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 PNP Plug & Play support is enabled. 51 PPC PowerPC architecture is enabled. 52 PPT Parallel port support is enabled. 53 PS2 Appropriate PS/2 support is enabled. 54 PV_OPS A paravirtualized kernel is enabled. 55 RAM RAM disk support is enabled. 56 RDT Intel Resource Director Technology. 57 RISCV RISCV architecture is enabled. 58 S390 S390 architecture is enabled. 59 SCSI Appropriate SCSI support is enabled. 60 A lot of drivers have their options described inside 61 the Documentation/scsi/ sub-directory. 62 SDW SoundWire support is enabled. 63 SECURITY Different security models are enabled. 64 SELINUX SELinux support is enabled. 65 SERIAL Serial support is enabled. 66 SH SuperH architecture is enabled. 67 SMP The kernel is an SMP kernel. 68 SPARC Sparc architecture is enabled. 69 SUSPEND System suspend states are enabled. 70 SWSUSP Software suspend (hibernation) is enabled. 71 TPM TPM drivers are enabled. 72 UMS USB Mass Storage support is enabled. 73 USB USB support is enabled. 74 NVME NVMe support is enabled 75 USBHID USB Human Interface Device support is enabled. 76 V4L Video For Linux support is enabled. 77 VGA The VGA console has been enabled. 78 VMMIO Driver for memory mapped virtio devices is enabled. 79 VT Virtual terminal support is enabled. 80 WDT Watchdog support is enabled. 81 X86-32 X86-32, aka i386 architecture is enabled. 82 X86-64 X86-64 architecture is enabled. 83 X86 Either 32-bit or 64-bit x86 (same as X86-32+X86-64) 84 X86_UV SGI UV support is enabled. 85 XEN Xen support is enabled 86 XTENSA xtensa architecture is enabled. 87 88In addition, the following text indicates that the option 89 90 BOOT Is a boot loader parameter. 91 BUGS= Relates to possible processor bugs on the said processor. 92 KNL Is a kernel start-up parameter. 93 94 95Kernel parameters 96 97 accept_memory= [MM] 98 Format: { eager | lazy } 99 default: lazy 100 By default, unaccepted memory is accepted lazily to 101 avoid prolonged boot times. The lazy option will add 102 some runtime overhead until all memory is eventually 103 accepted. In most cases the overhead is negligible. 104 For some workloads or for debugging purposes 105 accept_memory=eager can be used to accept all memory 106 at once during boot. 107 108 acpi= [HW,ACPI,X86,ARM64,RISCV64,EARLY] 109 Advanced Configuration and Power Interface 110 Format: { force | on | off | strict | noirq | rsdt | 111 copy_dsdt | nospcr } 112 force -- enable ACPI if default was off 113 on -- enable ACPI but allow fallback to DT [arm64,riscv64] 114 off -- disable ACPI if default was on 115 noirq -- do not use ACPI for IRQ routing 116 strict -- Be less tolerant of platforms that are not 117 strictly ACPI specification compliant. 118 rsdt -- prefer RSDT over (default) XSDT 119 copy_dsdt -- copy DSDT to memory 120 nocmcff -- Disable firmware first mode for corrected 121 errors. This disables parsing the HEST CMC error 122 source to check if firmware has set the FF flag. This 123 may result in duplicate corrected error reports. 124 nospcr -- disable console in ACPI SPCR table as 125 default _serial_ console on ARM64 126 spcr -- enable console in ACPI SPCR table as 127 default _serial_ console on x86 128 For ARM64, ONLY "acpi=off", "acpi=on", "acpi=force" or 129 "acpi=nospcr" are available 130 For RISCV64, ONLY "acpi=off", "acpi=on" or "acpi=force" 131 are available 132 133 See also Documentation/power/runtime_pm.rst, pci=noacpi 134 135 acpi_apic_instance= [ACPI,IOAPIC,EARLY] 136 Format: <int> 137 2: use 2nd APIC table, if available 138 1,0: use 1st APIC table 139 default: 0 140 141 acpi_backlight= [HW,ACPI] 142 { vendor | video | native | none } 143 If set to vendor, prefer vendor-specific driver 144 (e.g. thinkpad_acpi, sony_acpi, etc.) instead 145 of the ACPI video.ko driver. 146 If set to video, use the ACPI video.ko driver. 147 If set to native, use the device's native backlight mode. 148 If set to none, disable the ACPI backlight interface. 149 150 acpi_force_32bit_fadt_addr [ACPI,EARLY] 151 force FADT to use 32 bit addresses rather than the 152 64 bit X_* addresses. Some firmware have broken 64 153 bit addresses for force ACPI ignore these and use 154 the older legacy 32 bit addresses. 155 156 acpica_no_return_repair [HW, ACPI] 157 Disable AML predefined validation mechanism 158 This mechanism can repair the evaluation result to make 159 the return objects more ACPI specification compliant. 160 This option is useful for developers to identify the 161 root cause of an AML interpreter issue when the issue 162 has something to do with the repair mechanism. 163 164 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG] 165 acpi.debug_level= [HW,ACPI,ACPI_DEBUG] 166 Format: <int> 167 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI 168 debug output. Bits in debug_layer correspond to a 169 _COMPONENT in an ACPI source file, e.g., 170 #define _COMPONENT ACPI_EVENTS 171 Bits in debug_level correspond to a level in 172 ACPI_DEBUG_PRINT statements, e.g., 173 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ... 174 The debug_level mask defaults to "info". See 175 Documentation/firmware-guide/acpi/debug.rst for more information about 176 debug layers and levels. 177 178 Enable processor driver info messages: 179 acpi.debug_layer=0x20000000 180 Enable AML "Debug" output, i.e., stores to the Debug 181 object while interpreting AML: 182 acpi.debug_layer=0xffffffff acpi.debug_level=0x2 183 Enable all messages related to ACPI hardware: 184 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff 185 186 Some values produce so much output that the system is 187 unusable. The "log_buf_len" parameter may be useful 188 if you need to capture more output. 189 190 acpi.poweroff_on_fatal= [ACPI] 191 {0 | 1} 192 Causes the system to poweroff when the ACPI bytecode signals 193 a fatal error. The default value of this setting is 1. 194 Overriding this value should only be done for diagnosing 195 ACPI firmware problems, as the system might behave erratically 196 after having encountered a fatal ACPI error. 197 198 acpi_enforce_resources= [ACPI] 199 { strict | lax | no } 200 Check for resource conflicts between native drivers 201 and ACPI OperationRegions (SystemIO and SystemMemory 202 only). IO ports and memory declared in ACPI might be 203 used by the ACPI subsystem in arbitrary AML code and 204 can interfere with legacy drivers. 205 strict (default): access to resources claimed by ACPI 206 is denied; legacy drivers trying to access reserved 207 resources will fail to bind to device using them. 208 lax: access to resources claimed by ACPI is allowed; 209 legacy drivers trying to access reserved resources 210 will bind successfully but a warning message is logged. 211 no: ACPI OperationRegions are not marked as reserved, 212 no further checks are performed. 213 214 acpi_force_table_verification [HW,ACPI,EARLY] 215 Enable table checksum verification during early stage. 216 By default, this is disabled due to x86 early mapping 217 size limitation. 218 219 acpi_irq_balance [HW,ACPI] 220 ACPI will balance active IRQs 221 default in APIC mode 222 223 acpi_irq_nobalance [HW,ACPI] 224 ACPI will not move active IRQs (default) 225 default in PIC mode 226 227 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 228 Format: <irq>,<irq>... 229 230 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 231 use by PCI 232 Format: <irq>,<irq>... 233 234 acpi_mask_gpe= [HW,ACPI] 235 Due to the existence of _Lxx/_Exx, some GPEs triggered 236 by unsupported hardware/firmware features can result in 237 GPE floodings that cannot be automatically disabled by 238 the GPE dispatcher. 239 This facility can be used to prevent such uncontrolled 240 GPE floodings. 241 Format: <byte> or <bitmap-list> 242 243 acpi_no_auto_serialize [HW,ACPI] 244 Disable auto-serialization of AML methods 245 AML control methods that contain the opcodes to create 246 named objects will be marked as "Serialized" by the 247 auto-serialization feature. 248 This feature is enabled by default. 249 This option allows to turn off the feature. 250 251 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 252 kernels. 253 254 acpi_no_static_ssdt [HW,ACPI,EARLY] 255 Disable installation of static SSDTs at early boot time 256 By default, SSDTs contained in the RSDT/XSDT will be 257 installed automatically and they will appear under 258 /sys/firmware/acpi/tables. 259 This option turns off this feature. 260 Note that specifying this option does not affect 261 dynamic table installation which will install SSDT 262 tables to /sys/firmware/acpi/tables/dynamic. 263 264 acpi_no_watchdog [HW,ACPI,WDT] 265 Ignore the ACPI-based watchdog interface (WDAT) and let 266 a native driver control the watchdog device instead. 267 268 acpi_rsdp= [ACPI,EFI,KEXEC,EARLY] 269 Pass the RSDP address to the kernel, mostly used 270 on machines running EFI runtime service to boot the 271 second kernel for kdump. 272 273 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 274 Format: To spoof as Windows 98: ="Microsoft Windows" 275 276 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 277 of 2 which is mandated by ACPI 6) as the supported ACPI 278 specification revision (when using this switch, it may 279 be necessary to carry out a cold reboot _twice_ in a 280 row to make it take effect on the platform firmware). 281 282 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 283 acpi_osi="string1" # add string1 284 acpi_osi="!string2" # remove string2 285 acpi_osi=!* # remove all strings 286 acpi_osi=! # disable all built-in OS vendor 287 strings 288 acpi_osi=!! # enable all built-in OS vendor 289 strings 290 acpi_osi= # disable all strings 291 292 'acpi_osi=!' can be used in combination with single or 293 multiple 'acpi_osi="string1"' to support specific OS 294 vendor string(s). Note that such command can only 295 affect the default state of the OS vendor strings, thus 296 it cannot affect the default state of the feature group 297 strings and the current state of the OS vendor strings, 298 specifying it multiple times through kernel command line 299 is meaningless. This command is useful when one do not 300 care about the state of the feature group strings which 301 should be controlled by the OSPM. 302 Examples: 303 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 304 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 305 can make '_OSI("Windows 2000")' TRUE. 306 307 'acpi_osi=' cannot be used in combination with other 308 'acpi_osi=' command lines, the _OSI method will not 309 exist in the ACPI namespace. NOTE that such command can 310 only affect the _OSI support state, thus specifying it 311 multiple times through kernel command line is also 312 meaningless. 313 Examples: 314 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 315 FALSE. 316 317 'acpi_osi=!*' can be used in combination with single or 318 multiple 'acpi_osi="string1"' to support specific 319 string(s). Note that such command can affect the 320 current state of both the OS vendor strings and the 321 feature group strings, thus specifying it multiple times 322 through kernel command line is meaningful. But it may 323 still not able to affect the final state of a string if 324 there are quirks related to this string. This command 325 is useful when one want to control the state of the 326 feature group strings to debug BIOS issues related to 327 the OSPM features. 328 Examples: 329 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 330 '_OSI("Module Device")' FALSE. 331 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 332 '_OSI("Module Device")' TRUE. 333 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 334 equivalent to 335 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 336 and 337 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 338 they all will make '_OSI("Windows 2000")' TRUE. 339 340 acpi_pm_good [X86] 341 Override the pmtimer bug detection: force the kernel 342 to assume that this machine's pmtimer latches its value 343 and always returns good values. 344 345 acpi_sci= [HW,ACPI,EARLY] ACPI System Control Interrupt trigger mode 346 Format: { level | edge | high | low } 347 348 acpi_skip_timer_override [HW,ACPI,EARLY] 349 Recognize and ignore IRQ0/pin2 Interrupt Override. 350 For broken nForce2 BIOS resulting in XT-PIC timer. 351 352 acpi_sleep= [HW,ACPI] Sleep options 353 Format: { s3_bios, s3_mode, s3_beep, s4_hwsig, 354 s4_nohwsig, old_ordering, nonvs, 355 sci_force_enable, nobl } 356 See Documentation/power/video.rst for information on 357 s3_bios and s3_mode. 358 s3_beep is for debugging; it makes the PC's speaker beep 359 as soon as the kernel's real-mode entry point is called. 360 s4_hwsig causes the kernel to check the ACPI hardware 361 signature during resume from hibernation, and gracefully 362 refuse to resume if it has changed. This complies with 363 the ACPI specification but not with reality, since 364 Windows does not do this and many laptops do change it 365 on docking. So the default behaviour is to allow resume 366 and simply warn when the signature changes, unless the 367 s4_hwsig option is enabled. 368 s4_nohwsig prevents ACPI hardware signature from being 369 used (or even warned about) during resume. 370 old_ordering causes the ACPI 1.0 ordering of the _PTS 371 control method, with respect to putting devices into 372 low power states, to be enforced (the ACPI 2.0 ordering 373 of _PTS is used by default). 374 nonvs prevents the kernel from saving/restoring the 375 ACPI NVS memory during suspend/hibernation and resume. 376 sci_force_enable causes the kernel to set SCI_EN directly 377 on resume from S1/S3 (which is against the ACPI spec, 378 but some broken systems don't work without it). 379 nobl causes the internal blacklist of systems known to 380 behave incorrectly in some ways with respect to system 381 suspend and resume to be ignored (use wisely). 382 383 acpi_use_timer_override [HW,ACPI,EARLY] 384 Use timer override. For some broken Nvidia NF5 boards 385 that require a timer override, but don't have HPET 386 387 add_efi_memmap [EFI,X86,EARLY] Include EFI memory map in 388 kernel's map of available physical RAM. 389 390 agp= [AGP] 391 { off | try_unsupported } 392 off: disable AGP support 393 try_unsupported: try to drive unsupported chipsets 394 (may crash computer or cause data corruption) 395 396 ALSA [HW,ALSA] 397 See Documentation/sound/alsa-configuration.rst 398 399 alignment= [KNL,ARM] 400 Allow the default userspace alignment fault handler 401 behaviour to be specified. Bit 0 enables warnings, 402 bit 1 enables fixups, and bit 2 sends a segfault. 403 404 align_va_addr= [X86-64] 405 Align virtual addresses by clearing slice [14:12] when 406 allocating a VMA at process creation time. This option 407 gives you up to 3% performance improvement on AMD F15h 408 machines (where it is enabled by default) for a 409 CPU-intensive style benchmark, and it can vary highly in 410 a microbenchmark depending on workload and compiler. 411 412 32: only for 32-bit processes 413 64: only for 64-bit processes 414 on: enable for both 32- and 64-bit processes 415 off: disable for both 32- and 64-bit processes 416 417 alloc_snapshot [FTRACE] 418 Allocate the ftrace snapshot buffer on boot up when the 419 main buffer is allocated. This is handy if debugging 420 and you need to use tracing_snapshot() on boot up, and 421 do not want to use tracing_snapshot_alloc() as it needs 422 to be done where GFP_KERNEL allocations are allowed. 423 424 allow_mismatched_32bit_el0 [ARM64,EARLY] 425 Allow execve() of 32-bit applications and setting of the 426 PER_LINUX32 personality on systems where only a strict 427 subset of the CPUs support 32-bit EL0. When this 428 parameter is present, the set of CPUs supporting 32-bit 429 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0 430 and hot-unplug operations may be restricted. 431 432 See Documentation/arch/arm64/asymmetric-32bit.rst for more 433 information. 434 435 amd_iommu= [HW,X86-64] 436 Pass parameters to the AMD IOMMU driver in the system. 437 Possible values are: 438 fullflush - Deprecated, equivalent to iommu.strict=1 439 off - do not initialize any AMD IOMMU found in 440 the system 441 force_isolation - Force device isolation for all 442 devices. The IOMMU driver is not 443 allowed anymore to lift isolation 444 requirements as needed. This option 445 does not override iommu=pt 446 force_enable - Force enable the IOMMU on platforms known 447 to be buggy with IOMMU enabled. Use this 448 option with care. 449 pgtbl_v1 - Use v1 page table for DMA-API (Default). 450 pgtbl_v2 - Use v2 page table for DMA-API. 451 irtcachedis - Disable Interrupt Remapping Table (IRT) caching. 452 nohugepages - Limit page-sizes used for v1 page-tables 453 to 4 KiB. 454 v2_pgsizes_only - Limit page-sizes used for v1 page-tables 455 to 4KiB/2Mib/1GiB. 456 457 458 amd_iommu_dump= [HW,X86-64] 459 Enable AMD IOMMU driver option to dump the ACPI table 460 for AMD IOMMU. With this option enabled, AMD IOMMU 461 driver will print ACPI tables for AMD IOMMU during 462 IOMMU initialization. 463 464 amd_iommu_intr= [HW,X86-64] 465 Specifies one of the following AMD IOMMU interrupt 466 remapping modes: 467 legacy - Use legacy interrupt remapping mode. 468 vapic - Use virtual APIC mode, which allows IOMMU 469 to inject interrupts directly into guest. 470 This mode requires kvm-amd.avic=1. 471 (Default when IOMMU HW support is present.) 472 473 amd_pstate= [X86,EARLY] 474 disable 475 Do not enable amd_pstate as the default 476 scaling driver for the supported processors 477 passive 478 Use amd_pstate with passive mode as a scaling driver. 479 In this mode autonomous selection is disabled. 480 Driver requests a desired performance level and platform 481 tries to match the same performance level if it is 482 satisfied by guaranteed performance level. 