166 i2c_touchscreen_props=MSSL1680:touchscreen-size-x=1920: 2167 touchscreen-size-y=1080:touchscreen-inverted-y: 2168 firmware-name=gsl1680-vendor-model.fw:silead,home-button 2169 2170 i8042.debug [HW] Toggle i8042 debug mode 2171 i8042.unmask_kbd_data 2172 [HW] Enable printing of interrupt data from the KBD port 2173 (disabled by default, and as a pre-condition 2174 requires that i8042.debug=1 be enabled) 2175 i8042.direct [HW] Put keyboard port into non-translated mode 2176 i8042.dumbkbd [HW] Pretend that controller can only read data from 2177 keyboard and cannot control its state 2178 (Don't attempt to blink the leds) 2179 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 2180 i8042.nokbd [HW] Don't check/create keyboard port 2181 i8042.noloop [HW] Disable the AUX Loopback command while probing 2182 for the AUX port 2183 i8042.nomux [HW] Don't check presence of an active multiplexing 2184 controller 2185 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 2186 controllers 2187 i8042.notimeout [HW] Ignore timeout condition signalled by controller 2188 i8042.reset [HW] Reset the controller during init, cleanup and 2189 suspend-to-ram transitions, only during s2r 2190 transitions, or never reset 2191 Format: { 1 | Y | y | 0 | N | n } 2192 1, Y, y: always reset controller 2193 0, N, n: don't ever reset controller 2194 Default: only on s2r transitions on x86; most other 2195 architectures force reset to be always executed 2196 i8042.unlock [HW] Unlock (ignore) the keylock 2197 i8042.kbdreset [HW] Reset device connected to KBD port 2198 i8042.probe_defer 2199 [HW] Allow deferred probing upon i8042 probe errors 2200 2201 i810= [HW,DRM] 2202 2203 i915.invert_brightness= 2204 [DRM] Invert the sense of the variable that is used to 2205 set the brightness of the panel backlight. Normally a 2206 brightness value of 0 indicates backlight switched off, 2207 and the maximum of the brightness value sets the backlight 2208 to maximum brightness. If this parameter is set to 0 2209 (default) and the machine requires it, or this parameter 2210 is set to 1, a brightness value of 0 sets the backlight 2211 to maximum brightness, and the maximum of the brightness 2212 value switches the backlight off. 2213 -1 -- never invert brightness 2214 0 -- machine default 2215 1 -- force brightness inversion 2216 2217 ia32_emulation= [X86-64] 2218 Format: <bool> 2219 When true, allows loading 32-bit programs and executing 32-bit 2220 syscalls, essentially overriding IA32_EMULATION_DEFAULT_DISABLED at 2221 boot time. When false, unconditionally disables IA32 emulation. 2222 2223 2224 idle= [X86,EARLY] 2225 Format: idle=poll, idle=halt, idle=nomwait 2226 2227 idle=poll: Don't do power saving in the idle loop 2228 using HLT, but poll for rescheduling event. This will 2229 make the CPUs eat a lot more power, but may be useful 2230 to get slightly better performance in multiprocessor 2231 benchmarks. It also makes some profiling using 2232 performance counters more accurate. Please note that 2233 on systems with MONITOR/MWAIT support (like Intel 2234 EM64T CPUs) this option has no performance advantage 2235 over the normal idle loop. It may also interact badly 2236 with hyperthreading. 2237 2238 idle=halt: Halt is forced to be used for CPU idle. 2239 In such case C2/C3 won't be used again. 2240 2241 idle=nomwait: Disable mwait for CPU C-states 2242 2243 [ARM64,EARLY] 2244 Format: idle=wfi, idle=yield, idle=nop 2245 2246 idle=wfi: Use the WFI (Wait For Interrupt) hint 2247 instruction in the idle loop. This is the default and 2248 allows the CPU to enter a low-power state until an 2249 interrupt arrives. 2250 2251 idle=yield: Use the YIELD hint instruction instead of 2252 WFI. CPUs supporting simultaneous multi-threading (SMT), 2253 can continue executing another thread when the current 2254 thread reaches the idle loop. This will make the CPUs 2255 eat more power, but may be useful to get slightly better 2256 performance in some applications, since the CPUs will 2257 not enter a low-power state. 2258 2259 idle=nop: Do not execute any idle instruction in the 2260 idle loop. This is useful on platforms where WFI 2261 misbehaves, leading to system instability or loss of CPU 2262 state. This will make the CPUs eat more power, but may 2263 give slightly better performance in some applications, 2264 since the CPUs will not enter a low-power state. 2265 2266 idxd.sva= [HW] 2267 Format: <bool> 2268 Allow force disabling of Shared Virtual Memory (SVA) 2269 support for the idxd driver. By default it is set to 2270 true (1). 2271 2272 idxd.tc_override= [HW] 2273 Format: <bool> 2274 Allow override of default traffic class configuration 2275 for the device. By default it is set to false (0). 2276 2277 ieee754= [MIPS] Select IEEE Std 754 conformance mode 2278 Format: { strict | legacy | 2008 | relaxed | emulated } 2279 Default: strict 2280 2281 Choose which programs will be accepted for execution 2282 based on the IEEE 754 NaN encoding(s) supported by 2283 the FPU and the NaN encoding requested with the value 2284 of an ELF file header flag individually set by each 2285 binary. Hardware implementations are permitted to 2286 support either or both of the legacy and the 2008 NaN 2287 encoding mode. 2288 2289 Available settings are as follows: 2290 strict accept binaries that request a NaN encoding 2291 supported by the FPU 2292 legacy only accept legacy-NaN binaries, if supported 2293 by the FPU 2294 2008 only accept 2008-NaN binaries, if supported 2295 by the FPU 2296 relaxed accept any binaries regardless of whether 2297 supported by the FPU 2298 emulated accept any binaries but enable FPU emulator 2299 if binary mode is unsupported by the FPU. 2300 2301 The FPU emulator is always able to support both NaN 2302 encodings, so if no FPU hardware is present or it has 2303 been disabled with 'nofpu', then the settings of 2304 'legacy' and '2008' strap the emulator accordingly, 2305 'relaxed' straps the emulator for both legacy-NaN and 2306 2008-NaN, whereas 'strict' enables legacy-NaN only on 2307 legacy processors and both NaN encodings on MIPS32 or 2308 MIPS64 CPUs. 2309 2310 The setting for ABS.fmt/NEG.fmt instruction execution 2311 mode generally follows that for the NaN encoding, 2312 except where unsupported by hardware. 2313 2314 ignore_loglevel [KNL,EARLY] 2315 Ignore loglevel setting - this will print /all/ 2316 kernel messages to the console. Useful for debugging. 2317 We also add it as printk module parameter, so users 2318 could change it dynamically, usually by 2319 /sys/module/printk/parameters/ignore_loglevel. 2320 2321 ignore_rlimit_data 2322 Ignore RLIMIT_DATA setting for data mappings, 2323 print warning at first misuse. Can be changed via 2324 /sys/module/kernel/parameters/ignore_rlimit_data. 2325 2326 ihash_entries= [KNL] 2327 Set number of hash buckets for inode cache. 2328 2329 ima_appraise= [IMA] appraise integrity measurements 2330 Format: { "off" | "enforce" | "fix" | "log" } 2331 default: "enforce" 2332 2333 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 2334 The builtin appraise policy appraises all files 2335 owned by uid=0. 2336 2337 ima_canonical_fmt [IMA] 2338 Use the canonical format for the binary runtime 2339 measurements, instead of host native format. 2340 2341 ima_flush_htable [IMA] 2342 Flush the IMA hash table when deleting all the 2343 staged measurement records, to achieve maximum 2344 memory saving at the cost of having duplicate 2345 records across the staged measurement lists. 2346 2347 ima_hash= [IMA] 2348 Format: { md5 | sha1 | rmd160 | sha256 | sha384 2349 | sha512 | ... } 2350 default: "sha1" 2351 2352 The list of supported hash algorithms is defined 2353 in crypto/hash_info.h. 2354 2355 ima_policy= [IMA] 2356 The builtin policies to load during IMA setup. 2357 Format: "tcb | appraise_tcb | secure_boot | 2358 fail_securely | critical_data" 2359 2360 The "tcb" policy measures all programs exec'd, files 2361 mmap'd for exec, and all files opened with the read 2362 mode bit set by either the effective uid (euid=0) or 2363 uid=0. 2364 2365 The "appraise_tcb" policy appraises the integrity of 2366 all files owned by root. 2367 2368 The "secure_boot" policy appraises the integrity 2369 of files (eg. kexec kernel image, kernel modules, 2370 firmware, policy, etc) based on file signatures. 2371 2372 The "fail_securely" policy forces file signature 2373 verification failure also on privileged mounted 2374 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 2375 flag. 2376 2377 The "critical_data" policy measures kernel integrity 2378 critical data. 2379 2380 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 2381 Load a policy which meets the needs of the Trusted 2382 Computing Base. This means IMA will measure all 2383 programs exec'd, files mmap'd for exec, and all files 2384 opened for read by uid=0. 2385 2386 ima_template= [IMA] 2387 Select one of defined IMA measurements template formats. 2388 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" | 2389 "ima-sigv2" } 2390 Default: "ima-ng" 2391 2392 ima_template_fmt= 2393 [IMA] Define a custom template format. 2394 Format: { "field1|...|fieldN" } 2395 2396 ima= [IMA] Enable or disable IMA 2397 Format: { "off" | "on" } 2398 Default: "on" 2399 Note that disabling IMA is limited to kdump kernel. 2400 2401 indirect_target_selection= [X86,Intel] Mitigation control for Indirect 2402 Target Selection(ITS) bug in Intel CPUs. Updated 2403 microcode is also required for a fix in IBPB. 2404 2405 on: Enable mitigation (default). 2406 off: Disable mitigation. 2407 force: Force the ITS bug and deploy default 2408 mitigation. 2409 vmexit: Only deploy mitigation if CPU is affected by 2410 guest/host isolation part of ITS. 2411 stuff: Deploy RSB-fill mitigation when retpoline is 2412 also deployed. Otherwise, deploy the default 2413 mitigation. 2414 2415 For details see: 2416 Documentation/admin-guide/hw-vuln/indirect-target-selection.rst 2417 2418 init= [KNL] 2419 Format: <full_path> 2420 Run specified binary instead of /sbin/init as init 2421 process. 2422 2423 initcall_debug [KNL] Trace initcalls as they are executed. Useful 2424 for working out where the kernel is dying during 2425 startup. 2426 2427 initcall_blacklist= [KNL] Do not execute a comma-separated list of 2428 initcall functions. Useful for debugging built-in 2429 modules and initcalls. 2430 2431 initramfs_async= [KNL] 2432 Format: <bool> 2433 Default: 1 2434 This parameter controls whether the initramfs 2435 image is unpacked asynchronously, concurrently 2436 with devices being probed and 2437 initialized. This should normally just work, 2438 but as a debugging aid, one can get the 2439 historical behaviour of the initramfs 2440 unpacking being completed before device_ and 2441 late_ initcalls. 2442 2443 initrd= [BOOT,EARLY] Specify the location of the initial ramdisk 2444 2445 initrdmem= [KNL,EARLY] Specify a physical address and size from which to 2446 load the initrd. If an initrd is compiled in or 2447 specified in the bootparams, it takes priority over this 2448 setting. 2449 Format: ss[KMG],nn[KMG] 2450 Default is 0, 0 2451 2452 init_on_alloc= [MM,EARLY] Fill newly allocated pages and heap objects with 2453 zeroes. 2454 Format: 0 | 1 2455 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 2456 2457 init_on_free= [MM,EARLY] Fill freed pages and heap objects with zeroes. 2458 Format: 0 | 1 2459 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 2460 2461 init_pkru= [X86] Specify the default memory protection keys rights 2462 register contents for all processes. 0x55555554 by 2463 default (disallow access to all but pkey 0). Can 2464 override in debugfs after boot. 2465 2466 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 2467 Format: <irq> 2468 2469 int_pln_enable [X86] Enable power limit notification interrupt 2470 2471 integrity_audit=[IMA] 2472 Format: { "0" | "1" } 2473 0 -- basic integrity auditing messages. (Default) 2474 1 -- additional integrity auditing messages. 2475 2476 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 2477 on 2478 Enable intel iommu driver. 2479 off 2480 Disable intel iommu driver. 2481 igfx_off [Default Off] 2482 By default, gfx is mapped as normal device. If a gfx 2483 device has a dedicated DMAR unit, the DMAR unit is 2484 bypassed by not enabling DMAR with this option. In 2485 this case, gfx device will use physical address for 2486 DMA. 2487 strict [Default Off] 2488 Deprecated, equivalent to iommu.strict=1. 2489 sp_off [Default Off] 2490 By default, super page will be supported if Intel IOMMU 2491 has the capability. With this option, super page will 2492 not be supported. 2493 sm_on 2494 Enable the Intel IOMMU scalable mode if the hardware 2495 advertises that it has support for the scalable mode 2496 translation. 2497 sm_off 2498 Disallow use of the Intel IOMMU scalable mode. 2499 tboot_noforce [Default Off] 2500 Do not force the Intel IOMMU enabled under tboot. 2501 By default, tboot will force Intel IOMMU on, which 2502 could harm performance of some high-throughput 2503 devices like 40GBit network cards, even if identity 2504 mapping is enabled. 2505 Note that using this option lowers the security 2506 provided by tboot because it makes the system 2507 vulnerable to DMA attacks. 2508 2509 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 2510 0 disables intel_idle and fall back on acpi_idle. 2511 1 to 9 specify maximum depth of C-state. 2512 2513 intel_pstate= [X86,EARLY] 2514 disable 2515 Do not enable intel_pstate as the default 2516 scaling driver for the supported processors 2517 active 2518 Use intel_pstate driver to bypass the scaling 2519 governors layer of cpufreq and provides it own 2520 algorithms for p-state selection. There are two 2521 P-state selection algorithms provided by 2522 intel_pstate in the active mode: powersave and 2523 performance. The way they both operate depends 2524 on whether or not the hardware managed P-states 2525 (HWP) feature has been enabled in the processor 2526 and possibly on the processor model. 2527 passive 2528 Use intel_pstate as a scaling driver, but configure it 2529 to work with generic cpufreq governors (instead of 2530 enabling its internal governor). This mode cannot be 2531 used along with the hardware-managed P-states (HWP) 2532 feature. 2533 force 2534 Enable intel_pstate on systems that prohibit it by default 2535 in favor of acpi-cpufreq. Forcing the intel_pstate driver 2536 instead of acpi-cpufreq may disable platform features, such 2537 as thermal controls and power capping, that rely on ACPI 2538 P-States information being indicated to OSPM and therefore 2539 should be used with caution. This option does not work with 2540 processors that aren't supported by the intel_pstate driver 2541 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 2542 no_hwp 2543 Do not enable hardware P state control (HWP) 2544 if available. 2545 hwp_only 2546 Only load intel_pstate on systems which support 2547 hardware P state control (HWP) if available. 2548 support_acpi_ppc 2549 Enforce ACPI _PPC performance limits. If the Fixed ACPI 2550 Description Table, specifies preferred power management 2551 profile as "Enterprise Server" or "Performance Server", 2552 then this feature is turned on by default. 2553 per_cpu_perf_limits 2554 Allow per-logical-CPU P-State performance control limits using 2555 cpufreq sysfs interface 2556 no_cas 2557 Do not enable capacity-aware scheduling (CAS) on 2558 hybrid systems 2559 2560 intremap= [X86-64,Intel-IOMMU,EARLY] 2561 on enable Interrupt Remapping (default) 2562 off disable Interrupt Remapping 2563 nosid disable Source ID checking 2564 no_x2apic_optout 2565 BIOS x2APIC opt-out request will be ignored 2566 nopost disable Interrupt Posting 2567 posted_msi 2568 enable MSIs delivered as posted interrupts 2569 2570 iomem= Disable strict checking of access to MMIO memory 2571 strict regions from userspace. 2572 relaxed 2573 2574 iommu= [X86,EARLY] 2575 2576 off 2577 Don't initialize and use any kind of IOMMU. 2578 2579 force 2580 Force the use of the hardware IOMMU even when 2581 it is not actually needed (e.g. because < 3 GB 2582 memory). 2583 2584 noforce 2585 Don't force hardware IOMMU usage when it is not 2586 needed. (default). 2587 2588 merge 2589 Do scatter-gather (SG) merging. Implies "force" 2590 (experimental). 2591 2592 nomerge 2593 Don't do scatter-gather (SG) merging. 2594 2595 biomerge 2596 Do scatter-gather (SG) merging. Implies "force" 2597 (experimental). [same as "merge"] 2598 2599 panic 2600 Always panic when IOMMU overflows. 2601 2602 nopanic 2603 Don't panic on IOMMU overflows. 2604 2605 pt 2606 Use passththrough mode by default 2607 (Equivalent to iommu.passthrough=1) 2608 2609 nopt 2610 Use translated mode for DMA by default 2611 (Equivalent to iommu.passthrough=0) 2612 2613 soft 2614 Use software bounce buffering (SWIOTLB) (default for 2615 Intel machines). This can be used to prevent the usage 2616 of an available hardware IOMMU. 2617 2618 usedac 2619 Use the DAC on VIA PCI bridge 2620 (default: disable the VIA PCI bridge DAC) 2621 2622 AMD Gart HW IOMMU-specific options: (CONFIG_GART_IOMMU) 2623 2624 <size> 2625 Set the size of the remapping area in bytes. 2626 2627 allowed 2628 Overwrite iommu off workarounds for specific chipsets 2629 2630 force 2631 Overwrite iommu off workarounds for specific chipsets 2632 2633 fullflush 2634 Flush IOMMU on each allocation (default). 2635 2636 nofullflush 2637 Don't use IOMMU fullflush. 2638 2639 memaper[=<order>] 2640 Allocate an own aperture over RAM with size 2641 32MB<<order. (default: order=1, i.e. 64MB) 2642 2643 noaperture 2644 Ask the IOMMU not to touch the aperture for AGP. 2645 2646 noagp 2647 Don't initialize the AGP driver and use full aperture. 2648 2649 iommu= [PPC/POWERNV] 2650 2651 nobypass 2652 Disable IOMMU bypass, using IOMMU for PCI devices. 2653 2654 iommu.forcedac= [ARM64,X86,EARLY] Control IOVA allocation for PCI devices. 2655 Format: { "0" | "1" } 2656 0 - Try to allocate a 32-bit DMA address first, before 2657 falling back to the full range if needed. 2658 1 - Allocate directly from the full usable range, 2659 forcing Dual Address Cycle for PCI cards supporting 2660 greater than 32-bit addressing. 2661 2662 iommu.strict= [ARM64,X86,S390,EARLY] Configure TLB invalidation behaviour 2663 Format: { "0" | "1" } 2664 0 - Lazy mode. 2665 Request that DMA unmap operations use deferred 2666 invalidation of hardware TLBs, for increased 2667 throughput at the cost of reduced device isolation. 2668 Will fall back to strict mode if not supported by 2669 the relevant IOMMU driver. 2670 1 - Strict mode. 2671 DMA unmap operations invalidate IOMMU hardware TLBs 2672 synchronously. 2673 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}. 2674 Note: on x86, strict mode specified via one of the 2675 legacy driver-specific options takes precedence. 2676 2677 iommu.passthrough= 2678 [ARM64,X86,EARLY] Configure DMA to bypass the IOMMU by default. 2679 Format: { "0" | "1" } 2680 0 - Use IOMMU translation for DMA. 2681 1 - Bypass the IOMMU for DMA. 2682 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2683 2684 iommu.debug_pagealloc= 2685 [KNL,EARLY] When CONFIG_IOMMU_DEBUG_PAGEALLOC is set, this 2686 parameter enables the feature at boot time. By default, it 2687 is disabled and the system behaves the same way as a kernel 2688 built without CONFIG_IOMMU_DEBUG_PAGEALLOC. 2689 Format: { "0" | "1" } 2690 0 - Sanitizer disabled. 2691 1 - Sanitizer enabled, expect runtime overhead. 2692 2693 io7= [HW] IO7 for Marvel-based Alpha systems 2694 See comment before marvel_specify_io7 in 2695 arch/alpha/kernel/core_marvel.c. 2696 2697 io_delay= [X86,EARLY] I/O delay method 2698 0x80 2699 Standard port 0x80 based delay 2700 0xed 2701 Alternate port 0xed based delay (needed on some systems) 2702 udelay 2703 Simple two microseconds delay 2704 none 2705 No delay 2706 2707 ip= [IP_PNP] 2708 See Documentation/admin-guide/nfs/nfsroot.rst. 2709 2710 ipcmni_extend [KNL,EARLY] Extend the maximum number of unique System V 2711 IPC identifiers from 32,768 to 16,777,216. 2712 2713 ipe.enforce= [IPE] 2714 Format: <bool> 2715 Determine whether IPE starts in permissive (0) or 2716 enforce (1) mode. The default is enforce. 2717 2718 ipe.success_audit= 2719 [IPE] 2720 Format: <bool> 2721 Start IPE with success auditing enabled, emitting 2722 an audit event when a binary is allowed. The default 2723 is 0. 2724 2725 irqaffinity= [SMP] Set the default irq affinity mask 2726 The argument is a cpu list, as described above. 2727 2728 irqchip.gicv2_force_probe= 2729 [ARM,ARM64,EARLY] 2730 Format: <bool> 2731 Force the kernel to look for the second 4kB page 2732 of a GICv2 controller even if the memory range 2733 exposed by the device tree is too small. 2734 2735 irqchip.gicv3_nolpi= 2736 [ARM,ARM64,EARLY] 2737 Force the kernel to ignore the availability of 2738 LPIs (and by consequence ITSs). Intended for system 2739 that use the kernel as a bootloader, and thus want 2740 to let secondary kernels in charge of setting up 2741 LPIs. 2742 2743 irqchip.gicv3_pseudo_nmi= [ARM64,EARLY] 2744 Enables support for pseudo-NMIs in the kernel. This 2745 requires the kernel to be built with 2746 CONFIG_ARM64_PSEUDO_NMI. 2747 2748 irqchip.riscv_imsic_noipi 2749 [RISC-V,EARLY] 2750 Force the kernel to not use IMSIC software injected MSIs 2751 as IPIs. Intended for system where IMSIC is trap-n-emulated, 2752 and thus want to reduce MMIO traps when triggering IPIs 2753 to multiple harts. 2754 2755 irqfixup [HW] 2756 When an interrupt is not handled search all handlers 2757 for it. Intended to get systems with badly broken 2758 firmware running. 2759 2760 irqhandler.duration_warn_us= [KNL] 2761 Warn if an IRQ handler exceeds the specified duration 2762 threshold in microseconds. Useful for identifying 2763 long-running IRQs in the system. 2764 2765 irqpoll [HW] 2766 When an interrupt is not handled search all handlers 2767 for it. Also check all handlers each timer 2768 interrupt. Intended to get systems with badly broken 2769 firmware running. 2770 2771 isapnp= [ISAPNP] 2772 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2773 2774 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2775 Format: [flag-list,]<cpu-list> 2776 2777 Specify one or more CPUs to isolate from disturbances 2778 specified in the flag list (default: domain): 2779 2780 nohz 2781 Disable the tick when a single task runs as well as 2782 disabling other kernel noises like having RCU callbacks 2783 offloaded. This is equivalent to the nohz_full parameter. 2784 2785 A residual 1Hz tick is offloaded to workqueues, which you 2786 need to affine to housekeeping through the global 2787 workqueue's affinity configured via the 2788 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2789 by using the 'domain' flag described below. 2790 2791 NOTE: by default the global workqueue runs on all CPUs, 2792 so to protect individual CPUs the 'cpumask' file has to 2793 be configured manually after bootup. 2794 2795 domain 2796 Isolate from the general SMP balancing and scheduling 2797 algorithms. Note that performing domain isolation this way 2798 is irreversible: it's not possible to bring back a CPU to 2799 the domains once isolated through this boot time 2800 configuration. Use cpusets for a dynamic configuration 2801 which can be altered at runtime. For details see 2802 Documentation/admin-guide/cpu-isolation.rst. 2803 2804 You can move a process onto or off an "isolated" CPU via 2805 the CPU affinity syscalls or cpuset. 2806 <cpu number> begins at 0 and the maximum value is 2807 "number of CPUs in system - 1". 2808 2809 managed_irq 2810 2811 Isolate from being targeted by managed interrupts 2812 which have an interrupt mask containing isolated 2813 CPUs. The affinity of managed interrupts is 2814 handled by the kernel and cannot be changed via 2815 the /proc/irq/* interfaces. 2816 2817 This isolation is best effort and only effective 2818 if the automatically assigned interrupt mask of a 2819 device queue contains isolated and housekeeping 2820 CPUs. If housekeeping CPUs are online then such 2821 interrupts are directed to the housekeeping CPU 2822 so that IO submitted on the housekeeping CPU 2823 cannot disturb the isolated CPU. 2824 2825 If a queue's affinity mask contains only isolated 2826 CPUs then this parameter has no effect on the 2827 interrupt routing decision, though interrupts are 2828 only delivered when tasks running on those 2829 isolated CPUs submit IO. IO submitted on 2830 housekeeping CPUs has no influence on those 2831 queues. 2832 2833 The format of <cpu-list> is described above. 2834 2835 iucv= [HW,NET] 2836 2837 ivrs_ioapic [HW,X86-64] 2838 Provide an override to the IOAPIC-ID<->DEVICE-ID 2839 mapping provided in the IVRS ACPI table. 2840 By default, PCI segment is 0, and can be omitted. 2841 2842 For example, to map IOAPIC-ID decimal 10 to 2843 PCI segment 0x1 and PCI device 00:14.0, 2844 write the parameter as: 2845 ivrs_ioapic=10@0001:00:14.0 2846 2847 Deprecated formats: 2848 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0 2849 write the parameter as: 2850 ivrs_ioapic[10]=00:14.0 2851 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and 2852 PCI device 00:14.0 write the parameter as: 2853 ivrs_ioapic[10]=0001:00:14.0 2854 2855 ivrs_hpet [HW,X86-64] 2856 Provide an override to the HPET-ID<->DEVICE-ID 2857 mapping provided in the IVRS ACPI table. 2858 By default, PCI segment is 0, and can be omitted. 2859 2860 For example, to map HPET-ID decimal 10 to 2861 PCI segment 0x1 and PCI device 00:14.0, 2862 write the parameter as: 2863 ivrs_hpet=10@0001:00:14.0 2864 2865 Deprecated formats: 2866 * To map HPET-ID decimal 0 to PCI device 00:14.0 2867 write the parameter as: 2868 ivrs_hpet[0]=00:14.0 2869 * To map HPET-ID decimal 10 to PCI segment 0x1 and 2870 PCI device 00:14.0 write the parameter as: 2871 ivrs_ioapic[10]=0001:00:14.0 2872 2873 ivrs_acpihid [HW,X86-64] 2874 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2875 mapping provided in the IVRS ACPI table. 2876 By default, PCI segment is 0, and can be omitted. 2877 2878 For example, to map UART-HID:UID AMD0020:0 to 2879 PCI segment 0x1 and PCI device ID 00:14.5, 2880 write the parameter as: 2881 ivrs_acpihid=AMD0020:0@0001:00:14.5 2882 2883 Deprecated formats: 2884 * To map UART-HID:UID AMD0020:0 to PCI segment is 0, 2885 PCI device ID 00:14.5, write the parameter as: 2886 ivrs_acpihid[00:14.5]=AMD0020:0 2887 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and 2888 PCI device ID 00:14.5, write the parameter as: 2889 ivrs_acpihid[0001:00:14.5]=AMD0020:0 2890 2891 js= [HW,JOY] Analog joystick 2892 See Documentation/input/joydev/joystick.rst. 2893 2894 kasan_multi_shot 2895 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2896 report on every invalid memory access. Without this 2897 parameter KASAN will print report only for the first 2898 invalid access. 2899 2900 keep_bootcon [KNL,EARLY] 2901 Do not unregister boot console at start. This is only 2902 useful for debugging when something happens in the window 2903 between unregistering the boot console and initializing 2904 the real console. 2905 2906 keepinitrd [HW,ARM] See retain_initrd. 2907 2908 kernelcore= [KNL,X86,PPC,EARLY] 2909 Format: nn[KMGTPE] | nn% | "mirror" 2910 This parameter specifies the amount of memory usable by 2911 the kernel for non-movable allocations. The requested 2912 amount is spread evenly throughout all nodes in the 2913 system as ZONE_NORMAL. The remaining memory is used for 2914 movable memory in its own zone, ZONE_MOVABLE. In the 2915 event, a node is too small to have both ZONE_NORMAL and 2916 ZONE_MOVABLE, kernelcore memory will take priority and 2917 other nodes will have a larger ZONE_MOVABLE. 2918 2919 ZONE_MOVABLE is used for the allocation of pages that 2920 may be reclaimed or moved by the page migration 2921 subsystem. Note that allocations like PTEs-from-HighMem 2922 still use the HighMem zone if it exists, and the Normal 2923 zone if it does not. 2924 2925 It is possible to specify the exact amount of memory in 2926 the form of "nn[KMGTPE]", a percentage of total system 2927 memory in the form of "nn%", or "mirror". If "mirror" 2928 option is specified, mirrored (reliable) memory is used 2929 for non-movable allocations and remaining memory is used 2930 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2931 are exclusive, so you cannot specify multiple forms. 2932 2933 kfence.burst= [MM,KFENCE] The number of additional successive 2934 allocations to be attempted through KFENCE for each 2935 sample interval. 2936 Format: <unsigned integer> 2937 Default: 0 2938 2939 kfence.check_on_panic= 2940 [MM,KFENCE] Whether to check all KFENCE-managed objects' 2941 canaries on panic. 2942 Format: <bool> 2943 Default: false 2944 2945 kfence.deferrable= 2946 [MM,KFENCE] Whether to use a deferrable timer to trigger 2947 allocations. This avoids forcing CPU wake-ups if the 2948 system is idle, at the risk of a less predictable 2949 sample interval. 2950 Format: <bool> 2951 Default: CONFIG_KFENCE_DEFERRABLE 2952 2953 kfence.fault= [MM,KFENCE] Controls the behavior when a KFENCE 2954 error is detected. 2955 report - print the error report and continue (default). 2956 oops - print the error report and oops. 2957 panic - print the error report and panic. 2958 2959 kfence.sample_interval= 2960 [MM,KFENCE] KFENCE's sample interval in milliseconds. 2961 Format: <unsigned integer> 2962 0 - Disable KFENCE. 2963 >0 - Enabled KFENCE with given sample interval. 2964 Default: CONFIG_KFENCE_SAMPLE_INTERVAL 2965 2966 kfence.skip_covered_thresh= 2967 [MM,KFENCE] If pool utilization reaches this threshold 2968 (pool usage%), KFENCE limits currently covered 2969 allocations of the same source from further filling 2970 up the pool. 2971 Format: <unsigned integer> 2972 Default: 75 2973 2974 kgdbdbgp= [KGDB,HW,EARLY] kgdb over EHCI usb debug port. 2975 Format: <Controller#>[,poll interval] 2976 The controller # is the number of the ehci usb debug 2977 port as it is probed via PCI. The poll interval is 2978 optional and is the number seconds in between 2979 each poll cycle to the debug port in case you need 2980 the functionality for interrupting the kernel with 2981 gdb or control-c on the dbgp connection. When 2982 not using this parameter you use sysrq-g to break into 2983 the kernel debugger. 2984 2985 kgdboc= [KGDB,HW] kgdb over consoles. 2986 Requires a tty driver that supports console polling, 2987 or a supported polling keyboard driver (non-usb). 