483 active 484 Use amd_pstate_epp driver instance as the scaling driver, 485 driver provides a hint to the hardware if software wants 486 to bias toward performance (0x0) or energy efficiency (0xff) 487 to the CPPC firmware. then CPPC power algorithm will 488 calculate the runtime workload and adjust the realtime cores 489 frequency. 490 guided 491 Activate guided autonomous mode. Driver requests minimum and 492 maximum performance level and the platform autonomously 493 selects a performance level in this range and appropriate 494 to the current workload. 495 496 amd_prefcore= 497 [X86] 498 disable 499 Disable amd-pstate preferred core. 500 501 amd_dynamic_epp= 502 [X86] 503 disable 504 Disable amd-pstate dynamic EPP. 505 enable 506 Enable amd-pstate dynamic EPP. 507 508 amijoy.map= [HW,JOY] Amiga joystick support 509 Map of devices attached to JOY0DAT and JOY1DAT 510 Format: <a>,<b> 511 See also Documentation/input/joydev/joystick.rst 512 513 analog.map= [HW,JOY] Analog joystick and gamepad support 514 Specifies type or capabilities of an analog joystick 515 connected to one of 16 gameports 516 Format: <type1>,<type2>,..<type16> 517 518 apc= [HW,SPARC] 519 Power management functions (SPARCstation-4/5 + deriv.) 520 Format: noidle 521 Disable APC CPU standby support. SPARCstation-Fox does 522 not play well with APC CPU idle - disable it if you have 523 APC and your system crashes randomly. 524 525 apic [APIC,X86-64] Use IO-APIC. Default. 526 527 apic= [APIC,X86,EARLY] Advanced Programmable Interrupt Controller 528 Change the output verbosity while booting 529 Format: { quiet (default) | verbose | debug } 530 Change the amount of debugging information output 531 when initialising the APIC and IO-APIC components. 532 533 apic_extnmi= [APIC,X86,EARLY] External NMI delivery setting 534 Format: { bsp (default) | all | none } 535 bsp: External NMI is delivered only to CPU 0 536 all: External NMIs are broadcast to all CPUs as a 537 backup of CPU 0 538 none: External NMI is masked for all CPUs. This is 539 useful so that a dump capture kernel won't be 540 shot down by NMI 541 542 apicpmtimer Do APIC timer calibration using the pmtimer. Implies 543 apicmaintimer. Useful when your PIT timer is totally 544 broken. 545 546 autoconf= [IPV6] 547 See Documentation/networking/ipv6.rst. 548 549 apm= [APM] Advanced Power Management 550 See header of arch/x86/kernel/apm_32.c. 551 552 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 553 Format: { "0" | "1" } 554 See security/apparmor/Kconfig help text 555 0 -- disable. 556 1 -- enable. 557 Default value is set via kernel config option. 558 559 arm64.no32bit_el0 [ARM64] Unconditionally disable the execution of 560 32 bit applications. 561 562 arm64.nobti [ARM64] Unconditionally disable Branch Target 563 Identification support 564 565 arm64.nogcs [ARM64] Unconditionally disable Guarded Control Stack 566 support 567 568 arm64.nomops [ARM64] Unconditionally disable Memory Copy and Memory 569 Set instructions support 570 571 arm64.nompam [ARM64] Unconditionally disable Memory Partitioning And 572 Monitoring support 573 574 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension 575 support 576 577 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 578 support 579 580 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix 581 Extension support 582 583 arm64.nosve [ARM64] Unconditionally disable Scalable Vector 584 Extension support 585 586 ataflop= [HW,M68k] 587 588 atarimouse= [HW,MOUSE] Atari Mouse 589 590 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 591 EzKey and similar keyboards 592 593 atkbd.reset= [HW] Reset keyboard during initialization 594 595 atkbd.set= [HW] Select keyboard code set 596 Format: <int> (2 = AT (default), 3 = PS/2) 597 598 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 599 keyboards 600 601 atkbd.softraw= [HW] Choose between synthetic and real raw mode 602 Format: <bool> (0 = real, 1 = synthetic (default)) 603 604 atkbd.softrepeat= [HW] 605 Use software keyboard repeat 606 607 audit= [KNL] Enable the audit sub-system 608 Format: { "0" | "1" | "off" | "on" } 609 0 | off - kernel audit is disabled and can not be 610 enabled until the next reboot 611 unset - kernel audit is initialized but disabled and 612 will be fully enabled by the userspace auditd. 613 1 | on - kernel audit is initialized and partially 614 enabled, storing at most audit_backlog_limit 615 messages in RAM until it is fully enabled by the 616 userspace auditd. 617 Default: unset 618 619 audit_backlog_limit= [KNL] Set the audit queue size limit. 620 Format: <int> (must be >=0) 621 Default: 64 622 623 bau= [X86_UV] Enable the BAU on SGI UV. The default 624 behavior is to disable the BAU (i.e. bau=0). 625 Format: { "0" | "1" } 626 0 - Disable the BAU. 627 1 - Enable the BAU. 628 unset - Disable the BAU. 629 630 bdev_allow_write_mounted= 631 Format: <bool> 632 Control the ability to open a mounted block device 633 for writing, i.e., allow / disallow writes that bypass 634 the FS. This was implemented as a means to prevent 635 fuzzers from crashing the kernel by overwriting the 636 metadata underneath a mounted FS without its awareness. 637 This also prevents destructive formatting of mounted 638 filesystems by naive storage tooling that don't use 639 O_EXCL. Default is Y and can be changed through the 640 Kconfig option CONFIG_BLK_DEV_WRITE_MOUNTED. 641 642 bert_disable [ACPI] 643 Disable BERT OS support on buggy BIOSes. 644 645 bgrt_disable [ACPI,X86,EARLY] 646 Disable BGRT to avoid flickering OEM logo. 647 648 blkdevparts= Manual partition parsing of block device(s) for 649 embedded devices based on command line input. 650 See Documentation/block/cmdline-partition.rst 651 652 boot_delay= [KNL,EARLY] 653 Milliseconds to delay each printk during boot. 654 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled, 655 and you may also have to specify "lpj=". Boot_delay 656 values larger than 10 seconds (10000) are assumed 657 erroneous and ignored. 658 Format: integer 659 660 bootconfig [KNL,EARLY] 661 Extended command line options can be added to an initrd 662 and this will cause the kernel to look for it. 663 664 See Documentation/admin-guide/bootconfig.rst 665 666 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 667 bttv.radio= Most important insmod options are available as 668 kernel args too. 669 bttv.pll= See Documentation/admin-guide/media/bttv.rst 670 bttv.tuner= 671 672 bulk_remove=off [PPC] This parameter disables the use of the pSeries 673 firmware feature for flushing multiple hpte entries 674 at a time. 675 676 c101= [NET] Moxa C101 synchronous serial card 677 678 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 679 Sometimes CPU hardware bugs make them report the cache 680 size incorrectly. The kernel will attempt work arounds 681 to fix known problems, but for some CPUs it is not 682 possible to determine what the correct size should be. 683 This option provides an override for these situations. 684 685 carrier_timeout= 686 [NET] Specifies amount of time (in seconds) that 687 the kernel should wait for a network carrier. By default 688 it waits 120 seconds. 689 690 ca_keys= [KEYS] This parameter identifies a specific key(s) on 691 the system trusted keyring to be used for certificate 692 trust validation. 693 format: { id:<keyid> | builtin } 694 695 cca= [MIPS,EARLY] Override the kernel pages' cache coherency 696 algorithm. Accepted values range from 0 to 7 697 inclusive. See arch/mips/include/asm/pgtable-bits.h 698 for platform specific values (SB1, Loongson3 and 699 others). 700 701 ccw_timeout_log [S390] 702 See Documentation/arch/s390/common_io.rst for details. 703 704 cfi= [X86-64] Set Control Flow Integrity checking features 705 when CONFIG_FINEIBT is enabled. 706 Format: feature[,feature...] 707 Default: auto 708 709 auto: Use FineIBT if IBT available, otherwise kCFI. 710 Under FineIBT, enable "paranoid" mode when 711 FRED is not available. 712 off: Turn off CFI checking. 713 kcfi: Use kCFI (disable FineIBT). 714 fineibt: Use FineIBT (even if IBT not available). 715 norand: Do not re-randomize CFI hashes. 716 paranoid: Add caller hash checking under FineIBT. 717 bhi: Enable register poisoning to stop speculation 718 across FineIBT. (Disabled by default.) 719 warn: Do not enforce CFI checking: warn only. 720 debug: Report CFI initialization details. 721 722 cgroup_disable= [KNL] Disable a particular controller or optional feature 723 Format: {name of the controller(s) or feature(s) to disable} 724 The effects of cgroup_disable=foo are: 725 - foo isn't auto-mounted if you mount all cgroups in 726 a single hierarchy 727 - foo isn't visible as an individually mountable 728 subsystem 729 - if foo is an optional feature then the feature is 730 disabled and corresponding cgroup files are not 731 created 732 {Currently only "memory" controller deal with this and 733 cut the overhead, others just disable the usage. So 734 only cgroup_disable=memory is actually worthy} 735 Specifying "pressure" disables per-cgroup pressure 736 stall information accounting feature 737 738 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 739 Format: { { controller | "all" | "named" } 740 [,{ controller | "all" | "named" }...] } 741 Like cgroup_disable, but only applies to cgroup v1; 742 the blacklisted controllers remain available in cgroup2. 743 "all" blacklists all controllers and "named" disables 744 named mounts. Specifying both "all" and "named" disables 745 all v1 hierarchies. 746 747 cgroup_v1_proc= [KNL] Show also missing controllers in /proc/cgroups 748 Format: { "true" | "false" } 749 /proc/cgroups lists only v1 controllers by default. 750 This compatibility option enables listing also v2 751 controllers (whose v1 code is not compiled!), so that 752 semi-legacy software can check this file to decide 753 about usage of v2 (sic) controllers. 754 755 cgroup_favordynmods= [KNL] Enable or Disable favordynmods. 756 Format: { "true" | "false" } 757 Defaults to the value of CONFIG_CGROUP_FAVOR_DYNMODS. 758 759 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 760 Format: <string> 761 nosocket -- Disable socket memory accounting. 762 nokmem -- Disable kernel memory accounting. 763 nobpf -- Disable BPF memory accounting. 764 765 check_pages= [MM,EARLY] Enable sanity checking of pages after 766 allocations / before freeing. This adds checks to catch 767 double-frees, use-after-frees, and other sources of 768 page corruption by inspecting page internals (flags, 769 mapcount/refcount, memcg_data, etc.). 770 Format: { "0" | "1" } 771 Default: 0 (1 if CONFIG_DEBUG_VM is set) 772 773 checkreqprot= [SELINUX] Set initial checkreqprot flag value. 774 Format: { "0" | "1" } 775 See security/selinux/Kconfig help text. 776 0 -- check protection applied by kernel (includes 777 any implied execute protection). 778 1 -- check protection requested by application. 779 Default value is set via a kernel config option. 780 Value can be changed at runtime via 781 /sys/fs/selinux/checkreqprot. 782 Setting checkreqprot to 1 is deprecated. 783 784 cio_ignore= [S390] 785 See Documentation/arch/s390/common_io.rst for details. 786 787 clk_ignore_unused 788 [CLK] 789 Prevents the clock framework from automatically gating 790 clocks that have not been explicitly enabled by a Linux 791 device driver but are enabled in hardware at reset or 792 by the bootloader/firmware. Note that this does not 793 force such clocks to be always-on nor does it reserve 794 those clocks in any way. This parameter is useful for 795 debug and development, but should not be needed on a 796 platform with proper driver support. For more 797 information, see Documentation/driver-api/clk.rst. 798 799 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 800 [Deprecated] 801 Forces specified clocksource (if available) to be used 802 when calculating gettimeofday(). If specified 803 clocksource is not available, it defaults to PIT. 804 Format: { pit | tsc | cyclone | pmtmr } 805 806 clocksource= Override the default clocksource 807 Format: <string> 808 Override the default clocksource and use the clocksource 809 with the name specified. 810 Some clocksource names to choose from, depending on 811 the platform: 812 [all] jiffies (this is the base, fallback clocksource) 813 [ACPI] acpi_pm 814 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 815 pxa_timer,timer3,32k_counter,timer0_1 816 [X86-32] pit,hpet,tsc; 817 scx200_hrt on Geode; cyclone on IBM x440 818 [MIPS] MIPS 819 [PARISC] cr16 820 [S390] tod 821 [SH] SuperH 822 [SPARC64] tick 823 [X86-64] hpet,tsc 824 825 clocksource.arm_arch_timer.evtstrm= 826 [ARM,ARM64,EARLY] 827 Format: <bool> 828 Enable/disable the eventstream feature of the ARM 829 architected timer so that code using WFE-based polling 830 loops can be debugged more effectively on production 831 systems. 832 833 clocksource.verify_n_cpus= [KNL] 834 Limit the number of CPUs checked for clocksources 835 marked with CLOCK_SOURCE_VERIFY_PERCPU that 836 are marked unstable due to excessive skew. 837 A negative value says to check all CPUs, while 838 zero says not to check any. Values larger than 839 nr_cpu_ids are silently truncated to nr_cpu_ids. 840 The actual CPUs are chosen randomly, with 841 no replacement if the same CPU is chosen twice. 842 843 clocksource-wdtest.holdoff= [KNL] 844 Set the time in seconds that the clocksource 845 watchdog test waits before commencing its tests. 846 Defaults to zero when built as a module and to 847 10 seconds when built into the kernel. 848 849 cma=nn[MG]@[start[MG][-end[MG]]] 850 [KNL,CMA,EARLY] 851 Sets the size of kernel global memory area for 852 contiguous memory allocations and optionally the 853 placement constraint by the physical address range of 854 memory allocations. A value of 0 disables CMA 855 altogether. For more information, see 856 kernel/dma/contiguous.c 857 858 cma_pernuma=nn[MG] 859 [KNL,CMA,EARLY] 860 Sets the size of kernel per-numa memory area for 861 contiguous memory allocations. A value of 0 disables 862 per-numa CMA altogether. And If this option is not 863 specified, the default value is 0. 864 With per-numa CMA enabled, DMA users on node nid will 865 first try to allocate buffer from the pernuma area 866 which is located in node nid, if the allocation fails, 867 they will fallback to the global default memory area. 868 869 numa_cma=<node>:nn[MG][,<node>:nn[MG]] 870 [KNL,CMA,EARLY] 871 Sets the size of kernel numa memory area for 872 contiguous memory allocations. It will reserve CMA 873 area for the specified node. 874 875 If it is setup together with upper 'cmd_pernuma=' 876 (unlikely), its size setting takes priority for the 877 specified numa nodes. 878 879 With numa CMA enabled, DMA users on node nid will 880 first try to allocate buffer from the numa area 881 which is located in node nid, if the allocation fails, 882 they will fallback to the global default memory area. 883 884 cmo_free_hint= [PPC] Format: { yes | no } 885 Specify whether pages are marked as being inactive 886 when they are freed. This is used in CMO environments 887 to determine OS memory pressure for page stealing by 888 a hypervisor. 889 Default: yes 890 891 coherent_pool=nn[KMG] [ARM,KNL,EARLY] 892 Sets the size of memory pool for coherent, atomic dma 893 allocations, by default set to 256K. 894 895 condev= [HW,S390] console device 896 conmode= 897 898 con3215_drop= [S390,EARLY] 3215 console drop mode. 899 Format: y|n|Y|N|1|0 900 When set to true, drop data on the 3215 console when 901 the console buffer is full. In this case the 902 operator using a 3270 terminal emulator (for example 903 x3270) does not have to enter the clear key for the 904 console output to advance and the kernel to continue. 905 This leads to a much faster boot time when a 3270 906 terminal emulator is active. If no 3270 terminal 907 emulator is used, this parameter has no effect. 908 909 console= [KNL] Output console device and options. 910 911 tty<n> Use the virtual console device <n>. 912 913 ttyS<n>[,options] 914 ttyUSB0[,options] 915 Use the specified serial port. The options are of 916 the form "bbbbpnf", where "bbbb" is the baud rate, 917 "p" is parity ("n", "o", or "e"), "n" is number of 918 bits, and "f" is flow control ("r" for RTS or 919 omit it). Default is "9600n8". 920 921 See Documentation/admin-guide/serial-console.rst for more 922 information. See 923 Documentation/networking/netconsole.rst for an 924 alternative. 925 926 <DEVNAME>:<n>.<n>[,options] 927 Use the specified serial port on the serial core bus. 928 The addressing uses DEVNAME of the physical serial port 929 device, followed by the serial core controller instance, 930 and the serial port instance. The options are the same 931 as documented for the ttyS addressing above. 932 933 The mapping of the serial ports to the tty instances 934 can be viewed with: 935 936 $ ls -d /sys/bus/serial-base/devices/*:*.*/tty/* 937 /sys/bus/serial-base/devices/00:04:0.0/tty/ttyS0 938 939 In the above example, the console can be addressed with 940 console=00:04:0.0. Note that a console addressed this 941 way will only get added when the related device driver 942 is ready. The use of an earlycon parameter in addition to 943 the console may be desired for console output early on. 944 945 uart[8250],io,<addr>[,options] 946 uart[8250],mmio,<addr>[,options] 947 uart[8250],mmio16,<addr>[,options] 948 uart[8250],mmio32,<addr>[,options] 949 uart[8250],0x<addr>[,options] 950 Start an early, polled-mode console on the 8250/16550 951 UART at the specified I/O port or MMIO address, 952 switching to the matching ttyS device later. 953 MMIO inter-register address stride is either 8-bit 954 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 955 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 956 to be equivalent to 'mmio'. 'options' are specified in 957 the same format described for ttyS above; if unspecified, 958 the h/w is not re-initialized. 959 960 hvc<n> Use the hypervisor console device <n>. This is for 961 both Xen and PowerPC hypervisors. 962 963 { null | "" } 964 Use to disable console output, i.e., to have kernel 965 console messages discarded. 966 This must be the only console= parameter used on the 967 kernel command line. 968 969 If the device connected to the port is not a TTY but a braille 970 device, prepend "brl," before the device type, for instance 971 console=brl,ttyS0 972 For now, only VisioBraille is supported. 973 974 console_msg_format= 975 [KNL] Change console messages format 976 default 977 By default we print messages on consoles in 978 "[time stamp] text\n" format (time stamp may not be 979 printed, depending on CONFIG_PRINTK_TIME or 980 `printk_time' param). 981 syslog 982 Switch to syslog format: "<%u>[time stamp] text\n" 983 IOW, each message will have a facility and loglevel 984 prefix. The format is similar to one used by syslog() 985 syscall, or to executing "dmesg -S --raw" or to reading 986 from /proc/kmsg. 987 988 consoleblank= [KNL] The console blank (screen saver) timeout in 989 seconds. A value of 0 disables the blank timer. 990 Defaults to 0. 991 992 coredump_filter= 993 [KNL] Change the default value for 994 /proc/<pid>/coredump_filter. 995 See also Documentation/filesystems/proc.rst. 996 997 coresight_cpu_debug.enable 998 [ARM,ARM64] 999 Format: <bool> 1000 Enable/disable the CPU sampling based debugging. 1001 0: default value, disable debugging 1002 1: enable debugging at boot time 1003 1004 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 1005 Format: 1006 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 1007 1008 cpuidle.off=1 [CPU_IDLE] 1009 disable the cpuidle sub-system 1010 1011 cpuidle.governor= 1012 [CPU_IDLE] Name of the cpuidle governor to use. 1013 1014 cpufreq.off=1 [CPU_FREQ] 1015 disable the cpufreq sub-system 1016 1017 cpufreq.default_governor= 1018 [CPU_FREQ] Name of the default cpufreq governor or 1019 policy to use. This governor must be registered in the 1020 kernel before the cpufreq driver probes. 1021 1022 cpu_init_udelay=N 1023 [X86,EARLY] Delay for N microsec between assert and de-assert 1024 of APIC INIT to start processors. This delay occurs 1025 on every CPU online, such as boot, and resume from suspend. 1026 Default: 10000 1027 1028 cpuhp.parallel= 1029 [SMP] Enable/disable parallel bringup of secondary CPUs 1030 Format: <bool> 1031 Default is enabled if CONFIG_HOTPLUG_PARALLEL=y. Otherwise 1032 the parameter has no effect. 1033 1034 crash_kexec_post_notifiers 1035 Only jump to kdump kernel after running the panic 1036 notifiers and dumping kmsg. This option increases 1037 the risks of a kdump failure, since some panic 1038 notifiers can make the crashed kernel more unstable. 1039 In configurations where kdump may not be reliable, 1040 running the panic notifiers could allow collecting 1041 more data on dmesg, like stack traces from other CPUS 1042 or extra data dumped by panic_print. Note that some 1043 configurations enable this option unconditionally, 1044 like Hyper-V, PowerPC (fadump) and AMD SEV-SNP. 1045 1046 crashkernel=size[KMG][@offset[KMG]] 1047 [KNL,EARLY] Using kexec, Linux can switch to a 'crash kernel' 1048 upon panic. This parameter reserves the physical 1049 memory region [offset, offset + size] for that kernel 1050 image. If '@offset' is omitted, then a suitable offset 1051 is selected automatically. 1052 [KNL, X86-64, ARM64, RISCV, LoongArch] Select a region 1053 under 4G first, and fall back to reserve region above 1054 4G when '@offset' hasn't been specified. 1055 See Documentation/admin-guide/kdump/kdump.rst for further details. 1056 1057 crashkernel=range1:size1[,range2:size2,...][@offset] 1058 [KNL] Same as above, but depends on the memory 1059 in the running system. The syntax of range is 1060 start-[end] where start and end are both 1061 a memory unit (amount[KMG]). See also 1062 Documentation/admin-guide/kdump/kdump.rst for an example. 1063 1064 crashkernel=size[KMG],high 1065 [KNL, X86-64, ARM64, RISCV, LoongArch] range could be 1066 above 4G. 1067 Allow kernel to allocate physical memory region from top, 1068 so could be above 4G if system have more than 4G ram 1069 installed. Otherwise memory region will be allocated 1070 below 4G, if available. 1071 It will be ignored if crashkernel=X is specified. 1072 crashkernel=size[KMG],low 1073 [KNL, X86-64, ARM64, RISCV, LoongArch] range under 4G. 1074 When crashkernel=X,high is passed, kernel could allocate 1075 physical memory region above 4G, that cause second kernel 1076 crash on system that require some amount of low memory, 1077 e.g. swiotlb requires at least 64M+32K low memory, also 1078 enough extra low memory is needed to make sure DMA buffers 1079 for 32-bit devices won't run out. Kernel would try to allocate 1080 default size of memory below 4G automatically. The default 1081 size is platform dependent. 1082 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB) 1083 --> arm64: 128MiB 1084 --> riscv: 128MiB 1085 --> loongarch: 128MiB 1086 This one lets the user specify own low range under 4G 1087 for second kernel instead. 1088 0: to disable low allocation. 1089 It will be ignored when crashkernel=X,high is not used 1090 or memory reserved is below 4G. 1091 crashkernel=size[KMG],cma 1092 [KNL, X86, ppc] Reserve additional crash kernel memory from 1093 CMA. This reservation is usable by the first system's 1094 userspace memory and kernel movable allocations (memory 1095 balloon, zswap). Pages allocated from this memory range 1096 will not be included in the vmcore so this should not 1097 be used if dumping of userspace memory is intended and 1098 it has to be expected that some movable kernel pages 1099 may be missing from the dump. 1100 1101 A standard crashkernel reservation, as described above, 1102 is still needed to hold the crash kernel and initrd. 1103 1104 This option increases the risk of a kdump failure: DMA 1105 transfers configured by the first kernel may end up 1106 corrupting the second kernel's memory. 1107 1108 This reservation method is intended for systems that 1109 can't afford to sacrifice enough memory for standard 1110 crashkernel reservation and where less reliable and 1111 possibly incomplete kdump is preferable to no kdump at 1112 all. 1113 1114 cryptomgr.notests 1115 [KNL] Disable crypto self-tests 1116 1117 cs89x0_dma= [HW,NET] 1118 Format: <dma> 1119 1120 cs89x0_media= [HW,NET] 1121 Format: { rj45 | aui | bnc } 1122 1123 csdlock_debug= [KNL] Enable or disable debug add-ons of cross-CPU 1124 function call handling. When switched on, 1125 additional debug data is printed to the console 1126 in case a hanging CPU is detected, and that 1127 CPU is pinged again in order to try to resolve 1128 the hang situation. The default value of this 1129 option depends on the CSD_LOCK_WAIT_DEBUG_DEFAULT 1130 Kconfig option. 1131 1132 dasd= [HW,NET] 1133 See header of drivers/s390/block/dasd_devmap.c. 1134 1135 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 1136 (one device per port) 1137 Format: <port#>,<type> 1138 See also Documentation/input/devices/joystick-parport.rst 1139 1140 debug [KNL,EARLY] Enable kernel debugging (events log level). 1141 1142 debug_boot_weak_hash 1143 [KNL,EARLY] Enable printing [hashed] pointers early in the 1144 boot sequence. If enabled, we use a weak hash instead 1145 of siphash to hash pointers. Use this option if you are 1146 seeing instances of '(___ptrval___)') and need to see a 1147 value (hashed pointer) instead. Cryptographically 1148 insecure, please do not use on production kernels. 1149 1150 debug_locks_verbose= 1151 [KNL] verbose locking self-tests 1152 Format: <int> 1153 Print debugging info while doing the locking API 1154 self-tests. 1155 Bitmask for the various LOCKTYPE_ tests. Defaults to 0 1156 (no extra messages), setting it to -1 (all bits set) 1157 will print _a_lot_ more information - normally only 1158 useful to lockdep developers. 1159 1160 debug_objects [KNL,EARLY] Enable object debugging 1161 1162 debug_guardpage_minorder= 1163 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this 1164 parameter allows control of the order of pages that will 1165 be intentionally kept free (and hence protected) by the 1166 buddy allocator. Bigger value increase the probability 1167 of catching random memory corruption, but reduce the 1168 amount of memory for normal system use. The maximum 1169 possible value is MAX_PAGE_ORDER/2. Setting this 1170 parameter to 1 or 2 should be enough to identify most 1171 random memory corruption problems caused by bugs in 1172 kernel or driver code when a CPU writes to (or reads 1173 from) a random memory location. Note that there exists 1174 a class of memory corruptions problems caused by buggy 1175 H/W or F/W or by drivers badly programming DMA 1176 (basically when memory is written at bus level and the 1177 CPU MMU is bypassed) which are not detectable by 1178 CONFIG_DEBUG_PAGEALLOC, hence this option will not 1179 help tracking down these problems. 1180 1181 debug_pagealloc= 1182 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 1183 enables the feature at boot time. By default, it is 1184 disabled and the system will work mostly the same as a 1185 kernel built without CONFIG_DEBUG_PAGEALLOC. 1186 Note: to get most of debug_pagealloc error reports, it's 1187 useful to also enable the page_owner functionality. 1188 on: enable the feature 1189 1190 debugfs= [KNL,EARLY] This parameter enables what is exposed to 1191 userspace and debugfs internal clients. 1192 Format: { on, off } 1193 on: All functions are enabled. 1194 off: Filesystem is not registered and clients 1195 get a -EPERM as result when trying to register files 1196 or directories within debugfs. 1197 This is equivalent of the runtime functionality if 1198 debugfs was not enabled in the kernel at all. 1199 Default value is set in build-time with a kernel configuration. 1200 1201 debugpat [X86] Enable PAT debugging 1202 1203 default_hugepagesz= 1204 [HW] The size of the default HugeTLB page. This is 1205 the size represented by the legacy /proc/ hugepages 1206 APIs. In addition, this is the default hugetlb size 1207 used for shmget(), mmap() and mounting hugetlbfs 1208 filesystems. If not specified, defaults to the 1209 architecture's default huge page size. Huge page 1210 sizes are architecture dependent. See also 1211 Documentation/admin-guide/mm/hugetlbpage.rst. 1212 Format: size[KMG] 1213 1214 deferred_probe_timeout= 1215 [KNL] Debugging option to set a timeout in seconds for 1216 deferred probe to give up waiting on dependencies to 1217 probe. Only specific dependencies (subsystems or 1218 drivers) that have opted in will be ignored. A timeout 1219 of 0 will timeout at the end of initcalls; a negative 1220 value is treated as an infinite timeout value. If the 1221 timeout hasn't expired, it'll be restarted by each 1222 successful driver registration. This option will also 1223 dump out devices still on the deferred probe list after 1224 retrying. 1225 1226 delayacct [KNL] Enable per-task delay accounting 1227 1228 dell_smm_hwmon.ignore_dmi= 1229 [HW] Continue probing hardware even if DMI data 1230 indicates that the driver is running on unsupported 1231 hardware. 1232 1233 dell_smm_hwmon.force= 1234 [HW] Activate driver even if SMM BIOS signature does 1235 not match list of supported models and enable otherwise 1236 blacklisted features. 1237 1238 dell_smm_hwmon.power_status= 1239 [HW] Report power status in /proc/i8k 1240 (disabled by default). 1241 1242 dell_smm_hwmon.restricted= 1243 [HW] Allow controlling fans only if SYS_ADMIN 1244 capability is set. 1245 1246 dell_smm_hwmon.fan_mult= 1247 [HW] Factor to multiply fan speed with. 1248 1249 dell_smm_hwmon.fan_max= 1250 [HW] Maximum configurable fan speed. 1251 1252 dfltcc= [HW,S390] 1253 Format: { on | off | def_only | inf_only | always } 1254 on: s390 zlib hardware support for compression on 1255 level 1 and decompression (default) 1256 off: No s390 zlib hardware support 1257 def_only: s390 zlib hardware support for deflate 1258 only (compression on level 1) 1259 inf_only: s390 zlib hardware support for inflate 1260 only (decompression) 1261 always: Same as 'on' but ignores the selected compression 1262 level always using hardware support (used for debugging) 1263 1264 dhash_entries= [KNL] 1265 Set number of hash buckets for dentry cache. 1266 1267 disable_1tb_segments [PPC,EARLY] 1268 Disables the use of 1TB hash page table segments. This 1269 causes the kernel to fall back to 256MB segments which 1270 can be useful when debugging issues that require an SLB 1271 miss to occur. 1272 1273 disable= [IPV6] 1274 See Documentation/networking/ipv6.rst. 1275 1276 disable_radix [PPC,EARLY] 1277 Disable RADIX MMU mode on POWER9 1278 1279 disable_tlbie [PPC] 1280 Disable TLBIE instruction. Currently does not work 1281 with KVM, with HASH MMU, or with coherent accelerators. 1282 1283 disable_ddw [PPC/PSERIES,EARLY] 1284 Disable Dynamic DMA Window support. Use this 1285 to workaround buggy firmware. 1286 1287 disable_ipv6= [IPV6] 1288 See Documentation/networking/ipv6.rst. 1289 1290 disable_mtrr_cleanup [X86,EARLY] 1291 The kernel tries to adjust MTRR layout from continuous 1292 to discrete, to make X server driver able to add WB 1293 entry later. This parameter disables that. 1294 1295 disable_mtrr_trim [X86, Intel and AMD only,EARLY] 1296 By default the kernel will trim any uncacheable 1297 memory out of your available memory pool based on 1298 MTRR settings. This parameter disables that behavior, 1299 possibly causing your machine to run very slowly. 1300 1301 disable_timer_pin_1 [X86,EARLY] 1302 Disable PIN 1 of APIC timer 1303 Can be useful to work around chipset bugs. 1304 1305 dis_ucode_ldr [X86] Disable the microcode loader. 1306 1307 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 1308 this option disables the debugging code at boot. 1309 1310 dma_debug_entries=<number> 1311 This option allows to tune the number of preallocated 1312 entries for DMA-API debugging code. One entry is 1313 required per DMA-API allocation. Use this if the 1314 DMA-API debugging code disables itself because the 1315 architectural default is too low. 1316 1317 dma_debug_driver=<driver_name> 1318 With this option the DMA-API debugging driver 1319 filter feature can be enabled at boot time. Just 1320 pass the driver to filter for as the parameter. 1321 The filter can be disabled or changed to another 1322 driver later using sysfs. 1323 1324 reg_file_data_sampling= 1325 [X86] Controls mitigation for Register File Data 1326 Sampling (RFDS) vulnerability. RFDS is a CPU 1327 vulnerability which may allow userspace to infer 1328 kernel data values previously stored in floating point 1329 registers, vector registers, or integer registers. 1330 RFDS only affects Intel Atom processors. 1331 1332 on: Turns ON the mitigation. 1333 off: Turns OFF the mitigation. 1334 1335 This parameter overrides the compile time default set 1336 by CONFIG_MITIGATION_RFDS. Mitigation cannot be 1337 disabled when other VERW based mitigations (like MDS) 1338 are enabled. In order to disable RFDS mitigation all 1339 VERW based mitigations need to be disabled. 1340 1341 For details see: 1342 Documentation/admin-guide/hw-vuln/reg-file-data-sampling.rst 1343 1344 dm_verity.keyring_unsealed= 1345 [KNL] When set to 1, leave the dm-verity keyring 1346 unsealed after initialization so userspace can 1347 provision keys. Once the keyring is restricted 1348 it becomes active and is searched during signature 1349 verification. 1350 1351 driver_async_probe= [KNL] 1352 List of driver names to be probed asynchronously. * 1353 matches with all driver names. If * is specified, the 1354 rest of the listed driver names are those that will NOT 1355 match the *. 1356 Format: <driver_name1>,<driver_name2>... 1357 1358 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1359 Broken monitors, graphic adapters, KVMs and EDIDless 1360 panels may send no or incorrect EDID data sets. 1361 This parameter allows to specify an EDID data sets 1362 in the /lib/firmware directory that are used instead. 1363 An EDID data set will only be used for a particular 1364 connector, if its name and a colon are prepended to 1365 the EDID name. Each connector may use a unique EDID 1366 data set by separating the files with a comma. An EDID 1367 data set with no connector name will be used for 1368 any connectors not explicitly specified. 1369 1370 dscc4.setup= [NET] 1371 1372 dt_cpu_ftrs= [PPC,EARLY] 1373 Format: {"off" | "known"} 1374 Control how the dt_cpu_ftrs device-tree binding is 1375 used for CPU feature discovery and setup (if it 1376 exists). 1377 off: Do not use it, fall back to legacy cpu table. 1378 known: Do not pass through unknown features to guests 1379 or userspace, only those that the kernel is aware of. 1380 1381 dump_apple_properties [X86] 1382 Dump name and content of EFI device properties on 1383 x86 Macs. Useful for driver authors to determine 1384 what data is available or for reverse-engineering. 1385 1386 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1387 <module>.dyndbg[="val"] 1388 Enable debug messages at boot time. See 1389 Documentation/admin-guide/dynamic-debug-howto.rst 1390 for details. 1391 1392 early_ioremap_debug [KNL,EARLY] 1393 Enable debug messages in early_ioremap support. This 1394 is useful for tracking down temporary early mappings 1395 which are not unmapped. 1396 1397 earlycon= [KNL,EARLY] Output early console device and options. 1398 1399 When used with no options, the early console is 1400 determined by stdout-path property in device tree's 1401 chosen node or the ACPI SPCR table if supported by 1402 the platform. 