2988 Serial only format: <serial_device>[,baud] 2989 keyboard only format: kbd 2990 keyboard and serial format: kbd,<serial_device>[,baud] 2991 Optional Kernel mode setting: 2992 kms, kbd format: kms,kbd 2993 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2994 2995 kgdboc_earlycon= [KGDB,HW,EARLY] 2996 If the boot console provides the ability to read 2997 characters and can work in polling mode, you can use 2998 this parameter to tell kgdb to use it as a backend 2999 until the normal console is registered. Intended to 3000 be used together with the kgdboc parameter which 3001 specifies the normal console to transition to. 3002 3003 The name of the early console should be specified 3004 as the value of this parameter. Note that the name of 3005 the early console might be different than the tty 3006 name passed to kgdboc. It's OK to leave the value 3007 blank and the first boot console that implements 3008 read() will be picked. 3009 3010 kgdbwait [KGDB,EARLY] Stop kernel execution and enter the 3011 kernel debugger at the earliest opportunity. 3012 3013 kho= [KEXEC,EARLY] 3014 Format: { "0" | "1" | "off" | "on" | "y" | "n" } 3015 Enables or disables Kexec HandOver. 3016 "0" | "off" | "n" - kexec handover is disabled 3017 "1" | "on" | "y" - kexec handover is enabled 3018 3019 kho_scratch= [KEXEC,EARLY] 3020 Format: ll[KMG],mm[KMG],nn[KMG] | nn% 3021 Defines the size of the KHO scratch region. The KHO 3022 scratch regions are physically contiguous memory 3023 ranges that can only be used for non-kernel 3024 allocations. That way, even when memory is heavily 3025 fragmented with handed over memory, the kexeced 3026 kernel will always have enough contiguous ranges to 3027 bootstrap itself. 3028 3029 It is possible to specify the exact amount of 3030 memory in the form of "ll[KMG],mm[KMG],nn[KMG]" 3031 where the first parameter defines the size of a low 3032 memory scratch area, the second parameter defines 3033 the size of a global scratch area and the third 3034 parameter defines the size of additional per-node 3035 scratch areas. The form "nn%" defines scale factor 3036 (in percents) of memory that was used during boot. 3037 3038 kmac= [MIPS] Korina ethernet MAC address. 3039 Configure the RouterBoard 532 series on-chip 3040 Ethernet adapter MAC address. 3041 3042 kmemleak= [KNL,EARLY] Boot-time kmemleak enable/disable 3043 Valid arguments: on, off 3044 Default: on 3045 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 3046 the default is off. 3047 3048 kprobe_event=[probe-list] 3049 [FTRACE] Add kprobe events and enable at boot time. 3050 The probe-list is a semicolon delimited list of probe 3051 definitions. Each definition is same as kprobe_events 3052 interface, but the parameters are comma delimited. 3053 For example, to add a kprobe event on vfs_read with 3054 arg1 and arg2, add to the command line; 3055 3056 kprobe_event=p,vfs_read,$arg1,$arg2 3057 3058 See also Documentation/trace/kprobetrace.rst "Kernel 3059 Boot Parameter" section. 3060 3061 kpti= [ARM64,EARLY] Control page table isolation of 3062 user and kernel address spaces. 3063 Default: enabled on cores which need mitigation. 3064 0: force disabled 3065 1: force enabled 3066 3067 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires 3068 CONFIG_KUNIT to be set to be fully enabled. The 3069 default value can be overridden via 3070 KUNIT_DEFAULT_ENABLED. 3071 Default is 1 (enabled) 3072 3073 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 3074 Default is 0 (don't ignore, but inject #GP) 3075 3076 kvm.eager_page_split= 3077 [KVM,X86] Controls whether or not KVM will try to 3078 proactively split all huge pages during dirty logging. 3079 Eager page splitting reduces interruptions to vCPU 3080 execution by eliminating the write-protection faults 3081 and MMU lock contention that would otherwise be 3082 required to split huge pages lazily. 3083 3084 VM workloads that rarely perform writes or that write 3085 only to a small region of VM memory may benefit from 3086 disabling eager page splitting to allow huge pages to 3087 still be used for reads. 3088 3089 The behavior of eager page splitting depends on whether 3090 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If 3091 disabled, all huge pages in a memslot will be eagerly 3092 split when dirty logging is enabled on that memslot. If 3093 enabled, eager page splitting will be performed during 3094 the KVM_CLEAR_DIRTY ioctl, and only for the pages being 3095 cleared. 3096 3097 Eager page splitting is only supported when kvm.tdp_mmu=Y. 3098 3099 Default is Y (on). 3100 3101 kvm.enable_pmu=[KVM,X86] 3102 If enabled, KVM will virtualize PMU functionality based 3103 on the virtual CPU model defined by userspace. This 3104 can be overridden on a per-VM basis via 3105 KVM_CAP_PMU_CAPABILITY. 3106 3107 If disabled, KVM will not virtualize PMU functionality, 3108 e.g. MSRs, PMCs, PMIs, etc., even if userspace defines 3109 a virtual CPU model that contains PMU assets. 3110 3111 Note, KVM's vPMU support implicitly requires running 3112 with an in-kernel local APIC, e.g. to deliver PMIs to 3113 the guest. Running without an in-kernel local APIC is 3114 not supported, though KVM will allow such a combination 3115 (with severely degraded functionality). 3116 3117 See also enable_mediated_pmu. 3118 3119 Default is Y (on). 3120 3121 kvm.enable_virt_at_load=[KVM,ARM64,LOONGARCH,MIPS,RISCV,X86] 3122 If enabled, KVM will enable virtualization in hardware 3123 when KVM is loaded, and disable virtualization when KVM 3124 is unloaded (if KVM is built as a module). 3125 3126 If disabled, KVM will dynamically enable and disable 3127 virtualization on-demand when creating and destroying 3128 VMs, i.e. on the 0=>1 and 1=>0 transitions of the 3129 number of VMs. 3130 3131 Enabling virtualization at module load avoids potential 3132 latency for creation of the 0=>1 VM, as KVM serializes 3133 virtualization enabling across all online CPUs. The 3134 "cost" of enabling virtualization when KVM is loaded, 3135 is that doing so may interfere with using out-of-tree 3136 hypervisors that want to "own" virtualization hardware. 3137 3138 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 3139 Default is false (don't support). 3140 3141 kvm.nx_huge_pages= 3142 [KVM] Controls the software workaround for the 3143 X86_BUG_ITLB_MULTIHIT bug. 3144 force : Always deploy workaround. 3145 off : Never deploy workaround. 3146 auto : Deploy workaround based on the presence of 3147 X86_BUG_ITLB_MULTIHIT. 3148 3149 Default is 'auto'. 3150 3151 If the software workaround is enabled for the host, 3152 guests do need not to enable it for nested guests. 3153 3154 kvm.nx_huge_pages_recovery_ratio= 3155 [KVM] Controls how many 4KiB pages are periodically zapped 3156 back to huge pages. 0 disables the recovery, otherwise if 3157 the value is N KVM will zap 1/Nth of the 4KiB pages every 3158 period (see below). The default is 60. 3159 3160 kvm.nx_huge_pages_recovery_period_ms= 3161 [KVM] Controls the time period at which KVM zaps 4KiB pages 3162 back to huge pages. If the value is a non-zero N, KVM will 3163 zap a portion (see ratio above) of the pages every N msecs. 3164 If the value is 0 (the default), KVM will pick a period based 3165 on the ratio, such that a page is zapped after 1 hour on average. 3166 3167 kvm-{amd,intel}.enable_mediated_pmu=[KVM,AMD,INTEL] 3168 If enabled, KVM will provide a mediated virtual PMU, 3169 instead of the default perf-based virtual PMU (if 3170 kvm.enable_pmu is true and PMU is enumerated via the 3171 virtual CPU model). 3172 3173 With a perf-based vPMU, KVM operates as a user of perf, 3174 i.e. emulates guest PMU counters using perf events. 3175 KVM-created perf events are managed by perf as regular 3176 (guest-only) events, e.g. are scheduled in/out, contend 3177 for hardware resources, etc. Using a perf-based vPMU 3178 allows guest and host usage of the PMU to co-exist, but 3179 incurs non-trivial overhead and can result in silently 3180 dropped guest events (due to resource contention). 3181 3182 With a mediated vPMU, hardware PMU state is context 3183 switched around the world switch to/from the guest. 3184 KVM mediates which events the guest can utilize, but 3185 gives the guest direct access to all other PMU assets 3186 when possible (KVM may intercept some accesses if the 3187 virtual CPU model provides a subset of hardware PMU 3188 functionality). Using a mediated vPMU significantly 3189 reduces PMU virtualization overhead and eliminates lost 3190 guest events, but is mutually exclusive with using perf 3191 to profile KVM guests and adds latency to most VM-Exits 3192 (to context switch PMU state). 3193 3194 Default is N (off). 3195 3196 kvm-amd.nested= [KVM,AMD] Control nested virtualization feature in 3197 KVM/SVM. Default is 1 (enabled). 3198 3199 kvm-amd.npt= [KVM,AMD] Control KVM's use of Nested Page Tables, 3200 a.k.a. Two-Dimensional Page Tables. Default is 1 3201 (enabled). Disable by KVM if hardware lacks support 3202 for NPT. 3203 3204 kvm-amd.ciphertext_hiding_asids= 3205 [KVM,AMD] Ciphertext hiding prevents disallowed accesses 3206 to SNP private memory from reading ciphertext. Instead, 3207 reads will see constant default values (0xff). 3208 3209 If ciphertext hiding is enabled, the joint SEV-ES and 3210 SEV-SNP ASID space is partitioned into separate SEV-ES 3211 and SEV-SNP ASID ranges, with the SEV-SNP range being 3212 [1..max_snp_asid] and the SEV-ES range being 3213 (max_snp_asid..min_sev_asid), where min_sev_asid is 3214 enumerated by CPUID.0x.8000_001F[EDX]. 3215 3216 A non-zero value enables SEV-SNP ciphertext hiding and 3217 adjusts the ASID ranges for SEV-ES and SEV-SNP guests. 3218 KVM caps the number of SEV-SNP ASIDs at the maximum 3219 possible value, e.g. specifying -1u will assign all 3220 joint SEV-ES and SEV-SNP ASIDs to SEV-SNP. Note, 3221 assigning all joint ASIDs to SEV-SNP, i.e. configuring 3222 max_snp_asid == min_sev_asid-1, will effectively make 3223 SEV-ES unusable. 3224 3225 kvm-arm.mode= 3226 [KVM,ARM,EARLY] Select one of KVM/arm64's modes of 3227 operation. 3228 3229 none: Forcefully disable KVM. 3230 3231 nvhe: Standard nVHE-based mode, without support for 3232 protected guests. 3233 3234 protected: Mode with support for guests whose state is 3235 kept private from the host, using VHE or 3236 nVHE depending on HW support. 3237 3238 nested: VHE-based mode with support for nested 3239 virtualization. Requires at least ARMv8.4 3240 hardware (with FEAT_NV2). 3241 3242 Defaults to VHE/nVHE based on hardware support. Setting 3243 mode to "protected" will disable kexec and hibernation 3244 for the host. To force nVHE on VHE hardware, add 3245 "arm64_sw.hvhe=0 id_aa64mmfr1.vh=0" to the 3246 command-line. 3247 "nested" and "protected" are experimental and should be 3248 used with extreme caution. 3249 3250 kvm-arm.vgic_v3_group0_trap= 3251 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-0 3252 system registers 3253 3254 kvm-arm.vgic_v3_group1_trap= 3255 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-1 3256 system registers 3257 3258 kvm-arm.vgic_v3_common_trap= 3259 [KVM,ARM,EARLY] Trap guest accesses to GICv3 common 3260 system registers 3261 3262 kvm-arm.vgic_v4_enable= 3263 [KVM,ARM,EARLY] Allow use of GICv4 for direct 3264 injection of LPIs. 3265 3266 kvm-arm.wfe_trap_policy= 3267 [KVM,ARM] Control when to set WFE instruction trap for 3268 KVM VMs. Traps are allowed but not guaranteed by the 3269 CPU architecture. 3270 3271 trap: set WFE instruction trap 3272 3273 notrap: clear WFE instruction trap 3274 3275 kvm-arm.wfi_trap_policy= 3276 [KVM,ARM] Control when to set WFI instruction trap for 3277 KVM VMs. Traps are allowed but not guaranteed by the 3278 CPU architecture. 3279 3280 trap: set WFI instruction trap 3281 3282 notrap: clear WFI instruction trap 3283 3284 kvm_cma_resv_ratio=n [PPC,EARLY] 3285 Reserves given percentage from system memory area for 3286 contiguous memory allocation for KVM hash pagetable 3287 allocation. 3288 By default it reserves 5% of total system memory. 3289 Format: <integer> 3290 Default: 5 3291 3292 kvm-intel.ept= [KVM,Intel] Control KVM's use of Extended Page Tables, 3293 a.k.a. Two-Dimensional Page Tables. Default is 1 3294 (enabled). Disable by KVM if hardware lacks support 3295 for EPT. 3296 3297 kvm-intel.emulate_invalid_guest_state= 3298 [KVM,Intel] Control whether to emulate invalid guest 3299 state. Ignored if kvm-intel.enable_unrestricted_guest=1, 3300 as guest state is never invalid for unrestricted 3301 guests. This param doesn't apply to nested guests (L2), 3302 as KVM never emulates invalid L2 guest state. 3303 Default is 1 (enabled). 3304 3305 kvm-intel.flexpriority= 3306 [KVM,Intel] Control KVM's use of FlexPriority feature 3307 (TPR shadow). Default is 1 (enabled). Disable by KVM if 3308 hardware lacks support for it. 3309 3310 kvm-intel.nested= 3311 [KVM,Intel] Control nested virtualization feature in 3312 KVM/VMX. Default is 1 (enabled). 3313 3314 kvm-intel.unrestricted_guest= 3315 [KVM,Intel] Control KVM's use of unrestricted guest 3316 feature (virtualized real and unpaged mode). Default 3317 is 1 (enabled). Disable by KVM if EPT is disabled or 3318 hardware lacks support for it. 3319 3320 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 3321 CVE-2018-3620. 3322 3323 Valid arguments: never, cond, always 3324 3325 always: L1D cache flush on every VMENTER. 3326 cond: Flush L1D on VMENTER only when the code between 3327 VMEXIT and VMENTER can leak host memory. 3328 never: Disables the mitigation 3329 3330 Default is cond (do L1 cache flush in specific instances) 3331 3332 kvm-intel.vpid= [KVM,Intel] Control KVM's use of Virtual Processor 3333 Identification feature (tagged TLBs). Default is 1 3334 (enabled). Disable by KVM if hardware lacks support 3335 for it. 3336 3337 l1d_flush= [X86,INTEL,EARLY] 3338 Control mitigation for L1D based snooping vulnerability. 3339 3340 Certain CPUs are vulnerable to an exploit against CPU 3341 internal buffers which can forward information to a 3342 disclosure gadget under certain conditions. 3343 3344 In vulnerable processors, the speculatively 3345 forwarded data can be used in a cache side channel 3346 attack, to access data to which the attacker does 3347 not have direct access. 3348 3349 This parameter controls the mitigation. The 3350 options are: 3351 3352 on - enable the interface for the mitigation 3353 3354 l1tf= [X86,EARLY] Control mitigation of the L1TF vulnerability on 3355 affected CPUs 3356 3357 The kernel PTE inversion protection is unconditionally 3358 enabled and cannot be disabled. 3359 3360 full 3361 Provides all available mitigations for the 3362 L1TF vulnerability. Disables SMT and 3363 enables all mitigations in the 3364 hypervisors, i.e. unconditional L1D flush. 3365 3366 SMT control and L1D flush control via the 3367 sysfs interface is still possible after 3368 boot. Hypervisors will issue a warning 3369 when the first VM is started in a 3370 potentially insecure configuration, 3371 i.e. SMT enabled or L1D flush disabled. 3372 3373 full,force 3374 Same as 'full', but disables SMT and L1D 3375 flush runtime control. Implies the 3376 'nosmt=force' command line option. 3377 (i.e. sysfs control of SMT is disabled.) 3378 3379 flush 3380 Leaves SMT enabled and enables the default 3381 hypervisor mitigation, i.e. conditional 3382 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 flush,nosmt 3392 3393 Disables SMT and enables the default 3394 hypervisor mitigation. 3395 3396 SMT control and L1D flush control via the 3397 sysfs interface is still possible after 3398 boot. Hypervisors will issue a warning 3399 when the first VM is started in a 3400 potentially insecure configuration, 3401 i.e. SMT enabled or L1D flush disabled. 3402 3403 flush,nowarn 3404 Same as 'flush', but hypervisors will not 3405 warn when a VM is started in a potentially 3406 insecure configuration. 3407 3408 off 3409 Disables hypervisor mitigations and doesn't 3410 emit any warnings. 3411 It also drops the swap size and available 3412 RAM limit restriction on both hypervisor and 3413 bare metal. 3414 3415 Default is 'flush'. 3416 3417 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 3418 3419 l2cr= [PPC] 3420 3421 l3cr= [PPC] 3422 3423 lapic [X86-32,APIC,EARLY] Enable the local APIC even if BIOS 3424 disabled it. 3425 3426 lapic= [X86,APIC] Do not use TSC deadline 3427 value for LAPIC timer one-shot implementation. Default 3428 back to the programmable timer unit in the LAPIC. 3429 Format: notscdeadline 3430 3431 lapic_timer_c2_ok [X86,APIC,EARLY] trust the local apic timer 3432 in C2 power state. 3433 3434 libata.dma= [LIBATA] DMA control 3435 libata.dma=0 Disable all PATA and SATA DMA 3436 libata.dma=1 PATA and SATA Disk DMA only 3437 libata.dma=2 ATAPI (CDROM) DMA only 3438 libata.dma=4 Compact Flash DMA only 3439 Combinations also work, so libata.dma=3 enables DMA 3440 for disks and CDROMs, but not CFs. 3441 3442 libata.ignore_hpa= [LIBATA] Ignore HPA limit 3443 libata.ignore_hpa=0 keep BIOS limits (default) 3444 libata.ignore_hpa=1 ignore limits, using full disk 3445 3446 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 3447 when set. 3448 Format: <int> 3449 3450 libata.force= [LIBATA] Force configurations. The format is a comma- 3451 separated list of "[ID:]VAL" where ID is PORT[.DEVICE]. 3452 PORT and DEVICE are decimal numbers matching port, link 3453 or device. Basically, it matches the ATA ID string 3454 printed on console by libata. If the whole ID part is 3455 omitted, the last PORT and DEVICE values are used. If 3456 ID hasn't been specified yet, the configuration applies 3457 to all ports, links and devices. 3458 3459 If only DEVICE is omitted, the parameter applies to 3460 the port and all links and devices behind it. DEVICE 3461 number of 0 either selects the first device or the 3462 first fan-out link behind PMP device. It does not 3463 select the host link. DEVICE number of 15 selects the 3464 host link and device attached to it. 3465 3466 The VAL specifies the configuration to force. As long 3467 as there is no ambiguity, shortcut notation is allowed. 3468 For example, both 1.5 and 1.5G would work for 1.5Gbps. 3469 The following configurations can be forced. 3470 3471 * Cable type: 40c, 80c, short40c, unk, ign or sata. 3472 Any ID with matching PORT is used. 3473 3474 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 3475 3476 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 3477 udma[/][16,25,33,44,66,100,133] notation is also 3478 allowed. 3479 3480 * nohrst, nosrst, norst: suppress hard, soft and both 3481 resets. 3482 3483 * rstonce: only attempt one reset during hot-unplug 3484 link recovery. 3485 3486 * [no]dbdelay: Enable or disable the extra 200ms delay 3487 before debouncing a link PHY and device presence 3488 detection. 3489 3490 * [no]ncq: Turn on or off NCQ. 3491 3492 * [no]ncqtrim: Enable or disable queued DSM TRIM. 3493 3494 * [no]ncqati: Enable or disable NCQ trim on ATI chipset. 3495 3496 * [no]trim: Enable or disable (unqueued) TRIM. 3497 3498 * trim_zero: Indicate that TRIM command zeroes data. 3499 3500 * max_trim_128m: Set 128M maximum trim size limit. 3501 3502 * [no]dma: Turn on or off DMA transfers. 3503 3504 * atapi_dmadir: Enable ATAPI DMADIR bridge support. 3505 3506 * atapi_mod16_dma: Enable the use of ATAPI DMA for 3507 commands that are not a multiple of 16 bytes. 3508 3509 * [no]dmalog: Enable or disable the use of the 3510 READ LOG DMA EXT command to access logs. 3511 3512 * [no]iddevlog: Enable or disable access to the 3513 identify device data log. 3514 3515 * [no]logdir: Enable or disable access to the general 3516 purpose log directory. 3517 3518 * max_sec=<sectors>: Set the transfer size limit, in 3519 number of 512-byte sectors, to the value specified in 3520 <sectors>. The value specified in <sectors> has to be 3521 a non-zero positive integer. 3522 3523 * max_sec_128: Set transfer size limit to 128 sectors. 3524 3525 * max_sec_1024: Set or clear transfer size limit to 3526 1024 sectors. 3527 3528 * max_sec_lba48: Set or clear transfer size limit to 3529 65535 sectors. 3530 3531 * external: Mark port as external (hotplug-capable). 3532 3533 * [no]lpm: Enable or disable link power management. 3534 3535 * [no]setxfer: Indicate if transfer speed mode setting 3536 should be skipped. 3537 3538 * [no]fua: Disable or enable FUA (Force Unit Access) 3539 support for devices supporting this feature. 3540 3541 * dump_id: Dump IDENTIFY data. 3542 3543 * disable: Disable this device. 3544 3545 If there are multiple matching configurations changing 3546 the same attribute, the last one is used. 3547 3548 liveupdate= [KNL,EARLY] 3549 Format: <bool> 3550 Enable Live Update Orchestrator (LUO). 3551 Default: off. 3552 3553 lockd.nlm_grace_period=P [NFS] Assign grace period. 3554 Format: <integer> 3555 3556 lockd.nlm_tcpport=N [NFS] Assign TCP port. 3557 Format: <integer> 3558 3559 lockd.nlm_timeout=T [NFS] Assign timeout value. 3560 Format: <integer> 3561 3562 lockd.nlm_udpport=M [NFS] Assign UDP port. 3563 Format: <integer> 3564 3565 lockdown= [SECURITY,EARLY] 3566 { integrity | confidentiality } 3567 Enable the kernel lockdown feature. If set to 3568 integrity, kernel features that allow userland to 3569 modify the running kernel are disabled. If set to 3570 confidentiality, kernel features that allow userland 3571 to extract confidential information from the kernel 3572 are also disabled. 3573 3574 locktorture.acq_writer_lim= [KNL] 3575 Set the time limit in jiffies for a lock 3576 acquisition. Acquisitions exceeding this limit 3577 will result in a splat once they do complete. 3578 3579 locktorture.bind_readers= [KNL] 3580 Specify the list of CPUs to which the readers are 3581 to be bound. 3582 3583 locktorture.bind_writers= [KNL] 3584 Specify the list of CPUs to which the writers are 3585 to be bound. 3586 3587 locktorture.call_rcu_chains= [KNL] 3588 Specify the number of self-propagating call_rcu() 3589 chains to set up. These are used to ensure that 3590 there is a high probability of an RCU grace period 3591 in progress at any given time. Defaults to 0, 3592 which disables these call_rcu() chains. 3593 3594 locktorture.long_hold= [KNL] 3595 Specify the duration in milliseconds for the 3596 occasional long-duration lock hold time. Defaults 3597 to 100 milliseconds. Select 0 to disable. 3598 3599 locktorture.nested_locks= [KNL] 3600 Specify the maximum lock nesting depth that 3601 locktorture is to exercise, up to a limit of 8 3602 (MAX_NESTED_LOCKS). Specify zero to disable. 3603 Note that this parameter is ineffective on types 3604 of locks that do not support nested acquisition. 3605 3606 locktorture.nreaders_stress= [KNL] 3607 Set the number of locking read-acquisition kthreads. 3608 Defaults to being automatically set based on the 3609 number of online CPUs. 3610 3611 locktorture.nwriters_stress= [KNL] 3612 Set the number of locking write-acquisition kthreads. 3613 3614 locktorture.onoff_holdoff= [KNL] 3615 Set time (s) after boot for CPU-hotplug testing. 3616 3617 locktorture.onoff_interval= [KNL] 3618 Set time (s) between CPU-hotplug operations, or 3619 zero to disable CPU-hotplug testing. 3620 3621 locktorture.rt_boost= [KNL] 3622 Do periodic testing of real-time lock priority 3623 boosting. Select 0 to disable, 1 to boost 3624 only rt_mutex, and 2 to boost unconditionally. 3625 Defaults to 2, which might seem to be an 3626 odd choice, but which should be harmless for 3627 non-real-time spinlocks, due to their disabling 3628 of preemption. Note that non-realtime mutexes 3629 disable boosting. 3630 3631 locktorture.rt_boost_factor= [KNL] 3632 Number that determines how often and for how 3633 long priority boosting is exercised. This is 3634 scaled down by the number of writers, so that the 3635 number of boosts per unit time remains roughly 3636 constant as the number of writers increases. 3637 On the other hand, the duration of each boost 3638 increases with the number of writers. 3639 3640 locktorture.shuffle_interval= [KNL] 3641 Set task-shuffle interval (jiffies). Shuffling 3642 tasks allows some CPUs to go into dyntick-idle 3643 mode during the locktorture test. 3644 3645 locktorture.shutdown_secs= [KNL] 3646 Set time (s) after boot system shutdown. This 3647 is useful for hands-off automated testing. 3648 3649 locktorture.stat_interval= [KNL] 3650 Time (s) between statistics printk()s. 3651 3652 locktorture.stutter= [KNL] 3653 Time (s) to stutter testing, for example, 3654 specifying five seconds causes the test to run for 3655 five seconds, wait for five seconds, and so on. 3656 This tests the locking primitive's ability to 3657 transition abruptly to and from idle. 3658 3659 locktorture.torture_type= [KNL] 3660 Specify the locking implementation to test. 3661 3662 locktorture.verbose= [KNL] 3663 Enable additional printk() statements. 3664 3665 locktorture.writer_fifo= [KNL] 3666 Run the write-side locktorture kthreads at 3667 sched_set_fifo() real-time priority. 3668 3669 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 3670 Format: <irq> 3671 3672 loglevel= [KNL,EARLY] 3673 All Kernel Messages with a loglevel smaller than the 3674 console loglevel will be printed to the console. It can 3675 also be changed with klogd or other programs. The 3676 loglevels are defined as follows: 3677 3678 0 (KERN_EMERG) system is unusable 3679 1 (KERN_ALERT) action must be taken immediately 3680 2 (KERN_CRIT) critical conditions 3681 3 (KERN_ERR) error conditions 3682 4 (KERN_WARNING) warning conditions 3683 5 (KERN_NOTICE) normal but significant condition 3684 6 (KERN_INFO) informational 3685 7 (KERN_DEBUG) debug-level messages 3686 3687 log_buf_len=n[KMG] [KNL,EARLY] 3688 Sets the size of the printk ring buffer, in bytes. 3689 n must be a power of two and greater than the 3690 minimal size. The minimal size is defined by 3691 LOG_BUF_SHIFT kernel config parameter. There 3692 is also CONFIG_LOG_CPU_MAX_BUF_SHIFT config 3693 parameter that allows to increase the default size 3694 depending on the number of CPUs. See init/Kconfig 3695 for more details. 3696 3697 logo.nologo [FB] Disables display of the built-in Linux logo. 3698 This may be used to provide more screen space for 3699 kernel log messages and is useful when debugging 3700 kernel boot problems. 3701 3702 lp=0 [LP] Specify parallel ports to use, e.g, 3703 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 3704 lp=reset first parallel port). 'lp=0' disables the 3705 lp=auto printer driver. 'lp=reset' (which can be 3706 specified in addition to the ports) causes 3707 attached printers to be reset. Using 3708 lp=port1,port2,... specifies the parallel ports 3709 to associate lp devices with, starting with 3710 lp0. A port specification may be 'none' to skip 3711 that lp device, or a parport name such as 3712 'parport0'. Specifying 'lp=auto' instead of a 3713 port specification list means that device IDs 3714 from each port should be examined, to see if 3715 an IEEE 1284-compliant printer is attached; if 3716 so, the driver will manage that printer. 3717 See also header of drivers/char/lp.c. 3718 3719 lpj=n [KNL] 3720 Sets loops_per_jiffy to given constant, thus avoiding 3721 time-consuming boot-time autodetection (up to 250 ms per 3722 CPU). 0 enables autodetection (default). To determine 3723 the correct value for your kernel, boot with normal 3724 autodetection and see what value is printed. Note that 3725 on SMP systems the preset will be applied to all CPUs, 3726 which is likely to cause problems if your CPUs need 3727 significantly divergent settings. An incorrect value 3728 will cause delays in the kernel to be wrong, leading to 3729 unpredictable I/O errors and other breakage. Although 3730 unlikely, in the extreme case this might damage your 3731 hardware. 3732 3733 lsm.debug [SECURITY] Enable LSM initialization debugging output. 3734 3735 lsm=lsm1,...,lsmN 3736 [SECURITY] Choose order of LSM initialization. This 3737 overrides CONFIG_LSM, and the "security=" parameter. 3738 3739 machtype= [Loongson] Share the same kernel image file between 3740 different yeeloong laptops. 3741 Example: machtype=lemote-yeeloong-2f-7inch 3742 3743 maxcpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 3744 will bring up during bootup. maxcpus=n : n >= 0 limits 3745 the kernel to bring up 'n' processors. Surely after 3746 bootup you can bring up the other plugged cpu by executing 3747 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 3748 only takes effect during system bootup. 3749 While n=0 is a special case, it is equivalent to "nosmp", 3750 which also disables the IO APIC. 3751 3752 max_loop= [LOOP] The number of loop block devices that get 3753 (loop.max_loop) unconditionally pre-created at init time. The default 3754 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 3755 of statically allocating a predefined number, loop 3756 devices can be requested on-demand with the 3757 /dev/loop-control interface. 3758 3759 mce= [X86-{32,64}] 3760 3761 Please see Documentation/arch/x86/x86_64/machinecheck.rst for sysfs runtime tunables. 3762 3763 off 3764 disable machine check 3765 3766 no_cmci 3767 disable CMCI(Corrected Machine Check Interrupt) that 3768 Intel processor supports. Usually this disablement is 3769 not recommended, but it might be handy if your 3770 hardware is misbehaving. 3771 3772 Note that you'll get more problems without CMCI than 3773 with due to the shared banks, i.e. you might get 3774 duplicated error logs. 3775 3776 dont_log_ce 3777 don't make logs for corrected errors. All events 3778 reported as corrected are silently cleared by OS. This 3779 option will be useful if you have no interest in any 3780 of corrected errors. 3781 3782 ignore_ce 3783 disable features for corrected errors, e.g. 3784 polling timer and CMCI. All events reported as 3785 corrected are not cleared by OS and remained in its 3786 error banks. 3787 3788 Usually this disablement is not recommended, however 3789 if there is an agent checking/clearing corrected 3790 errors (e.g. BIOS or hardware monitoring 3791 applications), conflicting with OS's error handling, 3792 and you cannot deactivate the agent, then this option 3793 will be a help. 3794 3795 no_lmce 3796 do not opt-in to Local MCE delivery. Use legacy method 3797 to broadcast MCEs. 3798 3799 bootlog 3800 enable logging of machine checks left over from 3801 booting. Disabled by default on AMD Fam10h and older 3802 because some BIOS leave bogus ones. 3803 3804 If your BIOS doesn't do that it's a good idea to 3805 enable though to make sure you log even machine check 3806 events that result in a reboot. On Intel systems it is 3807 enabled by default. 