1403 1404 cdns,<addr>[,options] 1405 Start an early, polled-mode console on a Cadence 1406 (xuartps) serial port at the specified address. Only 1407 supported option is baud rate. If baud rate is not 1408 specified, the serial port must already be setup and 1409 configured. 1410 1411 uart[8250],io,<addr>[,options[,uartclk]] 1412 uart[8250],mmio,<addr>[,options[,uartclk]] 1413 uart[8250],mmio32,<addr>[,options[,uartclk]] 1414 uart[8250],mmio32be,<addr>[,options[,uartclk]] 1415 uart[8250],0x<addr>[,options] 1416 Start an early, polled-mode console on the 8250/16550 1417 UART at the specified I/O port or MMIO address. 1418 MMIO inter-register address stride is either 8-bit 1419 (mmio) or 32-bit (mmio32 or mmio32be). 1420 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1421 to be equivalent to 'mmio'. 'options' are specified 1422 in the same format described for "console=ttyS<n>"; if 1423 unspecified, the h/w is not initialized. 'uartclk' is 1424 the uart clock frequency; if unspecified, it is set 1425 to 'BASE_BAUD' * 16. 1426 1427 pl011,<addr> 1428 pl011,mmio32,<addr> 1429 Start an early, polled-mode console on a pl011 serial 1430 port at the specified address. The pl011 serial port 1431 must already be setup and configured. Options are not 1432 yet supported. If 'mmio32' is specified, then only 1433 the driver will use only 32-bit accessors to read/write 1434 the device registers. 1435 1436 liteuart,<addr> 1437 Start an early console on a litex serial port at the 1438 specified address. The serial port must already be 1439 setup and configured. Options are not yet supported. 1440 1441 meson,<addr> 1442 Start an early, polled-mode console on a meson serial 1443 port at the specified address. The serial port must 1444 already be setup and configured. Options are not yet 1445 supported. 1446 1447 msm_serial,<addr> 1448 Start an early, polled-mode console on an msm serial 1449 port at the specified address. The serial port 1450 must already be setup and configured. Options are not 1451 yet supported. 1452 1453 msm_serial_dm,<addr> 1454 Start an early, polled-mode console on an msm serial 1455 dm port at the specified address. The serial port 1456 must already be setup and configured. Options are not 1457 yet supported. 1458 1459 owl,<addr> 1460 Start an early, polled-mode console on a serial port 1461 of an Actions Semi SoC, such as S500 or S900, at the 1462 specified address. The serial port must already be 1463 setup and configured. Options are not yet supported. 1464 1465 rda,<addr> 1466 Start an early, polled-mode console on a serial port 1467 of an RDA Micro SoC, such as RDA8810PL, at the 1468 specified address. The serial port must already be 1469 setup and configured. Options are not yet supported. 1470 1471 sbi 1472 Use RISC-V SBI (Supervisor Binary Interface) for early 1473 console. 1474 1475 smh Use ARM semihosting calls for early console. 1476 1477 s3c2410,<addr> 1478 s3c2412,<addr> 1479 s3c2440,<addr> 1480 s3c6400,<addr> 1481 s5pv210,<addr> 1482 exynos4210,<addr> 1483 Use early console provided by serial driver available 1484 on Samsung SoCs, requires selecting proper type and 1485 a correct base address of the selected UART port. The 1486 serial port must already be setup and configured. 1487 Options are not yet supported. 1488 1489 lantiq,<addr> 1490 Start an early, polled-mode console on a lantiq serial 1491 (lqasc) port at the specified address. The serial port 1492 must already be setup and configured. Options are not 1493 yet supported. 1494 1495 lpuart,<addr> 1496 lpuart32,<addr> 1497 Use early console provided by Freescale LP UART driver 1498 found on Freescale Vybrid and QorIQ LS1021A processors. 1499 A valid base address must be provided, and the serial 1500 port must already be setup and configured. 1501 1502 ec_imx21,<addr> 1503 ec_imx6q,<addr> 1504 Start an early, polled-mode, output-only console on the 1505 Freescale i.MX UART at the specified address. The UART 1506 must already be setup and configured. 1507 1508 ar3700_uart,<addr> 1509 Start an early, polled-mode console on the 1510 Armada 3700 serial port at the specified 1511 address. The serial port must already be setup 1512 and configured. Options are not yet supported. 1513 1514 qcom_geni,<addr> 1515 Start an early, polled-mode console on a Qualcomm 1516 Generic Interface (GENI) based serial port at the 1517 specified address. The serial port must already be 1518 setup and configured. Options are not yet supported. 1519 1520 efifb,[options] 1521 Start an early, unaccelerated console on the EFI 1522 memory mapped framebuffer (if available). On cache 1523 coherent non-x86 systems that use system memory for 1524 the framebuffer, pass the 'ram' option so that it is 1525 mapped with the correct attributes. 1526 1527 linflex,<addr> 1528 Use early console provided by Freescale LINFlexD UART 1529 serial driver for NXP S32V234 SoCs. A valid base 1530 address must be provided, and the serial port must 1531 already be setup and configured. 1532 1533 earlyprintk= [X86,SH,ARM,M68k,S390,UM,EARLY] 1534 earlyprintk=vga 1535 earlyprintk=sclp 1536 earlyprintk=xen 1537 earlyprintk=serial[,ttySn[,baudrate]] 1538 earlyprintk=serial[,0x...[,baudrate]] 1539 earlyprintk=ttySn[,baudrate] 1540 earlyprintk=dbgp[debugController#] 1541 earlyprintk=mmio32,membase[,{nocfg|baudrate}] 1542 earlyprintk=pciserial[,force],bus:device.function[,{nocfg|baudrate}] 1543 earlyprintk=xdbc[xhciController#] 1544 earlyprintk=bios 1545 1546 earlyprintk is useful when the kernel crashes before 1547 the normal console is initialized. It is not enabled by 1548 default because it has some cosmetic problems. 1549 1550 Use "nocfg" to skip UART configuration, assume 1551 BIOS/firmware has configured UART correctly. 1552 1553 Append ",keep" to not disable it when the real console 1554 takes over. 1555 1556 Only one of vga, serial, or usb debug port can 1557 be used at a time. 1558 1559 Currently only ttyS0 and ttyS1 may be specified by 1560 name. Other I/O ports may be explicitly specified 1561 on some architectures (x86 and arm at least) by 1562 replacing ttySn with an I/O port address, like this: 1563 earlyprintk=serial,0x1008,115200 1564 You can find the port for a given device in 1565 /proc/tty/driver/serial: 1566 2: uart:ST16650V2 port:00001008 irq:18 ... 1567 1568 Interaction with the standard serial driver is not 1569 very good. 1570 1571 The VGA output is eventually overwritten by 1572 the real console. 1573 1574 The xen option can only be used in Xen domains. 1575 1576 The sclp output can only be used on s390. 1577 1578 The bios output can only be used on SuperH. 1579 1580 The optional "force" to "pciserial" enables use of a 1581 PCI device even when its classcode is not of the 1582 UART class. 1583 1584 edac_report= [HW,EDAC] Control how to report EDAC event 1585 Format: {"on" | "off" | "force"} 1586 on: enable EDAC to report H/W event. May be overridden 1587 by other higher priority error reporting module. 1588 off: disable H/W event reporting through EDAC. 1589 force: enforce the use of EDAC to report H/W event. 1590 default: on. 1591 1592 edd= [EDD] 1593 Format: {"off" | "on" | "skip[mbr]"} 1594 1595 efi= [EFI,EARLY] 1596 Format: { "debug", "disable_early_pci_dma", 1597 "nochunk", "noruntime", "nosoftreserve", 1598 "novamap", "no_disable_early_pci_dma" } 1599 debug: enable misc debug output. 1600 disable_early_pci_dma: disable the busmaster bit on all 1601 PCI bridges while in the EFI boot stub. 1602 nochunk: disable reading files in "chunks" in the EFI 1603 boot stub, as chunking can cause problems with some 1604 firmware implementations. 1605 noruntime : disable EFI runtime services support 1606 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1607 attribute may cause the kernel to reserve the 1608 memory range for a memory mapping driver to 1609 claim. Specify efi=nosoftreserve to disable this 1610 reservation and treat the memory by its base type 1611 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1612 novamap: do not call SetVirtualAddressMap(). 1613 no_disable_early_pci_dma: Leave the busmaster bit set 1614 on all PCI bridges while in the EFI boot stub 1615 1616 efi_no_storage_paranoia [EFI,X86,EARLY] 1617 The kernel reserves 5KB of EFI variable storage for 1618 safety, because some UEFI implementation may fail to 1619 boot if there's insufficient space in the EFI variable 1620 storage. 1621 1622 Using this parameter, you can use the 5KB reservation 1623 in the EFI variable storage. 1624 1625 However, Use this parameter only if you are really 1626 sure that your UEFI does sane gc and fulfills the spec 1627 otherwise your board may brick. 1628 1629 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1630 that is to be dynamically loaded by Linux. If there are 1631 multiple variables with the same name but with different 1632 vendor GUIDs, all of them will be loaded. See 1633 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1634 1635 1636 eisa_irq_edge= [PARISC,HW] 1637 See header of drivers/parisc/eisa.c. 1638 1639 ekgdboc= [X86,KGDB,EARLY] Allow early kernel console debugging 1640 Format: ekgdboc=kbd 1641 1642 This is designed to be used in conjunction with 1643 the boot argument: earlyprintk=vga 1644 1645 This parameter works in place of the kgdboc parameter 1646 but can only be used if the backing tty is available 1647 very early in the boot process. For early debugging 1648 via a serial port see kgdboc_earlycon instead. 1649 1650 elfcorehdr=[size[KMG]@]offset[KMG] [PPC,SH,X86,S390,EARLY] 1651 Specifies physical address of start of kernel core 1652 image elf header and optionally the size. Generally 1653 kexec loader will pass this option to capture kernel. 1654 See Documentation/admin-guide/kdump/kdump.rst for details. 1655 1656 enable_mtrr_cleanup [X86,EARLY] 1657 The kernel tries to adjust MTRR layout from continuous 1658 to discrete, to make X server driver able to add WB 1659 entry later. This parameter enables that. 1660 1661 enable_timer_pin_1 [X86] 1662 Enable PIN 1 of APIC timer 1663 Can be useful to work around chipset bugs 1664 (in particular on some ATI chipsets). 1665 The kernel tries to set a reasonable default. 1666 1667 enforcing= [SELINUX] Set initial enforcing status. 1668 Format: {"0" | "1"} 1669 See security/selinux/Kconfig help text. 1670 0 -- permissive (log only, no denials). 1671 1 -- enforcing (deny and log). 1672 Default value is 0. 1673 Value can be changed at runtime via 1674 /sys/fs/selinux/enforce. 1675 1676 erst_disable [ACPI] 1677 Disable Error Record Serialization Table (ERST) 1678 support. 1679 1680 ether= [HW,NET] Ethernet cards parameters 1681 This option is obsoleted by the "netdev=" option, which 1682 has equivalent usage. See its documentation for details. 1683 1684 evm= [EVM] 1685 Format: { "fix" } 1686 Permit 'security.evm' to be updated regardless of 1687 current integrity status. 1688 1689 early_page_ext [KNL,EARLY] Enforces page_ext initialization to earlier 1690 stages so cover more early boot allocations. 1691 Please note that as side effect some optimizations 1692 might be disabled to achieve that (e.g. parallelized 1693 memory initialization is disabled) so the boot process 1694 might take longer, especially on systems with a lot of 1695 memory. Available with CONFIG_PAGE_EXTENSION=y. 1696 1697 failslab= 1698 fail_usercopy= 1699 fail_page_alloc= 1700 fail_skb_realloc= 1701 fail_make_request=[KNL] 1702 General fault injection mechanism. 1703 Format: <interval>,<probability>,<space>,<times> 1704 See also Documentation/fault-injection/. 1705 1706 fb_tunnels= [NET] 1707 Format: { initns | none } 1708 See Documentation/admin-guide/sysctl/net.rst for 1709 fb_tunnels_only_for_init_ns 1710 1711 floppy= [HW] 1712 See Documentation/admin-guide/blockdev/floppy.rst. 1713 1714 forcepae [X86-32] 1715 Forcefully enable Physical Address Extension (PAE). 1716 Many Pentium M systems disable PAE but may have a 1717 functionally usable PAE implementation. 1718 Warning: use of this parameter will taint the kernel 1719 and may cause unknown problems. 1720 1721 fred= [X86-64] 1722 Enable/disable Flexible Return and Event Delivery. 1723 Format: { on | off } 1724 on: enable FRED when it's present, the default setting. 1725 off: disable FRED. 1726 1727 ftrace=[tracer] 1728 [FTRACE] will set and start the specified tracer 1729 as early as possible in order to facilitate early 1730 boot debugging. 1731 1732 ftrace_boot_snapshot 1733 [FTRACE] On boot up, a snapshot will be taken of the 1734 ftrace ring buffer that can be read at: 1735 /sys/kernel/tracing/snapshot. 1736 This is useful if you need tracing information from kernel 1737 boot up that is likely to be overridden by user space 1738 start up functionality. 1739 1740 Optionally, the snapshot can also be defined for a tracing 1741 instance that was created by the trace_instance= command 1742 line parameter. 1743 1744 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo 1745 1746 The above will cause the "foo" tracing instance to trigger 1747 a snapshot at the end of boot up. 1748 1749 ftrace_dump_on_oops[=2(orig_cpu) | =<instance>][,<instance> | 1750 ,<instance>=2(orig_cpu)] 1751 [FTRACE] will dump the trace buffers on oops. 1752 If no parameter is passed, ftrace will dump global 1753 buffers of all CPUs, if you pass 2 or orig_cpu, it 1754 will dump only the buffer of the CPU that triggered 1755 the oops, or the specific instance will be dumped if 1756 its name is passed. Multiple instance dump is also 1757 supported, and instances are separated by commas. Each 1758 instance supports only dump on CPU that triggered the 1759 oops by passing 2 or orig_cpu to it. 1760 1761 ftrace_dump_on_oops=foo=orig_cpu 1762 1763 The above will dump only the buffer of "foo" instance 1764 on CPU that triggered the oops. 1765 1766 ftrace_dump_on_oops,foo,bar=orig_cpu 1767 1768 The above will dump global buffer on all CPUs, the 1769 buffer of "foo" instance on all CPUs and the buffer 1770 of "bar" instance on CPU that triggered the oops. 1771 1772 ftrace_filter=[function-list] 1773 [FTRACE] Limit the functions traced by the function 1774 tracer at boot up. function-list is a comma-separated 1775 list of functions. This list can be changed at run 1776 time by the set_ftrace_filter file in the debugfs 1777 tracing directory. 1778 1779 ftrace_notrace=[function-list] 1780 [FTRACE] Do not trace the functions specified in 1781 function-list. This list can be changed at run time 1782 by the set_ftrace_notrace file in the debugfs 1783 tracing directory. 1784 1785 ftrace_graph_filter=[function-list] 1786 [FTRACE] Limit the top level callers functions traced 1787 by the function graph tracer at boot up. 1788 function-list is a comma-separated list of functions 1789 that can be changed at run time by the 1790 set_graph_function file in the debugfs tracing directory. 1791 1792 ftrace_graph_notrace=[function-list] 1793 [FTRACE] Do not trace from the functions specified in 1794 function-list. This list is a comma-separated list of 1795 functions that can be changed at run time by the 1796 set_graph_notrace file in the debugfs tracing directory. 1797 1798 ftrace_graph_max_depth=<uint> 1799 [FTRACE] Used with the function graph tracer. This is 1800 the max depth it will trace into a function. This value 1801 can be changed at run time by the max_graph_depth file 1802 in the tracefs tracing directory. default: 0 (no limit) 1803 1804 fw_devlink= [KNL,EARLY] Create device links between consumer and supplier 1805 devices by scanning the firmware to infer the 1806 consumer/supplier relationships. This feature is 1807 especially useful when drivers are loaded as modules as 1808 it ensures proper ordering of tasks like device probing 1809 (suppliers first, then consumers), supplier boot state 1810 clean up (only after all consumers have probed), 1811 suspend/resume & runtime PM (consumers first, then 1812 suppliers). 1813 Format: { off | permissive | on | rpm } 1814 off -- Don't create device links from firmware info. 1815 permissive -- Create device links from firmware info 1816 but use it only for ordering boot state clean 1817 up (sync_state() calls). 1818 on -- Create device links from firmware info and use it 1819 to enforce probe and suspend/resume ordering. 1820 rpm -- Like "on", but also use to order runtime PM. 1821 1822 fw_devlink.strict=<bool> 1823 [KNL,EARLY] Treat all inferred dependencies as mandatory 1824 dependencies. This only applies for fw_devlink=on|rpm. 1825 Format: <bool> 1826 1827 fw_devlink.sync_state = 1828 [KNL,EARLY] When all devices that could probe have finished 1829 probing, this parameter controls what to do with 1830 devices that haven't yet received their sync_state() 1831 calls. 1832 Format: { strict | timeout } 1833 strict -- Default. Continue waiting on consumers to 1834 probe successfully. 1835 timeout -- Give up waiting on consumers and call 1836 sync_state() on any devices that haven't yet 1837 received their sync_state() calls after 1838 deferred_probe_timeout has expired or by 1839 late_initcall() if !CONFIG_MODULES. 1840 1841 gamecon.map[2|3]= 1842 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1843 support via parallel port (up to 5 devices per port) 1844 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1845 See also Documentation/input/devices/joystick-parport.rst 1846 1847 gamma= [HW,DRM] 1848 1849 gart_fix_e820= [X86-64,EARLY] disable the fix e820 for K8 GART 1850 Format: off | on 1851 default: on 1852 1853 gather_data_sampling= 1854 [X86,INTEL,EARLY] Control the Gather Data Sampling (GDS) 1855 mitigation. 1856 1857 Gather Data Sampling is a hardware vulnerability which 1858 allows unprivileged speculative access to data which was 1859 previously stored in vector registers. 