3808 3809 nobootlog 3810 disable boot machine check logging. 3811 3812 print_all 3813 print all machine check logs to the console. 3814 3815 monarchtimeout (number) 3816 sets the time in us to wait for other CPUs on machine 3817 checks. 0 to disable. 3818 3819 bios_cmci_threshold 3820 don't overwrite the bios-set CMCI threshold. This boot 3821 option prevents Linux from overwriting the CMCI 3822 threshold set by the bios. Without this option, Linux 3823 always sets the CMCI threshold to 1. Enabling this may 3824 make memory predictive failure analysis less effective 3825 if the bios sets thresholds for memory errors since we 3826 will not see details for all errors. 3827 3828 recovery 3829 force-enable recoverable machine check code paths 3830 3831 Everything else is in sysfs now. 3832 3833 3834 md= [HW] RAID subsystems devices and level 3835 See Documentation/admin-guide/md.rst. 3836 3837 mds= [X86,INTEL,EARLY] 3838 Control mitigation for the Micro-architectural Data 3839 Sampling (MDS) vulnerability. 3840 3841 Certain CPUs are vulnerable to an exploit against CPU 3842 internal buffers which can forward information to a 3843 disclosure gadget under certain conditions. 3844 3845 In vulnerable processors, the speculatively 3846 forwarded data can be used in a cache side channel 3847 attack, to access data to which the attacker does 3848 not have direct access. 3849 3850 This parameter controls the MDS mitigation. The 3851 options are: 3852 3853 full - Enable MDS mitigation on vulnerable CPUs 3854 full,nosmt - Enable MDS mitigation and disable 3855 SMT on vulnerable CPUs 3856 off - Unconditionally disable MDS mitigation 3857 3858 On TAA-affected machines, mds=off can be prevented by 3859 an active TAA mitigation as both vulnerabilities are 3860 mitigated with the same mechanism so in order to disable 3861 this mitigation, you need to specify tsx_async_abort=off 3862 too. 3863 3864 Not specifying this option is equivalent to 3865 mds=full. 3866 3867 For details see: Documentation/admin-guide/hw-vuln/mds.rst 3868 3869 mem=nn[KMG] [HEXAGON,EARLY] Set the memory size. 3870 Must be specified, otherwise memory size will be 0. 3871 3872 mem=nn[KMG] [KNL,BOOT,EARLY] Force usage of a specific amount 3873 of memory Amount of memory to be used in cases 3874 as follows: 3875 3876 1 for test; 3877 2 when the kernel is not able to see the whole system memory; 3878 3 memory that lies after 'mem=' boundary is excluded from 3879 the hypervisor, then assigned to KVM guests. 3880 4 to limit the memory available for kdump kernel. 3881 3882 [ARC,MICROBLAZE] - the limit applies only to low memory, 3883 high memory is not affected. 3884 3885 [ARM64] - only limits memory covered by the linear 3886 mapping. The NOMAP regions are not affected. 3887 3888 [X86] Work as limiting max address. Use together 3889 with memmap= to avoid physical address space collisions. 3890 Without memmap= PCI devices could be placed at addresses 3891 belonging to unused RAM. 3892 3893 Note that this only takes effects during boot time since 3894 in above case 3, memory may need be hot added after boot 3895 if system memory of hypervisor is not sufficient. 3896 3897 mem=nn[KMG]@ss[KMG] 3898 [ARM,MIPS,EARLY] - override the memory layout 3899 reported by firmware. 3900 Define a memory region of size nn[KMG] starting at 3901 ss[KMG]. 3902 Multiple different regions can be specified with 3903 multiple mem= parameters on the command line. 3904 3905 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 3906 memory. 3907 3908 memblock=debug [KNL,EARLY] Enable memblock debug messages. 3909 3910 memchunk=nn[KMG] 3911 [KNL,SH] Allow user to override the default size for 3912 per-device physically contiguous DMA buffers. 3913 3914 memhp_default_state=online/offline/online_kernel/online_movable 3915 [KNL] Set the initial state for the memory hotplug 3916 onlining policy. If not specified, the default value is 3917 set according to the 3918 CONFIG_MHP_DEFAULT_ONLINE_TYPE kernel config 3919 options. 3920 See Documentation/admin-guide/mm/memory-hotplug.rst. 3921 3922 memmap=exactmap [KNL,X86,EARLY] Enable setting of an exact 3923 E820 memory map, as specified by the user. 3924 Such memmap=exactmap lines can be constructed based on 3925 BIOS output or other requirements. See the memmap=nn@ss 3926 option description. 3927 3928 memmap=nn[KMG]@ss[KMG] 3929 [KNL, X86,MIPS,XTENSA,EARLY] Force usage of a specific region of memory. 3930 Region of memory to be used is from ss to ss+nn. 3931 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 3932 which limits max address to nn[KMG]. 3933 Multiple different regions can be specified, 3934 comma delimited. 3935 Example: 3936 memmap=100M@2G,100M#3G,1G!1024G 3937 3938 memmap=nn[KMG]#ss[KMG] 3939 [KNL,ACPI,EARLY] Mark specific memory as ACPI data. 3940 Region of memory to be marked is from ss to ss+nn. 3941 3942 memmap=nn[KMG]$ss[KMG] 3943 [KNL,ACPI,EARLY] Mark specific memory as reserved. 3944 Region of memory to be reserved is from ss to ss+nn. 3945 Example: Exclude memory from 0x18690000-0x1869ffff 3946 memmap=64K$0x18690000 3947 or 3948 memmap=0x10000$0x18690000 3949 Some bootloaders may need an escape character before '$', 3950 like Grub2, otherwise '$' and the following number 3951 will be eaten. 3952 3953 memmap=nn[KMG]!ss[KMG,EARLY] 3954 [KNL,X86] Mark specific memory as protected. 3955 Region of memory to be used, from ss to ss+nn. 3956 The memory region may be marked as e820 type 12 (0xc) 3957 and is NVDIMM or ADR memory. 3958 3959 memmap=<size>%<offset>-<oldtype>+<newtype> 3960 [KNL,ACPI,EARLY] Convert memory within the specified region 3961 from <oldtype> to <newtype>. If "-<oldtype>" is left 3962 out, the whole region will be marked as <newtype>, 3963 even if previously unavailable. If "+<newtype>" is left 3964 out, matching memory will be removed. Types are 3965 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 3966 3 = ACPI, 12 = PRAM. 3967 3968 memory_corruption_check=0/1 [X86,EARLY] 3969 Some BIOSes seem to corrupt the first 64k of 3970 memory when doing things like suspend/resume. 3971 Setting this option will scan the memory 3972 looking for corruption. Enabling this will 3973 both detect corruption and prevent the kernel 3974 from using the memory being corrupted. 3975 However, it's intended as a diagnostic tool; if 3976 repeatable BIOS-originated corruption always 3977 affects the same memory, you can use memmap= 3978 to prevent the kernel from using that memory. 3979 3980 memory_corruption_check_size=size [X86,EARLY] 3981 By default it checks for corruption in the low 3982 64k, making this memory unavailable for normal 3983 use. Use this parameter to scan for 3984 corruption in more or less memory. 3985 3986 memory_corruption_check_period=seconds [X86,EARLY] 3987 By default it checks for corruption every 60 3988 seconds. Use this parameter to check at some 3989 other rate. 0 disables periodic checking. 3990 3991 memory_hotplug.memmap_on_memory 3992 [KNL,X86,ARM] Boolean flag to enable this feature. 3993 Format: {on | off (default)} 3994 When enabled, runtime hotplugged memory will 3995 allocate its internal metadata (struct pages, 3996 those vmemmap pages cannot be optimized even 3997 if hugetlb_free_vmemmap is enabled) from the 3998 hotadded memory which will allow to hotadd a 3999 lot of memory without requiring additional 4000 memory to do so. 4001 This feature is disabled by default because it 4002 has some implication on large (e.g. GB) 4003 allocations in some configurations (e.g. small 4004 memory blocks). 4005 The state of the flag can be read in 4006 /sys/module/memory_hotplug/parameters/memmap_on_memory. 4007 Note that even when enabled, there are a few cases where 4008 the feature is not effective. 4009 4010 memtest= [KNL,X86,ARM,M68K,PPC,RISCV,EARLY] Enable memtest 4011 Format: <integer> 4012 default : 0 <disable> 4013 Specifies the number of memtest passes to be 4014 performed. Each pass selects another test 4015 pattern from a given set of patterns. Memtest 4016 fills the memory with this pattern, validates 4017 memory contents and reserves bad memory 4018 regions that are detected. 4019 4020 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 4021 Valid arguments: on, off 4022 Default: off 4023 mem_encrypt=on: Activate SME 4024 mem_encrypt=off: Do not activate SME 4025 4026 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst 4027 for details on when memory encryption can be activated. 4028 4029 mem_sleep_default= [SUSPEND] Default system suspend mode: 4030 s2idle - Suspend-To-Idle 4031 shallow - Power-On Suspend or equivalent (if supported) 4032 deep - Suspend-To-RAM or equivalent (if supported) 4033 See Documentation/admin-guide/pm/sleep-states.rst. 4034 4035 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 4036 the BIOS has incorrectly applied a workaround. TinyBIOS 4037 version 0.98 is known to be affected, 0.99 fixes the 4038 problem by letting the user disable the workaround. 4039 4040 mga= [HW,DRM] 4041 4042 microcode= [X86] Control the behavior of the microcode loader. 4043 Available options, comma separated: 4044 4045 base_rev=X - with <X> with format: <u32> 4046 Set the base microcode revision of each thread when in 4047 debug mode. 4048 4049 dis_ucode_ldr: disable the microcode loader 4050 4051 force_minrev: 4052 Enable or disable the microcode minimal revision 4053 enforcement for the runtime microcode loader. 4054 4055 mini2440= [ARM,HW,KNL] 4056 Format:[0..2][b][c][t] 4057 Default: "0tb" 4058 MINI2440 configuration specification: 4059 0 - The attached screen is the 3.5" TFT 4060 1 - The attached screen is the 7" TFT 4061 2 - The VGA Shield is attached (1024x768) 4062 Leaving out the screen size parameter will not load 4063 the TFT driver, and the framebuffer will be left 4064 unconfigured. 4065 b - Enable backlight. The TFT backlight pin will be 4066 linked to the kernel VESA blanking code and a GPIO 4067 LED. This parameter is not necessary when using the 4068 VGA shield. 4069 c - Enable the s3c camera interface. 4070 t - Reserved for enabling touchscreen support. The 4071 touchscreen support is not enabled in the mainstream 4072 kernel as of 2.6.30, a preliminary port can be found 4073 in the "bleeding edge" mini2440 support kernel at 4074 https://repo.or.cz/w/linux-2.6/mini2440.git 4075 4076 mitigations= 4077 [X86,PPC,S390,ARM64,EARLY] Control optional mitigations for 4078 CPU vulnerabilities. This is a set of curated, 4079 arch-independent options, each of which is an 4080 aggregation of existing arch-specific options. 4081 4082 Note, "mitigations" is supported if and only if the 4083 kernel was built with CPU_MITIGATIONS=y. 4084 4085 off 4086 Disable all optional CPU mitigations. This 4087 improves system performance, but it may also 4088 expose users to several CPU vulnerabilities. 4089 Equivalent to: if nokaslr then kpti=0 [ARM64] 4090 gather_data_sampling=off [X86] 4091 indirect_target_selection=off [X86] 4092 kvm.nx_huge_pages=off [X86] 4093 l1tf=off [X86] 4094 mds=off [X86] 4095 mmio_stale_data=off [X86] 4096 no_entry_flush [PPC] 4097 no_uaccess_flush [PPC] 4098 nobp=0 [S390] 4099 nopti [X86,PPC] 4100 nospectre_bhb [ARM64] 4101 nospectre_v1 [X86,PPC] 4102 nospectre_v2 [X86,PPC,S390,ARM64] 4103 reg_file_data_sampling=off [X86] 4104 retbleed=off [X86] 4105 spec_rstack_overflow=off [X86] 4106 spec_store_bypass_disable=off [X86,PPC] 4107 spectre_bhi=off [X86] 4108 spectre_v2_user=off [X86] 4109 srbds=off [X86,INTEL] 4110 ssbd=force-off [ARM64] 4111 tsa=off [X86,AMD] 4112 tsx_async_abort=off [X86] 4113 vmscape=off [X86] 4114 4115 Exceptions: 4116 This does not have any effect on 4117 kvm.nx_huge_pages when 4118 kvm.nx_huge_pages=force. 4119 4120 auto (default) 4121 Mitigate all CPU vulnerabilities, but leave SMT 4122 enabled, even if it's vulnerable. This is for 4123 users who don't want to be surprised by SMT 4124 getting disabled across kernel upgrades, or who 4125 have other ways of avoiding SMT-based attacks. 4126 Equivalent to: (default behavior) 4127 4128 auto,nosmt 4129 Mitigate all CPU vulnerabilities, disabling SMT 4130 if needed. This is for users who always want to 4131 be fully mitigated, even if it means losing SMT. 4132 Equivalent to: l1tf=flush,nosmt [X86] 4133 mds=full,nosmt [X86] 4134 tsx_async_abort=full,nosmt [X86] 4135 mmio_stale_data=full,nosmt [X86] 4136 retbleed=auto,nosmt [X86] 4137 4138 [X86] After one of the above options, additionally 4139 supports attack-vector based controls as documented in 4140 Documentation/admin-guide/hw-vuln/attack_vector_controls.rst 4141 4142 mminit_loglevel= 4143 [KNL,EARLY] When CONFIG_DEBUG_MEMORY_INIT is set, this 4144 parameter allows control of the logging verbosity for 4145 the additional memory initialisation checks. A value 4146 of 0 disables mminit logging and a level of 4 will 4147 log everything. Information is printed at KERN_DEBUG 4148 so loglevel=8 may also need to be specified. 4149 4150 mmio_stale_data= 4151 [X86,INTEL,EARLY] Control mitigation for the Processor 4152 MMIO Stale Data vulnerabilities. 4153 4154 Processor MMIO Stale Data is a class of 4155 vulnerabilities that may expose data after an MMIO 4156 operation. Exposed data could originate or end in 4157 the same CPU buffers as affected by MDS and TAA. 4158 Therefore, similar to MDS and TAA, the mitigation 4159 is to clear the affected CPU buffers. 4160 4161 This parameter controls the mitigation. The 4162 options are: 4163 4164 full - Enable mitigation on vulnerable CPUs 4165 4166 full,nosmt - Enable mitigation and disable SMT on 4167 vulnerable CPUs. 4168 4169 off - Unconditionally disable mitigation 4170 4171 On MDS or TAA affected machines, 4172 mmio_stale_data=off can be prevented by an active 4173 MDS or TAA mitigation as these vulnerabilities are 4174 mitigated with the same mechanism so in order to 4175 disable this mitigation, you need to specify 4176 mds=off and tsx_async_abort=off too. 4177 4178 Not specifying this option is equivalent to 4179 mmio_stale_data=full. 4180 4181 For details see: 4182 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst 4183 4184 <module>.async_probe[=<bool>] [KNL] 4185 If no <bool> value is specified or if the value 4186 specified is not a valid <bool>, enable asynchronous 4187 probe on this module. Otherwise, enable/disable 4188 asynchronous probe on this module as indicated by the 4189 <bool> value. See also: module.async_probe 4190 4191 module.async_probe=<bool> 4192 [KNL] When set to true, modules will use async probing 4193 by default. To enable/disable async probing for a 4194 specific module, use the module specific control that 4195 is documented under <module>.async_probe. When both 4196 module.async_probe and <module>.async_probe are 4197 specified, <module>.async_probe takes precedence for 4198 the specific module. 4199 4200 module.enable_dups_trace 4201 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set, 4202 this means that duplicate request_module() calls will 4203 trigger a WARN_ON() instead of a pr_warn(). Note that 4204 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s 4205 will always be issued and this option does nothing. 4206 module.sig_enforce 4207 [KNL] When CONFIG_MODULE_SIG is set, this means that 4208 modules without (valid) signatures will fail to load. 4209 Note that if CONFIG_MODULE_SIG_FORCE is set, that 4210 is always true, so this option does nothing. 4211 4212 module_blacklist= [KNL] Do not load a comma-separated list of 4213 modules. Useful for debugging problem modules. 4214 4215 mousedev.tap_time= 4216 [MOUSE] Maximum time between finger touching and 4217 leaving touchpad surface for touch to be considered 4218 a tap and be reported as a left button click (for 4219 touchpads working in absolute mode only). 4220 Format: <msecs> 4221 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 4222 reporting absolute coordinates, such as tablets 4223 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 4224 reporting absolute coordinates, such as tablets 4225 4226 movablecore= [KNL,X86,PPC,EARLY] 4227 Format: nn[KMGTPE] | nn% 4228 This parameter is the complement to kernelcore=, it 4229 specifies the amount of memory used for migratable 4230 allocations. If both kernelcore and movablecore is 4231 specified, then kernelcore will be at *least* the 4232 specified value but may be more. If movablecore on its 4233 own is specified, the administrator must be careful 4234 that the amount of memory usable for all allocations 4235 is not too small. 4236 4237 movable_node [KNL,EARLY] Boot-time switch to make hotplugable memory 4238 NUMA nodes to be movable. This means that the memory 4239 of such nodes will be usable only for movable 4240 allocations which rules out almost all kernel 4241 allocations. Use with caution! 4242 4243 MTD_Partition= [MTD] 4244 Format: <name>,<region-number>,<size>,<offset> 4245 4246 MTD_Region= [MTD] Format: 4247 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 4248 4249 mtdparts= [MTD] 4250 See drivers/mtd/parsers/cmdlinepart.c 4251 4252 mtouchusb.raw_coordinates= 4253 [HW] Make the MicroTouch USB driver use raw coordinates 4254 ('y', default) or cooked coordinates ('n') 4255 4256 mtrr=debug [X86,EARLY] 4257 Enable printing debug information related to MTRR 4258 registers at boot time. 4259 4260 mtrr_chunk_size=nn[KMG,X86,EARLY] 4261 used for mtrr cleanup. It is largest continuous chunk 4262 that could hold holes aka. UC entries. 4263 4264 mtrr_gran_size=nn[KMG,X86,EARLY] 4265 Used for mtrr cleanup. It is granularity of mtrr block. 4266 Default is 1. 4267 Large value could prevent small alignment from 4268 using up MTRRs. 4269 4270 mtrr_spare_reg_nr=n [X86,EARLY] 4271 Format: <integer> 4272 Range: 0,7 : spare reg number 4273 Default : 1 4274 Used for mtrr cleanup. It is spare mtrr entries number. 4275 Set to 2 or more if your graphical card needs more. 4276 4277 multitce=off [PPC] This parameter disables the use of the pSeries 4278 firmware feature for updating multiple TCE entries 4279 at a time. 4280 4281 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 4282 4283 netdev= [NET] NE2000 ISA network devices parameters 4284 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 4285 4286 netpoll.carrier_timeout= 4287 [NET] Specifies amount of time (in seconds) that 4288 netpoll should wait for a carrier. By default netpoll 4289 waits 4 seconds. 4290 4291 nf_conntrack.acct= 4292 [NETFILTER] Enable connection tracking flow accounting 4293 0 to disable accounting 4294 1 to enable accounting 4295 Default value is 0. 4296 4297 nfs.cache_getent= 4298 [NFS] sets the pathname to the program which is used 4299 to update the NFS client cache entries. 4300 4301 nfs.cache_getent_timeout= 4302 [NFS] sets the timeout after which an attempt to 4303 update a cache entry is deemed to have failed. 4304 4305 nfs.callback_nr_threads= 4306 [NFSv4] set the total number of threads that the 4307 NFS client will assign to service NFSv4 callback 4308 requests. 4309 4310 nfs.callback_tcpport= 4311 [NFS] set the TCP port on which the NFSv4 callback 4312 channel should listen. 4313 4314 nfs.delay_retrans= 4315 [NFS] specifies the number of times the NFSv4 client 4316 retries the request before returning an EAGAIN error, 4317 after a reply of NFS4ERR_DELAY from the server. 4318 Only applies if the softerr mount option is enabled, 4319 and the specified value is >= 0. 4320 4321 nfs.idmap_cache_timeout= 4322 [NFS] set the maximum lifetime for idmapper cache 4323 entries. 4324 4325 nfs.max_session_cb_slots= 4326 [NFSv4.1] Sets the maximum number of session 4327 slots the client will assign to the callback 4328 channel. This determines the maximum number of 4329 callbacks the client will process in parallel for 4330 a particular server. 4331 4332 nfs.max_session_slots= 4333 [NFSv4.1] Sets the maximum number of session slots 4334 the client will attempt to negotiate with the server. 4335 This limits the number of simultaneous RPC requests 4336 that the client can send to the NFSv4.1 server. 4337 Note that there is little point in setting this 4338 value higher than the max_tcp_slot_table_limit. 4339 4340 nfs.nfs4_disable_idmapping= 4341 [NFSv4] When set to the default of '1', this option 4342 ensures that both the RPC level authentication 4343 scheme and the NFS level operations agree to use 4344 numeric uids/gids if the mount is using the 4345 'sec=sys' security flavour. In effect it is 4346 disabling idmapping, which can make migration from 4347 legacy NFSv2/v3 systems to NFSv4 easier. 4348 Servers that do not support this mode of operation 4349 will be autodetected by the client, and it will fall 4350 back to using the idmapper. 4351 To turn off this behaviour, set the value to '0'. 4352 4353 nfs.nfs4_unique_id= 4354 [NFS4] Specify an additional fixed unique ident- 4355 ification string that NFSv4 clients can insert into 4356 their nfs_client_id4 string. This is typically a 4357 UUID that is generated at system install time. 4358 4359 nfs.recover_lost_locks= 4360 [NFSv4] Attempt to recover locks that were lost due 4361 to a lease timeout on the server. Please note that 4362 doing this risks data corruption, since there are 4363 no guarantees that the file will remain unchanged 4364 after the locks are lost. 4365 If you want to enable the kernel legacy behaviour of 4366 attempting to recover these locks, then set this 4367 parameter to '1'. 4368 The default parameter value of '0' causes the kernel 4369 not to attempt recovery of lost locks. 4370 4371 nfs.send_implementation_id= 4372 [NFSv4.1] Send client implementation identification 4373 information in exchange_id requests. 4374 If zero, no implementation identification information 4375 will be sent. 4376 The default is to send the implementation identification 4377 information. 4378 4379 nfs4.layoutstats_timer= 4380 [NFSv4.2] Change the rate at which the kernel sends 4381 layoutstats to the pNFS metadata server. 4382 4383 Setting this to value to 0 causes the kernel to use 4384 whatever value is the default set by the layout 4385 driver. A non-zero value sets the minimum interval 4386 in seconds between layoutstats transmissions. 4387 4388 nfsd.inter_copy_offload_enable= 4389 [NFSv4.2] When set to 1, the server will support 4390 server-to-server copies for which this server is 4391 the destination of the copy. 4392 4393 nfsd.nfs4_disable_idmapping= 4394 [NFSv4] When set to the default of '1', the NFSv4 4395 server will return only numeric uids and gids to 4396 clients using auth_sys, and will accept numeric uids 4397 and gids from such clients. This is intended to ease 4398 migration from NFSv2/v3. 4399 4400 nfsd.nfsd4_ssc_umount_timeout= 4401 [NFSv4.2] When used as the destination of a 4402 server-to-server copy, knfsd temporarily mounts 4403 the source server. It caches the mount in case 4404 it will be needed again, and discards it if not 4405 used for the number of milliseconds specified by 4406 this parameter. 4407 4408 nfsaddrs= [NFS] Deprecated. Use ip= instead. 4409 See Documentation/admin-guide/nfs/nfsroot.rst. 4410 4411 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 4412 See Documentation/admin-guide/nfs/nfsroot.rst. 4413 4414 nfsrootdebug [NFS] enable nfsroot debugging messages. 4415 See Documentation/admin-guide/nfs/nfsroot.rst. 4416 4417 nmi_backtrace.backtrace_idle [KNL] 4418 Dump stacks even of idle CPUs in response to an 4419 NMI stack-backtrace request. 4420 4421 nmi_debug= [KNL,SH] Specify one or more actions to take 4422 when a NMI is triggered. 4423 Format: [state][,regs][,debounce][,die] 4424 4425 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 4426 Format: [panic,][nopanic,][rNNN,][num] 4427 Valid num: 0 or 1 4428 0 - turn hardlockup detector in nmi_watchdog off 4429 1 - turn hardlockup detector in nmi_watchdog on 4430 rNNN - configure the watchdog with raw perf event 0xNNN 4431 4432 When panic is specified, panic when an NMI watchdog 4433 timeout occurs (or 'nopanic' to not panic on an NMI 4434 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 4435 To disable both hard and soft lockup detectors, 4436 please see 'nowatchdog'. 4437 This is useful when you use a panic=... timeout and 4438 need the box quickly up again. 4439 4440 These settings can be accessed at runtime via 4441 the nmi_watchdog and hardlockup_panic sysctls. 4442 4443 no4lvl [RISCV,EARLY] Disable 4-level and 5-level paging modes. 4444 Forces kernel to use 3-level paging instead. 4445 4446 no5lvl [X86-64,RISCV,EARLY] Disable 5-level paging mode. Forces 4447 kernel to use 4-level paging instead. 4448 4449 noalign [KNL,ARM] 4450 4451 noapic [SMP,APIC,EARLY] Tells the kernel to not make use of any 4452 IOAPICs that may be present in the system. 4453 4454 noapictimer [APIC,X86] Don't set up the APIC timer 4455 4456 noautogroup Disable scheduler automatic task group creation. 4457 4458 nocache [ARM,EARLY] 4459 4460 no_console_suspend 4461 [HW] Never suspend the console 4462 Disable suspending of consoles during suspend and 4463 hibernate operations. Once disabled, debugging 4464 messages can reach various consoles while the rest 4465 of the system is being put to sleep (ie, while 4466 debugging driver suspend/resume hooks). This may 4467 not work reliably with all consoles, but is known 4468 to work with serial and VGA consoles. 4469 To facilitate more flexible debugging, we also add 4470 console_suspend, a printk module parameter to control 4471 it. Users could use console_suspend (usually 4472 /sys/module/printk/parameters/console_suspend) to 4473 turn on/off it dynamically. 4474 4475 no_debug_objects 4476 [KNL,EARLY] Disable object debugging 4477 4478 nodsp [SH] Disable hardware DSP at boot time. 4479 4480 noefi [EFI,EARLY] Disable EFI runtime services support. 4481 4482 no_entry_flush [PPC,EARLY] Don't flush the L1-D cache when entering the kernel. 4483 4484 noexec32 [X86-64] 4485 This affects only 32-bit executables. 4486 noexec32=on: enable non-executable mappings (default) 4487 read doesn't imply executable mappings 4488 noexec32=off: disable non-executable mappings 4489 read implies executable mappings 4490 4491 no_file_caps Tells the kernel not to honor file capabilities. The 4492 only way then for a file to be executed with privilege 4493 is to be setuid root or executed by root. 4494 4495 nofpu [MIPS,SH] Disable hardware FPU at boot time. 4496 4497 nofsgsbase [X86] Disables FSGSBASE instructions. 4498 4499 nofxsr [BUGS=X86-32] Disables x86 floating point extended 4500 register save and restore. The kernel will only save 4501 legacy floating-point registers on task switch. 4502 4503 nogbpages [X86] Do not use GB pages for kernel direct mappings. 4504 4505 no_hash_pointers 4506 [KNL,EARLY] 4507 Alias for "hash_pointers=never". 4508 4509 nohibernate [HIBERNATION] Disable hibernation and resume. 4510 4511 nohlt [ARM,ARM64,MICROBLAZE,MIPS,PPC,RISCV,SH] Forces the kernel to 4512 busy wait in do_idle() and not use the arch_cpu_idle() 4513 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP 4514 to be effective. This is useful on platforms where the 4515 sleep(SH) or wfi(ARM,ARM64) instructions do not work 4516 correctly or when doing power measurements to evaluate 4517 the impact of the sleep instructions. This is also 4518 useful when using JTAG debugger. 4519 4520 nohpet [X86] Don't use the HPET timer. 4521 4522 nohugeiomap [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge I/O mappings. 4523 4524 nohugevmalloc [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge vmalloc mappings. 4525 4526 nohz= [KNL] Boottime enable/disable dynamic ticks 4527 Valid arguments: on, off 4528 Default: on 4529 4530 nohz_full= [KNL,BOOT,SMP,ISOL] 4531 The argument is a cpu list, as described above. 4532 In kernels built with CONFIG_NO_HZ_FULL=y, set 4533 the specified list of CPUs whose tick will be stopped 4534 whenever possible. The boot CPU will be forced outside 4535 the range to maintain the timekeeping. Any CPUs 4536 in this list will have their RCU callbacks offloaded, 4537 just as if they had also been called out in the 4538 rcu_nocbs= boot parameter. 4539 4540 Note that this argument takes precedence over 4541 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 4542 4543 noinitrd [Deprecated,RAM] Tells the kernel not to load any configured 4544 initial RAM disk. Currently this parameter applies to 4545 initrd only, not to initramfs. But it applies to both 4546 in EFI mode. 4547 4548 nointremap [X86-64,Intel-IOMMU,EARLY] Do not enable interrupt 4549 remapping. 4550 [Deprecated - use intremap=off] 4551 4552 noinvpcid [X86,EARLY] Disable the INVPCID cpu feature. 4553 4554 noiotrap [SH] Disables trapped I/O port accesses. 4555 4556 noirqdebug [X86-32] Disables the code which attempts to detect and 4557 disable unhandled interrupt sources. 4558 4559 noisapnp [ISAPNP] Disables ISA PnP code. 4560 4561 nokaslr [KNL,EARLY] 4562 When CONFIG_RANDOMIZE_BASE is set, this disables 4563 kernel and module base offset ASLR (Address Space 4564 Layout Randomization). 4565 4566 no-kvmapf [X86,KVM,EARLY] Disable paravirtualized asynchronous page 4567 fault handling. 4568 4569 no-kvmclock [X86,KVM,EARLY] Disable paravirtualized KVM clock driver 4570 4571 nolapic [X86-32,APIC,EARLY] Do not enable or use the local APIC. 4572 4573 nolapic_timer [X86-32,APIC,EARLY] Do not use the local APIC timer. 4574 4575 nomce [X86-32] Disable Machine Check Exception 4576 4577 nomfgpt [X86-32] Disable Multi-Function General Purpose 4578 Timer usage (for AMD Geode machines). 4579 4580 nomodeset Disable kernel modesetting. Most systems' firmware 4581 sets up a display mode and provides framebuffer memory 4582 for output. With nomodeset, DRM and fbdev drivers will 4583 not load if they could possibly displace the pre- 4584 initialized output. Only the system framebuffer will 4585 be available for use. The respective drivers will not 4586 perform display-mode changes or accelerated rendering. 4587 4588 Useful as error fallback, or for testing and debugging. 4589 4590 nomodule Disable module load 4591 4592 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 4593 shutdown the other cpus. Instead use the REBOOT_VECTOR 4594 irq. 4595 4596 nopat [X86,EARLY] Disable PAT (page attribute table extension of 4597 pagetables) support. 4598 4599 nopcid [X86-64,EARLY] Disable the PCID cpu feature. 4600 4601 nopku [X86] Disable Memory Protection Keys CPU feature found 4602 in some Intel CPUs. 4603 4604 nopti [X86-64,EARLY] 4605 Equivalent to pti=off 4606 4607 nopv= [X86,XEN,KVM,HYPER_V,VMWARE,EARLY] 4608 Disables the PV optimizations forcing the guest to run 4609 as generic guest with no PV drivers. Currently support 4610 XEN HVM, KVM, HYPER_V and VMWARE guest. 4611 4612 nopvspin [X86,XEN,KVM,EARLY] 4613 Disables the qspinlock slow path using PV optimizations 4614 which allow the hypervisor to 'idle' the guest on lock 4615 contention. 4616 4617 norandmaps Don't use address space randomization. Equivalent to 4618 echo 0 > /proc/sys/kernel/randomize_va_space 4619 4620 noresume [SWSUSP] Disables resume and restores original swap 4621 space. 4622 4623 no-scroll [VGA] Disables scrollback. 4624 This is required for the Braillex ib80-piezo Braille 4625 reader made by F.H. Papenmeier (Germany). 4626 4627 nosgx [X86-64,SGX,EARLY] Disables Intel SGX kernel support. 4628 4629 nosmap [PPC,EARLY] 4630 Disable SMAP (Supervisor Mode Access Prevention) 4631 even if it is supported by processor. 4632 4633 nosmep [PPC64s,EARLY] 4634 Disable SMEP (Supervisor Mode Execution Prevention) 4635 even if it is supported by processor. 4636 4637 nosmp [SMP,EARLY] Tells an SMP kernel to act as a UP kernel, 4638 and disable the IO APIC. legacy for "maxcpus=0". 4639 4640 nosmt [KNL,MIPS,PPC,EARLY] Disable symmetric multithreading (SMT). 4641 Equivalent to smt=1. 4642 4643 [KNL,LOONGARCH,X86,ARM64,PPC,S390] Disable symmetric multithreading (SMT). 4644 nosmt=force: Force disable SMT, cannot be undone 4645 via the sysfs control file. 4646 4647 nosoftlockup [KNL] Disable the soft-lockup detector. 4648 4649 nospec_store_bypass_disable 4650 [HW,EARLY] Disable all mitigations for the Speculative 4651 Store Bypass vulnerability 4652 4653 nospectre_bhb [ARM64,EARLY] Disable all mitigations for Spectre-BHB (branch 4654 history injection) vulnerability. System may allow data leaks 4655 with this option. 4656 4657 nospectre_v1 [X86,PPC,EARLY] Disable mitigations for Spectre Variant 1 4658 (bounds check bypass). With this option data leaks are 4659 possible in the system. 4660 4661 nospectre_v2 [X86,PPC_E500,ARM64,EARLY] Disable all mitigations 4662 for the Spectre variant 2 (indirect branch 4663 prediction) vulnerability. System may allow data 4664 leaks with this option. 4665 4666 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES,RISCV,LOONGARCH,EARLY] 4667 Disable paravirtualized steal time accounting. steal time 4668 is computed, but won't influence scheduler behaviour 4669 4670 nosync [HW,M68K] Disables sync negotiation for all devices. 4671 4672 no_timer_check [X86,APIC] Disables the code which tests for broken 4673 timer IRQ sources, i.e., the IO-APIC timer. This can 4674 work around problems with incorrect timer 4675 initialization on some boards. 4676 4677 no_uaccess_flush 4678 [PPC,EARLY] Don't flush the L1-D cache after accessing user data. 4679 4680 novmcoredd [KNL,KDUMP] 4681 Disable device dump. Device dump allows drivers to 4682 append dump data to vmcore so you can collect driver 4683 specified debug info. Drivers can append the data 4684 without any limit and this data is stored in memory, 4685 so this may cause significant memory stress. Disabling 4686 device dump can help save memory but the driver debug 4687 data will be no longer available. This parameter 4688 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 4689 is set. 4690 4691 no-vmw-sched-clock 4692 [X86,PV_OPS,EARLY] Disable paravirtualized VMware 4693 scheduler clock and use the default one. 4694 4695 nowatchdog [KNL] Disable both lockup detectors, i.e. 4696 soft-lockup and NMI watchdog (hard-lockup). 4697 4698 nowb [ARM,EARLY] 4699 4700 nox2apic [X86-64,APIC,EARLY] Do not enable x2APIC mode. 4701 4702 NOTE: this parameter will be ignored on systems with the 4703 LEGACY_XAPIC_DISABLED bit set in the 4704 IA32_XAPIC_DISABLE_STATUS MSR. 4705 4706 noxsave [BUGS=X86] Disables x86 extended register state save 4707 and restore using xsave. The kernel will fallback to 4708 enabling legacy floating-point and sse state. 4709 4710 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 4711 register states. The kernel will fall back to use 4712 xsave to save the states. By using this parameter, 4713 performance of saving the states is degraded because 4714 xsave doesn't support modified optimization while 4715 xsaveopt supports it on xsaveopt enabled systems. 4716 4717 noxsaves [X86] Disables xsaves and xrstors used in saving and 4718 restoring x86 extended register state in compacted 4719 form of xsave area. The kernel will fall back to use 4720 xsaveopt and xrstor to save and restore the states 4721 in standard form of xsave area. By using this 4722 parameter, xsave area per process might occupy more 4723 memory on xsaves enabled systems. 4724 4725 nr_cpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 4726 could support. nr_cpus=n : n >= 1 limits the kernel to 4727 support 'n' processors. It could be larger than the 4728 number of already plugged CPU during bootup, later in 4729 runtime you can physically add extra cpu until it reaches 4730 n. So during boot up some boot time memory for per-cpu 4731 variables need be pre-allocated for later physical cpu 4732 hot plugging. 4733 4734 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 4735 4736 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86, EARLY] 4737 Disable NUMA, Only set up a single NUMA node 4738 spanning all memory. 4739 4740 numa=fake=<size>[MG] 4741 [KNL, ARM64, RISCV, X86, EARLY] 4742 If given as a memory unit, fills all system RAM with 4743 nodes of size interleaved over physical nodes. 4744 4745 numa=fake=<N> 4746 [KNL, ARM64, RISCV, X86, EARLY] 4747 If given as an integer, fills all system RAM with N 4748 fake nodes interleaved over physical nodes. 4749 4750 numa=fake=<N>U 4751 [KNL, ARM64, RISCV, X86, EARLY] 4752 If given as an integer followed by 'U', it will 4753 divide each physical node into N emulated nodes. 4754 4755 numa=noacpi [X86] Don't parse the SRAT table for NUMA setup 4756 4757 numa=nohmat [X86] Don't parse the HMAT table for NUMA setup, or 4758 soft-reserved memory partitioning. 4759 4760 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic 4761 NUMA balancing. 4762 Allowed values are enable and disable 4763 4764 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 4765 'node', 'default' can be specified 4766 This can be set from sysctl after boot. 4767 See Documentation/admin-guide/sysctl/vm.rst for details. 4768 4769 nvme.quirks= [NVME] A list of quirk entries to augment the built-in 4770 nvme quirk list. List entries are separated by a 4771 '-' character. 4772 Each entry has the form VendorID:ProductID:quirk_names. 4773 The IDs are 4-digits hex numbers and quirk_names is a 4774 list of quirk names separated by commas. A quirk name 4775 can be prefixed by '^', meaning that the specified 4776 quirk must be disabled. 4777 4778 Example: 4779 nvme.quirks=7710:2267:bogus_nid,^identify_cns-9900:7711:broken_msi 4780 4781 ohci1394_dma=early [HW,EARLY] enable debugging via the ohci1394 driver. 4782 See Documentation/core-api/debugging-via-ohci1394.rst for more 4783 info. 4784 4785 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 4786 Rather than timing out after 20 ms if an EC 4787 command is not properly ACKed, override the length 4788 of the timeout. We have interrupts disabled while 4789 waiting for the ACK, so if this is set too high 4790 interrupts *may* be lost! 4791 4792 omap_mux= [OMAP] Override bootloader pin multiplexing. 4793 Format: <mux_mode0.mode_name=value>... 4794 For example, to override I2C bus2: 4795 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 4796 4797 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 4798 4799 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 4800 4801 boundary - index of last SLC block on Flex-OneNAND. 4802 The remaining blocks are configured as MLC blocks. 4803 lock - Configure if Flex-OneNAND boundary should be locked. 4804 Once locked, the boundary cannot be changed. 4805 1 indicates lock status, 0 indicates unlock status. 4806 4807 oops=panic [KNL,EARLY] 4808 Always panic on oopses. Default is to just kill the 4809 process, but there is a small probability of 4810 deadlocking the machine. 4811 This will also cause panics on machine check exceptions. 4812 Useful together with panic=30 to trigger a reboot. 4813 4814 page_alloc.shuffle= 4815 [KNL] Boolean flag to control whether the page allocator 4816 should randomize its free lists. This parameter can be 4817 used to enable/disable page randomization. The state of 4818 the flag can be read from sysfs at: 4819 /sys/module/page_alloc/parameters/shuffle. 4820 This parameter is only available if CONFIG_SHUFFLE_PAGE_ALLOCATOR=y. 4821 4822 page_owner= [KNL,EARLY] Boot-time page_owner enabling option. 4823 Storage of the information about who allocated 4824 each page is disabled in default. With this switch, 4825 we can turn it on. 4826 on: enable the feature 4827 4828 page_poison= [KNL,EARLY] Boot-time parameter changing the state of 4829 poisoning on the buddy allocator, available with 4830 CONFIG_PAGE_POISONING=y. 4831 off: turn off poisoning (default) 4832 on: turn on poisoning 4833 4834 page_reporting.page_reporting_order= 4835 [KNL] Minimal page reporting order 4836 Format: <integer> 4837 Adjust the minimal page reporting order. The page 4838 reporting is disabled when it exceeds MAX_PAGE_ORDER. 4839 4840 panic= [KNL] Kernel behaviour on panic: delay <timeout> 4841 timeout > 0: seconds before rebooting 4842 timeout = 0: wait forever 4843 timeout < 0: reboot immediately 4844 Format: <timeout> 4845 4846 panic_on_taint= [KNL,EARLY] 4847 Bitmask for conditionally calling panic() in add_taint() 4848 Format: <hex>[,nousertaint] 4849 Hexadecimal bitmask representing the set of TAINT flags 4850 that will cause the kernel to panic when add_taint() is 4851 called with any of the flags in this set. 4852 The optional switch "nousertaint" can be utilized to 4853 prevent userspace forced crashes by writing to sysctl 4854 /proc/sys/kernel/tainted any flagset matching with the 4855 bitmask set on panic_on_taint. 4856 See Documentation/admin-guide/tainted-kernels.rst for 4857 extra details on the taint flags that users can pick 4858 to compose the bitmask to assign to panic_on_taint. 4859 4860 panic_on_warn=1 panic() instead of WARN(). Useful to cause kdump 4861 on a WARN(). 4862 4863 panic_force_cpu= 4864 [KNL,SMP] Force panic handling to execute on a specific CPU. 4865 Format: <cpu number> 4866 Some platforms require panic handling to occur on a 4867 specific CPU for the crash kernel to function correctly. 4868 This can be due to firmware limitations, interrupt routing 4869 constraints, or platform-specific requirements where only 4870 a particular CPU can safely enter the crash kernel. 4871 When set, panic() will redirect execution to the specified 4872 CPU before proceeding with the normal panic and kexec flow. 4873 If the target CPU is offline or unavailable, panic proceeds 4874 on the current CPU. 4875 This option should only be used for systems with the above 4876 constraints as it might cause the panic operation to be less reliable. 4877 4878 panic_print= Bitmask for printing system info when panic happens. 4879 User can chose combination of the following bits: 4880 bit 0: print all tasks info 4881 bit 1: print system memory info 4882 bit 2: print timer info 4883 bit 3: print locks info if CONFIG_LOCKDEP is on 4884 bit 4: print ftrace buffer 4885 bit 5: replay all kernel messages on consoles at the end of panic 4886 bit 6: print all CPUs backtrace (if available in the arch) 4887 bit 7: print only tasks in uninterruptible (blocked) state 4888 *Be aware* that this option may print a _lot_ of lines, 4889 so there are risks of losing older messages in the log. 4890 Use this option carefully, maybe worth to setup a 4891 bigger log buffer with "log_buf_len" along with this. 4892 4893 panic_sys_info= A comma separated list of extra information to be dumped 4894 on panic. 4895 Format: val[,val...] 4896 Where @val can be any of the following: 4897 4898 tasks: print all tasks info 4899 mem: print system memory info 4900 timers: print timers info 4901 locks: print locks info if CONFIG_LOCKDEP is on 4902 ftrace: print ftrace buffer 4903 all_bt: print all CPUs backtrace (if available in the arch) 4904 blocked_tasks: print only tasks in uninterruptible (blocked) state 4905 4906 This is a human readable alternative to the 'panic_print' option. 4907 4908 panic_console_replay 4909 When panic happens, replay all kernel messages on 4910 consoles at the end of panic. 4911 4912 parkbd.port= [HW] Parallel port number the keyboard adapter is 4913 connected to, default is 0. 4914 Format: <parport#> 4915 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 4916 0 for XT, 1 for AT (default is AT). 4917 Format: <mode> 4918 4919 parport= [HW,PPT] Specify parallel ports. 0 disables. 4920 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 4921 Use 'auto' to force the driver to use any 4922 IRQ/DMA settings detected (the default is to 4923 ignore detected IRQ/DMA settings because of 4924 possible conflicts). You can specify the base 4925 address, IRQ, and DMA settings; IRQ and DMA 4926 should be numbers, or 'auto' (for using detected 4927 settings on that particular port), or 'nofifo' 4928 (to avoid using a FIFO even if it is detected). 4929 Parallel ports are assigned in the order they 4930 are specified on the command line, starting 4931 with parport0. 4932 4933 parport_init_mode= [HW,PPT] 4934 Configure VIA parallel port to operate in 4935 a specific mode. This is necessary on Pegasos 4936 computer where firmware has no options for setting 4937 up parallel port mode and sets it to spp. 4938 Currently this function knows 686a and 8231 chips. 4939 Format: [spp|ps2|epp|ecp|ecpepp] 4940 4941 pata_legacy.all= [HW,LIBATA] 4942 Format: <int> 4943 Set to non-zero to probe primary and secondary ISA 4944 port ranges on PCI systems where no PCI PATA device 4945 has been found at either range. Disabled by default. 4946 4947 pata_legacy.autospeed= [HW,LIBATA] 4948 Format: <int> 4949 Set to non-zero if a chip is present that snoops speed 4950 changes. Disabled by default. 4951 4952 pata_legacy.ht6560a= [HW,LIBATA] 4953 Format: <int> 4954 Set to 1, 2, or 3 for HT 6560A on the primary channel, 4955 the secondary channel, or both channels respectively. 4956 Disabled by default. 4957 4958 pata_legacy.ht6560b= [HW,LIBATA] 4959 Format: <int> 4960 Set to 1, 2, or 3 for HT 6560B on the primary channel, 4961 the secondary channel, or both channels respectively. 4962 Disabled by default. 4963 4964 pata_legacy.iordy_mask= [HW,LIBATA] 4965 Format: <int> 4966 IORDY enable mask. Set individual bits to allow IORDY 4967 for the respective channel. Bit 0 is for the first 4968 legacy channel handled by this driver, bit 1 is for 4969 the second channel, and so on. The sequence will often 4970 correspond to the primary legacy channel, the secondary 4971 legacy channel, and so on, but the handling of a PCI 4972 bus and the use of other driver options may interfere 4973 with the sequence. By default IORDY is allowed across 4974 all channels. 4975 4976 pata_legacy.opti82c46x= [HW,LIBATA] 4977 Format: <int> 4978 Set to 1, 2, or 3 for Opti 82c611A on the primary 4979 channel, the secondary channel, or both channels 4980 respectively. Disabled by default. 4981 4982 pata_legacy.opti82c611a= [HW,LIBATA] 4983 Format: <int> 4984 Set to 1, 2, or 3 for Opti 82c465MV on the primary 4985 channel, the secondary channel, or both channels 4986 respectively. Disabled by default. 4987 4988 pata_legacy.pio_mask= [HW,LIBATA] 4989 Format: <int> 4990 PIO mode mask for autospeed devices. Set individual 4991 bits to allow the use of the respective PIO modes. 4992 Bit 0 is for mode 0, bit 1 is for mode 1, and so on. 4993 All modes allowed by default. 4994 4995 pata_legacy.probe_all= [HW,LIBATA] 4996 Format: <int> 4997 Set to non-zero to probe tertiary and further ISA 4998 port ranges on PCI systems. Disabled by default. 4999 5000 pata_legacy.probe_mask= [HW,LIBATA] 5001 Format: <int> 5002 Probe mask for legacy ISA PATA ports. Depending on 5003 platform configuration and the use of other driver 5004 options up to 6 legacy ports are supported: 0x1f0, 5005 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing 5006 of individual ports can be disabled by setting the 5007 corresponding bits in the mask to 1. Bit 0 is for 5008 the first port in the list above (0x1f0), and so on. 5009 By default all supported ports are probed. 5010 5011 pata_legacy.qdi= [HW,LIBATA] 5012 Format: <int> 5013 Set to non-zero to probe QDI controllers. By default 5014 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise. 5015 5016 pata_legacy.winbond= [HW,LIBATA] 5017 Format: <int> 5018 Set to non-zero to probe Winbond controllers. Use 5019 the standard I/O port (0x130) if 1, otherwise the 5020 value given is the I/O port to use (typically 0x1b0). 5021 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE, 5022 0 otherwise. 5023 5024 pata_platform.pio_mask= [HW,LIBATA] 5025 Format: <int> 5026 Supported PIO mode mask. Set individual bits to allow 5027 the use of the respective PIO modes. Bit 0 is for 5028 mode 0, bit 1 is for mode 1, and so on. Mode 0 only 5029 allowed by default. 5030 5031 pause_on_oops=<int> 5032 Halt all CPUs after the first oops has been printed for 5033 the specified number of seconds. This is to be used if 5034 your oopses keep scrolling off the screen. 5035 5036 pci=option[,option...] [PCI,EARLY] various PCI subsystem options. 5037 5038 Some options herein operate on a specific device 5039 or a set of devices (<pci_dev>). These are 5040 specified in one of the following formats: 5041 5042 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 5043 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 5044 5045 Note: the first format specifies a PCI 5046 bus/device/function address which may change 5047 if new hardware is inserted, if motherboard 5048 firmware changes, or due to changes caused 5049 by other kernel parameters. If the 5050 domain is left unspecified, it is 5051 taken to be zero. Optionally, a path 5052 to a device through multiple device/function 5053 addresses can be specified after the base 5054 address (this is more robust against 5055 renumbering issues). The second format 5056 selects devices using IDs from the 5057 configuration space which may match multiple 5058 devices in the system. 5059 5060 earlydump dump PCI config space before the kernel 5061 changes anything 5062 off [X86] don't probe for the PCI bus 5063 bios [X86-32] force use of PCI BIOS, don't access 5064 the hardware directly. Use this if your machine 5065 has a non-standard PCI host bridge. 5066 nobios [X86-32] disallow use of PCI BIOS, only direct 5067 hardware access methods are allowed. Use this 5068 if you experience crashes upon bootup and you 5069 suspect they are caused by the BIOS. 5070 conf1 [X86] Force use of PCI Configuration Access 5071 Mechanism 1 (config address in IO port 0xCF8, 5072 data in IO port 0xCFC, both 32-bit). 5073 conf2 [X86] Force use of PCI Configuration Access 5074 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 5075 the function, IO port 0xCFA, also 8-bit, sets 5076 bus number. The config space is then accessed 5077 through ports 0xC000-0xCFFF). 5078 See http://wiki.osdev.org/PCI for more info 5079 on the configuration access mechanisms. 5080 noaer [PCIE] If the PCIEAER kernel config parameter is 5081 enabled, this kernel boot option can be used to 5082 disable the use of PCIE advanced error reporting. 5083 nodomains [PCI] Disable support for multiple PCI 5084 root domains (aka PCI segments, in ACPI-speak). 5085 nommconf [X86] Disable use of MMCONFIG for PCI 5086 Configuration 5087 check_enable_amd_mmconf [X86] check for and enable 5088 properly configured MMIO access to PCI 5089 config space on AMD family 10h CPU 5090 nomsi [MSI] If the PCI_MSI kernel config parameter is 5091 enabled, this kernel boot option can be used to 5092 disable the use of MSI interrupts system-wide. 5093 noioapicquirk [APIC] Disable all boot interrupt quirks. 5094 Safety option to keep boot IRQs enabled. This 5095 should never be necessary. 5096 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 5097 primary IO-APIC for bridges that cannot disable 5098 boot IRQs. This fixes a source of spurious IRQs 5099 when the system masks IRQs. 5100 noioapicreroute [APIC] Disable workaround that uses the 5101 boot IRQ equivalent of an IRQ that connects to 5102 a chipset where boot IRQs cannot be disabled. 5103 The opposite of ioapicreroute. 5104 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 5105 routing table. These calls are known to be buggy 5106 on several machines and they hang the machine 5107 when used, but on other computers it's the only 5108 way to get the interrupt routing table. Try 5109 this option if the kernel is unable to allocate 5110 IRQs or discover secondary PCI buses on your 5111 motherboard. 5112 rom [X86] Assign address space to expansion ROMs. 5113 Use with caution as certain devices share 5114 address decoders between ROMs and other 5115 resources. 5116 norom [X86] Do not assign address space to 5117 expansion ROMs that do not already have 5118 BIOS assigned address ranges. 5119 nobar [X86] Do not assign address space to the 5120 BARs that weren't assigned by the BIOS. 5121 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 5122 assigned automatically to PCI devices. You can 5123 make the kernel exclude IRQs of your ISA cards 5124 this way. 5125 pirqaddr=0xAAAAA [X86] Specify the physical address 5126 of the PIRQ table (normally generated 5127 by the BIOS) if it is outside the 5128 F0000h-100000h range. 5129 lastbus=N [X86] Scan all buses thru bus #N. Can be 5130 useful if the kernel is unable to find your 5131 secondary buses and you want to tell it 5132 explicitly which ones they are. 5133 assign-busses [X86] Always assign all PCI bus 5134 numbers ourselves, overriding 5135 whatever the firmware may have done. 5136 usepirqmask [X86] Honor the possible IRQ mask stored 5137 in the BIOS $PIR table. This is needed on 5138 some systems with broken BIOSes, notably 5139 some HP Pavilion N5400 and Omnibook XE3 5140 notebooks. This will have no effect if ACPI 5141 IRQ routing is enabled. 5142 noacpi [X86] Do not use ACPI for IRQ routing 5143 or for PCI scanning. 5144 use_crs [X86] Use PCI host bridge window information 5145 from ACPI. On BIOSes from 2008 or later, this 5146 is enabled by default. If you need to use this, 5147 please report a bug. 5148 nocrs [X86] Ignore PCI host bridge windows from ACPI. 5149 If you need to use this, please report a bug. 5150 use_e820 [X86] Use E820 reservations to exclude parts of 5151 PCI host bridge windows. This is a workaround 5152 for BIOS defects in host bridge _CRS methods. 5153 If you need to use this, please report a bug to 5154 <linux-pci@vger.kernel.org>. 5155 no_e820 [X86] Ignore E820 reservations for PCI host 5156 bridge windows. This is the default on modern 5157 hardware. If you need to use this, please report 5158 a bug to <linux-pci@vger.kernel.org>. 5159 routeirq Do IRQ routing for all PCI devices. 5160 This is normally done in pci_enable_device(), 5161 so this option is a temporary workaround 5162 for broken drivers that don't call it. 5163 skip_isa_align [X86] do not align io start addr, so can 5164 handle more pci cards 5165 noearly [X86] Don't do any early type 1 scanning. 5166 This might help on some broken boards which 5167 machine check when some devices' config space 5168 is read. But various workarounds are disabled 5169 and some IOMMU drivers will not work. 5170 bfsort Sort PCI devices into breadth-first order. 5171 This sorting is done to get a device 5172 order compatible with older (<= 2.4) kernels. 5173 nobfsort Don't sort PCI devices into breadth-first order. 5174 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 5175 tuning and use the BIOS-configured MPS defaults. 5176 pcie_bus_safe Set every device's MPS to the largest value 5177 supported by all devices below the root complex. 5178 pcie_bus_perf Set device MPS to the largest allowable MPS 5179 based on its parent bus. Also set MRRS (Max 5180 Read Request Size) to the largest supported 5181 value (no larger than the MPS that the device 5182 or bus can support) for best performance. 5183 pcie_bus_peer2peer Set every device's MPS to 128B, which 5184 every device is guaranteed to support. This 5185 configuration allows peer-to-peer DMA between 5186 any pair of devices, possibly at the cost of 5187 reduced performance. This also guarantees 5188 that hot-added devices will work. 5189 cbiosize=nn[KMG] The fixed amount of bus space which is 5190 reserved for the CardBus bridge's IO window. 5191 The default value is 256 bytes. 5192 cbmemsize=nn[KMG] The fixed amount of bus space which is 5193 reserved for the CardBus bridge's memory 5194 window. The default value is 64 megabytes. 5195 resource_alignment= 5196 Format: 5197 [<order of align>@]<pci_dev>[; ...] 5198 Specifies alignment and device to reassign 5199 aligned memory resources. How to 5200 specify the device is described above. 5201 If <order of align> is not specified, 5202 PAGE_SIZE is used as alignment. 5203 A PCI-PCI bridge can be specified if resource 5204 windows need to be expanded. 5205 To specify the alignment for several 5206 instances of a device, the PCI vendor, 5207 device, subvendor, and subdevice may be 5208 specified, e.g., 12@pci:8086:9c22:103c:198f 5209 for 4096-byte alignment. 5210 ecrc= Enable/disable PCIe ECRC (transaction layer 5211 end-to-end CRC checking). Only effective if 5212 OS has native AER control (either granted by 5213 ACPI _OSC or forced via "pcie_ports=native") 5214 bios: Use BIOS/firmware settings. This is the 5215 the default. 5216 off: Turn ECRC off 5217 on: Turn ECRC on. 5218 hpiosize=nn[KMG] The fixed amount of bus space which is 5219 reserved for hotplug bridge's IO window. 5220 Default size is 256 bytes. 5221 hpmmiosize=nn[KMG] The fixed amount of bus space which is 5222 reserved for hotplug bridge's MMIO window. 5223 Default size is 2 megabytes. 5224 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 5225 reserved for hotplug bridge's MMIO_PREF window. 5226 Default size is 2 megabytes. 5227 hpmemsize=nn[KMG] The fixed amount of bus space which is 5228 reserved for hotplug bridge's MMIO and 5229 MMIO_PREF window. 5230 Default size is 2 megabytes. 5231 hpbussize=nn The minimum amount of additional bus numbers 5232 reserved for buses below a hotplug bridge. 5233 Default is 1. 5234 realloc= Enable/disable reallocating PCI bridge resources 5235 if allocations done by BIOS are too small to 5236 accommodate resources required by all child 5237 devices. 5238 off: Turn realloc off 5239 on: Turn realloc on 5240 realloc same as realloc=on 5241 noari do not use PCIe ARI. 5242 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 5243 do not use PCIe ATS (and IOMMU device IOTLB). 5244 pcie_scan_all Scan all possible PCIe devices. Otherwise we 5245 only look for one device below a PCIe downstream 5246 port. 5247 big_root_window Try to add a big 64bit memory window to the PCIe 5248 root complex on AMD CPUs. Some GFX hardware 5249 can resize a BAR to allow access to all VRAM. 5250 Adding the window is slightly risky (it may 5251 conflict with unreported devices), so this 5252 taints the kernel. 5253 disable_acs_redir=<pci_dev>[; ...] 5254 Specify one or more PCI devices (in the format 5255 specified above) separated by semicolons. 5256 Each device specified will have the PCI ACS 5257 redirect capabilities forced off which will 5258 allow P2P traffic between devices through 5259 bridges without forcing it upstream. Note: 5260 this removes isolation between devices and 5261 may put more devices in an IOMMU group. 5262 config_acs= 5263 Format: 5264 <ACS flags>@<pci_dev>[; ...] 5265 Specify one or more PCI devices (in the format 5266 specified above) optionally prepended with flags 5267 and separated by semicolons. The respective 5268 capabilities will be enabled, disabled or 5269 unchanged based on what is specified in 5270 flags. 5271 5272 ACS Flags is defined as follows: 5273 bit-0 : ACS Source Validation 5274 bit-1 : ACS Translation Blocking 5275 bit-2 : ACS P2P Request Redirect 5276 bit-3 : ACS P2P Completion Redirect 5277 bit-4 : ACS Upstream Forwarding 5278 bit-5 : ACS P2P Egress Control 5279 bit-6 : ACS Direct Translated P2P 5280 Each bit can be marked as: 5281 '0' – force disabled 5282 '1' – force enabled 5283 'x' – unchanged 5284 For example, 5285 pci=config_acs=10x@pci:0:0 5286 would configure all devices that support 5287 ACS to enable P2P Request Redirect, disable 5288 Translation Blocking, and leave Source 5289 Validation unchanged from whatever power-up 5290 or firmware set it to. 5291 5292 Note: this may remove isolation between devices 5293 and may put more devices in an IOMMU group. 5294 force_floating [S390] Force usage of floating interrupts. 5295 nomio [S390] Do not use MIO instructions. 5296 norid [S390] ignore the RID field and force use of 5297 one PCI domain per PCI function 5298 notph [PCIE] If the PCIE_TPH kernel config parameter 5299 is enabled, this kernel boot option can be used 5300 to disable PCIe TLP Processing Hints support 5301 system-wide. 5302 5303 pcie_aspm= [PCIE] Forcibly enable or ignore PCIe Active State Power 5304 Management. 5305 off Don't touch ASPM configuration at all. Leave any 5306 configuration done by firmware unchanged. 5307 force Enable ASPM even on devices that claim not to support it. 5308 WARNING: Forcing ASPM on may cause system lockups. 5309 5310 pcie_ports= [PCIE] PCIe port services handling: 5311 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 5312 even if the platform doesn't give the OS permission to 5313 use them. This may cause conflicts if the platform 5314 also tries to use these services. 5315 dpc-native Use native PCIe service for DPC only. May 5316 cause conflicts if firmware uses AER or DPC. 5317 compat Disable native PCIe services (PME, AER, DPC, PCIe 5318 hotplug). 5319 5320 pcie_port_pm= [PCIE] PCIe port power management handling: 5321 off Disable power management of all PCIe ports 5322 force Forcibly enable power management of all PCIe ports 5323 5324 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 5325 nomsi Do not use MSI for native PCIe PME signaling (this makes 5326 all PCIe root ports use INTx for all services). 