1860 1861 This issue is mitigated by default in updated microcode. 1862 The mitigation may have a performance impact but can be 1863 disabled. On systems without the microcode mitigation 1864 disabling AVX serves as a mitigation. 1865 1866 force: Disable AVX to mitigate systems without 1867 microcode mitigation. No effect if the microcode 1868 mitigation is present. Known to cause crashes in 1869 userspace with buggy AVX enumeration. 1870 1871 off: Disable GDS mitigation. 1872 1873 gbpages [X86] Use GB pages for kernel direct mappings. 1874 1875 gcov_persist= [GCOV] When non-zero (default), profiling data for 1876 kernel modules is saved and remains accessible via 1877 debugfs, even when the module is unloaded/reloaded. 1878 When zero, profiling data is discarded and associated 1879 debugfs files are removed at module unload time. 1880 1881 goldfish [X86] Enable the goldfish android emulator platform. 1882 Don't use this when you are not running on the 1883 android emulator 1884 1885 gpio-mockup.gpio_mockup_ranges 1886 [HW] Sets the ranges of gpiochip of for this device. 1887 Format: <start1>,<end1>,<start2>,<end2>... 1888 gpio-mockup.gpio_mockup_named_lines 1889 [HW] Let the driver know GPIO lines should be named. 1890 1891 gpt [EFI] Forces disk with valid GPT signature but 1892 invalid Protective MBR to be treated as GPT. If the 1893 primary GPT is corrupted, it enables the backup/alternate 1894 GPT to be used instead. 1895 1896 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1897 the "Enable 0" bit of the configuration register. 1898 Format: 0 | 1 1899 Default: 0 1900 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1901 the "Enable 0" bit of the configuration register. 1902 Format: 0 | 1 1903 Default: 0 1904 grcan.select= [HW] Select which physical interface to use. 1905 Format: 0 | 1 1906 Default: 0 1907 grcan.txsize= [HW] Sets the size of the tx buffer. 1908 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1909 Default: 1024 1910 grcan.rxsize= [HW] Sets the size of the rx buffer. 1911 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1912 Default: 1024 1913 1914 hardened_usercopy= 1915 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 1916 hardening is enabled for this boot. Hardened 1917 usercopy checking is used to protect the kernel 1918 from reading or writing beyond known memory 1919 allocation boundaries as a proactive defense 1920 against bounds-checking flaws in the kernel's 1921 copy_to_user()/copy_from_user() interface. 1922 The default is determined by 1923 CONFIG_HARDENED_USERCOPY_DEFAULT_ON. 1924 on Perform hardened usercopy checks. 1925 off Disable hardened usercopy checks. 1926 1927 hardlockup_all_cpu_backtrace= 1928 [KNL] Should the hard-lockup detector generate 1929 backtraces on all cpus. 1930 Format: 0 | 1 1931 1932 hash_pointers= 1933 [KNL,EARLY] 1934 By default, when pointers are printed to the console 1935 or buffers via the %p format string, that pointer is 1936 "hashed", i.e. obscured by hashing the pointer value. 1937 This is a security feature that hides actual kernel 1938 addresses from unprivileged users, but it also makes 1939 debugging the kernel more difficult since unequal 1940 pointers can no longer be compared. The choices are: 1941 Format: { auto | always | never } 1942 Default: auto 1943 1944 auto - Hash pointers unless slab_debug is enabled. 1945 always - Always hash pointers (even if slab_debug is 1946 enabled). 1947 never - Never hash pointers. This option should only 1948 be specified when debugging the kernel. Do 1949 not use on production kernels. The boot 1950 param "no_hash_pointers" is an alias for 1951 this mode. 1952 1953 For controlling hashing dynamically at runtime, 1954 use the "kernel.kptr_restrict" sysctl instead. 1955 1956 hashdist= [KNL,NUMA] Large hashes allocated during boot 1957 are distributed across NUMA nodes. Defaults on 1958 for 64-bit NUMA, off otherwise. 1959 Format: 0 | 1 (for off | on) 1960 1961 hd= [EIDE] (E)IDE hard drive subsystem geometry 1962 Format: <cyl>,<head>,<sect> 1963 1964 hest_disable [ACPI] 1965 Disable Hardware Error Source Table (HEST) support; 1966 corresponding firmware-first mode error processing 1967 logic will be disabled. 1968 1969 hibernate= [HIBERNATION] 1970 noresume Don't check if there's a hibernation image 1971 present during boot. 1972 nocompress Don't compress/decompress hibernation images. 1973 no Disable hibernation and resume. 1974 protect_image Turn on image protection during restoration 1975 (that will set all pages holding image data 1976 during restoration read-only). 1977 1978 hibernate.compressor= [HIBERNATION] Compression algorithm to be 1979 used with hibernation. 1980 Format: { lzo | lz4 } 1981 Default: lzo 1982 1983 lzo: Select LZO compression algorithm to 1984 compress/decompress hibernation image. 1985 1986 lz4: Select LZ4 compression algorithm to 1987 compress/decompress hibernation image. 1988 1989 hibernate.pm_test_delay= 1990 [HIBERNATION] 1991 Sets the number of seconds to remain in a hibernation test 1992 mode before resuming the system (see 1993 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 1994 is set. Default value is 5. 1995 1996 hibernate_compression_threads= 1997 [HIBERNATION] 1998 Set the number of threads used for compressing or decompressing 1999 hibernation images. 2000 2001 Format: <integer> 2002 Default: 3 2003 Minimum: 1 2004 Example: hibernate_compression_threads=4 2005 2006 highmem=nn[KMG] [KNL,BOOT,EARLY] forces the highmem zone to have an exact 2007 size of <nn>. This works even on boxes that have no 2008 highmem otherwise. This also works to reduce highmem 2009 size on bigger boxes. 2010 2011 highres= [KNL] Enable/disable high resolution timer mode. 2012 Valid parameters: "on", "off" 2013 Default: "on" 2014 2015 hlt [BUGS=ARM,SH] 2016 2017 hostname= [KNL,EARLY] Set the hostname (aka UTS nodename). 2018 Format: <string> 2019 This allows setting the system's hostname during early 2020 startup. This sets the name returned by gethostname. 2021 Using this parameter to set the hostname makes it 2022 possible to ensure the hostname is correctly set before 2023 any userspace processes run, avoiding the possibility 2024 that a process may call gethostname before the hostname 2025 has been explicitly set, resulting in the calling 2026 process getting an incorrect result. The string must 2027 not exceed the maximum allowed hostname length (usually 2028 64 characters) and will be truncated otherwise. 2029 2030 hpet= [X86-32,HPET] option to control HPET usage 2031 Format: { enable (default) | disable | force | 2032 verbose } 2033 disable: disable HPET and use PIT instead 2034 force: allow force enabled of undocumented chips (ICH4, 2035 VIA, nVidia) 2036 verbose: show contents of HPET registers during setup 2037 2038 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 2039 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 2040 2041 hugepages= [HW,EARLY] Number of HugeTLB pages to allocate at boot. 2042 If this follows hugepagesz (below), it specifies 2043 the number of pages of hugepagesz to be allocated. 2044 If this is the first HugeTLB parameter on the command 2045 line, it specifies the number of pages to allocate for 2046 the default huge page size. If using node format, the 2047 number of pages to allocate per-node can be specified. 2048 See also Documentation/admin-guide/mm/hugetlbpage.rst. 2049 Format: <integer> or (node format) 2050 <node>:<integer>[,<node>:<integer>] 2051 2052 hugepagesz= 2053 [HW,EARLY] The size of the HugeTLB pages. This is 2054 used in conjunction with hugepages (above) to 2055 allocate huge pages of a specific size at boot. The 2056 pair hugepagesz=X hugepages=Y can be specified once 2057 for each supported huge page size. Huge page sizes 2058 are architecture dependent. See also 2059 Documentation/admin-guide/mm/hugetlbpage.rst. 2060 Format: size[KMG] 2061 2062 hugepage_alloc_threads= 2063 [HW] The number of threads that should be used to 2064 allocate hugepages during boot. This option can be 2065 used to improve system bootup time when allocating 2066 a large amount of huge pages. 2067 The default value is 25% of the available hardware threads. 2068 2069 Note that this parameter only applies to non-gigantic huge pages. 2070 2071 hugetlb_cma= [HW,CMA,EARLY] The size of a CMA area used for allocation 2072 of gigantic hugepages. Or using node format, the size 2073 of a CMA area per node can be specified. 2074 Format: nn[KMGTPE] or (node format) 2075 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]] 2076 2077 The size must be a multiple of the gigantic page size. 2078 When using node format, this applies to each per-node size. 2079 Missaligned values are dropped with a warning. 2080 2081 Reserve a CMA area of given size and allocate gigantic 2082 hugepages using the CMA allocator. If enabled, the 2083 boot-time allocation of gigantic hugepages is skipped. 2084 2085 hugetlb_cma_only= 2086 [HW,CMA,EARLY] When allocating new HugeTLB pages, only 2087 try to allocate from the CMA areas. 2088 2089 This option does nothing if hugetlb_cma= is not also 2090 specified. 2091 2092 hugetlb_free_vmemmap= 2093 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP 2094 enabled. 2095 Control if HugeTLB Vmemmap Optimization (HVO) is enabled. 2096 Allows heavy hugetlb users to free up some more 2097 memory (7 * PAGE_SIZE for each 2MB hugetlb page). 2098 Format: { on | off (default) } 2099 2100 on: enable HVO 2101 off: disable HVO 2102 2103 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y, 2104 the default is on. 2105 2106 Note that the vmemmap pages may be allocated from the added 2107 memory block itself when memory_hotplug.memmap_on_memory is 2108 enabled, those vmemmap pages cannot be optimized even if this 2109 feature is enabled. Other vmemmap pages not allocated from 2110 the added memory block itself do not be affected. 2111 2112 hung_task_panic= 2113 [KNL] Number of hung tasks to trigger kernel panic. 2114 Format: <int> 2115 2116 When set to a non-zero value, a kernel panic will be triggered if 2117 the number of detected hung tasks reaches this value. 2118 2119 0: don't panic 2120 1: panic immediately on first hung task 2121 N: panic after N hung tasks are detected in a single scan 2122 2123 The default value is controlled by the 2124 CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time option. The value 2125 selected by this boot parameter can be changed later by the 2126 kernel.hung_task_panic sysctl. 2127 2128 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 2129 terminal devices. Valid values: 0..8 2130 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 2131 If specified, z/VM IUCV HVC accepts connections 2132 from listed z/VM user IDs only. 2133 2134 hv_nopvspin [X86,HYPER_V,EARLY] 2135 Disables the paravirt spinlock optimizations 2136 which allow the hypervisor to 'idle' the guest 2137 on lock contention. 2138 2139 hw_protection= [HW] 2140 Format: reboot | shutdown 2141 2142 Hardware protection action taken on critical events like 2143 overtemperature or imminent voltage loss. 2144 2145 i2c_bus= [HW] Override the default board specific I2C bus speed 2146 or register an additional I2C bus that is not 2147 registered from board initialization code. 2148 Format: 2149 <bus_id>,<clkrate> 2150 2151 i2c_touchscreen_props= [HW,ACPI,X86] 2152 Set device-properties for ACPI-enumerated I2C-attached 2153 touchscreen, to e.g. fix coordinates of upside-down 2154 mounted touchscreens. If you need this option please 2155 submit a drivers/platform/x86/touchscreen_dmi.c patch 2156 adding a DMI quirk for this. 2157 2158 Format: 2159 <ACPI_HW_ID>:<prop_name>=<val>[:prop_name=val][:...] 2160 Where <val> is one of: 2161 Omit "=<val>" entirely Set a boolean device-property 2162 Unsigned number Set a u32 device-property 2163 Anything else Set a string device-property 2164 2165 Examples (split over multiple lines): 2166 i2c_touchscreen_props=GDIX1001:touchscreen-inverted-x: 2167 touchscreen-inverted-y 2168 2169 i2c_touchscreen_props=MSSL1680:touchscreen-size-x=1920: 2170 touchscreen-size-y=1080:touchscreen-inverted-y: 2171 firmware-name=gsl1680-vendor-model.fw:silead,home-button 2172 2173 i8042.debug [HW] Toggle i8042 debug mode 2174 i8042.unmask_kbd_data 2175 [HW] Enable printing of interrupt data from the KBD port 2176 (disabled by default, and as a pre-condition 2177 requires that i8042.debug=1 be enabled) 2178 i8042.direct [HW] Put keyboard port into non-translated mode 2179 i8042.dumbkbd [HW] Pretend that controller can only read data from 2180 keyboard and cannot control its state 2181 (Don't attempt to blink the leds) 2182 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 2183 i8042.nokbd [HW] Don't check/create keyboard port 2184 i8042.noloop [HW] Disable the AUX Loopback command while probing 2185 for the AUX port 2186 i8042.nomux [HW] Don't check presence of an active multiplexing 2187 controller 2188 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 2189 controllers 2190 i8042.notimeout [HW] Ignore timeout condition signalled by controller 2191 i8042.reset [HW] Reset the controller during init, cleanup and 2192 suspend-to-ram transitions, only during s2r 2193 transitions, or never reset 2194 Format: { 1 | Y | y | 0 | N | n } 2195 1, Y, y: always reset controller 2196 0, N, n: don't ever reset controller 2197 Default: only on s2r transitions on x86; most other 2198 architectures force reset to be always executed 2199 i8042.unlock [HW] Unlock (ignore) the keylock 2200 i8042.kbdreset [HW] Reset device connected to KBD port 2201 i8042.probe_defer 2202 [HW] Allow deferred probing upon i8042 probe errors 2203 2204 i810= [HW,DRM] 2205 2206 i915.invert_brightness= 2207 [DRM] Invert the sense of the variable that is used to 2208 set the brightness of the panel backlight. Normally a 2209 brightness value of 0 indicates backlight switched off, 2210 and the maximum of the brightness value sets the backlight 2211 to maximum brightness. If this parameter is set to 0 2212 (default) and the machine requires it, or this parameter 2213 is set to 1, a brightness value of 0 sets the backlight 2214 to maximum brightness, and the maximum of the brightness 2215 value switches the backlight off. 2216 -1 -- never invert brightness 2217 0 -- machine default 2218 1 -- force brightness inversion 2219 2220 ia32_emulation= [X86-64] 2221 Format: <bool> 2222 When true, allows loading 32-bit programs and executing 32-bit 2223 syscalls, essentially overriding IA32_EMULATION_DEFAULT_DISABLED at 2224 boot time. When false, unconditionally disables IA32 emulation. 2225 2226 2227 idle= [X86,EARLY] 2228 Format: idle=poll, idle=halt, idle=nomwait 2229 2230 idle=poll: Don't do power saving in the idle loop 2231 using HLT, but poll for rescheduling event. This will 2232 make the CPUs eat a lot more power, but may be useful 2233 to get slightly better performance in multiprocessor 2234 benchmarks. It also makes some profiling using 2235 performance counters more accurate. Please note that 2236 on systems with MONITOR/MWAIT support (like Intel 2237 EM64T CPUs) this option has no performance advantage 2238 over the normal idle loop. It may also interact badly 2239 with hyperthreading. 2240 2241 idle=halt: Halt is forced to be used for CPU idle. 2242 In such case C2/C3 won't be used again. 2243 2244 idle=nomwait: Disable mwait for CPU C-states 2245 2246 idxd.sva= [HW] 2247 Format: <bool> 2248 Allow force disabling of Shared Virtual Memory (SVA) 2249 support for the idxd driver. By default it is set to 2250 true (1). 2251 2252 idxd.tc_override= [HW] 2253 Format: <bool> 2254 Allow override of default traffic class configuration 2255 for the device. By default it is set to false (0). 2256 2257 ieee754= [MIPS] Select IEEE Std 754 conformance mode 2258 Format: { strict | legacy | 2008 | relaxed | emulated } 2259 Default: strict 2260 2261 Choose which programs will be accepted for execution 2262 based on the IEEE 754 NaN encoding(s) supported by 2263 the FPU and the NaN encoding requested with the value 2264 of an ELF file header flag individually set by each 2265 binary. Hardware implementations are permitted to 2266 support either or both of the legacy and the 2008 NaN 2267 encoding mode. 2268 2269 Available settings are as follows: 2270 strict accept binaries that request a NaN encoding 2271 supported by the FPU 2272 legacy only accept legacy-NaN binaries, if supported 2273 by the FPU 2274 2008 only accept 2008-NaN binaries, if supported 2275 by the FPU 2276 relaxed accept any binaries regardless of whether 2277 supported by the FPU 2278 emulated accept any binaries but enable FPU emulator 2279 if binary mode is unsupported by the FPU. 2280 2281 The FPU emulator is always able to support both NaN 2282 encodings, so if no FPU hardware is present or it has 2283 been disabled with 'nofpu', then the settings of 2284 'legacy' and '2008' strap the emulator accordingly, 2285 'relaxed' straps the emulator for both legacy-NaN and 2286 2008-NaN, whereas 'strict' enables legacy-NaN only on 2287 legacy processors and both NaN encodings on MIPS32 or 2288 MIPS64 CPUs. 2289 2290 The setting for ABS.fmt/NEG.fmt instruction execution 2291 mode generally follows that for the NaN encoding, 2292 except where unsupported by hardware. 2293 2294 ignore_loglevel [KNL,EARLY] 2295 Ignore loglevel setting - this will print /all/ 2296 kernel messages to the console. Useful for debugging. 2297 We also add it as printk module parameter, so users 2298 could change it dynamically, usually by 2299 /sys/module/printk/parameters/ignore_loglevel. 2300 2301 ignore_rlimit_data 2302 Ignore RLIMIT_DATA setting for data mappings, 2303 print warning at first misuse. Can be changed via 2304 /sys/module/kernel/parameters/ignore_rlimit_data. 2305 2306 ihash_entries= [KNL] 2307 Set number of hash buckets for inode cache. 