5327 5328 pcmv= [HW,PCMCIA] BadgePAD 4 5329 5330 pd_ignore_unused 5331 [PM] 5332 Keep all power-domains already enabled by bootloader on, 5333 even if no driver has claimed them. This is useful 5334 for debug and development, but should not be 5335 needed on a platform with proper driver support. 5336 5337 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 5338 boot time. 5339 Format: { 0 | 1 } 5340 See arch/parisc/kernel/pdc_chassis.c 5341 5342 percpu_alloc= [MM,EARLY] 5343 Select which percpu first chunk allocator to use. 5344 Currently supported values are "embed" and "page". 5345 Archs may support subset or none of the selections. 5346 See comments in mm/percpu.c for details on each 5347 allocator. This parameter is primarily for debugging 5348 and performance comparison. 5349 5350 pirq= [SMP,APIC] Manual mp-table setup 5351 See Documentation/arch/x86/i386/IO-APIC.rst. 5352 5353 plip= [PPT,NET] Parallel port network link 5354 Format: { parport<nr> | timid | 0 } 5355 See also Documentation/admin-guide/parport.rst. 5356 5357 pmtmr= [X86] Manual setup of pmtmr I/O Port. 5358 Override pmtimer IOPort with a hex value. 5359 e.g. pmtmr=0x508 5360 5361 pmu_override= [PPC] Override the PMU. 5362 This option takes over the PMU facility, so it is no 5363 longer usable by perf. Setting this option starts the 5364 PMU counters by setting MMCR0 to 0 (the FC bit is 5365 cleared). If a number is given, then MMCR1 is set to 5366 that number, otherwise (e.g., 'pmu_override=on'), MMCR1 5367 remains 0. 5368 5369 pm_async= [PM] 5370 Format: off 5371 This parameter sets the initial value of the 5372 /sys/power/pm_async sysfs knob at boot time. 5373 If set to "off", disables asynchronous suspend and 5374 resume of devices during system-wide power transitions. 5375 This can be useful on platforms where device 5376 dependencies are not well-defined, or for debugging 5377 power management issues. Asynchronous operations are 5378 enabled by default. 5379 5380 5381 pm_debug_messages [SUSPEND,KNL] 5382 Enable suspend/resume debug messages during boot up. 5383 5384 pm.dpm_watchdog_enabled= 5385 [PM] Enable or disable the DPM watchdog. Requires 5386 CONFIG_PM_SLEEP and CONFIG_DPM_WATCHDOG enabled. 5387 Format: <bool> 5388 Default value is set by CONFIG_DPM_WATCHDOG_ENABLED. 5389 5390 pnp.debug=1 [PNP] 5391 Enable PNP debug messages (depends on the 5392 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 5393 via /sys/module/pnp/parameters/debug. We always show 5394 current resource usage; turning this on also shows 5395 possible settings and some assignment information. 5396 5397 pnpacpi= [ACPI] 5398 { off } 5399 5400 pnpbios= [ISAPNP] 5401 { on | off | curr | res | no-curr | no-res } 5402 5403 pnp_reserve_irq= 5404 [ISAPNP] Exclude IRQs for the autoconfiguration 5405 5406 pnp_reserve_dma= 5407 [ISAPNP] Exclude DMAs for the autoconfiguration 5408 5409 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 5410 Ranges are in pairs (I/O port base and size). 5411 5412 pnp_reserve_mem= 5413 [ISAPNP] Exclude memory regions for the 5414 autoconfiguration. 5415 Ranges are in pairs (memory base and size). 5416 5417 ports= [IP_VS_FTP] IPVS ftp helper module 5418 Default is 21. 5419 Up to 8 (IP_VS_APP_MAX_PORTS) ports 5420 may be specified. 5421 Format: <port>,<port>.... 5422 5423 possible_cpus= [SMP,S390,X86] 5424 Format: <unsigned int> 5425 Set the number of possible CPUs, overriding the 5426 regular discovery mechanisms (such as ACPI/FW, etc). 5427 5428 powersave=off [PPC] This option disables power saving features. 5429 It specifically disables cpuidle and sets the 5430 platform machine description specific power_save 5431 function to NULL. On Idle the CPU just reduces 5432 execution priority. 5433 5434 ppc_strict_facility_enable 5435 [PPC,ENABLE] This option catches any kernel floating point, 5436 Altivec, VSX and SPE outside of regions specifically 5437 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 5438 There is some performance impact when enabling this. 5439 5440 ppc_tm= [PPC,EARLY] 5441 Format: {"off"} 5442 Disable Hardware Transactional Memory 5443 5444 preempt= [KNL] 5445 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC 5446 none - Limited to cond_resched() calls 5447 voluntary - Limited to cond_resched() and might_sleep() calls 5448 full - Any section that isn't explicitly preempt disabled 5449 can be preempted anytime. Tasks will also yield 5450 contended spinlocks (if the critical section isn't 5451 explicitly preempt disabled beyond the lock itself). 5452 lazy - Scheduler controlled. Similar to full but instead 5453 of preempting the task immediately, the task gets 5454 one HZ tick time to yield itself before the 5455 preemption will be forced. One preemption is when the 5456 task returns to user space. 5457 5458 print-fatal-signals= 5459 [KNL] debug: print fatal signals 5460 5461 If enabled, warn about various signal handling 5462 related application anomalies: too many signals, 5463 too many POSIX.1 timers, fatal signals causing a 5464 coredump - etc. 5465 5466 If you hit the warning due to signal overflow, 5467 you might want to try "ulimit -i unlimited". 5468 5469 default: off. 5470 5471 printk.always_kmsg_dump= 5472 Trigger kmsg_dump for cases other than kernel oops or 5473 panics 5474 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5475 default: disabled 5476 5477 printk.console_no_auto_verbose= 5478 Disable console loglevel raise on oops, panic 5479 or lockdep-detected issues (only if lock debug is on). 5480 With an exception to setups with low baudrate on 5481 serial console, keeping this 0 is a good choice 5482 in order to provide more debug information. 5483 Format: <bool> 5484 default: 0 (auto_verbose is enabled) 5485 5486 printk.debug_non_panic_cpus= 5487 Allows storing messages from non-panic CPUs into 5488 the printk log buffer during panic(). They are 5489 flushed to consoles by the panic-CPU on 5490 a best-effort basis. 5491 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5492 Default: disabled 5493 5494 printk.devkmsg={on,off,ratelimit} 5495 Control writing to /dev/kmsg. 5496 on - unlimited logging to /dev/kmsg from userspace 5497 off - logging to /dev/kmsg disabled 5498 ratelimit - ratelimit the logging 5499 Default: ratelimit 5500 5501 printk.time= Show timing data prefixed to each printk message line 5502 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5503 5504 proc_mem.force_override= [KNL] 5505 Format: {always | ptrace | never} 5506 Traditionally /proc/pid/mem allows memory permissions to be 5507 overridden without restrictions. This option may be set to 5508 restrict that. Can be one of: 5509 - 'always': traditional behavior always allows mem overrides. 5510 - 'ptrace': only allow mem overrides for active ptracers. 5511 - 'never': never allow mem overrides. 5512 If not specified, default is the CONFIG_PROC_MEM_* choice. 5513 5514 processor.max_cstate= [HW,ACPI] 5515 Limit processor to maximum C-state 5516 max_cstate=9 overrides any DMI blacklist limit. 5517 5518 processor.nocst [HW,ACPI] 5519 Ignore the _CST method to determine C-states, 5520 instead using the legacy FADT method 5521 5522 profile= [KNL] Enable kernel profiling via /proc/profile 5523 Format: [<profiletype>,]<number> 5524 Param: <profiletype>: "schedule" or "kvm" 5525 [defaults to kernel profiling] 5526 Param: "schedule" - profile schedule points. 5527 Param: "kvm" - profile VM exits. 5528 Param: <number> - step/bucket size as a power of 2 for 5529 statistical time based profiling. 5530 5531 prot_virt= [S390] enable hosting protected virtual machines 5532 isolated from the hypervisor (if hardware supports 5533 that). If enabled, the default kernel base address 5534 might be overridden even when Kernel Address Space 5535 Layout Randomization is disabled. 5536 Format: <bool> 5537 5538 psi= [KNL] Enable or disable pressure stall information 5539 tracking. 5540 Format: <bool> 5541 5542 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 5543 probe for; one of (bare|imps|exps|lifebook|any). 5544 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 5545 per second. 5546 psmouse.resetafter= [HW,MOUSE] 5547 Try to reset the device after so many bad packets 5548 (0 = never). 5549 psmouse.resolution= 5550 [HW,MOUSE] Set desired mouse resolution, in dpi. 5551 psmouse.smartscroll= 5552 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 5553 0 = disabled, 1 = enabled (default). 5554 5555 pstore.backend= Specify the name of the pstore backend to use 5556 5557 pti= [X86-64] Control Page Table Isolation of user and 5558 kernel address spaces. Disabling this feature 5559 removes hardening, but improves performance of 5560 system calls and interrupts. 5561 5562 on - unconditionally enable 5563 off - unconditionally disable 5564 auto - kernel detects whether your CPU model is 5565 vulnerable to issues that PTI mitigates 5566 5567 Not specifying this option is equivalent to pti=auto. 5568 5569 pty.legacy_count= 5570 [KNL] Number of legacy pty's. Overwrites compiled-in 5571 default number. 5572 5573 quiet [KNL,EARLY] Disable most log messages 5574 5575 r128= [HW,DRM] 5576 5577 radix_hcall_invalidate=on [PPC/PSERIES] 5578 Disable RADIX GTSE feature and use hcall for TLB 5579 invalidate. 5580 5581 raid= [HW,RAID] 5582 See Documentation/admin-guide/md.rst. 5583 5584 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 5585 See Documentation/admin-guide/blockdev/ramdisk.rst. 5586 5587 ramdisk_start= [Deprecated,RAM] RAM disk image start address 5588 5589 random.trust_cpu=off 5590 [KNL,EARLY] Disable trusting the use of the CPU's 5591 random number generator (if available) to 5592 initialize the kernel's RNG. 5593 5594 random.trust_bootloader=off 5595 [KNL,EARLY] Disable trusting the use of the a seed 5596 passed by the bootloader (if available) to 5597 initialize the kernel's RNG. 5598 5599 randomize_kstack_offset= 5600 [KNL,EARLY] Enable or disable kernel stack offset 5601 randomization, which provides roughly 5 bits of 5602 entropy, frustrating memory corruption attacks 5603 that depend on stack address determinism or 5604 cross-syscall address exposures. This is only 5605 available on architectures that have defined 5606 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET. 5607 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5608 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT. 5609 5610 ras=option[,option,...] [KNL] RAS-specific options 5611 5612 cec_disable [X86] 5613 Disable the Correctable Errors Collector, 5614 see CONFIG_RAS_CEC help text. 5615 5616 rcu_nocbs[=cpu-list] 5617 [KNL] The optional argument is a cpu list, 5618 as described above. 5619 5620 In kernels built with CONFIG_RCU_NOCB_CPU=y, 5621 enable the no-callback CPU mode, which prevents 5622 such CPUs' callbacks from being invoked in 5623 softirq context. Invocation of such CPUs' RCU 5624 callbacks will instead be offloaded to "rcuox/N" 5625 kthreads created for that purpose, where "x" is 5626 "p" for RCU-preempt, "s" for RCU-sched, and "g" 5627 for the kthreads that mediate grace periods; and 5628 "N" is the CPU number. This reduces OS jitter on 5629 the offloaded CPUs, which can be useful for HPC 5630 and real-time workloads. It can also improve 5631 energy efficiency for asymmetric multiprocessors. 5632 5633 If a cpulist is passed as an argument, the specified 5634 list of CPUs is set to no-callback mode from boot. 5635 5636 Otherwise, if the '=' sign and the cpulist 5637 arguments are omitted, no CPU will be set to 5638 no-callback mode from boot but the mode may be 5639 toggled at runtime via cpusets. 5640 5641 Note that this argument takes precedence over 5642 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 5643 5644 rcu_nocb_poll [KNL] 5645 Rather than requiring that offloaded CPUs 5646 (specified by rcu_nocbs= above) explicitly 5647 awaken the corresponding "rcuoN" kthreads, 5648 make these kthreads poll for callbacks. 5649 This improves the real-time response for the 5650 offloaded CPUs by relieving them of the need to 5651 wake up the corresponding kthread, but degrades 5652 energy efficiency by requiring that the kthreads 5653 periodically wake up to do the polling. 5654 5655 rcutree.blimit= [KNL] 5656 Set maximum number of finished RCU callbacks to 5657 process in one batch. 5658 5659 rcutree.csd_lock_suppress_rcu_stall= [KNL] 5660 Do only a one-line RCU CPU stall warning when 5661 there is an ongoing too-long CSD-lock wait. 5662 5663 rcutree.do_rcu_barrier= [KNL] 5664 Request a call to rcu_barrier(). This is 5665 throttled so that userspace tests can safely 5666 hammer on the sysfs variable if they so choose. 5667 If triggered before the RCU grace-period machinery 5668 is fully active, this will error out with EAGAIN. 5669 5670 rcutree.dump_tree= [KNL] 5671 Dump the structure of the rcu_node combining tree 5672 out at early boot. This is used for diagnostic 5673 purposes, to verify correct tree setup. 5674 5675 rcutree.gp_cleanup_delay= [KNL] 5676 Set the number of jiffies to delay each step of 5677 RCU grace-period cleanup. 5678 5679 rcutree.gp_init_delay= [KNL] 5680 Set the number of jiffies to delay each step of 5681 RCU grace-period initialization. 5682 5683 rcutree.gp_preinit_delay= [KNL] 5684 Set the number of jiffies to delay each step of 5685 RCU grace-period pre-initialization, that is, 5686 the propagation of recent CPU-hotplug changes up 5687 the rcu_node combining tree. 5688 5689 rcutree.jiffies_till_first_fqs= [KNL] 5690 Set delay from grace-period initialization to 5691 first attempt to force quiescent states. 5692 Units are jiffies, minimum value is zero, 5693 and maximum value is HZ. 5694 5695 rcutree.jiffies_till_next_fqs= [KNL] 5696 Set delay between subsequent attempts to force 5697 quiescent states. Units are jiffies, minimum 5698 value is one, and maximum value is HZ. 5699 5700 rcutree.jiffies_till_sched_qs= [KNL] 5701 Set required age in jiffies for a 5702 given grace period before RCU starts 5703 soliciting quiescent-state help from 5704 rcu_note_context_switch() and cond_resched(). 5705 If not specified, the kernel will calculate 5706 a value based on the most recent settings 5707 of rcutree.jiffies_till_first_fqs 5708 and rcutree.jiffies_till_next_fqs. 5709 This calculated value may be viewed in 5710 rcutree.jiffies_to_sched_qs. Any attempt to set 5711 rcutree.jiffies_to_sched_qs will be cheerfully 5712 overwritten. 5713 5714 rcutree.kthread_prio= [KNL,BOOT] 5715 Set the SCHED_FIFO priority of the RCU per-CPU 5716 kthreads (rcuc/N). This value is also used for 5717 the priority of the RCU boost threads (rcub/N) 5718 and for the RCU grace-period kthreads (rcu_bh, 5719 rcu_preempt, and rcu_sched). If RCU_BOOST is 5720 set, valid values are 1-99 and the default is 1 5721 (the least-favored priority). Otherwise, when 5722 RCU_BOOST is not set, valid values are 0-99 and 5723 the default is zero (non-realtime operation). 5724 When RCU_NOCB_CPU is set, also adjust the 5725 priority of NOCB callback kthreads. 5726 5727 rcutree.nocb_nobypass_lim_per_jiffy= [KNL] 5728 On callback-offloaded (rcu_nocbs) CPUs, 5729 RCU reduces the lock contention that would 5730 otherwise be caused by callback floods through 5731 use of the ->nocb_bypass list. However, in the 5732 common non-flooded case, RCU queues directly to 5733 the main ->cblist in order to avoid the extra 5734 overhead of the ->nocb_bypass list and its lock. 5735 But if there are too many callbacks queued during 5736 a single jiffy, RCU pre-queues the callbacks into 5737 the ->nocb_bypass queue. The definition of "too 5738 many" is supplied by this kernel boot parameter. 5739 5740 rcutree.nohz_full_patience_delay= [KNL] 5741 On callback-offloaded (rcu_nocbs) CPUs, avoid 5742 disturbing RCU unless the grace period has 5743 reached the specified age in milliseconds. 5744 Defaults to zero. Large values will be capped 5745 at five seconds. All values will be rounded down 5746 to the nearest value representable by jiffies. 5747 5748 rcutree.qhimark= [KNL] 5749 Set threshold of queued RCU callbacks beyond which 5750 batch limiting is disabled. 5751 5752 rcutree.qlowmark= [KNL] 5753 Set threshold of queued RCU callbacks below which 5754 batch limiting is re-enabled. 5755 5756 rcutree.qovld= [KNL] 5757 Set threshold of queued RCU callbacks beyond which 5758 RCU's force-quiescent-state scan will aggressively 5759 enlist help from cond_resched() and sched IPIs to 5760 help CPUs more quickly reach quiescent states. 5761 Set to less than zero to make this be set based 5762 on rcutree.qhimark at boot time and to zero to 5763 disable more aggressive help enlistment. 5764 5765 rcutree.rcu_delay_page_cache_fill_msec= [KNL] 5766 Set the page-cache refill delay (in milliseconds) 5767 in response to low-memory conditions. The range 5768 of permitted values is in the range 0:100000. 5769 5770 rcutree.rcu_divisor= [KNL] 5771 Set the shift-right count to use to compute 5772 the callback-invocation batch limit bl from 5773 the number of callbacks queued on this CPU. 5774 The result will be bounded below by the value of 5775 the rcutree.blimit kernel parameter. Every bl 5776 callbacks, the softirq handler will exit in 5777 order to allow the CPU to do other work. 5778 5779 Please note that this callback-invocation batch 5780 limit applies only to non-offloaded callback 5781 invocation. Offloaded callbacks are instead 5782 invoked in the context of an rcuoc kthread, which 5783 scheduler will preempt as it does any other task. 5784 5785 rcutree.rcu_fanout_exact= [KNL] 5786 Disable autobalancing of the rcu_node combining 5787 tree. This is used by rcutorture, and might 5788 possibly be useful for architectures having high 5789 cache-to-cache transfer latencies. 5790 5791 rcutree.rcu_fanout_leaf= [KNL] 5792 Change the number of CPUs assigned to each 5793 leaf rcu_node structure. Useful for very 5794 large systems, which will choose the value 64, 5795 and for NUMA systems with large remote-access 5796 latencies, which will choose a value aligned 5797 with the appropriate hardware boundaries. 5798 5799 rcutree.rcu_min_cached_objs= [KNL] 5800 Minimum number of objects which are cached and 5801 maintained per one CPU. Object size is equal 5802 to PAGE_SIZE. The cache allows to reduce the 5803 pressure to page allocator, also it makes the 5804 whole algorithm to behave better in low memory 5805 condition. 5806 5807 rcutree.rcu_nocb_gp_stride= [KNL] 5808 Set the number of NOCB callback kthreads in 5809 each group, which defaults to the square root 5810 of the number of CPUs. Larger numbers reduce 5811 the wakeup overhead on the global grace-period 5812 kthread, but increases that same overhead on 5813 each group's NOCB grace-period kthread. 5814 5815 rcutree.rcu_kick_kthreads= [KNL] 5816 Cause the grace-period kthread to get an extra 5817 wake_up() if it sleeps three times longer than 5818 it should at force-quiescent-state time. 5819 This wake_up() will be accompanied by a 5820 WARN_ONCE() splat and an ftrace_dump(). 5821 5822 rcutree.rcu_resched_ns= [KNL] 5823 Limit the time spend invoking a batch of RCU 5824 callbacks to the specified number of nanoseconds. 5825 By default, this limit is checked only once 5826 every 32 callbacks in order to limit the pain 5827 inflicted by local_clock() overhead. 5828 5829 rcutree.rcu_unlock_delay= [KNL] 5830 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 5831 this specifies an rcu_read_unlock()-time delay 5832 in microseconds. This defaults to zero. 5833 Larger delays increase the probability of 5834 catching RCU pointer leaks, that is, buggy use 5835 of RCU-protected pointers after the relevant 5836 rcu_read_unlock() has completed. 5837 5838 rcutree.sysrq_rcu= [KNL] 5839 Commandeer a sysrq key to dump out Tree RCU's 5840 rcu_node tree with an eye towards determining 5841 why a new grace period has not yet started. 5842 5843 rcutree.use_softirq= [KNL] 5844 If set to zero, move all RCU_SOFTIRQ processing to 5845 per-CPU rcuc kthreads. Defaults to a non-zero 5846 value, meaning that RCU_SOFTIRQ is used by default. 5847 Specify rcutree.use_softirq=0 to use rcuc kthreads. 5848 5849 But note that CONFIG_PREEMPT_RT=y kernels disable 5850 this kernel boot parameter, forcibly setting it 5851 to zero. 5852 5853 rcutree.enable_rcu_lazy= [KNL] 5854 To save power, batch RCU callbacks and flush after 5855 delay, memory pressure or callback list growing too 5856 big. 5857 5858 rcutree.rcu_normal_wake_from_gp= [KNL] 5859 Reduces a latency of synchronize_rcu() call. This approach 5860 maintains its own track of synchronize_rcu() callers, so it 5861 does not interact with regular callbacks because it does not 5862 use a call_rcu[_hurry]() path. Please note, this is for a 5863 normal grace period. 5864 5865 How to disable it: 5866 5867 echo 0 > /sys/module/rcutree/parameters/rcu_normal_wake_from_gp 5868 or pass a boot parameter "rcutree.rcu_normal_wake_from_gp=0" 5869 5870 Default is 1 if it is not explicitly disabled by the boot parameter 5871 passing 0. 5872 5873 rcuscale.gp_async= [KNL] 5874 Measure performance of asynchronous 5875 grace-period primitives such as call_rcu(). 5876 5877 rcuscale.gp_async_max= [KNL] 5878 Specify the maximum number of outstanding 5879 callbacks per writer thread. When a writer 5880 thread exceeds this limit, it invokes the 5881 corresponding flavor of rcu_barrier() to allow 5882 previously posted callbacks to drain. 5883 5884 rcuscale.gp_exp= [KNL] 5885 Measure performance of expedited synchronous 5886 grace-period primitives. 5887 5888 rcuscale.holdoff= [KNL] 5889 Set test-start holdoff period. The purpose of 5890 this parameter is to delay the start of the 5891 test until boot completes in order to avoid 5892 interference. 5893 5894 rcuscale.kfree_by_call_rcu= [KNL] 5895 In kernels built with CONFIG_RCU_LAZY=y, test 5896 call_rcu() instead of kfree_rcu(). 5897 5898 rcuscale.kfree_mult= [KNL] 5899 Instead of allocating an object of size kfree_obj, 5900 allocate one of kfree_mult * sizeof(kfree_obj). 5901 Defaults to 1. 5902 5903 rcuscale.kfree_rcu_test= [KNL] 5904 Set to measure performance of kfree_rcu() flooding. 5905 5906 rcuscale.kfree_rcu_test_double= [KNL] 5907 Test the double-argument variant of kfree_rcu(). 5908 If this parameter has the same value as 5909 rcuscale.kfree_rcu_test_single, both the single- 5910 and double-argument variants are tested. 5911 5912 rcuscale.kfree_rcu_test_single= [KNL] 5913 Test the single-argument variant of kfree_rcu(). 5914 If this parameter has the same value as 5915 rcuscale.kfree_rcu_test_double, both the single- 5916 and double-argument variants are tested. 5917 5918 rcuscale.kfree_nthreads= [KNL] 5919 The number of threads running loops of kfree_rcu(). 5920 5921 rcuscale.kfree_alloc_num= [KNL] 5922 Number of allocations and frees done in an iteration. 5923 5924 rcuscale.kfree_loops= [KNL] 5925 Number of loops doing rcuscale.kfree_alloc_num number 5926 of allocations and frees. 5927 5928 rcuscale.minruntime= [KNL] 5929 Set the minimum test run time in seconds. This 5930 does not affect the data-collection interval, 5931 but instead allows better measurement of things 5932 like CPU consumption. 5933 5934 rcuscale.nreaders= [KNL] 5935 Set number of RCU readers. The value -1 selects 5936 N, where N is the number of CPUs. A value 5937 "n" less than -1 selects N-n+1, where N is again 5938 the number of CPUs. For example, -2 selects N 5939 (the number of CPUs), -3 selects N+1, and so on. 5940 A value of "n" less than or equal to -N selects 5941 a single reader. 5942 5943 rcuscale.nwriters= [KNL] 5944 Set number of RCU writers. The values operate 5945 the same as for rcuscale.nreaders. 5946 N, where N is the number of CPUs 5947 5948 rcuscale.scale_type= [KNL] 5949 Specify the RCU implementation to test. 5950 5951 rcuscale.shutdown= [KNL] 5952 Shut the system down after performance tests 5953 complete. This is useful for hands-off automated 5954 testing. 5955 5956 rcuscale.verbose= [KNL] 5957 Enable additional printk() statements. 5958 5959 rcuscale.writer_holdoff= [KNL] 5960 Write-side holdoff between grace periods, 5961 in microseconds. The default of zero says 5962 no holdoff. 5963 5964 rcuscale.writer_holdoff_jiffies= [KNL] 5965 Additional write-side holdoff between grace 5966 periods, but in jiffies. The default of zero 5967 says no holdoff. 5968 5969 rcutorture.fqs_duration= [KNL] 5970 Set duration of force_quiescent_state bursts 5971 in microseconds. 5972 5973 rcutorture.fqs_holdoff= [KNL] 5974 Set holdoff time within force_quiescent_state bursts 5975 in microseconds. 5976 5977 rcutorture.fqs_stutter= [KNL] 5978 Set wait time between force_quiescent_state bursts 5979 in seconds. 5980 5981 rcutorture.fwd_progress= [KNL] 5982 Specifies the number of kthreads to be used 5983 for RCU grace-period forward-progress testing 5984 for the types of RCU supporting this notion. 5985 Defaults to 1 kthread, values less than zero or 5986 greater than the number of CPUs cause the number 5987 of CPUs to be used. 5988 5989 rcutorture.fwd_progress_div= [KNL] 5990 Specify the fraction of a CPU-stall-warning 5991 period to do tight-loop forward-progress testing. 5992 5993 rcutorture.fwd_progress_holdoff= [KNL] 5994 Number of seconds to wait between successive 5995 forward-progress tests. 5996 5997 rcutorture.fwd_progress_need_resched= [KNL] 5998 Enclose cond_resched() calls within checks for 5999 need_resched() during tight-loop forward-progress 6000 testing. 6001 6002 rcutorture.gp_cond= [KNL] 6003 Use conditional/asynchronous update-side 6004 normal-grace-period primitives, if available. 6005 6006 rcutorture.gp_cond_exp= [KNL] 6007 Use conditional/asynchronous update-side 6008 expedited-grace-period primitives, if available. 6009 6010 rcutorture.gp_cond_full= [KNL] 6011 Use conditional/asynchronous update-side 6012 normal-grace-period primitives that also take 6013 concurrent expedited grace periods into account, 6014 if available. 6015 6016 rcutorture.gp_cond_exp_full= [KNL] 6017 Use conditional/asynchronous update-side 6018 expedited-grace-period primitives that also take 6019 concurrent normal grace periods into account, 6020 if available. 6021 6022 rcutorture.gp_cond_wi= [KNL] 6023 Nominal wait interval for normal conditional 6024 grace periods (specified by rcutorture's 6025 gp_cond and gp_cond_full module parameters), 6026 in microseconds. The actual wait interval will 6027 be randomly selected to nanosecond granularity up 6028 to this wait interval. Defaults to 16 jiffies, 6029 for example, 16,000 microseconds on a system 6030 with HZ=1000. 6031 6032 rcutorture.gp_cond_wi_exp= [KNL] 6033 Nominal wait interval for expedited conditional 6034 grace periods (specified by rcutorture's 6035 gp_cond_exp and gp_cond_exp_full module 6036 parameters), in microseconds. The actual wait 6037 interval will be randomly selected to nanosecond 6038 granularity up to this wait interval. Defaults to 6039 128 microseconds. 6040 6041 rcutorture.gp_exp= [KNL] 6042 Use expedited update-side primitives, if available. 6043 6044 rcutorture.gp_normal= [KNL] 6045 Use normal (non-expedited) asynchronous 6046 update-side primitives, if available. 6047 6048 rcutorture.gp_poll= [KNL] 6049 Use polled update-side normal-grace-period 6050 primitives, if available. 6051 6052 rcutorture.gp_poll_exp= [KNL] 6053 Use polled update-side expedited-grace-period 6054 primitives, if available. 6055 6056 rcutorture.gp_poll_full= [KNL] 6057 Use polled update-side normal-grace-period 6058 primitives that also take concurrent expedited 6059 grace periods into account, if available. 6060 6061 rcutorture.gp_poll_exp_full= [KNL] 6062 Use polled update-side expedited-grace-period 6063 primitives that also take concurrent normal 6064 grace periods into account, if available. 6065 6066 rcutorture.gp_poll_wi= [KNL] 6067 Nominal wait interval for normal conditional 6068 grace periods (specified by rcutorture's 6069 gp_poll and gp_poll_full module parameters), 6070 in microseconds. The actual wait interval will 6071 be randomly selected to nanosecond granularity up 6072 to this wait interval. Defaults to 16 jiffies, 6073 for example, 16,000 microseconds on a system 6074 with HZ=1000. 6075 6076 rcutorture.gp_poll_wi_exp= [KNL] 6077 Nominal wait interval for expedited conditional 6078 grace periods (specified by rcutorture's 6079 gp_poll_exp and gp_poll_exp_full module 6080 parameters), in microseconds. The actual wait 6081 interval will be randomly selected to nanosecond 6082 granularity up to this wait interval. Defaults to 6083 128 microseconds. 6084 6085 rcutorture.gp_sync= [KNL] 6086 Use normal (non-expedited) synchronous 6087 update-side primitives, if available. If all 6088 of rcutorture.gp_cond=, rcutorture.gp_exp=, 6089 rcutorture.gp_normal=, and rcutorture.gp_sync= 6090 are zero, rcutorture acts as if is interpreted 6091 they are all non-zero. 6092 6093 rcutorture.gpwrap_lag= [KNL] 6094 Enable grace-period wrap lag testing. Setting 6095 to false prevents the gpwrap lag test from 6096 running. Default is true. 6097 6098 rcutorture.gpwrap_lag_gps= [KNL] 6099 Set the value for grace-period wrap lag during 6100 active lag testing periods. This controls how many 6101 grace periods differences we tolerate between 6102 rdp and rnp's gp_seq before setting overflow flag. 6103 The default is always set to 8. 6104 6105 rcutorture.gpwrap_lag_cycle_mins= [KNL] 6106 Set the total cycle duration for gpwrap lag 6107 testing in minutes. This is the total time for 6108 one complete cycle of active and inactive 6109 testing periods. Default is 30 minutes. 6110 6111 rcutorture.gpwrap_lag_active_mins= [KNL] 6112 Set the duration for which gpwrap lag is active 6113 within each cycle, in minutes. During this time, 6114 the grace-period wrap lag will be set to the 6115 value specified by gpwrap_lag_gps. Default is 6116 5 minutes. 6117 6118 rcutorture.irqreader= [KNL] 6119 Run RCU readers from irq handlers, or, more 6120 accurately, from a timer handler. Not all RCU 6121 flavors take kindly to this sort of thing. 6122 6123 rcutorture.leakpointer= [KNL] 6124 Leak an RCU-protected pointer out of the reader. 6125 This can of course result in splats, and is 6126 intended to test the ability of things like 6127 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 6128 such leaks. 6129 6130 rcutorture.n_barrier_cbs= [KNL] 6131 Set callbacks/threads for rcu_barrier() testing. 6132 6133 rcutorture.nfakewriters= [KNL] 6134 Set number of concurrent RCU writers. These just 6135 stress RCU, they don't participate in the actual 6136 test, hence the "fake". 6137 6138 rcutorture.nocbs_nthreads= [KNL] 6139 Set number of RCU callback-offload togglers. 6140 Zero (the default) disables toggling. 6141 6142 rcutorture.nocbs_toggle= [KNL] 6143 Set the delay in milliseconds between successive 6144 callback-offload toggling attempts. 6145 6146 rcutorture.nreaders= [KNL] 6147 Set number of RCU readers. The value -1 selects 6148 N-1, where N is the number of CPUs. A value 6149 "n" less than -1 selects N-n-2, where N is again 6150 the number of CPUs. For example, -2 selects N 6151 (the number of CPUs), -3 selects N+1, and so on. 6152 6153 rcutorture.object_debug= [KNL] 6154 Enable debug-object double-call_rcu() testing. 6155 6156 rcutorture.onoff_holdoff= [KNL] 6157 Set time (s) after boot for CPU-hotplug testing. 6158 6159 rcutorture.onoff_interval= [KNL] 6160 Set time (jiffies) between CPU-hotplug operations, 6161 or zero to disable CPU-hotplug testing. 6162 6163 rcutorture.preempt_duration= [KNL] 6164 Set duration (in milliseconds) of preemptions 6165 by a high-priority FIFO real-time task. Set to 6166 zero (the default) to disable. The CPUs to 6167 preempt are selected randomly from the set that 6168 are online at a given point in time. Races with 6169 CPUs going offline are ignored, with that attempt 6170 at preemption skipped. 6171 6172 rcutorture.preempt_interval= [KNL] 6173 Set interval (in milliseconds, defaulting to one 6174 second) between preemptions by a high-priority 6175 FIFO real-time task. This delay is mediated 6176 by an hrtimer and is further fuzzed to avoid 6177 inadvertent synchronizations. 6178 6179 rcutorture.read_exit_burst= [KNL] 6180 The number of times in a given read-then-exit 6181 episode that a set of read-then-exit kthreads 6182 is spawned. 6183 6184 rcutorture.read_exit_delay= [KNL] 6185 The delay, in seconds, between successive 6186 read-then-exit testing episodes. 6187 6188 rcutorture.reader_flavor= [KNL] 6189 A bit mask indicating which readers to use. 6190 If there is more than one bit set, the readers 6191 are entered from low-order bit up, and are 6192 exited in the opposite order. For SRCU, the 6193 0x1 bit is normal readers, 0x2 NMI-safe readers, 6194 and 0x4 light-weight readers. 6195 6196 rcutorture.shuffle_interval= [KNL] 6197 Set task-shuffle interval (s). Shuffling tasks 6198 allows some CPUs to go into dyntick-idle mode 6199 during the rcutorture test. 6200 6201 rcutorture.shutdown_secs= [KNL] 6202 Set time (s) after boot system shutdown. This 6203 is useful for hands-off automated testing. 6204 6205 rcutorture.stall_cpu= [KNL] 6206 Duration of CPU stall (s) to test RCU CPU stall 6207 warnings, zero to disable. 6208 6209 rcutorture.stall_cpu_block= [KNL] 6210 Sleep while stalling if set. This will result 6211 in warnings from preemptible RCU in addition to 6212 any other stall-related activity. Note that 6213 in kernels built with CONFIG_PREEMPTION=n and 6214 CONFIG_PREEMPT_COUNT=y, this parameter will 6215 cause the CPU to pass through a quiescent state. 6216 Given CONFIG_PREEMPTION=n, this will suppress 6217 RCU CPU stall warnings, but will instead result 6218 in scheduling-while-atomic splats. 6219 6220 Use of this module parameter results in splats. 6221 6222 6223 rcutorture.stall_cpu_holdoff= [KNL] 6224 Time to wait (s) after boot before inducing stall. 6225 6226 rcutorture.stall_cpu_irqsoff= [KNL] 6227 Disable interrupts while stalling if set, but only 6228 on the first stall in the set. 6229 6230 rcutorture.stall_cpu_repeat= [KNL] 6231 Number of times to repeat the stall sequence, 6232 so that rcutorture.stall_cpu_repeat=3 will result 6233 in four stall sequences. 6234 6235 rcutorture.stall_gp_kthread= [KNL] 6236 Duration (s) of forced sleep within RCU 6237 grace-period kthread to test RCU CPU stall 6238 warnings, zero to disable. If both stall_cpu 6239 and stall_gp_kthread are specified, the 6240 kthread is starved first, then the CPU. 6241 6242 rcutorture.stat_interval= [KNL] 6243 Time (s) between statistics printk()s. 6244 6245 rcutorture.stutter= [KNL] 6246 Time (s) to stutter testing, for example, specifying 6247 five seconds causes the test to run for five seconds, 6248 wait for five seconds, and so on. This tests RCU's 6249 ability to transition abruptly to and from idle. 6250 6251 rcutorture.test_boost= [KNL] 6252 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 6253 "Maybe" means test if the RCU implementation 6254 under test support RCU priority boosting. 6255 6256 rcutorture.test_boost_duration= [KNL] 6257 Duration (s) of each individual boost test. 6258 6259 rcutorture.test_boost_holdoff= [KNL] 6260 Holdoff time (s) from start of test to the start 6261 of RCU priority-boost testing. Defaults to zero, 6262 that is, no holdoff. 6263 6264 rcutorture.test_boost_interval= [KNL] 6265 Interval (s) between each boost test. 6266 6267 rcutorture.test_no_idle_hz= [KNL] 6268 Test RCU's dyntick-idle handling. See also the 6269 rcutorture.shuffle_interval parameter. 6270 6271 rcutorture.torture_type= [KNL] 6272 Specify the RCU implementation to test. 6273 6274 rcutorture.verbose= [KNL] 6275 Enable additional printk() statements. 6276 6277 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 6278 Dump ftrace buffer after reporting RCU CPU 6279 stall warning. 6280 6281 rcupdate.rcu_cpu_stall_notifiers= [KNL] 6282 Provide RCU CPU stall notifiers, but see the 6283 warnings in the RCU_CPU_STALL_NOTIFIER Kconfig 6284 option's help text. TL;DR: You almost certainly 6285 do not want rcupdate.rcu_cpu_stall_notifiers. 6286 6287 rcupdate.rcu_cpu_stall_suppress= [KNL] 6288 Suppress RCU CPU stall warning messages. 6289 6290 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 6291 Suppress RCU CPU stall warning messages and 6292 rcutorture writer stall warnings that occur 6293 during early boot, that is, during the time 6294 before the init task is spawned. 6295 6296 rcupdate.rcu_cpu_stall_timeout= [KNL] 6297 Set timeout for RCU CPU stall warning messages. 6298 The value is in seconds and the maximum allowed 6299 value is 300 seconds. 6300 6301 rcupdate.rcu_exp_cpu_stall_timeout= [KNL] 6302 Set timeout for expedited RCU CPU stall warning 6303 messages. The value is in milliseconds 6304 and the maximum allowed value is 21000 6305 milliseconds. Please note that this value is 6306 adjusted to an arch timer tick resolution. 6307 Setting this to zero causes the value from 6308 rcupdate.rcu_cpu_stall_timeout to be used (after 6309 conversion from seconds to milliseconds). 6310 6311 rcupdate.rcu_cpu_stall_cputime= [KNL] 6312 Provide statistics on the cputime and count of 6313 interrupts and tasks during the sampling period. For 6314 multiple continuous RCU stalls, all sampling periods 6315 begin at half of the first RCU stall timeout. 6316 6317 rcupdate.rcu_exp_stall_task_details= [KNL] 6318 Print stack dumps of any tasks blocking the 6319 current expedited RCU grace period during an 6320 expedited RCU CPU stall warning. 6321 6322 rcupdate.rcu_expedited= [KNL] 6323 Use expedited grace-period primitives, for 6324 example, synchronize_rcu_expedited() instead 6325 of synchronize_rcu(). This reduces latency, 6326 but can increase CPU utilization, degrade 6327 real-time latency, and degrade energy efficiency. 6328 No effect on CONFIG_TINY_RCU kernels. 6329 6330 rcupdate.rcu_normal= [KNL] 6331 Use only normal grace-period primitives, 6332 for example, synchronize_rcu() instead of 6333 synchronize_rcu_expedited(). This improves 6334 real-time latency, CPU utilization, and 6335 energy efficiency, but can expose users to 6336 increased grace-period latency. This parameter 6337 overrides rcupdate.rcu_expedited. No effect on 6338 CONFIG_TINY_RCU kernels. 6339 6340 rcupdate.rcu_normal_after_boot= [KNL] 6341 Once boot has completed (that is, after 6342 rcu_end_inkernel_boot() has been invoked), use 6343 only normal grace-period primitives. No effect 6344 on CONFIG_TINY_RCU kernels. 6345 6346 But note that CONFIG_PREEMPT_RT=y kernels enables 6347 this kernel boot parameter, forcibly setting 6348 it to the value one, that is, converting any 6349 post-boot attempt at an expedited RCU grace 6350 period to instead use normal non-expedited 6351 grace-period processing. 6352 6353 rcupdate.rcu_task_collapse_lim= [KNL] 6354 Set the maximum number of callbacks present 6355 at the beginning of a grace period that allows 6356 the RCU Tasks flavors to collapse back to using 6357 a single callback queue. This switching only 6358 occurs when rcupdate.rcu_task_enqueue_lim is 6359 set to the default value of -1. 6360 6361 rcupdate.rcu_task_contend_lim= [KNL] 6362 Set the minimum number of callback-queuing-time 6363 lock-contention events per jiffy required to 6364 cause the RCU Tasks flavors to switch to per-CPU 6365 callback queuing. This switching only occurs 6366 when rcupdate.rcu_task_enqueue_lim is set to 6367 the default value of -1. 6368 6369 rcupdate.rcu_task_enqueue_lim= [KNL] 6370 Set the number of callback queues to use for the 6371 RCU Tasks family of RCU flavors. The default 6372 of -1 allows this to be automatically (and 6373 dynamically) adjusted. This parameter is intended 6374 for use in testing. 6375 6376 rcupdate.rcu_task_lazy_lim= [KNL] 6377 Number of callbacks on a given CPU that will 6378 cancel laziness on that CPU. Use -1 to disable 6379 cancellation of laziness, but be advised that 6380 doing so increases the danger of OOM due to 6381 callback flooding. 6382 6383 rcupdate.rcu_task_stall_info= [KNL] 6384 Set initial timeout in jiffies for RCU task stall 6385 informational messages, which give some indication 6386 of the problem for those not patient enough to 6387 wait for ten minutes. Informational messages are 6388 only printed prior to the stall-warning message 6389 for a given grace period. Disable with a value 6390 less than or equal to zero. Defaults to ten 6391 seconds. A change in value does not take effect 6392 until the beginning of the next grace period. 6393 6394 rcupdate.rcu_task_stall_info_mult= [KNL] 6395 Multiplier for time interval between successive 6396 RCU task stall informational messages for a given 6397 RCU tasks grace period. This value is clamped 6398 to one through ten, inclusive. It defaults to 6399 the value three, so that the first informational 6400 message is printed 10 seconds into the grace 6401 period, the second at 40 seconds, the third at 6402 160 seconds, and then the stall warning at 600 6403 seconds would prevent a fourth at 640 seconds. 6404 6405 rcupdate.rcu_task_stall_timeout= [KNL] 6406 Set timeout in jiffies for RCU task stall 6407 warning messages. Disable with a value less 6408 than or equal to zero. Defaults to ten minutes. 6409 A change in value does not take effect until 6410 the beginning of the next grace period. 6411 6412 rcupdate.rcu_tasks_lazy_ms= [KNL] 6413 Set timeout in milliseconds RCU Tasks asynchronous 6414 callback batching for call_rcu_tasks(). 6415 A negative value will take the default. A value 6416 of zero will disable batching. Batching is 6417 always disabled for synchronize_rcu_tasks(). 6418 6419 rcupdate.rcu_self_test= [KNL] 6420 Run the RCU early boot self tests 6421 6422 rdinit= [KNL] 6423 Format: <full_path> 6424 Run specified binary instead of /init from the ramdisk, 6425 used for early userspace startup. See initrd. 6426 6427 rdrand= [X86,EARLY] 6428 force - Override the decision by the kernel to hide the 6429 advertisement of RDRAND support (this affects 6430 certain AMD processors because of buggy BIOS 6431 support, specifically around the suspend/resume 6432 path). 6433 6434 rdt= [HW,X86,RDT] 6435 Turn on/off individual RDT features. List is: 6436 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 6437 mba, smba, bmec, abmc, sdciae, energy[:guid], 6438 perf[:guid]. 6439 E.g. to turn on cmt and turn off mba use: 6440 rdt=cmt,!mba 6441 To turn off all energy telemetry monitoring and ensure that 6442 perf telemetry monitoring associated with guid 0x12345 6443 is enabled use: 6444 rdt=!energy,perf:0x12345 6445 6446 reboot= [KNL] 6447 Format (x86 or x86_64): 6448 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \ 6449 [[,]s[mp]#### \ 6450 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 6451 [[,]f[orce] 6452 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 6453 (prefix with 'panic_' to set mode for panic 6454 reboot only), 6455 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 6456 reboot_force is either force or not specified, 6457 reboot_cpu is s[mp]#### with #### being the processor 6458 to be used for rebooting. 6459 6460 acpi 6461 Use the ACPI RESET_REG in the FADT. If ACPI is not 6462 configured or the ACPI reset does not work, the reboot 6463 path attempts the reset using the keyboard controller. 6464 6465 bios 6466 Use the CPU reboot vector for warm reset 6467 6468 cold 6469 Set the cold reboot flag 6470 6471 default 6472 There are some built-in platform specific "quirks" 6473 - you may see: "reboot: <name> series board detected. 6474 Selecting <type> for reboots." In the case where you 6475 think the quirk is in error (e.g. you have newer BIOS, 6476 or newer board) using this option will ignore the 6477 built-in quirk table, and use the generic default 6478 reboot actions. 6479 6480 efi 6481 Use efi reset_system runtime service. If EFI is not 6482 configured or the EFI reset does not work, the reboot 6483 path attempts the reset using the keyboard controller. 6484 6485 force 6486 Don't stop other CPUs on reboot. This can make reboot 6487 more reliable in some cases. 6488 6489 kbd 6490 Use the keyboard controller. cold reset (default) 6491 6492 pci 6493 Use a write to the PCI config space register 0xcf9 to 6494 trigger reboot. 6495 6496 triple 6497 Force a triple fault (init) 6498 6499 warm 6500 Don't set the cold reboot flag 6501 6502 Using warm reset will be much faster especially on big 6503 memory systems because the BIOS will not go through 6504 the memory check. Disadvantage is that not all 6505 hardware will be completely reinitialized on reboot so 6506 there may be boot problems on some systems. 6507 6508 6509 refscale.holdoff= [KNL] 6510 Set test-start holdoff period. The purpose of 6511 this parameter is to delay the start of the 6512 test until boot completes in order to avoid 6513 interference. 6514 6515 refscale.lookup_instances= [KNL] 6516 Number of data elements to use for the forms of 6517 SLAB_TYPESAFE_BY_RCU testing. A negative number 6518 is negated and multiplied by nr_cpu_ids, while 6519 zero specifies nr_cpu_ids. 6520 6521 refscale.loops= [KNL] 6522 Set the number of loops over the synchronization 6523 primitive under test. Increasing this number 6524 reduces noise due to loop start/end overhead, 6525 but the default has already reduced the per-pass 6526 noise to a handful of picoseconds on ca. 2020 6527 x86 laptops. 6528 6529 refscale.nreaders= [KNL] 6530 Set number of readers. The default value of -1 6531 selects N, where N is roughly 75% of the number 6532 of CPUs. A value of zero is an interesting choice. 6533 6534 refscale.nruns= [KNL] 6535 Set number of runs, each of which is dumped onto 6536 the console log. 6537 6538 refscale.readdelay= [KNL] 6539 Set the read-side critical-section duration, 6540 measured in microseconds. 6541 6542 refscale.scale_type= [KNL] 6543 Specify the read-protection implementation to test. 6544 6545 refscale.shutdown= [KNL] 6546 Shut down the system at the end of the performance 6547 test. This defaults to 1 (shut it down) when 6548 refscale is built into the kernel and to 0 (leave 6549 it running) when refscale is built as a module. 6550 6551 refscale.verbose= [KNL] 6552 Enable additional printk() statements. 6553 6554 refscale.verbose_batched= [KNL] 6555 Batch the additional printk() statements. If zero 6556 (the default) or negative, print everything. Otherwise, 6557 print every Nth verbose statement, where N is the value 6558 specified. 6559 6560 regulator_ignore_unused 6561 [REGULATOR] 6562 Prevents regulator framework from disabling regulators 6563 that are unused, due no driver claiming them. This may 6564 be useful for debug and development, but should not be 6565 needed on a platform with proper driver support. 6566 6567 relax_domain_level= 6568 [KNL, SMP] Set scheduler's default relax_domain_level. 6569 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 6570 6571 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 6572 Format: <base1>,<size1>[,<base2>,<size2>,...] 6573 Reserve I/O ports or memory so the kernel won't use 6574 them. If <base> is less than 0x10000, the region 6575 is assumed to be I/O ports; otherwise it is memory. 6576 6577 reserve_mem= [RAM] 6578 Format: nn[KMG]:<align>:<label> 6579 Reserve physical memory and label it with a name that 6580 other subsystems can use to access it. This is typically 6581 used for systems that do not wipe the RAM, and this command 6582 line will try to reserve the same physical memory on 6583 soft reboots. Note, it is not guaranteed to be the same 6584 location. For example, if anything about the system changes 6585 or if booting a different kernel. It can also fail if KASLR 6586 places the kernel at the location of where the RAM reservation 6587 was from a previous boot, the new reservation will be at a 6588 different location. 6589 Any subsystem using this feature must add a way to verify 6590 that the contents of the physical memory is from a previous 6591 boot, as there may be cases where the memory will not be 6592 located at the same location. 6593 6594 The format is size:align:label for example, to request 6595 12 megabytes of 4096 alignment for ramoops: 6596 6597 reserve_mem=12M:4096:oops ramoops.mem_name=oops 6598 6599 reservetop= [X86-32,EARLY] 6600 Format: nn[KMG] 6601 Reserves a hole at the top of the kernel virtual 6602 address space. 6603 6604 reset_devices [KNL] Force drivers to reset the underlying device 6605 during initialization. 6606 6607 resume= [SWSUSP] 6608 Specify the partition device for software suspend 6609 Format: 6610 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 6611 6612 resume_offset= [SWSUSP] 6613 Specify the offset from the beginning of the partition 6614 given by "resume=" at which the swap header is located, 6615 in <PAGE_SIZE> units (needed only for swap files). 6616 See Documentation/power/swsusp-and-swap-files.rst 6617 6618 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 6619 read the resume files 6620 6621 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 6622 Useful for devices that are detected asynchronously 6623 (e.g. USB and MMC devices). 6624 6625 retain_initrd [RAM] Keep initrd memory after extraction. After boot, it will 6626 be accessible via /sys/firmware/initrd. 6627 6628 retbleed= [X86] Control mitigation of RETBleed (Arbitrary 6629 Speculative Code Execution with Return Instructions) 6630 vulnerability. 6631 6632 AMD-based UNRET and IBPB mitigations alone do not stop 6633 sibling threads from influencing the predictions of other 6634 sibling threads. For that reason, STIBP is used on pro- 6635 cessors that support it, and mitigate SMT on processors 6636 that don't. 6637 6638 off - no mitigation 6639 auto - automatically select a mitigation 6640 auto,nosmt - automatically select a mitigation, 6641 disabling SMT if necessary for 6642 the full mitigation (only on Zen1 6643 and older without STIBP). 6644 ibpb - On AMD, mitigate short speculation 6645 windows on basic block boundaries too. 6646 Safe, highest perf impact. It also 6647 enables STIBP if present. Not suitable 6648 on Intel. 6649 ibpb,nosmt - Like "ibpb" above but will disable SMT 6650 when STIBP is not available. This is 6651 the alternative for systems which do not 6652 have STIBP. 6653 unret - Force enable untrained return thunks, 6654 only effective on AMD f15h-f17h based 6655 systems. 6656 unret,nosmt - Like unret, but will disable SMT when STIBP 6657 is not available. This is the alternative for 6658 systems which do not have STIBP. 6659 6660 Selecting 'auto' will choose a mitigation method at run 6661 time according to the CPU. 6662 6663 Not specifying this option is equivalent to retbleed=auto. 6664 6665 rfkill.default_state= 6666 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 6667 etc. communication is blocked by default. 6668 1 Unblocked. 6669 6670 rfkill.master_switch_mode= 6671 0 The "airplane mode" button does nothing. 6672 1 The "airplane mode" button toggles between everything 6673 blocked and the previous configuration. 6674 2 The "airplane mode" button toggles between everything 6675 blocked and everything unblocked. 6676 6677 ring3mwait=disable 6678 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 6679 CPUs. 6680 6681 riscv_isa_fallback [RISCV,EARLY] 6682 When CONFIG_RISCV_ISA_FALLBACK is not enabled, permit 6683 falling back to detecting extension support by parsing 6684 "riscv,isa" property on devicetree systems when the 6685 replacement properties are not found. See the Kconfig 6686 entry for RISCV_ISA_FALLBACK. 6687 6688 riscv_nousercfi= 6689 all Disable user CFI ABI to userspace even if cpu extension 6690 are available. 6691 bcfi Disable user backward CFI ABI to userspace even if 6692 the shadow stack extension is available. 6693 fcfi Disable user forward CFI ABI to userspace even if the 6694 landing pad extension is available. 6695 6696 ro [KNL] Mount root device read-only on boot 6697 6698 rodata= [KNL,EARLY] 6699 on Mark read-only kernel memory as read-only (default). 6700 off Leave read-only kernel memory writable for debugging. 6701 noalias Mark read-only kernel memory as read-only but retain 6702 writable aliases in the direct map for regions outside 6703 of the kernel image. [arm64] 6704 6705 rockchip.usb_uart 6706 [EARLY] 6707 Enable the uart passthrough on the designated usb port 6708 on Rockchip SoCs. When active, the signals of the 6709 debug-uart get routed to the D+ and D- pins of the usb 6710 port and the regular usb controller gets disabled. 6711 6712 root= [KNL] Root filesystem 6713 Usually this is a block device specifier of some kind, 6714 see the early_lookup_bdev comment in 6715 block/early-lookup.c for details. 6716 Alternatively this can be "ram" for the legacy initial 6717 ramdisk, "nfs" and "cifs" for root on a network file 6718 system, or "mtd" and "ubi" for mounting from raw flash. 6719 6720 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 6721 mount the root filesystem 6722 6723 rootflags= [KNL] Set root filesystem mount option string 6724 6725 rseq_slice_ext= [KNL] RSEQ based time slice extension 6726 Format: boolean 6727 Control enablement of RSEQ based time slice extension. 6728 Default is 'on'. 6729 6730 initramfs_options= [KNL] 6731 Specify mount options for the initramfs mount. 6732 6733 rootfstype= [KNL] Set root filesystem type 6734 6735 rootwait [KNL] Wait (indefinitely) for root device to show up. 6736 Useful for devices that are detected asynchronously 6737 (e.g. USB and MMC devices). 6738 6739 rootwait= [KNL] Maximum time (in seconds) to wait for root device 6740 to show up before attempting to mount the root 6741 filesystem. 6742 6743 rproc_mem=nn[KMG][@address] 6744 [KNL,ARM,CMA] Remoteproc physical memory block. 6745 Memory area to be used by remote processor image, 6746 managed by CMA. 6747 6748 rseq_debug= [KNL] Enable or disable restartable sequence 6749 debug mode. Defaults to CONFIG_RSEQ_DEBUG_DEFAULT_ENABLE. 6750 Format: <bool> 6751 6752 rt_group_sched= [KNL] Enable or disable SCHED_RR/FIFO group scheduling 6753 when CONFIG_RT_GROUP_SCHED=y. Defaults to 6754 !CONFIG_RT_GROUP_SCHED_DEFAULT_DISABLED. 6755 Format: <bool> 6756 6757 rw [KNL] Mount root device read-write on boot 6758 6759 S [KNL] Run init in single mode 6760 6761 s390_iommu= [HW,S390] 6762 Set s390 IOTLB flushing mode 6763 strict 6764 With strict flushing every unmap operation will result 6765 in an IOTLB flush. Default is lazy flushing before 6766 reuse, which is faster. Deprecated, equivalent to 6767 iommu.strict=1. 6768 6769 s390_iommu_aperture= [KNL,S390] 6770 Specifies the size of the per device DMA address space 6771 accessible through the DMA and IOMMU APIs as a decimal 6772 factor of the size of main memory. 6773 The default is 1 meaning that one can concurrently use 6774 as many DMA addresses as physical memory is installed, 6775 if supported by hardware, and thus map all of memory 6776 once. With a value of 2 one can map all of memory twice 6777 and so on. As a special case a factor of 0 imposes no 6778 restrictions other than those given by hardware at the 6779 cost of significant additional memory use for tables. 6780 6781 6782 sched_proxy_exec= [KNL] 6783 Enables or disables "proxy execution" style 6784 solution to mutex-based priority inversion. 6785 Format: <bool> 6786 6787 sched_verbose [KNL,EARLY] Enables verbose scheduler debug messages. 6788 6789 schedstats= [KNL,X86] Enable or disable scheduled statistics. 6790 Allowed values are enable and disable. This feature 6791 incurs a small amount of overhead in the scheduler 6792 but is useful for debugging and performance tuning. 6793 6794 sched_thermal_decay_shift= 6795 [Deprecated] 6796 [KNL, SMP] Set a decay shift for scheduler thermal 6797 pressure signal. Thermal pressure signal follows the 6798 default decay period of other scheduler pelt 6799 signals(usually 32 ms but configurable). Setting 6800 sched_thermal_decay_shift will left shift the decay 6801 period for the thermal pressure signal by the shift 6802 value. 6803 i.e. with the default pelt decay period of 32 ms 6804 sched_thermal_decay_shift thermal pressure decay pr 6805 1 64 ms 6806 2 128 ms 6807 and so on. 6808 Format: integer between 0 and 10 6809 Default is 0. 6810 6811 scftorture.holdoff= [KNL] 6812 Number of seconds to hold off before starting 6813 test. Defaults to zero for module insertion and 6814 to 10 seconds for built-in smp_call_function() 6815 tests. 6816 6817 scftorture.longwait= [KNL] 6818 Request ridiculously long waits randomly selected 6819 up to the chosen limit in seconds. Zero (the 6820 default) disables this feature. Please note 6821 that requesting even small non-zero numbers of 6822 seconds can result in RCU CPU stall warnings, 6823 softlockup complaints, and so on. 6824 6825 scftorture.nthreads= [KNL] 6826 Number of kthreads to spawn to invoke the 6827 smp_call_function() family of functions. 6828 The default of -1 specifies a number of kthreads 6829 equal to the number of CPUs. 6830 6831 scftorture.onoff_holdoff= [KNL] 6832 Number seconds to wait after the start of the 6833 test before initiating CPU-hotplug operations. 6834 6835 scftorture.onoff_interval= [KNL] 6836 Number seconds to wait between successive 6837 CPU-hotplug operations. Specifying zero (which 6838 is the default) disables CPU-hotplug operations. 6839 6840 scftorture.shutdown_secs= [KNL] 6841 The number of seconds following the start of the 6842 test after which to shut down the system. The 6843 default of zero avoids shutting down the system. 6844 Non-zero values are useful for automated tests. 6845 6846 scftorture.stat_interval= [KNL] 6847 The number of seconds between outputting the 6848 current test statistics to the console. A value 6849 of zero disables statistics output. 6850 6851 scftorture.stutter_cpus= [KNL] 6852 The number of jiffies to wait between each change 6853 to the set of CPUs under test. 6854 6855 scftorture.use_cpus_read_lock= [KNL] 6856 Use use_cpus_read_lock() instead of the default 6857 preempt_disable() to disable CPU hotplug 6858 while invoking one of the smp_call_function*() 6859 functions. 6860 6861 scftorture.verbose= [KNL] 6862 Enable additional printk() statements. 6863 6864 scftorture.weight_single= [KNL] 6865 The probability weighting to use for the 6866 smp_call_function_single() function with a zero 6867 "wait" parameter. A value of -1 selects the 6868 default if all other weights are -1. However, 6869 if at least one weight has some other value, a 6870 value of -1 will instead select a weight of zero. 6871 6872 scftorture.weight_single_wait= [KNL] 6873 The probability weighting to use for the 6874 smp_call_function_single() function with a 6875 non-zero "wait" parameter. See weight_single. 6876 6877 scftorture.weight_many= [KNL] 6878 The probability weighting to use for the 6879 smp_call_function_many() function with a zero 6880 "wait" parameter. See weight_single. 6881 Note well that setting a high probability for 6882 this weighting can place serious IPI load 6883 on the system. 6884 6885 scftorture.weight_many_wait= [KNL] 6886 The probability weighting to use for the 6887 smp_call_function_many() function with a 6888 non-zero "wait" parameter. See weight_single 6889 and weight_many. 6890 6891 scftorture.weight_all= [KNL] 6892 The probability weighting to use for the 6893 smp_call_function_all() function with a zero 6894 "wait" parameter. See weight_single and 6895 weight_many. 6896 6897 scftorture.weight_all_wait= [KNL] 6898 The probability weighting to use for the 6899 smp_call_function_all() function with a 6900 non-zero "wait" parameter. See weight_single 6901 and weight_many. 6902 6903 sdw_mclk_divider=[SDW] 6904 Specify the MCLK divider for Intel SoundWire buses in 6905 case the BIOS does not provide the clock rate properly. 6906 6907 skew_tick= [KNL,EARLY] Offset the periodic timer tick per cpu to mitigate 6908 xtime_lock contention on larger systems, and/or RCU lock 6909 contention on all systems with CONFIG_MAXSMP set. 6910 Format: { "0" | "1" } 6911 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 6912 1 -- enable. 6913 Note: increases power consumption, thus should only be 6914 enabled if running jitter sensitive (HPC/RT) workloads. 6915 6916 security= [SECURITY] Choose a legacy "major" security module to 6917 enable at boot. This has been deprecated by the 6918 "lsm=" parameter. 6919 6920 selinux= [SELINUX] Disable or enable SELinux at boot time. 6921 Format: { "0" | "1" } 6922 See security/selinux/Kconfig help text. 6923 0 -- disable. 6924 1 -- enable. 6925 Default value is 1. 6926 6927 serialnumber [BUGS=X86-32] 6928 6929 sev=option[,option...] [X86-64] 6930 6931 debug 6932 Enable debug messages. 6933 6934 nosnp 6935 Do not enable SEV-SNP (applies to host/hypervisor 6936 only). Setting 'nosnp' avoids the RMP check overhead 6937 in memory accesses when users do not want to run 6938 SEV-SNP guests. 6939 6940 shapers= [NET] 6941 Maximal number of shapers. 6942 6943 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 6944 Limit apic dumping. The parameter defines the maximal 6945 number of local apics being dumped. Also it is possible 6946 to set it to "all" by meaning -- no limit here. 6947 Format: { 1 (default) | 2 | ... | all }. 6948 The parameter valid if only apic=debug or 6949 apic=verbose is specified. 