2308 2309 ima_appraise= [IMA] appraise integrity measurements 2310 Format: { "off" | "enforce" | "fix" | "log" } 2311 default: "enforce" 2312 2313 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 2314 The builtin appraise policy appraises all files 2315 owned by uid=0. 2316 2317 ima_canonical_fmt [IMA] 2318 Use the canonical format for the binary runtime 2319 measurements, instead of host native format. 2320 2321 ima_flush_htable [IMA] 2322 Flush the IMA hash table when deleting all the 2323 staged measurement records, to achieve maximum 2324 memory saving at the cost of having duplicate 2325 records across the staged measurement lists. 2326 2327 ima_hash= [IMA] 2328 Format: { md5 | sha1 | rmd160 | sha256 | sha384 2329 | sha512 | ... } 2330 default: "sha1" 2331 2332 The list of supported hash algorithms is defined 2333 in crypto/hash_info.h. 2334 2335 ima_policy= [IMA] 2336 The builtin policies to load during IMA setup. 2337 Format: "tcb | appraise_tcb | secure_boot | 2338 fail_securely | critical_data" 2339 2340 The "tcb" policy measures all programs exec'd, files 2341 mmap'd for exec, and all files opened with the read 2342 mode bit set by either the effective uid (euid=0) or 2343 uid=0. 2344 2345 The "appraise_tcb" policy appraises the integrity of 2346 all files owned by root. 2347 2348 The "secure_boot" policy appraises the integrity 2349 of files (eg. kexec kernel image, kernel modules, 2350 firmware, policy, etc) based on file signatures. 2351 2352 The "fail_securely" policy forces file signature 2353 verification failure also on privileged mounted 2354 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 2355 flag. 2356 2357 The "critical_data" policy measures kernel integrity 2358 critical data. 2359 2360 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 2361 Load a policy which meets the needs of the Trusted 2362 Computing Base. This means IMA will measure all 2363 programs exec'd, files mmap'd for exec, and all files 2364 opened for read by uid=0. 2365 2366 ima_template= [IMA] 2367 Select one of defined IMA measurements template formats. 2368 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" | 2369 "ima-sigv2" } 2370 Default: "ima-ng" 2371 2372 ima_template_fmt= 2373 [IMA] Define a custom template format. 2374 Format: { "field1|...|fieldN" } 2375 2376 ima= [IMA] Enable or disable IMA 2377 Format: { "off" | "on" } 2378 Default: "on" 2379 Note that disabling IMA is limited to kdump kernel. 2380 2381 indirect_target_selection= [X86,Intel] Mitigation control for Indirect 2382 Target Selection(ITS) bug in Intel CPUs. Updated 2383 microcode is also required for a fix in IBPB. 2384 2385 on: Enable mitigation (default). 2386 off: Disable mitigation. 2387 force: Force the ITS bug and deploy default 2388 mitigation. 2389 vmexit: Only deploy mitigation if CPU is affected by 2390 guest/host isolation part of ITS. 2391 stuff: Deploy RSB-fill mitigation when retpoline is 2392 also deployed. Otherwise, deploy the default 2393 mitigation. 2394 2395 For details see: 2396 Documentation/admin-guide/hw-vuln/indirect-target-selection.rst 2397 2398 init= [KNL] 2399 Format: <full_path> 2400 Run specified binary instead of /sbin/init as init 2401 process. 2402 2403 initcall_debug [KNL] Trace initcalls as they are executed. Useful 2404 for working out where the kernel is dying during 2405 startup. 2406 2407 initcall_blacklist= [KNL] Do not execute a comma-separated list of 2408 initcall functions. Useful for debugging built-in 2409 modules and initcalls. 2410 2411 initramfs_async= [KNL] 2412 Format: <bool> 2413 Default: 1 2414 This parameter controls whether the initramfs 2415 image is unpacked asynchronously, concurrently 2416 with devices being probed and 2417 initialized. This should normally just work, 2418 but as a debugging aid, one can get the 2419 historical behaviour of the initramfs 2420 unpacking being completed before device_ and 2421 late_ initcalls. 2422 2423 initrd= [BOOT,EARLY] Specify the location of the initial ramdisk 2424 2425 initrdmem= [KNL,EARLY] Specify a physical address and size from which to 2426 load the initrd. If an initrd is compiled in or 2427 specified in the bootparams, it takes priority over this 2428 setting. 2429 Format: ss[KMG],nn[KMG] 2430 Default is 0, 0 2431 2432 init_on_alloc= [MM,EARLY] Fill newly allocated pages and heap objects with 2433 zeroes. 2434 Format: 0 | 1 2435 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 2436 2437 init_on_free= [MM,EARLY] Fill freed pages and heap objects with zeroes. 2438 Format: 0 | 1 2439 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 2440 2441 init_pkru= [X86] Specify the default memory protection keys rights 2442 register contents for all processes. 0x55555554 by 2443 default (disallow access to all but pkey 0). Can 2444 override in debugfs after boot. 2445 2446 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 2447 Format: <irq> 2448 2449 int_pln_enable [X86] Enable power limit notification interrupt 2450 2451 integrity_audit=[IMA] 2452 Format: { "0" | "1" } 2453 0 -- basic integrity auditing messages. (Default) 2454 1 -- additional integrity auditing messages. 2455 2456 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 2457 on 2458 Enable intel iommu driver. 2459 off 2460 Disable intel iommu driver. 2461 igfx_off [Default Off] 2462 By default, gfx is mapped as normal device. If a gfx 2463 device has a dedicated DMAR unit, the DMAR unit is 2464 bypassed by not enabling DMAR with this option. In 2465 this case, gfx device will use physical address for 2466 DMA. 2467 strict [Default Off] 2468 Deprecated, equivalent to iommu.strict=1. 2469 sp_off [Default Off] 2470 By default, super page will be supported if Intel IOMMU 2471 has the capability. With this option, super page will 2472 not be supported. 2473 sm_on 2474 Enable the Intel IOMMU scalable mode if the hardware 2475 advertises that it has support for the scalable mode 2476 translation. 2477 sm_off 2478 Disallow use of the Intel IOMMU scalable mode. 2479 tboot_noforce [Default Off] 2480 Do not force the Intel IOMMU enabled under tboot. 2481 By default, tboot will force Intel IOMMU on, which 2482 could harm performance of some high-throughput 2483 devices like 40GBit network cards, even if identity 2484 mapping is enabled. 2485 Note that using this option lowers the security 2486 provided by tboot because it makes the system 2487 vulnerable to DMA attacks. 2488 2489 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 2490 0 disables intel_idle and fall back on acpi_idle. 2491 1 to 9 specify maximum depth of C-state. 2492 2493 intel_pstate= [X86,EARLY] 2494 disable 2495 Do not enable intel_pstate as the default 2496 scaling driver for the supported processors 2497 active 2498 Use intel_pstate driver to bypass the scaling 2499 governors layer of cpufreq and provides it own 2500 algorithms for p-state selection. There are two 2501 P-state selection algorithms provided by 2502 intel_pstate in the active mode: powersave and 2503 performance. The way they both operate depends 2504 on whether or not the hardware managed P-states 2505 (HWP) feature has been enabled in the processor 2506 and possibly on the processor model. 2507 passive 2508 Use intel_pstate as a scaling driver, but configure it 2509 to work with generic cpufreq governors (instead of 2510 enabling its internal governor). This mode cannot be 2511 used along with the hardware-managed P-states (HWP) 2512 feature. 2513 force 2514 Enable intel_pstate on systems that prohibit it by default 2515 in favor of acpi-cpufreq. Forcing the intel_pstate driver 2516 instead of acpi-cpufreq may disable platform features, such 2517 as thermal controls and power capping, that rely on ACPI 2518 P-States information being indicated to OSPM and therefore 2519 should be used with caution. This option does not work with 2520 processors that aren't supported by the intel_pstate driver 2521 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 2522 no_hwp 2523 Do not enable hardware P state control (HWP) 2524 if available. 2525 hwp_only 2526 Only load intel_pstate on systems which support 2527 hardware P state control (HWP) if available. 2528 support_acpi_ppc 2529 Enforce ACPI _PPC performance limits. If the Fixed ACPI 2530 Description Table, specifies preferred power management 2531 profile as "Enterprise Server" or "Performance Server", 2532 then this feature is turned on by default. 2533 per_cpu_perf_limits 2534 Allow per-logical-CPU P-State performance control limits using 2535 cpufreq sysfs interface 2536 no_cas 2537 Do not enable capacity-aware scheduling (CAS) on 2538 hybrid systems 2539 2540 intremap= [X86-64,Intel-IOMMU,EARLY] 2541 on enable Interrupt Remapping (default) 2542 off disable Interrupt Remapping 2543 nosid disable Source ID checking 2544 no_x2apic_optout 2545 BIOS x2APIC opt-out request will be ignored 2546 nopost disable Interrupt Posting 2547 posted_msi 2548 enable MSIs delivered as posted interrupts 2549 2550 iomem= Disable strict checking of access to MMIO memory 2551 strict regions from userspace. 2552 relaxed 2553 2554 iommu= [X86,EARLY] 2555 2556 off 2557 Don't initialize and use any kind of IOMMU. 2558 2559 force 2560 Force the use of the hardware IOMMU even when 2561 it is not actually needed (e.g. because < 3 GB 2562 memory). 2563 2564 noforce 2565 Don't force hardware IOMMU usage when it is not 2566 needed. (default). 2567 2568 merge 2569 Do scatter-gather (SG) merging. Implies "force" 2570 (experimental). 2571 2572 nomerge 2573 Don't do scatter-gather (SG) merging. 2574 2575 biomerge 2576 Do scatter-gather (SG) merging. Implies "force" 2577 (experimental). [same as "merge"] 2578 2579 panic 2580 Always panic when IOMMU overflows. 2581 2582 nopanic 2583 Don't panic on IOMMU overflows. 2584 2585 pt 2586 Use passththrough mode by default 2587 (Equivalent to iommu.passthrough=1) 2588 2589 nopt 2590 Use translated mode for DMA by default 2591 (Equivalent to iommu.passthrough=0) 2592 2593 soft 2594 Use software bounce buffering (SWIOTLB) (default for 2595 Intel machines). This can be used to prevent the usage 2596 of an available hardware IOMMU. 2597 2598 usedac 2599 Use the DAC on VIA PCI bridge 2600 (default: disable the VIA PCI bridge DAC) 2601 2602 AMD Gart HW IOMMU-specific options: (CONFIG_GART_IOMMU) 2603 2604 <size> 2605 Set the size of the remapping area in bytes. 2606 2607 allowed 2608 Overwrite iommu off workarounds for specific chipsets 2609 2610 force 2611 Overwrite iommu off workarounds for specific chipsets 2612 2613 fullflush 2614 Flush IOMMU on each allocation (default). 2615 2616 nofullflush 2617 Don't use IOMMU fullflush. 2618 2619 memaper[=<order>] 2620 Allocate an own aperture over RAM with size 2621 32MB<<order. (default: order=1, i.e. 64MB) 2622 2623 noaperture 2624 Ask the IOMMU not to touch the aperture for AGP. 2625 2626 noagp 2627 Don't initialize the AGP driver and use full aperture. 2628 2629 iommu= [PPC/POWERNV] 2630 2631 nobypass 2632 Disable IOMMU bypass, using IOMMU for PCI devices. 2633 2634 iommu.forcedac= [ARM64,X86,EARLY] Control IOVA allocation for PCI devices. 2635 Format: { "0" | "1" } 2636 0 - Try to allocate a 32-bit DMA address first, before 2637 falling back to the full range if needed. 2638 1 - Allocate directly from the full usable range, 2639 forcing Dual Address Cycle for PCI cards supporting 2640 greater than 32-bit addressing. 2641 2642 iommu.strict= [ARM64,X86,S390,EARLY] Configure TLB invalidation behaviour 2643 Format: { "0" | "1" } 2644 0 - Lazy mode. 2645 Request that DMA unmap operations use deferred 2646 invalidation of hardware TLBs, for increased 2647 throughput at the cost of reduced device isolation. 2648 Will fall back to strict mode if not supported by 2649 the relevant IOMMU driver. 2650 1 - Strict mode. 2651 DMA unmap operations invalidate IOMMU hardware TLBs 2652 synchronously. 2653 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}. 2654 Note: on x86, strict mode specified via one of the 2655 legacy driver-specific options takes precedence. 2656 2657 iommu.passthrough= 2658 [ARM64,X86,EARLY] Configure DMA to bypass the IOMMU by default. 2659 Format: { "0" | "1" } 2660 0 - Use IOMMU translation for DMA. 2661 1 - Bypass the IOMMU for DMA. 2662 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2663 2664 iommu.debug_pagealloc= 2665 [KNL,EARLY] When CONFIG_IOMMU_DEBUG_PAGEALLOC is set, this 2666 parameter enables the feature at boot time. By default, it 2667 is disabled and the system behaves the same way as a kernel 2668 built without CONFIG_IOMMU_DEBUG_PAGEALLOC. 2669 Format: { "0" | "1" } 2670 0 - Sanitizer disabled. 2671 1 - Sanitizer enabled, expect runtime overhead. 2672 2673 io7= [HW] IO7 for Marvel-based Alpha systems 2674 See comment before marvel_specify_io7 in 2675 arch/alpha/kernel/core_marvel.c. 2676 2677 io_delay= [X86,EARLY] I/O delay method 2678 0x80 2679 Standard port 0x80 based delay 2680 0xed 2681 Alternate port 0xed based delay (needed on some systems) 2682 udelay 2683 Simple two microseconds delay 2684 none 2685 No delay 2686 2687 ip= [IP_PNP] 2688 See Documentation/admin-guide/nfs/nfsroot.rst. 2689 2690 ipcmni_extend [KNL,EARLY] Extend the maximum number of unique System V 2691 IPC identifiers from 32,768 to 16,777,216. 2692 2693 ipe.enforce= [IPE] 2694 Format: <bool> 2695 Determine whether IPE starts in permissive (0) or 2696 enforce (1) mode. The default is enforce. 2697 2698 ipe.success_audit= 2699 [IPE] 2700 Format: <bool> 2701 Start IPE with success auditing enabled, emitting 2702 an audit event when a binary is allowed. The default 2703 is 0. 2704 2705 irqaffinity= [SMP] Set the default irq affinity mask 2706 The argument is a cpu list, as described above. 2707 2708 irqchip.gicv2_force_probe= 2709 [ARM,ARM64,EARLY] 2710 Format: <bool> 2711 Force the kernel to look for the second 4kB page 2712 of a GICv2 controller even if the memory range 2713 exposed by the device tree is too small. 2714 2715 irqchip.gicv3_nolpi= 2716 [ARM,ARM64,EARLY] 2717 Force the kernel to ignore the availability of 2718 LPIs (and by consequence ITSs). Intended for system 2719 that use the kernel as a bootloader, and thus want 2720 to let secondary kernels in charge of setting up 2721 LPIs. 2722 2723 irqchip.gicv3_pseudo_nmi= [ARM64,EARLY] 2724 Enables support for pseudo-NMIs in the kernel. This 2725 requires the kernel to be built with 2726 CONFIG_ARM64_PSEUDO_NMI. 2727 2728 irqchip.riscv_imsic_noipi 2729 [RISC-V,EARLY] 2730 Force the kernel to not use IMSIC software injected MSIs 2731 as IPIs. Intended for system where IMSIC is trap-n-emulated, 2732 and thus want to reduce MMIO traps when triggering IPIs 2733 to multiple harts. 2734 2735 irqfixup [HW] 2736 When an interrupt is not handled search all handlers 2737 for it. Intended to get systems with badly broken 2738 firmware running. 2739 2740 irqhandler.duration_warn_us= [KNL] 2741 Warn if an IRQ handler exceeds the specified duration 2742 threshold in microseconds. Useful for identifying 2743 long-running IRQs in the system. 2744 2745 irqpoll [HW] 2746 When an interrupt is not handled search all handlers 2747 for it. Also check all handlers each timer 2748 interrupt. Intended to get systems with badly broken 2749 firmware running. 2750 2751 isapnp= [ISAPNP] 2752 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2753 2754 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2755 Format: [flag-list,]<cpu-list> 2756 2757 Specify one or more CPUs to isolate from disturbances 2758 specified in the flag list (default: domain): 2759 2760 nohz 2761 Disable the tick when a single task runs as well as 2762 disabling other kernel noises like having RCU callbacks 2763 offloaded. This is equivalent to the nohz_full parameter. 2764 2765 A residual 1Hz tick is offloaded to workqueues, which you 2766 need to affine to housekeeping through the global 2767 workqueue's affinity configured via the 2768 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2769 by using the 'domain' flag described below. 2770 2771 NOTE: by default the global workqueue runs on all CPUs, 2772 so to protect individual CPUs the 'cpumask' file has to 2773 be configured manually after bootup. 2774 2775 domain 2776 Isolate from the general SMP balancing and scheduling 2777 algorithms. Note that performing domain isolation this way 2778 is irreversible: it's not possible to bring back a CPU to 2779 the domains once isolated through this boot time 2780 configuration. Use cpusets for a dynamic configuration 2781 which can be altered at runtime. For details see 2782 Documentation/admin-guide/cpu-isolation.rst. 2783 2784 You can move a process onto or off an "isolated" CPU via 2785 the CPU affinity syscalls or cpuset. 2786 <cpu number> begins at 0 and the maximum value is 2787 "number of CPUs in system - 1". 2788 2789 managed_irq 2790 2791 Isolate from being targeted by managed interrupts 2792 which have an interrupt mask containing isolated 2793 CPUs. The affinity of managed interrupts is 2794 handled by the kernel and cannot be changed via 2795 the /proc/irq/* interfaces. 2796 2797 This isolation is best effort and only effective 2798 if the automatically assigned interrupt mask of a 2799 device queue contains isolated and housekeeping 2800 CPUs. If housekeeping CPUs are online then such 2801 interrupts are directed to the housekeeping CPU 2802 so that IO submitted on the housekeeping CPU 2803 cannot disturb the isolated CPU. 2804 2805 If a queue's affinity mask contains only isolated 2806 CPUs then this parameter has no effect on the 2807 interrupt routing decision, though interrupts are 2808 only delivered when tasks running on those 2809 isolated CPUs submit IO. IO submitted on 2810 housekeeping CPUs has no influence on those 2811 queues. 2812 2813 The format of <cpu-list> is described above. 2814 2815 iucv= [HW,NET] 2816 2817 ivrs_ioapic [HW,X86-64] 2818 Provide an override to the IOAPIC-ID<->DEVICE-ID 2819 mapping provided in the IVRS ACPI table. 2820 By default, PCI segment is 0, and can be omitted. 2821 2822 For example, to map IOAPIC-ID decimal 10 to 2823 PCI segment 0x1 and PCI device 00:14.0, 2824 write the parameter as: 2825 ivrs_ioapic=10@0001:00:14.0 2826 2827 Deprecated formats: 2828 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0 2829 write the parameter as: 2830 ivrs_ioapic[10]=00:14.0 2831 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and 2832 PCI device 00:14.0 write the parameter as: 2833 ivrs_ioapic[10]=0001:00:14.0 2834 2835 ivrs_hpet [HW,X86-64] 2836 Provide an override to the HPET-ID<->DEVICE-ID 2837 mapping provided in the IVRS ACPI table. 2838 By default, PCI segment is 0, and can be omitted. 2839 2840 For example, to map HPET-ID decimal 10 to 2841 PCI segment 0x1 and PCI device 00:14.0, 2842 write the parameter as: 2843 ivrs_hpet=10@0001:00:14.0 2844 2845 Deprecated formats: 2846 * To map HPET-ID decimal 0 to PCI device 00:14.0 2847 write the parameter as: 2848 ivrs_hpet[0]=00:14.0 2849 * To map HPET-ID decimal 10 to PCI segment 0x1 and 2850 PCI device 00:14.0 write the parameter as: 2851 ivrs_ioapic[10]=0001:00:14.0 2852 2853 ivrs_acpihid [HW,X86-64] 2854 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2855 mapping provided in the IVRS ACPI table. 2856 By default, PCI segment is 0, and can be omitted. 2857 2858 For example, to map UART-HID:UID AMD0020:0 to 2859 PCI segment 0x1 and PCI device ID 00:14.5, 2860 write the parameter as: 2861 ivrs_acpihid=AMD0020:0@0001:00:14.5 2862 2863 Deprecated formats: 2864 * To map UART-HID:UID AMD0020:0 to PCI segment is 0, 2865 PCI device ID 00:14.5, write the parameter as: 2866 ivrs_acpihid[00:14.5]=AMD0020:0 2867 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and 2868 PCI device ID 00:14.5, write the parameter as: 2869 ivrs_acpihid[0001:00:14.5]=AMD0020:0 2870 2871 js= [HW,JOY] Analog joystick 2872 See Documentation/input/joydev/joystick.rst. 2873 2874 kasan_multi_shot 2875 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2876 report on every invalid memory access. Without this 2877 parameter KASAN will print report only for the first 2878 invalid access. 2879 2880 keep_bootcon [KNL,EARLY] 2881 Do not unregister boot console at start. This is only 2882 useful for debugging when something happens in the window 2883 between unregistering the boot console and initializing 2884 the real console. 2885 2886 keepinitrd [HW,ARM] See retain_initrd. 2887 2888 kernelcore= [KNL,X86,PPC,EARLY] 2889 Format: nn[KMGTPE] | nn% | "mirror" 2890 This parameter specifies the amount of memory usable by 2891 the kernel for non-movable allocations. The requested 2892 amount is spread evenly throughout all nodes in the 2893 system as ZONE_NORMAL. The remaining memory is used for 2894 movable memory in its own zone, ZONE_MOVABLE. In the 2895 event, a node is too small to have both ZONE_NORMAL and 2896 ZONE_MOVABLE, kernelcore memory will take priority and 2897 other nodes will have a larger ZONE_MOVABLE. 2898 2899 ZONE_MOVABLE is used for the allocation of pages that 2900 may be reclaimed or moved by the page migration 2901 subsystem. Note that allocations like PTEs-from-HighMem 2902 still use the HighMem zone if it exists, and the Normal 2903 zone if it does not. 2904 2905 It is possible to specify the exact amount of memory in 2906 the form of "nn[KMGTPE]", a percentage of total system 2907 memory in the form of "nn%", or "mirror". If "mirror" 2908 option is specified, mirrored (reliable) memory is used 2909 for non-movable allocations and remaining memory is used 2910 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2911 are exclusive, so you cannot specify multiple forms. 2912 2913 kfence.burst= [MM,KFENCE] The number of additional successive 2914 allocations to be attempted through KFENCE for each 2915 sample interval. 2916 Format: <unsigned integer> 2917 Default: 0 2918 2919 kfence.check_on_panic= 2920 [MM,KFENCE] Whether to check all KFENCE-managed objects' 2921 canaries on panic. 2922 Format: <bool> 2923 Default: false 2924 2925 kfence.deferrable= 2926 [MM,KFENCE] Whether to use a deferrable timer to trigger 2927 allocations. This avoids forcing CPU wake-ups if the 2928 system is idle, at the risk of a less predictable 2929 sample interval. 2930 Format: <bool> 2931 Default: CONFIG_KFENCE_DEFERRABLE 2932 2933 kfence.fault= [MM,KFENCE] Controls the behavior when a KFENCE 2934 error is detected. 2935 report - print the error report and continue (default). 2936 oops - print the error report and oops. 2937 panic - print the error report and panic. 2938 2939 kfence.sample_interval= 2940 [MM,KFENCE] KFENCE's sample interval in milliseconds. 2941 Format: <unsigned integer> 2942 0 - Disable KFENCE. 2943 >0 - Enabled KFENCE with given sample interval. 2944 Default: CONFIG_KFENCE_SAMPLE_INTERVAL 2945 2946 kfence.skip_covered_thresh= 2947 [MM,KFENCE] If pool utilization reaches this threshold 2948 (pool usage%), KFENCE limits currently covered 2949 allocations of the same source from further filling 2950 up the pool. 2951 Format: <unsigned integer> 2952 Default: 75 2953 2954 kgdbdbgp= [KGDB,HW,EARLY] kgdb over EHCI usb debug port. 2955 Format: <Controller#>[,poll interval] 2956 The controller # is the number of the ehci usb debug 2957 port as it is probed via PCI. The poll interval is 2958 optional and is the number seconds in between 2959 each poll cycle to the debug port in case you need 2960 the functionality for interrupting the kernel with 2961 gdb or control-c on the dbgp connection. When 2962 not using this parameter you use sysrq-g to break into 2963 the kernel debugger. 2964 2965 kgdboc= [KGDB,HW] kgdb over consoles. 2966 Requires a tty driver that supports console polling, 2967 or a supported polling keyboard driver (non-usb). 2968 Serial only format: <serial_device>[,baud] 2969 keyboard only format: kbd 2970 keyboard and serial format: kbd,<serial_device>[,baud] 2971 Optional Kernel mode setting: 2972 kms, kbd format: kms,kbd 2973 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2974 2975 kgdboc_earlycon= [KGDB,HW,EARLY] 2976 If the boot console provides the ability to read 2977 characters and can work in polling mode, you can use 2978 this parameter to tell kgdb to use it as a backend 2979 until the normal console is registered. Intended to 2980 be used together with the kgdboc parameter which 2981 specifies the normal console to transition to. 2982 2983 The name of the early console should be specified 2984 as the value of this parameter. Note that the name of 2985 the early console might be different than the tty 2986 name passed to kgdboc. It's OK to leave the value 2987 blank and the first boot console that implements 2988 read() will be picked. 2989 2990 kgdbwait [KGDB,EARLY] Stop kernel execution and enter the 2991 kernel debugger at the earliest opportunity. 2992 2993 kho= [KEXEC,EARLY] 2994 Format: { "0" | "1" | "off" | "on" | "y" | "n" } 2995 Enables or disables Kexec HandOver. 2996 "0" | "off" | "n" - kexec handover is disabled 2997 "1" | "on" | "y" - kexec handover is enabled 2998 2999 kho_scratch= [KEXEC,EARLY] 3000 Format: ll[KMG],mm[KMG],nn[KMG] | nn% 3001 Defines the size of the KHO scratch region. The KHO 3002 scratch regions are physically contiguous memory 3003 ranges that can only be used for non-kernel 3004 allocations. That way, even when memory is heavily 3005 fragmented with handed over memory, the kexeced 3006 kernel will always have enough contiguous ranges to 3007 bootstrap itself. 3008 3009 It is possible to specify the exact amount of 3010 memory in the form of "ll[KMG],mm[KMG],nn[KMG]" 3011 where the first parameter defines the size of a low 3012 memory scratch area, the second parameter defines 3013 the size of a global scratch area and the third 3014 parameter defines the size of additional per-node 3015 scratch areas. The form "nn%" defines scale factor 3016 (in percents) of memory that was used during boot. 3017 3018 kmac= [MIPS] Korina ethernet MAC address. 3019 Configure the RouterBoard 532 series on-chip 3020 Ethernet adapter MAC address. 3021 3022 kmemleak= [KNL,EARLY] Boot-time kmemleak enable/disable 3023 Valid arguments: on, off 3024 Default: on 3025 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 3026 the default is off. 3027 3028 kprobe_event=[probe-list] 3029 [FTRACE] Add kprobe events and enable at boot time. 3030 The probe-list is a semicolon delimited list of probe 3031 definitions. Each definition is same as kprobe_events 3032 interface, but the parameters are comma delimited. 3033 For example, to add a kprobe event on vfs_read with 3034 arg1 and arg2, add to the command line; 3035 3036 kprobe_event=p,vfs_read,$arg1,$arg2 3037 3038 See also Documentation/trace/kprobetrace.rst "Kernel 3039 Boot Parameter" section. 3040 3041 kpti= [ARM64,EARLY] Control page table isolation of 3042 user and kernel address spaces. 3043 Default: enabled on cores which need mitigation. 3044 0: force disabled 3045 1: force enabled 3046 3047 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires 3048 CONFIG_KUNIT to be set to be fully enabled. The 3049 default value can be overridden via 3050 KUNIT_DEFAULT_ENABLED. 3051 Default is 1 (enabled) 3052 3053 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 3054 Default is 0 (don't ignore, but inject #GP) 3055 3056 kvm.eager_page_split= 3057 [KVM,X86] Controls whether or not KVM will try to 3058 proactively split all huge pages during dirty logging. 3059 Eager page splitting reduces interruptions to vCPU 3060 execution by eliminating the write-protection faults 3061 and MMU lock contention that would otherwise be 3062 required to split huge pages lazily. 3063 3064 VM workloads that rarely perform writes or that write 3065 only to a small region of VM memory may benefit from 3066 disabling eager page splitting to allow huge pages to 3067 still be used for reads. 3068 3069 The behavior of eager page splitting depends on whether 3070 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If 3071 disabled, all huge pages in a memslot will be eagerly 3072 split when dirty logging is enabled on that memslot. If 3073 enabled, eager page splitting will be performed during 3074 the KVM_CLEAR_DIRTY ioctl, and only for the pages being 3075 cleared. 3076 3077 Eager page splitting is only supported when kvm.tdp_mmu=Y. 3078 3079 Default is Y (on). 3080 3081 kvm.enable_pmu=[KVM,X86] 3082 If enabled, KVM will virtualize PMU functionality based 3083 on the virtual CPU model defined by userspace. This 3084 can be overridden on a per-VM basis via 3085 KVM_CAP_PMU_CAPABILITY. 3086 3087 If disabled, KVM will not virtualize PMU functionality, 3088 e.g. MSRs, PMCs, PMIs, etc., even if userspace defines 3089 a virtual CPU model that contains PMU assets. 3090 3091 Note, KVM's vPMU support implicitly requires running 3092 with an in-kernel local APIC, e.g. to deliver PMIs to 3093 the guest. Running without an in-kernel local APIC is 3094 not supported, though KVM will allow such a combination 3095 (with severely degraded functionality). 3096 3097 See also enable_mediated_pmu. 3098 3099 Default is Y (on). 3100 3101 kvm.enable_virt_at_load=[KVM,ARM64,LOONGARCH,MIPS,RISCV,X86] 3102 If enabled, KVM will enable virtualization in hardware 3103 when KVM is loaded, and disable virtualization when KVM 3104 is unloaded (if KVM is built as a module). 3105 3106 If disabled, KVM will dynamically enable and disable 3107 virtualization on-demand when creating and destroying 3108 VMs, i.e. on the 0=>1 and 1=>0 transitions of the 3109 number of VMs. 3110 3111 Enabling virtualization at module load avoids potential 3112 latency for creation of the 0=>1 VM, as KVM serializes 3113 virtualization enabling across all online CPUs. The 3114 "cost" of enabling virtualization when KVM is loaded, 3115 is that doing so may interfere with using out-of-tree 3116 hypervisors that want to "own" virtualization hardware. 3117 3118 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 3119 Default is false (don't support). 3120 3121 kvm.nx_huge_pages= 3122 [KVM] Controls the software workaround for the 3123 X86_BUG_ITLB_MULTIHIT bug. 3124 force : Always deploy workaround. 3125 off : Never deploy workaround. 3126 auto : Deploy workaround based on the presence of 3127 X86_BUG_ITLB_MULTIHIT. 3128 3129 Default is 'auto'. 3130 3131 If the software workaround is enabled for the host, 3132 guests do need not to enable it for nested guests. 3133 3134 kvm.nx_huge_pages_recovery_ratio= 3135 [KVM] Controls how many 4KiB pages are periodically zapped 3136 back to huge pages. 0 disables the recovery, otherwise if 3137 the value is N KVM will zap 1/Nth of the 4KiB pages every 3138 period (see below). The default is 60. 3139 3140 kvm.nx_huge_pages_recovery_period_ms= 3141 [KVM] Controls the time period at which KVM zaps 4KiB pages 3142 back to huge pages. If the value is a non-zero N, KVM will 3143 zap a portion (see ratio above) of the pages every N msecs. 3144 If the value is 0 (the default), KVM will pick a period based 3145 on the ratio, such that a page is zapped after 1 hour on average. 3146 3147 kvm-{amd,intel}.enable_mediated_pmu=[KVM,AMD,INTEL] 3148 If enabled, KVM will provide a mediated virtual PMU, 3149 instead of the default perf-based virtual PMU (if 3150 kvm.enable_pmu is true and PMU is enumerated via the 3151 virtual CPU model). 3152 3153 With a perf-based vPMU, KVM operates as a user of perf, 3154 i.e. emulates guest PMU counters using perf events. 3155 KVM-created perf events are managed by perf as regular 3156 (guest-only) events, e.g. are scheduled in/out, contend 3157 for hardware resources, etc. Using a perf-based vPMU 3158 allows guest and host usage of the PMU to co-exist, but 3159 incurs non-trivial overhead and can result in silently 3160 dropped guest events (due to resource contention). 3161 3162 With a mediated vPMU, hardware PMU state is context 3163 switched around the world switch to/from the guest. 3164 KVM mediates which events the guest can utilize, but 3165 gives the guest direct access to all other PMU assets 3166 when possible (KVM may intercept some accesses if the 3167 virtual CPU model provides a subset of hardware PMU 3168 functionality). Using a mediated vPMU significantly 3169 reduces PMU virtualization overhead and eliminates lost 3170 guest events, but is mutually exclusive with using perf 3171 to profile KVM guests and adds latency to most VM-Exits 3172 (to context switch PMU state). 3173 3174 Default is N (off). 3175 3176 kvm-amd.nested= [KVM,AMD] Control nested virtualization feature in 3177 KVM/SVM. Default is 1 (enabled). 3178 3179 kvm-amd.npt= [KVM,AMD] Control KVM's use of Nested Page Tables, 3180 a.k.a. Two-Dimensional Page Tables. Default is 1 3181 (enabled). Disable by KVM if hardware lacks support 3182 for NPT. 3183 3184 kvm-amd.ciphertext_hiding_asids= 3185 [KVM,AMD] Ciphertext hiding prevents disallowed accesses 3186 to SNP private memory from reading ciphertext. Instead, 3187 reads will see constant default values (0xff). 3188 3189 If ciphertext hiding is enabled, the joint SEV-ES and 3190 SEV-SNP ASID space is partitioned into separate SEV-ES 3191 and SEV-SNP ASID ranges, with the SEV-SNP range being 3192 [1..max_snp_asid] and the SEV-ES range being 3193 (max_snp_asid..min_sev_asid), where min_sev_asid is 3194 enumerated by CPUID.0x.8000_001F[EDX]. 3195 3196 A non-zero value enables SEV-SNP ciphertext hiding and 3197 adjusts the ASID ranges for SEV-ES and SEV-SNP guests. 3198 KVM caps the number of SEV-SNP ASIDs at the maximum 3199 possible value, e.g. specifying -1u will assign all 3200 joint SEV-ES and SEV-SNP ASIDs to SEV-SNP. Note, 3201 assigning all joint ASIDs to SEV-SNP, i.e. configuring 3202 max_snp_asid == min_sev_asid-1, will effectively make 3203 SEV-ES unusable. 3204 3205 kvm-arm.mode= 3206 [KVM,ARM,EARLY] Select one of KVM/arm64's modes of 3207 operation. 3208 3209 none: Forcefully disable KVM. 3210 3211 nvhe: Standard nVHE-based mode, without support for 3212 protected guests. 3213 3214 protected: Mode with support for guests whose state is 3215 kept private from the host, using VHE or 3216 nVHE depending on HW support. 3217 3218 nested: VHE-based mode with support for nested 3219 virtualization. Requires at least ARMv8.4 3220 hardware (with FEAT_NV2). 3221 3222 Defaults to VHE/nVHE based on hardware support. Setting 3223 mode to "protected" will disable kexec and hibernation 3224 for the host. To force nVHE on VHE hardware, add 3225 "arm64_sw.hvhe=0 id_aa64mmfr1.vh=0" to the 3226 command-line. 3227 "nested" and "protected" are experimental and should be 3228 used with extreme caution. 3229 3230 kvm-arm.vgic_v3_group0_trap= 3231 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-0 3232 system registers 3233 3234 kvm-arm.vgic_v3_group1_trap= 3235 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-1 3236 system registers 3237 3238 kvm-arm.vgic_v3_common_trap= 3239 [KVM,ARM,EARLY] Trap guest accesses to GICv3 common 3240 system registers 3241 3242 kvm-arm.vgic_v4_enable= 3243 [KVM,ARM,EARLY] Allow use of GICv4 for direct 3244 injection of LPIs. 3245 3246 kvm-arm.wfe_trap_policy= 3247 [KVM,ARM] Control when to set WFE instruction trap for 3248 KVM VMs. Traps are allowed but not guaranteed by the 3249 CPU architecture. 3250 3251 trap: set WFE instruction trap 3252 3253 notrap: clear WFE instruction trap 3254 3255 kvm-arm.wfi_trap_policy= 3256 [KVM,ARM] Control when to set WFI instruction trap for 3257 KVM VMs. Traps are allowed but not guaranteed by the 3258 CPU architecture. 