6950 Example: apic=debug show_lapic=all 6951 6952 slab_debug[=options[,slabs][;[options[,slabs]]...] [MM] 6953 Enabling slab_debug allows one to determine the 6954 culprit if slab objects become corrupted. Enabling 6955 slab_debug can create guard zones around objects and 6956 may poison objects when not in use. Also tracks the 6957 last alloc / free. For more information see 6958 Documentation/admin-guide/mm/slab.rst. 6959 (slub_debug legacy name also accepted for now) 6960 6961 Using this option implies the "no_hash_pointers" 6962 option which can be undone by adding the 6963 "hash_pointers=always" option. 6964 6965 slab_max_order= [MM] 6966 Determines the maximum allowed order for slabs. 6967 A high setting may cause OOMs due to memory 6968 fragmentation. For more information see 6969 Documentation/admin-guide/mm/slab.rst. 6970 (slub_max_order legacy name also accepted for now) 6971 6972 slab_merge [MM] 6973 Enable merging of slabs with similar size when the 6974 kernel is built without CONFIG_SLAB_MERGE_DEFAULT. 6975 (slub_merge legacy name also accepted for now) 6976 6977 slab_min_objects= [MM] 6978 The minimum number of objects per slab. SLUB will 6979 increase the slab order up to slab_max_order to 6980 generate a sufficiently large slab able to contain 6981 the number of objects indicated. The higher the number 6982 of objects the smaller the overhead of tracking slabs 6983 and the less frequently locks need to be acquired. 6984 For more information see 6985 Documentation/admin-guide/mm/slab.rst. 6986 (slub_min_objects legacy name also accepted for now) 6987 6988 slab_min_order= [MM] 6989 Determines the minimum page order for slabs. Must be 6990 lower or equal to slab_max_order. For more information see 6991 Documentation/admin-guide/mm/slab.rst. 6992 (slub_min_order legacy name also accepted for now) 6993 6994 slab_nomerge [MM] 6995 Disable merging of slabs with similar size. May be 6996 necessary if there is some reason to distinguish 6997 allocs to different slabs, especially in hardened 6998 environments where the risk of heap overflows and 6999 layout control by attackers can usually be 7000 frustrated by disabling merging. This will reduce 7001 most of the exposure of a heap attack to a single 7002 cache (risks via metadata attacks are mostly 7003 unchanged). Debug options disable merging on their 7004 own. 7005 For more information see 7006 Documentation/admin-guide/mm/slab.rst. 7007 (slub_nomerge legacy name also accepted for now) 7008 7009 slab_strict_numa [MM] 7010 Support memory policies on a per object level 7011 in the slab allocator. The default is for memory 7012 policies to be applied at the folio level when 7013 a new folio is needed or a partial folio is 7014 retrieved from the lists. Increases overhead 7015 in the slab fastpaths but gains more accurate 7016 NUMA kernel object placement which helps with slow 7017 interconnects in NUMA systems. 7018 7019 slram= [HW,MTD] 7020 7021 smart2= [HW] 7022 Format: <io1>[,<io2>[,...,<io8>]] 7023 7024 smp.csd_lock_timeout= [KNL] 7025 Specify the period of time in milliseconds 7026 that smp_call_function() and friends will wait 7027 for a CPU to release the CSD lock. This is 7028 useful when diagnosing bugs involving CPUs 7029 disabling interrupts for extended periods 7030 of time. Defaults to 5,000 milliseconds, and 7031 setting a value of zero disables this feature. 7032 This feature may be more efficiently disabled 7033 using the csdlock_debug- kernel parameter. 7034 7035 smp.panic_on_ipistall= [KNL] 7036 If a csd_lock_timeout extends for more than 7037 the specified number of milliseconds, panic the 7038 system. By default, let CSD-lock acquisition 7039 take as long as they take. Specifying 300,000 7040 for this value provides a 5-minute timeout. 7041 7042 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 7043 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 7044 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 7045 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 7046 smsc-ircc2.ircc_irq= [HW] IRQ line 7047 smsc-ircc2.ircc_dma= [HW] DMA channel 7048 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 7049 0: Toshiba Satellite 1800 (GP data pin select) 7050 1: Fast pin select (default) 7051 2: ATC IRMode 7052 7053 smt= [KNL,MIPS,S390,EARLY] Set the maximum number of threads 7054 (logical CPUs) to use per physical CPU on systems 7055 capable of symmetric multithreading (SMT). Will 7056 be capped to the actual hardware limit. 7057 Format: <integer> 7058 Default: -1 (no limit) 7059 7060 softlockup_panic= 7061 [KNL] Should the soft-lockup detector generate panics. 7062 Format: <int> 7063 7064 A value of non-zero instructs the soft-lockup detector 7065 to panic the machine when a soft-lockup duration exceeds 7066 N thresholds. It is also controlled by the kernel.softlockup_panic 7067 sysctl and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 7068 respective build-time switch to that functionality. 7069 7070 softlockup_all_cpu_backtrace= 7071 [KNL] Should the soft-lockup detector generate 7072 backtraces on all cpus. 7073 Format: 0 | 1 7074 7075 sonypi.*= [HW] Sony Programmable I/O Control Device driver 7076 See Documentation/admin-guide/laptops/sonypi.rst 7077 7078 spectre_bhi= [X86] Control mitigation of Branch History Injection 7079 (BHI) vulnerability. This setting affects the 7080 deployment of the HW BHI control and the SW BHB 7081 clearing sequence. 7082 7083 on - (default) Enable the HW or SW mitigation as 7084 needed. This protects the kernel from 7085 both syscalls and VMs. 7086 vmexit - On systems which don't have the HW mitigation 7087 available, enable the SW mitigation on vmexit 7088 ONLY. On such systems, the host kernel is 7089 protected from VM-originated BHI attacks, but 7090 may still be vulnerable to syscall attacks. 7091 off - Disable the mitigation. 7092 7093 spectre_v2= [X86,EARLY] Control mitigation of Spectre variant 2 7094 (indirect branch speculation) vulnerability. 7095 The default operation protects the kernel from 7096 user space attacks. 7097 7098 on - unconditionally enable, implies 7099 spectre_v2_user=on 7100 off - unconditionally disable, implies 7101 spectre_v2_user=off 7102 auto - kernel detects whether your CPU model is 7103 vulnerable 7104 7105 Selecting 'on' will, and 'auto' may, choose a 7106 mitigation method at run time according to the 7107 CPU, the available microcode, the setting of the 7108 CONFIG_MITIGATION_RETPOLINE configuration option, 7109 and the compiler with which the kernel was built. 7110 7111 Selecting 'on' will also enable the mitigation 7112 against user space to user space task attacks. 7113 Selecting specific mitigation does not force enable 7114 user mitigations. 7115 7116 Selecting 'off' will disable both the kernel and 7117 the user space protections. 7118 7119 Specific mitigations can also be selected manually: 7120 7121 retpoline - replace indirect branches 7122 retpoline,generic - Retpolines 7123 retpoline,lfence - LFENCE; indirect branch 7124 retpoline,amd - alias for retpoline,lfence 7125 eibrs - Enhanced/Auto IBRS 7126 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines 7127 eibrs,lfence - Enhanced/Auto IBRS + LFENCE 7128 ibrs - use IBRS to protect kernel 7129 7130 Not specifying this option is equivalent to 7131 spectre_v2=auto. 7132 7133 spectre_v2_user= 7134 [X86] Control mitigation of Spectre variant 2 7135 (indirect branch speculation) vulnerability between 7136 user space tasks 7137 7138 on - Unconditionally enable mitigations. Is 7139 enforced by spectre_v2=on 7140 7141 off - Unconditionally disable mitigations. Is 7142 enforced by spectre_v2=off 7143 7144 prctl - Indirect branch speculation is enabled, 7145 but mitigation can be enabled via prctl 7146 per thread. The mitigation control state 7147 is inherited on fork. 7148 7149 prctl,ibpb 7150 - Like "prctl" above, but only STIBP is 7151 controlled per thread. IBPB is issued 7152 always when switching between different user 7153 space processes. 7154 7155 seccomp 7156 - Same as "prctl" above, but all seccomp 7157 threads will enable the mitigation unless 7158 they explicitly opt out. 7159 7160 seccomp,ibpb 7161 - Like "seccomp" above, but only STIBP is 7162 controlled per thread. IBPB is issued 7163 always when switching between different 7164 user space processes. 7165 7166 auto - Kernel selects the mitigation depending on 7167 the available CPU features and vulnerability. 7168 7169 Default mitigation: "prctl" 7170 7171 Not specifying this option is equivalent to 7172 spectre_v2_user=auto. 7173 7174 spec_rstack_overflow= 7175 [X86,EARLY] Control RAS overflow mitigation on AMD Zen CPUs 7176 7177 off - Disable mitigation 7178 microcode - Enable microcode mitigation only 7179 safe-ret - Enable sw-only safe RET mitigation (default) 7180 ibpb - Enable mitigation by issuing IBPB on 7181 kernel entry 7182 ibpb-vmexit - Issue IBPB only on VMEXIT 7183 (cloud-specific mitigation) 7184 7185 spec_store_bypass_disable= 7186 [HW,EARLY] Control Speculative Store Bypass (SSB) Disable mitigation 7187 (Speculative Store Bypass vulnerability) 7188 7189 Certain CPUs are vulnerable to an exploit against a 7190 a common industry wide performance optimization known 7191 as "Speculative Store Bypass" in which recent stores 7192 to the same memory location may not be observed by 7193 later loads during speculative execution. The idea 7194 is that such stores are unlikely and that they can 7195 be detected prior to instruction retirement at the 7196 end of a particular speculation execution window. 7197 7198 In vulnerable processors, the speculatively forwarded 7199 store can be used in a cache side channel attack, for 7200 example to read memory to which the attacker does not 7201 directly have access (e.g. inside sandboxed code). 7202 7203 This parameter controls whether the Speculative Store 7204 Bypass optimization is used. 7205 7206 On x86 the options are: 7207 7208 on - Unconditionally disable Speculative Store Bypass 7209 off - Unconditionally enable Speculative Store Bypass 7210 auto - Kernel detects whether the CPU model contains an 7211 implementation of Speculative Store Bypass and 7212 picks the most appropriate mitigation. If the 7213 CPU is not vulnerable, "off" is selected. If the 7214 CPU is vulnerable the default mitigation is 7215 architecture and Kconfig dependent. See below. 7216 prctl - Control Speculative Store Bypass per thread 7217 via prctl. Speculative Store Bypass is enabled 7218 for a process by default. The state of the control 7219 is inherited on fork. 7220 seccomp - Same as "prctl" above, but all seccomp threads 7221 will disable SSB unless they explicitly opt out. 7222 7223 Default mitigations: 7224 X86: "prctl" 7225 7226 On powerpc the options are: 7227 7228 on,auto - On Power8 and Power9 insert a store-forwarding 7229 barrier on kernel entry and exit. On Power7 7230 perform a software flush on kernel entry and 7231 exit. 7232 off - No action. 7233 7234 Not specifying this option is equivalent to 7235 spec_store_bypass_disable=auto. 7236 7237 split_llc= 7238 [X86,EARLY] Split the LLC N-ways 7239 7240 When set, the LLC is split this many ways by matching 7241 'core_id % n'. This is setup before SMP bringup and 7242 used during SMP bringup before it knows the full 7243 topology. If your core count doesn't nicely divide by 7244 the number given, you get to keep the pieces. 7245 7246 This is mostly a debug feature to emulate multiple LLCs 7247 on hardware that only have a single LLC. 7248 7249 split_lock_detect= 7250 [X86] Enable split lock detection or bus lock detection 7251 7252 When enabled (and if hardware support is present), atomic 7253 instructions that access data across cache line 7254 boundaries will result in an alignment check exception 7255 for split lock detection or a debug exception for 7256 bus lock detection. 7257 7258 off - not enabled 7259 7260 warn - the kernel will emit rate-limited warnings 7261 about applications triggering the #AC 7262 exception or the #DB exception. This mode is 7263 the default on CPUs that support split lock 7264 detection or bus lock detection. Default 7265 behavior is by #AC if both features are 7266 enabled in hardware. 7267 7268 fatal - the kernel will send SIGBUS to applications 7269 that trigger the #AC exception or the #DB 7270 exception. Default behavior is by #AC if 7271 both features are enabled in hardware. 7272 7273 ratelimit:N - 7274 Set system wide rate limit to N bus locks 7275 per second for bus lock detection. 7276 0 < N <= 1000. 7277 7278 N/A for split lock detection. 7279 7280 7281 If an #AC exception is hit in the kernel or in 7282 firmware (i.e. not while executing in user mode) 7283 the kernel will oops in either "warn" or "fatal" 7284 mode. 7285 7286 #DB exception for bus lock is triggered only when 7287 CPL > 0. 7288 7289 srbds= [X86,INTEL,EARLY] 7290 Control the Special Register Buffer Data Sampling 7291 (SRBDS) mitigation. 7292 7293 Certain CPUs are vulnerable to an MDS-like 7294 exploit which can leak bits from the random 7295 number generator. 7296 7297 By default, this issue is mitigated by 7298 microcode. However, the microcode fix can cause 7299 the RDRAND and RDSEED instructions to become 7300 much slower. Among other effects, this will 7301 result in reduced throughput from /dev/urandom. 7302 7303 The microcode mitigation can be disabled with 7304 the following option: 7305 7306 off: Disable mitigation and remove 7307 performance impact to RDRAND and RDSEED 7308 7309 srcutree.big_cpu_lim [KNL] 7310 Specifies the number of CPUs constituting a 7311 large system, such that srcu_struct structures 7312 should immediately allocate an srcu_node array. 7313 This kernel-boot parameter defaults to 128, 7314 but takes effect only when the low-order four 7315 bits of srcutree.convert_to_big is equal to 3 7316 (decide at boot). 7317 7318 srcutree.convert_to_big [KNL] 7319 Specifies under what conditions an SRCU tree 7320 srcu_struct structure will be converted to big 7321 form, that is, with an rcu_node tree: 7322 7323 0: Never. 7324 1: At init_srcu_struct() time. 7325 2: When rcutorture decides to. 7326 3: Decide at boot time (default). 7327 0x1X: Above plus if high contention. 7328 7329 Either way, the srcu_node tree will be sized based 7330 on the actual runtime number of CPUs (nr_cpu_ids) 7331 instead of the compile-time CONFIG_NR_CPUS. 7332 7333 srcutree.counter_wrap_check [KNL] 7334 Specifies how frequently to check for 7335 grace-period sequence counter wrap for the 7336 srcu_data structure's ->srcu_gp_seq_needed field. 7337 The greater the number of bits set in this kernel 7338 parameter, the less frequently counter wrap will 7339 be checked for. Note that the bottom two bits 7340 are ignored. 7341 7342 srcutree.exp_holdoff [KNL] 7343 Specifies how many nanoseconds must elapse 7344 since the end of the last SRCU grace period for 7345 a given srcu_struct until the next normal SRCU 7346 grace period will be considered for automatic 7347 expediting. Set to zero to disable automatic 7348 expediting. 7349 7350 srcutree.srcu_max_nodelay [KNL] 7351 Specifies the number of no-delay instances 7352 per jiffy for which the SRCU grace period 7353 worker thread will be rescheduled with zero 7354 delay. Beyond this limit, worker thread will 7355 be rescheduled with a sleep delay of one jiffy. 7356 7357 srcutree.srcu_max_nodelay_phase [KNL] 7358 Specifies the per-grace-period phase, number of 7359 non-sleeping polls of readers. Beyond this limit, 7360 grace period worker thread will be rescheduled 7361 with a sleep delay of one jiffy, between each 7362 rescan of the readers, for a grace period phase. 7363 7364 srcutree.srcu_retry_check_delay [KNL] 7365 Specifies number of microseconds of non-sleeping 7366 delay between each non-sleeping poll of readers. 7367 7368 srcutree.small_contention_lim [KNL] 7369 Specifies the number of update-side contention 7370 events per jiffy will be tolerated before 7371 initiating a conversion of an srcu_struct 7372 structure to big form. Note that the value of 7373 srcutree.convert_to_big must have the 0x10 bit 7374 set for contention-based conversions to occur. 7375 7376 ssbd= [ARM64,HW,EARLY] 7377 Speculative Store Bypass Disable control 7378 7379 On CPUs that are vulnerable to the Speculative 7380 Store Bypass vulnerability and offer a 7381 firmware based mitigation, this parameter 7382 indicates how the mitigation should be used: 7383 7384 force-on: Unconditionally enable mitigation for 7385 for both kernel and userspace 7386 force-off: Unconditionally disable mitigation for 7387 for both kernel and userspace 7388 kernel: Always enable mitigation in the 7389 kernel, and offer a prctl interface 7390 to allow userspace to register its 7391 interest in being mitigated too. 7392 7393 stack_guard_gap= [MM] 7394 override the default stack gap protection. The value 7395 is in page units and it defines how many pages prior 7396 to (for stacks growing down) resp. after (for stacks 7397 growing up) the main stack are reserved for no other 7398 mapping. Default value is 256 pages. 7399 7400 stack_depot_disable= [KNL,EARLY] 7401 Setting this to true through kernel command line will 7402 disable the stack depot thereby saving the static memory 7403 consumed by the stack hash table. By default this is set 7404 to false. 7405 7406 stack_depot_max_pools= [KNL,EARLY] 7407 Specify the maximum number of pools to use for storing 7408 stack traces. Pools are allocated on-demand up to this 7409 limit. Default value is 8191 pools. 7410 7411 stacktrace [FTRACE] 7412 Enable the stack tracer on boot up. 7413 7414 stacktrace_filter=[function-list] 7415 [FTRACE] Limit the functions that the stack tracer 7416 will trace at boot up. function-list is a comma-separated 7417 list of functions. This list can be changed at run 7418 time by the stack_trace_filter file in the debugfs 7419 tracing directory. Note, this enables stack tracing 7420 and the stacktrace above is not needed. 7421 7422 sti= [PARISC,HW] 7423 Format: <num> 7424 Set the STI (builtin display/keyboard on the HP-PARISC 7425 machines) console (graphic card) which should be used 7426 as the initial boot-console. 7427 See also comment in drivers/video/sticore.c. 7428 7429 sti_font= [HW] 7430 See comment in drivers/video/sticore.c. 7431 7432 stifb= [HW] 7433 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 7434 7435 strict_sas_size= 7436 [X86] 7437 Format: <bool> 7438 Enable or disable strict sigaltstack size checks 7439 against the required signal frame size which 7440 depends on the supported FPU features. This can 7441 be used to filter out binaries which have 7442 not yet been made aware of AT_MINSIGSTKSZ. 7443 7444 stress_hpt [PPC,EARLY] 7445 Limits the number of kernel HPT entries in the hash 7446 page table to increase the rate of hash page table 7447 faults on kernel addresses. 7448 7449 stress_slb [PPC,EARLY] 7450 Limits the number of kernel SLB entries, and flushes 7451 them frequently to increase the rate of SLB faults 7452 on kernel addresses. 7453 7454 no_slb_preload [PPC,EARLY] 7455 Disables slb preloading for userspace. 7456 7457 sunrpc.min_resvport= 7458 sunrpc.max_resvport= 7459 [NFS,SUNRPC] 7460 SunRPC servers often require that client requests 7461 originate from a privileged port (i.e. a port in the 7462 range 0 < portnr < 1024). 7463 An administrator who wishes to reserve some of these 7464 ports for other uses may adjust the range that the 7465 kernel's sunrpc client considers to be privileged 7466 using these two parameters to set the minimum and 7467 maximum port values. 7468 7469 sunrpc.svc_rpc_per_connection_limit= 7470 [NFS,SUNRPC] 7471 Limit the number of requests that the server will 7472 process in parallel from a single connection. 7473 The default value is 0 (no limit). 7474 7475 sunrpc.pool_mode= 7476 [NFS] 7477 Control how the NFS server code allocates CPUs to 7478 service thread pools. Depending on how many NICs 7479 you have and where their interrupts are bound, this 7480 option will affect which CPUs will do NFS serving. 7481 Note: this parameter cannot be changed while the 7482 NFS server is running. 7483 7484 auto the server chooses an appropriate mode 7485 automatically using heuristics 7486 global a single global pool contains all CPUs 7487 percpu one pool for each CPU 7488 pernode one pool for each NUMA node (equivalent 7489 to global on non-NUMA machines) 7490 7491 sunrpc.tcp_slot_table_entries= 7492 sunrpc.udp_slot_table_entries= 7493 [NFS,SUNRPC] 7494 Sets the upper limit on the number of simultaneous 7495 RPC calls that can be sent from the client to a 7496 server. Increasing these values may allow you to 7497 improve throughput, but will also increase the 7498 amount of memory reserved for use by the client. 7499 7500 suspend.pm_test_delay= 7501 [SUSPEND] 7502 Sets the number of seconds to remain in a suspend test 7503 mode before resuming the system (see 7504 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 7505 is set. Default value is 5. 7506 7507 svm= [PPC] 7508 Format: { on | off | y | n | 1 | 0 } 7509 This parameter controls use of the Protected 7510 Execution Facility on pSeries. 7511 7512 swiotlb= [ARM,PPC,MIPS,X86,S390,EARLY] 7513 Format: { <int> [,<int>] | force | noforce } 7514 <int> -- Number of I/O TLB slabs 7515 <int> -- Second integer after comma. Number of swiotlb 7516 areas with their own lock. Will be rounded up 7517 to a power of 2. 7518 force -- force using of bounce buffers even if they 7519 wouldn't be automatically used by the kernel 7520 noforce -- Never use bounce buffers (for debugging) 7521 7522 switches= [HW,M68k,EARLY] 7523 7524 sysctl.*= [KNL] 7525 Set a sysctl parameter, right before loading the init 7526 process, as if the value was written to the respective 7527 /proc/sys/... file. Both '.' and '/' are recognized as 7528 separators. Unrecognized parameters and invalid values 7529 are reported in the kernel log. Sysctls registered 7530 later by a loaded module cannot be set this way. 7531 Example: sysctl.vm.swappiness=40 7532 7533 sysrq_always_enabled 7534 [KNL] 7535 Ignore sysrq setting - this boot parameter will 7536 neutralize any effect of /proc/sys/kernel/sysrq. 7537 Useful for debugging. 7538 7539 tcpmhash_entries= [KNL,NET] 7540 Set the number of tcp_metrics_hash slots. 7541 Default value is 8192 or 16384 depending on total 7542 ram pages. This is used to specify the TCP metrics 7543 cache size. See Documentation/networking/ip-sysctl.rst 7544 "tcp_no_metrics_save" section for more details. 7545 7546 tdfx= [HW,DRM] 7547 7548 test_suspend= [SUSPEND] 7549 Format: { "mem" | "standby" | "freeze" }[,N] 7550 Specify "mem" (for Suspend-to-RAM) or "standby" (for 7551 standby suspend) or "freeze" (for suspend type freeze) 7552 as the system sleep state during system startup with 7553 the optional capability to repeat N number of times. 7554 The system is woken from this state using a 7555 wakeup-capable RTC alarm. 7556 7557 thash_entries= [KNL,NET] 7558 Set number of hash buckets for TCP connection 7559 7560 thermal.act= [HW,ACPI] 7561 -1: disable all active trip points in all thermal zones 7562 <degrees C>: override all lowest active trip points 7563 7564 thermal.crt= [HW,ACPI] 7565 -1: disable all critical trip points in all thermal zones 7566 <degrees C>: override all critical trip points 7567 7568 thermal.off= [HW,ACPI] 7569 1: disable ACPI thermal control 7570 7571 thermal.psv= [HW,ACPI] 7572 -1: disable all passive trip points 7573 <degrees C>: override all passive trip points to this 7574 value 7575 7576 thermal.tzp= [HW,ACPI] 7577 Specify global default ACPI thermal zone polling rate 7578 <deci-seconds>: poll all this frequency 7579 0: no polling (default) 7580 7581 thp_anon= [KNL] 7582 Format: <size>[KMG],<size>[KMG]:<state>;<size>[KMG]-<size>[KMG]:<state> 7583 state is one of "always", "madvise", "never" or "inherit". 7584 Control the default behavior of the system with respect 7585 to anonymous transparent hugepages. 7586 Can be used multiple times for multiple anon THP sizes. 7587 See Documentation/admin-guide/mm/transhuge.rst for more 7588 details. 7589 7590 threadirqs [KNL,EARLY] 7591 Force threading of all interrupt handlers except those 7592 marked explicitly IRQF_NO_THREAD. 7593 7594 thp_shmem= [KNL] 7595 Format: <size>[KMG],<size>[KMG]:<policy>;<size>[KMG]-<size>[KMG]:<policy> 7596 Control the default policy of each hugepage size for the 7597 internal shmem mount. <policy> is one of policies available 7598 for the shmem mount ("always", "inherit", "never", "within_size", 7599 and "advise"). 7600 It can be used multiple times for multiple shmem THP sizes. 7601 See Documentation/admin-guide/mm/transhuge.rst for more 7602 details. 7603 7604 tlbi= [X86-64] 7605 Format: {ipi} 7606 ipi: switch to IPI-based TLB flushing 7607 7608 topology= [S390,EARLY] 7609 Format: {off | on} 7610 Specify if the kernel should make use of the cpu 7611 topology information if the hardware supports this. 7612 The scheduler will make use of this information and 7613 e.g. base its process migration decisions on it. 7614 Default is on. 7615 7616 torture.disable_onoff_at_boot= [KNL] 7617 Prevent the CPU-hotplug component of torturing 7618 until after init has spawned. 7619 7620 torture.ftrace_dump_at_shutdown= [KNL] 7621 Dump the ftrace buffer at torture-test shutdown, 7622 even if there were no errors. This can be a 7623 very costly operation when many torture tests 7624 are running concurrently, especially on systems 7625 with rotating-rust storage. 7626 7627 torture.verbose_sleep_frequency= [KNL] 7628 Specifies how many verbose printk()s should be 7629 emitted between each sleep. The default of zero 7630 disables verbose-printk() sleeping. 7631 7632 torture.verbose_sleep_duration= [KNL] 7633 Duration of each verbose-printk() sleep in jiffies. 7634 7635 tpm.disable_pcr_integrity= [HW,TPM] 7636 Do not protect PCR registers from unintended physical 7637 access, or interposers in the bus by the means of 7638 having an integrity protected session wrapped around 7639 TPM2_PCR_Extend command. Consider this in a situation 7640 where TPM is heavily utilized by IMA, thus protection 7641 causing a major performance hit, and the space where 7642 machines are deployed is by other means guarded. 7643 7644 tpm_crb_ffa.busy_timeout_ms= [ARM64,TPM] 7645 Maximum time in milliseconds to retry sending a message 7646 to the TPM service before giving up. This parameter controls 7647 how long the system will continue retrying when the TPM 7648 service is busy. 7649 Format: <unsigned int> 7650 Default: 2000 (2 seconds) 7651 7652 tpm_suspend_pcr=[HW,TPM] 7653 Format: integer pcr id 7654 Specify that at suspend time, the tpm driver 7655 should extend the specified pcr with zeros, 7656 as a workaround for some chips which fail to 7657 flush the last written pcr on TPM_SaveState. 7658 This will guarantee that all the other pcrs 7659 are saved. 7660 7661 tpm_tis.interrupts= [HW,TPM] 7662 Enable interrupts for the MMIO based physical layer 7663 for the FIFO interface. By default it is set to false 7664 (0). For more information about TPM hardware interfaces 7665 defined by Trusted Computing Group (TCG) see 7666 https://trustedcomputinggroup.org/resource/pc-client-platform-tpm-profile-ptp-specification/ 7667 7668 tp_printk [FTRACE] 7669 Have the tracepoints sent to printk as well as the 7670 tracing ring buffer. This is useful for early boot up 7671 where the system hangs or reboots and does not give the 7672 option for reading the tracing buffer or performing a 7673 ftrace_dump_on_oops. 7674 7675 To turn off having tracepoints sent to printk, 7676 echo 0 > /proc/sys/kernel/tracepoint_printk 7677 Note, echoing 1 into this file without the 7678 tp_printk kernel cmdline option has no effect. 7679 7680 The tp_printk_stop_on_boot (see below) can also be used 7681 to stop the printing of events to console at 7682 late_initcall_sync. 7683 7684 ** CAUTION ** 7685 7686 Having tracepoints sent to printk() and activating high 7687 frequency tracepoints such as irq or sched, can cause 7688 the system to live lock. 7689 7690 tp_printk_stop_on_boot [FTRACE] 7691 When tp_printk (above) is set, it can cause a lot of noise 7692 on the console. It may be useful to only include the 7693 printing of events during boot up, as user space may 7694 make the system inoperable. 7695 7696 This command line option will stop the printing of events 7697 to console at the late_initcall_sync() time frame. 7698 7699 trace_buf_size=nn[KMG] 7700 [FTRACE] will set tracing buffer size on each cpu. 7701 7702 trace_clock= [FTRACE] Set the clock used for tracing events 7703 at boot up. 7704 local - Use the per CPU time stamp counter 7705 (converted into nanoseconds). Fast, but 7706 depending on the architecture, may not be 7707 in sync between CPUs. 7708 global - Event time stamps are synchronized across 7709 CPUs. May be slower than the local clock, 7710 but better for some race conditions. 7711 counter - Simple counting of events (1, 2, ..) 7712 note, some counts may be skipped due to the 7713 infrastructure grabbing the clock more than 7714 once per event. 7715 uptime - Use jiffies as the time stamp. 7716 perf - Use the same clock that perf uses. 7717 mono - Use ktime_get_mono_fast_ns() for time stamps. 7718 mono_raw - Use ktime_get_raw_fast_ns() for time 7719 stamps. 7720 boot - Use ktime_get_boot_fast_ns() for time stamps. 7721 Architectures may add more clocks. See 7722 Documentation/trace/ftrace.rst for more details. 7723 7724 trace_event=[event-list] 7725 [FTRACE] Set and start specified trace events in order 7726 to facilitate early boot debugging. The event-list is a 7727 comma-separated list of trace events to enable. See 7728 also Documentation/trace/events.rst 7729 7730 To enable modules, use :mod: keyword: 7731 7732 trace_event=:mod:<module> 7733 7734 The value before :mod: will only enable specific events 7735 that are part of the module. See the above mentioned 7736 document for more information. 7737 7738 trace_instance=[instance-info] 7739 [FTRACE] Create a ring buffer instance early in boot up. 7740 This will be listed in: 7741 7742 /sys/kernel/tracing/instances 7743 7744 Events can be enabled at the time the instance is created 7745 via: 7746 7747 trace_instance=<name>,<system1>:<event1>,<system2>:<event2> 7748 7749 Note, the "<system*>:" portion is optional if the event is 7750 unique. 7751 7752 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall 7753 7754 will enable the "sched_switch" event (note, the "sched:" is optional, and 7755 the same thing would happen if it was left off). The irq_handler_entry 7756 event, and all events under the "initcall" system. 7757 7758 Flags can be added to the instance to modify its behavior when it is 7759 created. The flags are separated by '^'. 7760 7761 The available flags are: 7762 7763 traceoff - Have the tracing instance tracing disabled after it is created. 7764 traceprintk - Have trace_printk() write into this trace instance 7765 (note, "printk" and "trace_printk" can also be used) 7766 7767 trace_instance=foo^traceoff^traceprintk,sched,irq 7768 7769 The flags must come before the defined events. 7770 7771 If memory has been reserved (see memmap for x86), the instance 7772 can use that memory: 7773 7774 memmap=12M$0x284500000 trace_instance=boot_map@0x284500000:12M 7775 7776 The above will create a "boot_map" instance that uses the physical 7777 memory at 0x284500000 that is 12Megs. The per CPU buffers of that 7778 instance will be split up accordingly. 7779 7780 Alternatively, the memory can be reserved by the reserve_mem option: 7781 7782 reserve_mem=12M:4096:trace trace_instance=boot_map@trace 7783 7784 This will reserve 12 megabytes at boot up with a 4096 byte alignment 7785 and place the ring buffer in this memory. Note that due to KASLR, the 7786 memory may not be the same location each time, which will not preserve 7787 the buffer content. 7788 7789 Also note that the layout of the ring buffer data may change between 7790 kernel versions where the validator will fail and reset the ring buffer 7791 if the layout is not the same as the previous kernel. 7792 7793 If the ring buffer is used for persistent bootups and has events enabled, 7794 it is recommend to disable tracing so that events from a previous boot do not 7795 mix with events of the current boot (unless you are debugging a random crash 7796 at boot up). 7797 7798 reserve_mem=12M:4096:trace trace_instance=boot_map^traceoff^traceprintk@trace,sched,irq 7799 7800 Note, saving the trace buffer across reboots does require that the system 7801 is set up to not wipe memory. For instance, CONFIG_RESET_ATTACK_MITIGATION 7802 can force a memory reset on boot which will clear any trace that was stored. 7803 This is just one of many ways that can clear memory. Make sure your system 7804 keeps the content of memory across reboots before relying on this option. 7805 7806 NB: Both the mapped address and size must be page aligned for the architecture. 7807 7808 See also Documentation/trace/debugging.rst 7809 7810 7811 trace_options=[option-list] 7812 [FTRACE] Enable or disable tracer options at boot. 7813 The option-list is a comma delimited list of options 7814 that can be enabled or disabled just as if you were 7815 to echo the option name into 7816 7817 /sys/kernel/tracing/trace_options 7818 7819 For example, to enable stacktrace option (to dump the 7820 stack trace of each event), add to the command line: 7821 7822 trace_options=stacktrace 7823 7824 See also Documentation/trace/ftrace.rst "trace options" 7825 section. 7826 7827 trace_trigger=[trigger-list] 7828 [FTRACE] Add an event trigger on specific events. 7829 Set a trigger on top of a specific event, with an optional 7830 filter. 7831 7832 The format is "trace_trigger=<event>.<trigger>[ if <filter>],..." 7833 Where more than one trigger may be specified that are comma delimited. 7834 7835 For example: 7836 7837 trace_trigger="sched_switch.stacktrace if prev_state == 2" 7838 7839 The above will enable the "stacktrace" trigger on the "sched_switch" 7840 event but only trigger it if the "prev_state" of the "sched_switch" 7841 event is "2" (TASK_UNINTERRUPTIBLE). 7842 7843 See also "Event triggers" in Documentation/trace/events.rst 7844 7845 7846 traceoff_after_boot 7847 [FTRACE] Sometimes tracing is used to debug issues 7848 during the boot process. Since the trace buffer has a 7849 limited amount of storage, it may be prudent to 7850 disable tracing after the boot is finished, otherwise 7851 the critical information may be overwritten. With this 7852 option, the main tracing buffer will be turned off at 7853 the end of the boot process. 7854 7855 traceoff_on_warning 7856 [FTRACE] enable this option to disable tracing when a 7857 warning is hit. This turns off "tracing_on". Tracing can 7858 be enabled again by echoing '1' into the "tracing_on" 7859 file located in /sys/kernel/tracing/ 7860 7861 This option is useful, as it disables the trace before 7862 the WARNING dump is called, which prevents the trace to 7863 be filled with content caused by the warning output. 7864 7865 This option can also be set at run time via the sysctl 7866 option: kernel/traceoff_on_warning 7867 7868 transparent_hugepage= 7869 [KNL] 7870 Format: [always|madvise|never] 7871 Can be used to control the default behavior of the system 7872 with respect to transparent hugepages. 7873 See Documentation/admin-guide/mm/transhuge.rst 7874 for more details. 7875 7876 transparent_hugepage_shmem= [KNL] 7877 Format: [always|within_size|advise|never|deny|force] 7878 Can be used to control the hugepage allocation policy for 7879 the internal shmem mount. 7880 See Documentation/admin-guide/mm/transhuge.rst 7881 for more details. 7882 7883 transparent_hugepage_tmpfs= [KNL] 7884 Format: [always|within_size|advise|never] 7885 Can be used to control the default hugepage allocation policy 7886 for the tmpfs mount. 7887 See Documentation/admin-guide/mm/transhuge.rst 7888 for more details. 7889 7890 trusted.source= [KEYS] 7891 Format: <string> 7892 This parameter identifies the trust source as a backend 7893 for trusted keys implementation. Supported trust 7894 sources: 7895 - "tpm" 7896 - "tee" 7897 - "caam" 7898 - "dcp" 7899 - "pkwm" 7900 If not specified then it defaults to iterating through 7901 the trust source list starting with TPM and assigns the 7902 first trust source as a backend which is initialized 7903 successfully during iteration. 7904 7905 trusted.debug= [KEYS] 7906 Format: <bool> 7907 Enable trusted keys debug traces at runtime when 7908 CONFIG_TRUSTED_KEYS_DEBUG=y. 7909 7910 To make the traces visible after enabling the option, 7911 use trusted.dyndbg='+p' as needed. By convention, 7912 the subsystem uses pr_debug() for these traces. 7913 7914 SAFETY: The traces can leak sensitive data, so be 7915 cautious before enabling this. They remain inactive 7916 unless this parameter is set this option to a true 7917 value. 7918 7919 Default: false 7920 7921 trusted.rng= [KEYS] 7922 Format: <string> 7923 The RNG used to generate key material for trusted keys. 7924 Can be one of: 7925 - "kernel" 7926 - the same value as trusted.source: "tpm" or "tee" 7927 - "default" 7928 If not specified, "default" is used. In this case, 7929 the RNG's choice is left to each individual trust source. 7930 7931 trusted.dcp_use_otp_key 7932 This is intended to be used in combination with 7933 trusted.source=dcp and will select the DCP OTP key 7934 instead of the DCP UNIQUE key blob encryption. 7935 7936 trusted.dcp_skip_zk_test 7937 This is intended to be used in combination with 7938 trusted.source=dcp and will disable the check if the 7939 blob key is all zeros. This is helpful for situations where 7940 having this key zero'ed is acceptable. E.g. in testing 7941 scenarios. 7942 7943 tsa= [X86] Control mitigation for Transient Scheduler 7944 Attacks on AMD CPUs. Search the following in your 7945 favourite search engine for more details: 7946 7947 "Technical guidance for mitigating transient scheduler 7948 attacks". 7949 7950 off - disable the mitigation 7951 on - enable the mitigation (default) 7952 user - mitigate only user/kernel transitions 7953 vm - mitigate only guest/host transitions 7954 7955 7956 tsc= Disable clocksource stability checks for TSC. 7957 Format: <string> 7958 [x86] reliable: mark tsc clocksource as reliable, this 7959 disables clocksource verification at runtime, as well 7960 as the stability checks done at bootup. Used to enable 7961 high-resolution timer mode on older hardware, and in 7962 virtualized environment. 7963 [x86] noirqtime: Do not use TSC to do irq accounting. 7964 Used to run time disable IRQ_TIME_ACCOUNTING on any 7965 platforms where RDTSC is slow and this accounting 7966 can add overhead. 7967 [x86] unstable: mark the TSC clocksource as unstable, this 7968 marks the TSC unconditionally unstable at bootup and 7969 avoids any further wobbles once the TSC watchdog notices. 7970 [x86] nowatchdog: disable clocksource watchdog. Used 7971 in situations with strict latency requirements (where 7972 interruptions from clocksource watchdog are not 7973 acceptable). 7974 [x86] recalibrate: force recalibration against a HW timer 7975 (HPET or PM timer) on systems whose TSC frequency was 7976 obtained from HW or FW using either an MSR or CPUID(0x15). 7977 Warn if the difference is more than 500 ppm. 7978 [x86] watchdog: Enforce the clocksource watchdog on TSC 7979 7980 tsc_early_khz= [X86,EARLY] Skip early TSC calibration and use the given 7981 value instead. Useful when the early TSC frequency discovery 7982 procedure is not reliable, such as on overclocked systems 7983 with CPUID.16h support and partial CPUID.15h support. 7984 Format: <unsigned int> 7985 7986 tsx= [X86] Control Transactional Synchronization 7987 Extensions (TSX) feature in Intel processors that 7988 support TSX control. 7989 7990 This parameter controls the TSX feature. The options are: 7991 7992 on - Enable TSX on the system. Although there are 7993 mitigations for all known security vulnerabilities, 7994 TSX has been known to be an accelerator for 7995 several previous speculation-related CVEs, and 7996 so there may be unknown security risks associated 7997 with leaving it enabled. 7998 7999 off - Disable TSX on the system. (Note that this 8000 option takes effect only on newer CPUs which are 8001 not vulnerable to MDS, i.e., have 8002 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 8003 the new IA32_TSX_CTRL MSR through a microcode 8004 update. This new MSR allows for the reliable 8005 deactivation of the TSX functionality.) 8006 8007 auto - Disable TSX if X86_BUG_TAA is present, 8008 otherwise enable TSX on the system. 8009 8010 Not specifying this option is equivalent to tsx=off. 8011 8012 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 8013 for more details. 8014 8015 tsx_async_abort= [X86,INTEL,EARLY] Control mitigation for the TSX Async 8016 Abort (TAA) vulnerability. 8017 8018 Similar to Micro-architectural Data Sampling (MDS) 8019 certain CPUs that support Transactional 8020 Synchronization Extensions (TSX) are vulnerable to an 8021 exploit against CPU internal buffers which can forward 8022 information to a disclosure gadget under certain 8023 conditions. 8024 8025 In vulnerable processors, the speculatively forwarded 8026 data can be used in a cache side channel attack, to 8027 access data to which the attacker does not have direct 8028 access. 8029 8030 This parameter controls the TAA mitigation. The 8031 options are: 8032 8033 full - Enable TAA mitigation on vulnerable CPUs 8034 if TSX is enabled. 8035 8036 full,nosmt - Enable TAA mitigation and disable SMT on 8037 vulnerable CPUs. If TSX is disabled, SMT 8038 is not disabled because CPU is not 8039 vulnerable to cross-thread TAA attacks. 8040 off - Unconditionally disable TAA mitigation 8041 8042 On MDS-affected machines, tsx_async_abort=off can be 8043 prevented by an active MDS mitigation as both vulnerabilities 8044 are mitigated with the same mechanism so in order to disable 8045 this mitigation, you need to specify mds=off too. 8046 8047 Not specifying this option is equivalent to 8048 tsx_async_abort=full. On CPUs which are MDS affected 8049 and deploy MDS mitigation, TAA mitigation is not 8050 required and doesn't provide any additional 8051 mitigation. 8052 8053 For details see: 8054 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 8055 8056 turbografx.map[2|3]= [HW,JOY] 8057 TurboGraFX parallel port interface 8058 Format: 8059 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 8060 See also Documentation/input/devices/joystick-parport.rst 8061 8062 udbg-immortal [PPC] When debugging early kernel crashes that 8063 happen after console_init() and before a proper 8064 console driver takes over, this boot options might 8065 help "seeing" what's going on. 8066 8067 uhash_entries= [KNL,NET] 8068 Set number of hash buckets for UDP/UDP-Lite connections 8069 8070 uhci-hcd.ignore_oc= 8071 [USB] Ignore overcurrent events (default N). 8072 Some badly-designed motherboards generate lots of 8073 bogus events, for ports that aren't wired to 8074 anything. Set this parameter to avoid log spamming. 8075 Note that genuine overcurrent events won't be 8076 reported either. 8077 8078 unaligned_scalar_speed= 8079 [RISCV] 8080 Format: {slow | fast | unsupported} 8081 Allow skipping scalar unaligned access speed tests. This 8082 is useful for testing alternative code paths and to skip 8083 the tests in environments where they run too slowly. All 8084 CPUs must have the same scalar unaligned access speed. 8085 8086 unaligned_vector_speed= 8087 [RISCV] 8088 Format: {slow | fast | unsupported} 8089 Allow skipping vector 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 vector unaligned access speed. 8093 8094 unknown_nmi_panic 8095 [X86] Cause panic on unknown NMI. 8096 8097 unwind_debug [X86-64,EARLY] 8098 Enable unwinder debug output. This can be 8099 useful for debugging certain unwinder error 8100 conditions, including corrupt stacks and 8101 bad/missing unwinder metadata. 8102 8103 usbcore.authorized_default= 8104 [USB] Default USB device authorization: 8105 (default -1 = authorized (same as 1), 8106 0 = not authorized, 1 = authorized, 2 = authorized 8107 if device connected to internal port) 8108 8109 usbcore.autosuspend= 8110 [USB] The autosuspend time delay (in seconds) used 8111 for newly-detected USB devices (default 2). This 8112 is the time required before an idle device will be 8113 autosuspended. Devices for which the delay is set 8114 to a negative value won't be autosuspended at all. 8115 8116 usbcore.usbfs_snoop= 8117 [USB] Set to log all usbfs traffic (default 0 = off). 8118 8119 usbcore.usbfs_snoop_max= 8120 [USB] Maximum number of bytes to snoop in each URB 8121 (default = 65536). 8122 8123 usbcore.blinkenlights= 8124 [USB] Set to cycle leds on hubs (default 0 = off). 8125 8126 usbcore.old_scheme_first= 8127 [USB] Start with the old device initialization 8128 scheme (default 0 = off). 8129 8130 usbcore.usbfs_memory_mb= 8131 [USB] Memory limit (in MB) for buffers allocated by 8132 usbfs (default = 16, 0 = max = 2047). 8133 8134 usbcore.use_both_schemes= 8135 [USB] Try the other device initialization scheme 8136 if the first one fails (default 1 = enabled). 8137 8138 usbcore.initial_descriptor_timeout= 8139 [USB] Specifies timeout for the initial 64-byte 8140 USB_REQ_GET_DESCRIPTOR request in milliseconds 8141 (default 5000 = 5.0 seconds). 8142 8143 usbcore.nousb [USB] Disable the USB subsystem 8144 8145 usbcore.quirks= 8146 [USB] A list of quirk entries to augment the built-in 8147 usb core quirk list. List entries are separated by 8148 commas. Each entry has the form 8149 VendorID:ProductID:Flags. The IDs are 4-digit hex 8150 numbers and Flags is a set of letters. Each letter 8151 will change the built-in quirk; setting it if it is 8152 clear and clearing it if it is set. The letters have 8153 the following meanings: 8154 a = USB_QUIRK_STRING_FETCH_255 (string 8155 descriptors must not be fetched using 8156 a 255-byte read); 8157 b = USB_QUIRK_RESET_RESUME (device can't resume 8158 correctly so reset it instead); 8159 c = USB_QUIRK_NO_SET_INTF (device can't handle 8160 Set-Interface requests); 8161 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 8162 handle its Configuration or Interface 8163 strings); 8164 e = USB_QUIRK_RESET (device can't be reset 8165 (e.g morph devices), don't use reset); 8166 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 8167 more interface descriptions than the 8168 bNumInterfaces count, and can't handle 8169 talking to these interfaces); 8170 g = USB_QUIRK_DELAY_INIT (device needs a pause 8171 during initialization, after we read 8172 the device descriptor); 8173 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 8174 high speed and super speed interrupt 8175 endpoints, the USB 2.0 and USB 3.0 spec 8176 require the interval in microframes (1 8177 microframe = 125 microseconds) to be 8178 calculated as interval = 2 ^ 8179 (bInterval-1). 8180 Devices with this quirk report their 8181 bInterval as the result of this 8182 calculation instead of the exponent 8183 variable used in the calculation); 8184 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 8185 handle device_qualifier descriptor 8186 requests); 8187 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 8188 generates spurious wakeup, ignore 8189 remote wakeup capability); 8190 k = USB_QUIRK_NO_LPM (device can't handle Link 8191 Power Management); 8192 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 8193 (Device reports its bInterval as linear 8194 frames instead of the USB 2.0 8195 calculation); 8196 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 8197 to be disconnected before suspend to 8198 prevent spurious wakeup); 8199 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 8200 pause after every control message); 8201 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 8202 delay after resetting its port); 8203 p = USB_QUIRK_SHORT_SET_ADDRESS_REQ_TIMEOUT 8204 (Reduce timeout of the SET_ADDRESS 8205 request from 5000 ms to 500 ms); 8206 q = USB_QUIRK_FORCE_ONE_CONFIG (Device 8207 claims zero configurations, 8208 forcing to 1); 8209 r = USB_QUIRK_WINDOWS_CONFIG_REQ_SIZE (Device 8210 fails during initialization when asked for 8211 9-bytes configuration descriptor request. 8212 Ask for 255-bytes request instead to mirror 8213 Windows' behavior); 8214 Example: quirks=0781:5580:bk,0a5c:5834:gij 8215 8216 usbhid.mousepoll= 8217 [USBHID] The interval which mice are to be polled at. 8218 8219 usbhid.jspoll= 8220 [USBHID] The interval which joysticks are to be polled at. 8221 8222 usbhid.kbpoll= 8223 [USBHID] The interval which keyboards are to be polled at. 8224 8225 usb-storage.delay_use= 8226 [UMS] The delay in seconds before a new device is 8227 scanned for Logical Units (default 1). 8228 Optionally the delay in milliseconds if the value has 8229 suffix with "ms". 8230 Example: delay_use=2567ms 8231 8232 usb-storage.quirks= 8233 [UMS] A list of quirks entries to supplement or 8234 override the built-in unusual_devs list. List 8235 entries are separated by commas. Each entry has 8236 the form VID:PID:Flags where VID and PID are Vendor 8237 and Product ID values (4-digit hex numbers) and 8238 Flags is a set of characters, each corresponding 8239 to a common usb-storage quirk flag as follows: 8240 a = SANE_SENSE (collect more than 18 bytes 8241 of sense data, not on uas); 8242 b = BAD_SENSE (don't collect more than 18 8243 bytes of sense data, not on uas); 8244 c = FIX_CAPACITY (decrease the reported 8245 device capacity by one sector); 8246 d = NO_READ_DISC_INFO (don't use 8247 READ_DISC_INFO command, not on uas); 8248 e = NO_READ_CAPACITY_16 (don't use 8249 READ_CAPACITY_16 command); 8250 f = NO_REPORT_OPCODES (don't use report opcodes 8251 command, uas only); 8252 g = MAX_SECTORS_240 (don't transfer more than 8253 240 sectors at a time, uas only); 8254 h = CAPACITY_HEURISTICS (decrease the 8255 reported device capacity by one 8256 sector if the number is odd); 8257 i = IGNORE_DEVICE (don't bind to this 8258 device); 8259 j = NO_REPORT_LUNS (don't use report luns 8260 command, uas only); 8261 k = NO_SAME (do not use WRITE_SAME, uas only) 8262 l = NOT_LOCKABLE (don't try to lock and 8263 unlock ejectable media, not on uas); 8264 m = MAX_SECTORS_64 (don't transfer more 8265 than 64 sectors = 32 KB at a time, 8266 not on uas); 8267 n = INITIAL_READ10 (force a retry of the 8268 initial READ(10) command, not on uas); 8269 o = CAPACITY_OK (accept the capacity 8270 reported by the device, not on uas); 8271 p = WRITE_CACHE (the device cache is ON 8272 by default, not on uas); 8273 r = IGNORE_RESIDUE (the device reports 8274 bogus residue values, not on uas); 8275 s = SINGLE_LUN (the device has only one 8276 Logical Unit); 8277 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 8278 commands, uas only); 8279 u = IGNORE_UAS (don't bind to the uas driver); 8280 w = NO_WP_DETECT (don't test whether the 8281 medium is write-protected). 8282 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 8283 even if the device claims no cache, 8284 not on uas) 8285 Example: quirks=0419:aaf5:rl,0421:0433:rc 8286 8287 user_debug= [KNL,ARM] 8288 Format: <int> 8289 See arch/arm/Kconfig.debug help text. 8290 1 - undefined instruction events 8291 2 - system calls 8292 4 - invalid data aborts 8293 8 - SIGSEGV faults 8294 16 - SIGBUS faults 8295 Example: user_debug=31 8296 8297 vdso= [X86,SH,SPARC] 8298 On X86_32, this is an alias for vdso32=. Otherwise: 8299 8300 vdso=1: enable VDSO (the default) 8301 vdso=0: disable VDSO mapping 8302 8303 vdso32= [X86] Control the 32-bit vDSO 8304 vdso32=1: enable 32-bit VDSO 8305 vdso32=0 or vdso32=2: disable 32-bit VDSO 8306 8307 See the help text for CONFIG_COMPAT_VDSO for more 8308 details. If CONFIG_COMPAT_VDSO is set, the default is 8309 vdso32=0; otherwise, the default is vdso32=1. 8310 8311 For compatibility with older kernels, vdso32=2 is an 8312 alias for vdso32=0. 8313 8314 Try vdso32=0 if you encounter an error that says: 8315 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 8316 8317 video= [FB,EARLY] Frame buffer configuration 8318 See Documentation/fb/modedb.rst. 8319 8320 video.brightness_switch_enabled= [ACPI] 8321 Format: [0|1] 8322 If set to 1, on receiving an ACPI notify event 8323 generated by hotkey, video driver will adjust brightness 8324 level and then send out the event to user space through 8325 the allocated input device. If set to 0, video driver 8326 will only send out the event without touching backlight 8327 brightness level. 8328 default: 1 8329 8330 virtio_mmio.device= 8331 [VMMIO] Memory mapped virtio (platform) device. 8332 8333 <size>@<baseaddr>:<irq>[:<id>] 8334 where: 8335 <size> := size (can use standard suffixes 8336 like K, M and G) 8337 <baseaddr> := physical base address 8338 <irq> := interrupt number (as passed to 8339 request_irq()) 8340 <id> := (optional) platform device id 8341 example: 8342 virtio_mmio.device=1K@0x100b0000:48:7 8343 8344 Can be used multiple times for multiple devices. 8345 8346 vga= [BOOT,X86-32] Select a particular video mode 8347 See Documentation/arch/x86/boot.rst and 8348 Documentation/admin-guide/svga.rst. 8349 Use vga=ask for menu. 8350 This is actually a boot loader parameter; the value is 8351 passed to the kernel using a special protocol. 8352 8353 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 8354 May slow down system boot speed, especially when 8355 enabled on systems with a large amount of memory. 8356 All options are enabled by default, and this 8357 interface is meant to allow for selectively 8358 enabling or disabling specific virtual memory 8359 debugging features. 8360 8361 Available options are: 8362 P Enable page structure init time poisoning 8363 - Disable all of the above options 8364 8365 vmalloc=nn[KMG] [KNL,BOOT,EARLY] Forces the vmalloc area to have an 8366 exact size of <nn>. This can be used to increase 8367 the minimum size (128MB on x86, arm32 platforms). 8368 It can also be used to decrease the size and leave more room 8369 for directly mapped kernel RAM. Note that this parameter does 8370 not exist on many other platforms (including arm64, alpha, 8371 loongarch, arc, csky, hexagon, microblaze, mips, nios2, openrisc, 8372 parisc, m64k, powerpc, riscv, sh, um, xtensa, s390, sparc). 8373 8374 vmcp_cma=nn[MG] [KNL,S390,EARLY] 8375 Sets the memory size reserved for contiguous memory 8376 allocations for the vmcp device driver. 8377 8378 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 8379 Format: <command> 8380 8381 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 8382 Format: <command> 8383 8384 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 8385 Format: <command> 8386 8387 vmscape= [X86] Controls mitigation for VMscape attacks. 8388 VMscape attacks can leak information from a userspace 8389 hypervisor to a guest via speculative side-channels. 8390 8391 off - disable the mitigation 8392 ibpb - use Indirect Branch Prediction Barrier 8393 (IBPB) mitigation (default) 8394 force - force vulnerability detection even on 8395 unaffected processors 8396 8397 vsyscall= [X86-64,EARLY] 8398 Controls the behavior of vsyscalls (i.e. calls to 8399 fixed addresses of 0xffffffffff600x00 from legacy 8400 code). Most statically-linked binaries and older 8401 versions of glibc use these calls. Because these 8402 functions are at fixed addresses, they make nice 8403 targets for exploits that can control RIP. 8404 8405 emulate Vsyscalls turn into traps and are emulated 8406 reasonably safely. The vsyscall page is 8407 readable. This disables the Linear 8408 Address Space Separation (LASS) security 8409 feature and makes the system less secure. 8410 8411 xonly [default] Vsyscalls turn into traps and are 8412 emulated reasonably safely. The vsyscall 8413 page is not readable. 8414 8415 none Vsyscalls don't work at all. This makes 8416 them quite hard to use for exploits but 8417 might break your system. 8418 8419 vt.color= [VT] Default text color. 8420 Format: 0xYX, X = foreground, Y = background. 8421 Default: 0x07 = light gray on black. 8422 8423 vt.cur_default= [VT] Default cursor shape. 8424 Format: 0xCCBBAA, where AA, BB, and CC are the same as 8425 the parameters of the <Esc>[?A;B;Cc escape sequence; 8426 see vga-softcursor.rst. Default: 2 = underline. 8427 8428 vt.default_blu= [VT] 8429 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 8430 Change the default blue palette of the console. 8431 This is a 16-member array composed of values 8432 ranging from 0-255. 8433 8434 vt.default_grn= [VT] 8435 Format: <green0>,<green1>,<green2>,...,<green15> 8436 Change the default green palette of the console. 8437 This is a 16-member array composed of values 8438 ranging from 0-255. 8439 8440 vt.default_red= [VT] 8441 Format: <red0>,<red1>,<red2>,...,<red15> 8442 Change the default red palette of the console. 8443 This is a 16-member array composed of values 8444 ranging from 0-255. 8445 8446 vt.default_utf8= 8447 [VT] 8448 Format=<0|1> 8449 Set system-wide default UTF-8 mode for all tty's. 8450 Default is 1, i.e. UTF-8 mode is enabled for all 8451 newly opened terminals. 8452 8453 vt.global_cursor_default= 8454 [VT] 8455 Format=<-1|0|1> 8456 Set system-wide default for whether a cursor 8457 is shown on new VTs. Default is -1, 8458 i.e. cursors will be created by default unless 8459 overridden by individual drivers. 0 will hide 8460 cursors, 1 will display them. 8461 8462 vt.italic= [VT] Default color for italic text; 0-15. 8463 Default: 2 = green. 8464 8465 vt.underline= [VT] Default color for underlined text; 0-15. 8466 Default: 3 = cyan. 8467 8468 watchdog timers [HW,WDT] For information on watchdog timers, 8469 see Documentation/watchdog/watchdog-parameters.rst 8470 or other driver-specific files in the 8471 Documentation/watchdog/ directory. 8472 8473 watchdog_thresh= 8474 [KNL] 8475 Set the hard lockup detector stall duration 8476 threshold in seconds. The soft lockup detector 8477 threshold is set to twice the value. A value of 0 8478 disables both lockup detectors. Default is 10 8479 seconds. 8480 8481 workqueue.unbound_cpus= 8482 [KNL,SMP] Specify to constrain one or some CPUs 8483 to use in unbound workqueues. 8484 Format: <cpu-list> 8485 By default, all online CPUs are available for 8486 unbound workqueues. 8487 8488 workqueue.watchdog_thresh= 8489 If CONFIG_WQ_WATCHDOG is configured, workqueue can 8490 warn stall conditions and dump internal state to 8491 help debugging. 0 disables workqueue stall 8492 detection; otherwise, it's the stall threshold 8493 duration in seconds. The default value is 30 and 8494 it can be updated at runtime by writing to the 8495 corresponding sysfs file. 8496 8497 workqueue.panic_on_stall=<uint> 8498 Panic when workqueue stall is detected by 8499 CONFIG_WQ_WATCHDOG. It sets the number times of the 8500 stall to trigger panic. 8501 8502 The default is set by CONFIG_BOOTPARAM_WQ_STALL_PANIC, 8503 which is 0 (disabled) if not configured. 8504 8505 workqueue.panic_on_stall_time=<uint> 8506 Panic when a workqueue stall has been continuous for 8507 the specified number of seconds. Unlike panic_on_stall 8508 which counts accumulated stall events, this triggers 8509 based on the duration of a single continuous stall. 8510 8511 The default is 0, which disables the time-based panic. 8512 8513 workqueue.cpu_intensive_thresh_us= 8514 Per-cpu work items which run for longer than this 8515 threshold are automatically considered CPU intensive 8516 and excluded from concurrency management to prevent 8517 them from noticeably delaying other per-cpu work 8518 items. Default is 10000 (10ms). 8519 8520 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8521 will report the work functions which violate this 8522 threshold repeatedly. They are likely good 8523 candidates for using WQ_UNBOUND workqueues instead. 8524 8525 workqueue.cpu_intensive_warning_thresh=<uint> 8526 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8527 will report the work functions which violate the 8528 intensive_threshold_us repeatedly. In order to prevent 8529 spurious warnings, start printing only after a work 8530 function has violated this threshold number of times. 8531 8532 The default is 4 times. 0 disables the warning. 8533 8534 workqueue.power_efficient 8535 Per-cpu workqueues are generally preferred because 8536 they show better performance thanks to cache 8537 locality; unfortunately, per-cpu workqueues tend to 8538 be more power hungry than unbound workqueues. 8539 8540 Enabling this makes the per-cpu workqueues which 8541 were observed to contribute significantly to power 8542 consumption unbound, leading to measurably lower 8543 power usage at the cost of small performance 8544 overhead. 8545 8546 The default value of this parameter is determined by 8547 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 8548 8549 workqueue.default_affinity_scope= 8550 Select the default affinity scope to use for unbound 8551 workqueues. Can be one of "cpu", "smt", "cache", 8552 "cache_shard", "numa" and "system". Default is 8553 "cache_shard". 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