3259 3260 trap: set WFI instruction trap 3261 3262 notrap: clear WFI instruction trap 3263 3264 kvm_cma_resv_ratio=n [PPC,EARLY] 3265 Reserves given percentage from system memory area for 3266 contiguous memory allocation for KVM hash pagetable 3267 allocation. 3268 By default it reserves 5% of total system memory. 3269 Format: <integer> 3270 Default: 5 3271 3272 kvm-intel.ept= [KVM,Intel] Control KVM's use of Extended Page Tables, 3273 a.k.a. Two-Dimensional Page Tables. Default is 1 3274 (enabled). Disable by KVM if hardware lacks support 3275 for EPT. 3276 3277 kvm-intel.emulate_invalid_guest_state= 3278 [KVM,Intel] Control whether to emulate invalid guest 3279 state. Ignored if kvm-intel.enable_unrestricted_guest=1, 3280 as guest state is never invalid for unrestricted 3281 guests. This param doesn't apply to nested guests (L2), 3282 as KVM never emulates invalid L2 guest state. 3283 Default is 1 (enabled). 3284 3285 kvm-intel.flexpriority= 3286 [KVM,Intel] Control KVM's use of FlexPriority feature 3287 (TPR shadow). Default is 1 (enabled). Disable by KVM if 3288 hardware lacks support for it. 3289 3290 kvm-intel.nested= 3291 [KVM,Intel] Control nested virtualization feature in 3292 KVM/VMX. Default is 1 (enabled). 3293 3294 kvm-intel.unrestricted_guest= 3295 [KVM,Intel] Control KVM's use of unrestricted guest 3296 feature (virtualized real and unpaged mode). Default 3297 is 1 (enabled). Disable by KVM if EPT is disabled or 3298 hardware lacks support for it. 3299 3300 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 3301 CVE-2018-3620. 3302 3303 Valid arguments: never, cond, always 3304 3305 always: L1D cache flush on every VMENTER. 3306 cond: Flush L1D on VMENTER only when the code between 3307 VMEXIT and VMENTER can leak host memory. 3308 never: Disables the mitigation 3309 3310 Default is cond (do L1 cache flush in specific instances) 3311 3312 kvm-intel.vpid= [KVM,Intel] Control KVM's use of Virtual Processor 3313 Identification feature (tagged TLBs). Default is 1 3314 (enabled). Disable by KVM if hardware lacks support 3315 for it. 3316 3317 l1d_flush= [X86,INTEL,EARLY] 3318 Control mitigation for L1D based snooping vulnerability. 3319 3320 Certain CPUs are vulnerable to an exploit against CPU 3321 internal buffers which can forward information to a 3322 disclosure gadget under certain conditions. 3323 3324 In vulnerable processors, the speculatively 3325 forwarded data can be used in a cache side channel 3326 attack, to access data to which the attacker does 3327 not have direct access. 3328 3329 This parameter controls the mitigation. The 3330 options are: 3331 3332 on - enable the interface for the mitigation 3333 3334 l1tf= [X86,EARLY] Control mitigation of the L1TF vulnerability on 3335 affected CPUs 3336 3337 The kernel PTE inversion protection is unconditionally 3338 enabled and cannot be disabled. 3339 3340 full 3341 Provides all available mitigations for the 3342 L1TF vulnerability. Disables SMT and 3343 enables all mitigations in the 3344 hypervisors, i.e. unconditional L1D flush. 3345 3346 SMT control and L1D flush control via the 3347 sysfs interface is still possible after 3348 boot. Hypervisors will issue a warning 3349 when the first VM is started in a 3350 potentially insecure configuration, 3351 i.e. SMT enabled or L1D flush disabled. 3352 3353 full,force 3354 Same as 'full', but disables SMT and L1D 3355 flush runtime control. Implies the 3356 'nosmt=force' command line option. 3357 (i.e. sysfs control of SMT is disabled.) 3358 3359 flush 3360 Leaves SMT enabled and enables the default 3361 hypervisor mitigation, i.e. conditional 3362 L1D flush. 3363 3364 SMT control and L1D flush control via the 3365 sysfs interface is still possible after 3366 boot. Hypervisors will issue a warning 3367 when the first VM is started in a 3368 potentially insecure configuration, 3369 i.e. SMT enabled or L1D flush disabled. 3370 3371 flush,nosmt 3372 3373 Disables SMT and enables the default 3374 hypervisor mitigation. 3375 3376 SMT control and L1D flush control via the 3377 sysfs interface is still possible after 3378 boot. Hypervisors will issue a warning 3379 when the first VM is started in a 3380 potentially insecure configuration, 3381 i.e. SMT enabled or L1D flush disabled. 3382 3383 flush,nowarn 3384 Same as 'flush', but hypervisors will not 3385 warn when a VM is started in a potentially 3386 insecure configuration. 3387 3388 off 3389 Disables hypervisor mitigations and doesn't 3390 emit any warnings. 3391 It also drops the swap size and available 3392 RAM limit restriction on both hypervisor and 3393 bare metal. 3394 3395 Default is 'flush'. 3396 3397 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 3398 3399 l2cr= [PPC] 3400 3401 l3cr= [PPC] 3402 3403 lapic [X86-32,APIC,EARLY] Enable the local APIC even if BIOS 3404 disabled it. 3405 3406 lapic= [X86,APIC] Do not use TSC deadline 3407 value for LAPIC timer one-shot implementation. Default 3408 back to the programmable timer unit in the LAPIC. 3409 Format: notscdeadline 3410 3411 lapic_timer_c2_ok [X86,APIC,EARLY] trust the local apic timer 3412 in C2 power state. 3413 3414 libata.dma= [LIBATA] DMA control 3415 libata.dma=0 Disable all PATA and SATA DMA 3416 libata.dma=1 PATA and SATA Disk DMA only 3417 libata.dma=2 ATAPI (CDROM) DMA only 3418 libata.dma=4 Compact Flash DMA only 3419 Combinations also work, so libata.dma=3 enables DMA 3420 for disks and CDROMs, but not CFs. 3421 3422 libata.ignore_hpa= [LIBATA] Ignore HPA limit 3423 libata.ignore_hpa=0 keep BIOS limits (default) 3424 libata.ignore_hpa=1 ignore limits, using full disk 3425 3426 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 3427 when set. 3428 Format: <int> 3429 3430 libata.force= [LIBATA] Force configurations. The format is a comma- 3431 separated list of "[ID:]VAL" where ID is PORT[.DEVICE]. 3432 PORT and DEVICE are decimal numbers matching port, link 3433 or device. Basically, it matches the ATA ID string 3434 printed on console by libata. If the whole ID part is 3435 omitted, the last PORT and DEVICE values are used. If 3436 ID hasn't been specified yet, the configuration applies 3437 to all ports, links and devices. 3438 3439 If only DEVICE is omitted, the parameter applies to 3440 the port and all links and devices behind it. DEVICE 3441 number of 0 either selects the first device or the 3442 first fan-out link behind PMP device. It does not 3443 select the host link. DEVICE number of 15 selects the 3444 host link and device attached to it. 3445 3446 The VAL specifies the configuration to force. As long 3447 as there is no ambiguity, shortcut notation is allowed. 3448 For example, both 1.5 and 1.5G would work for 1.5Gbps. 3449 The following configurations can be forced. 3450 3451 * Cable type: 40c, 80c, short40c, unk, ign or sata. 3452 Any ID with matching PORT is used. 3453 3454 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 3455 3456 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 3457 udma[/][16,25,33,44,66,100,133] notation is also 3458 allowed. 3459 3460 * nohrst, nosrst, norst: suppress hard, soft and both 3461 resets. 3462 3463 * rstonce: only attempt one reset during hot-unplug 3464 link recovery. 3465 3466 * [no]dbdelay: Enable or disable the extra 200ms delay 3467 before debouncing a link PHY and device presence 3468 detection. 3469 3470 * [no]ncq: Turn on or off NCQ. 3471 3472 * [no]ncqtrim: Enable or disable queued DSM TRIM. 3473 3474 * [no]ncqati: Enable or disable NCQ trim on ATI chipset. 3475 3476 * [no]trim: Enable or disable (unqueued) TRIM. 3477 3478 * trim_zero: Indicate that TRIM command zeroes data. 3479 3480 * max_trim_128m: Set 128M maximum trim size limit. 3481 3482 * [no]dma: Turn on or off DMA transfers. 3483 3484 * atapi_dmadir: Enable ATAPI DMADIR bridge support. 3485 3486 * atapi_mod16_dma: Enable the use of ATAPI DMA for 3487 commands that are not a multiple of 16 bytes. 3488 3489 * [no]dmalog: Enable or disable the use of the 3490 READ LOG DMA EXT command to access logs. 3491 3492 * [no]iddevlog: Enable or disable access to the 3493 identify device data log. 3494 3495 * [no]logdir: Enable or disable access to the general 3496 purpose log directory. 3497 3498 * max_sec=<sectors>: Set the transfer size limit, in 3499 number of 512-byte sectors, to the value specified in 3500 <sectors>. The value specified in <sectors> has to be 3501 a non-zero positive integer. 3502 3503 * max_sec_128: Set transfer size limit to 128 sectors. 3504 3505 * max_sec_1024: Set or clear transfer size limit to 3506 1024 sectors. 3507 3508 * max_sec_lba48: Set or clear transfer size limit to 3509 65535 sectors. 3510 3511 * external: Mark port as external (hotplug-capable). 3512 3513 * [no]lpm: Enable or disable link power management. 3514 3515 * [no]setxfer: Indicate if transfer speed mode setting 3516 should be skipped. 3517 3518 * [no]fua: Disable or enable FUA (Force Unit Access) 3519 support for devices supporting this feature. 3520 3521 * dump_id: Dump IDENTIFY data. 3522 3523 * disable: Disable this device. 3524 3525 If there are multiple matching configurations changing 3526 the same attribute, the last one is used. 3527 3528 liveupdate= [KNL,EARLY] 3529 Format: <bool> 3530 Enable Live Update Orchestrator (LUO). 3531 Default: off. 3532 3533 lockd.nlm_grace_period=P [NFS] Assign grace period. 3534 Format: <integer> 3535 3536 lockd.nlm_tcpport=N [NFS] Assign TCP port. 3537 Format: <integer> 3538 3539 lockd.nlm_timeout=T [NFS] Assign timeout value. 3540 Format: <integer> 3541 3542 lockd.nlm_udpport=M [NFS] Assign UDP port. 3543 Format: <integer> 3544 3545 lockdown= [SECURITY,EARLY] 3546 { integrity | confidentiality } 3547 Enable the kernel lockdown feature. If set to 3548 integrity, kernel features that allow userland to 3549 modify the running kernel are disabled. If set to 3550 confidentiality, kernel features that allow userland 3551 to extract confidential information from the kernel 3552 are also disabled. 3553 3554 locktorture.acq_writer_lim= [KNL] 3555 Set the time limit in jiffies for a lock 3556 acquisition. Acquisitions exceeding this limit 3557 will result in a splat once they do complete. 3558 3559 locktorture.bind_readers= [KNL] 3560 Specify the list of CPUs to which the readers are 3561 to be bound. 3562 3563 locktorture.bind_writers= [KNL] 3564 Specify the list of CPUs to which the writers are 3565 to be bound. 3566 3567 locktorture.call_rcu_chains= [KNL] 3568 Specify the number of self-propagating call_rcu() 3569 chains to set up. These are used to ensure that 3570 there is a high probability of an RCU grace period 3571 in progress at any given time. Defaults to 0, 3572 which disables these call_rcu() chains. 3573 3574 locktorture.long_hold= [KNL] 3575 Specify the duration in milliseconds for the 3576 occasional long-duration lock hold time. Defaults 3577 to 100 milliseconds. Select 0 to disable. 3578 3579 locktorture.nested_locks= [KNL] 3580 Specify the maximum lock nesting depth that 3581 locktorture is to exercise, up to a limit of 8 3582 (MAX_NESTED_LOCKS). Specify zero to disable. 3583 Note that this parameter is ineffective on types 3584 of locks that do not support nested acquisition. 3585 3586 locktorture.nreaders_stress= [KNL] 3587 Set the number of locking read-acquisition kthreads. 3588 Defaults to being automatically set based on the 3589 number of online CPUs. 3590 3591 locktorture.nwriters_stress= [KNL] 3592 Set the number of locking write-acquisition kthreads. 3593 3594 locktorture.onoff_holdoff= [KNL] 3595 Set time (s) after boot for CPU-hotplug testing. 3596 3597 locktorture.onoff_interval= [KNL] 3598 Set time (s) between CPU-hotplug operations, or 3599 zero to disable CPU-hotplug testing. 3600 3601 locktorture.rt_boost= [KNL] 3602 Do periodic testing of real-time lock priority 3603 boosting. Select 0 to disable, 1 to boost 3604 only rt_mutex, and 2 to boost unconditionally. 3605 Defaults to 2, which might seem to be an 3606 odd choice, but which should be harmless for 3607 non-real-time spinlocks, due to their disabling 3608 of preemption. Note that non-realtime mutexes 3609 disable boosting. 3610 3611 locktorture.rt_boost_factor= [KNL] 3612 Number that determines how often and for how 3613 long priority boosting is exercised. This is 3614 scaled down by the number of writers, so that the 3615 number of boosts per unit time remains roughly 3616 constant as the number of writers increases. 3617 On the other hand, the duration of each boost 3618 increases with the number of writers. 3619 3620 locktorture.shuffle_interval= [KNL] 3621 Set task-shuffle interval (jiffies). Shuffling 3622 tasks allows some CPUs to go into dyntick-idle 3623 mode during the locktorture test. 3624 3625 locktorture.shutdown_secs= [KNL] 3626 Set time (s) after boot system shutdown. This 3627 is useful for hands-off automated testing. 3628 3629 locktorture.stat_interval= [KNL] 3630 Time (s) between statistics printk()s. 3631 3632 locktorture.stutter= [KNL] 3633 Time (s) to stutter testing, for example, 3634 specifying five seconds causes the test to run for 3635 five seconds, wait for five seconds, and so on. 3636 This tests the locking primitive's ability to 3637 transition abruptly to and from idle. 3638 3639 locktorture.torture_type= [KNL] 3640 Specify the locking implementation to test. 3641 3642 locktorture.verbose= [KNL] 3643 Enable additional printk() statements. 3644 3645 locktorture.writer_fifo= [KNL] 3646 Run the write-side locktorture kthreads at 3647 sched_set_fifo() real-time priority. 3648 3649 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 3650 Format: <irq> 3651 3652 loglevel= [KNL,EARLY] 3653 All Kernel Messages with a loglevel smaller than the 3654 console loglevel will be printed to the console. It can 3655 also be changed with klogd or other programs. The 3656 loglevels are defined as follows: 3657 3658 0 (KERN_EMERG) system is unusable 3659 1 (KERN_ALERT) action must be taken immediately 3660 2 (KERN_CRIT) critical conditions 3661 3 (KERN_ERR) error conditions 3662 4 (KERN_WARNING) warning conditions 3663 5 (KERN_NOTICE) normal but significant condition 3664 6 (KERN_INFO) informational 3665 7 (KERN_DEBUG) debug-level messages 3666 3667 log_buf_len=n[KMG] [KNL,EARLY] 3668 Sets the size of the printk ring buffer, in bytes. 3669 n must be a power of two and greater than the 3670 minimal size. The minimal size is defined by 3671 LOG_BUF_SHIFT kernel config parameter. There 3672 is also CONFIG_LOG_CPU_MAX_BUF_SHIFT config 3673 parameter that allows to increase the default size 3674 depending on the number of CPUs. See init/Kconfig 3675 for more details. 3676 3677 logo.nologo [FB] Disables display of the built-in Linux logo. 3678 This may be used to provide more screen space for 3679 kernel log messages and is useful when debugging 3680 kernel boot problems. 3681 3682 lp=0 [LP] Specify parallel ports to use, e.g, 3683 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 3684 lp=reset first parallel port). 'lp=0' disables the 3685 lp=auto printer driver. 'lp=reset' (which can be 3686 specified in addition to the ports) causes 3687 attached printers to be reset. Using 3688 lp=port1,port2,... specifies the parallel ports 3689 to associate lp devices with, starting with 3690 lp0. A port specification may be 'none' to skip 3691 that lp device, or a parport name such as 3692 'parport0'. Specifying 'lp=auto' instead of a 3693 port specification list means that device IDs 3694 from each port should be examined, to see if 3695 an IEEE 1284-compliant printer is attached; if 3696 so, the driver will manage that printer. 3697 See also header of drivers/char/lp.c. 3698 3699 lpj=n [KNL] 3700 Sets loops_per_jiffy to given constant, thus avoiding 3701 time-consuming boot-time autodetection (up to 250 ms per 3702 CPU). 0 enables autodetection (default). To determine 3703 the correct value for your kernel, boot with normal 3704 autodetection and see what value is printed. Note that 3705 on SMP systems the preset will be applied to all CPUs, 3706 which is likely to cause problems if your CPUs need 3707 significantly divergent settings. An incorrect value 3708 will cause delays in the kernel to be wrong, leading to 3709 unpredictable I/O errors and other breakage. Although 3710 unlikely, in the extreme case this might damage your 3711 hardware. 3712 3713 lsm.debug [SECURITY] Enable LSM initialization debugging output. 3714 3715 lsm=lsm1,...,lsmN 3716 [SECURITY] Choose order of LSM initialization. This 3717 overrides CONFIG_LSM, and the "security=" parameter. 3718 3719 machtype= [Loongson] Share the same kernel image file between 3720 different yeeloong laptops. 3721 Example: machtype=lemote-yeeloong-2f-7inch 3722 3723 maxcpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 3724 will bring up during bootup. maxcpus=n : n >= 0 limits 3725 the kernel to bring up 'n' processors. Surely after 3726 bootup you can bring up the other plugged cpu by executing 3727 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 3728 only takes effect during system bootup. 3729 While n=0 is a special case, it is equivalent to "nosmp", 3730 which also disables the IO APIC. 3731 3732 max_loop= [LOOP] The number of loop block devices that get 3733 (loop.max_loop) unconditionally pre-created at init time. The default 3734 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 3735 of statically allocating a predefined number, loop 3736 devices can be requested on-demand with the 3737 /dev/loop-control interface. 3738 3739 mce= [X86-{32,64}] 3740 3741 Please see Documentation/arch/x86/x86_64/machinecheck.rst for sysfs runtime tunables. 3742 3743 off 3744 disable machine check 3745 3746 no_cmci 3747 disable CMCI(Corrected Machine Check Interrupt) that 3748 Intel processor supports. Usually this disablement is 3749 not recommended, but it might be handy if your 3750 hardware is misbehaving. 3751 3752 Note that you'll get more problems without CMCI than 3753 with due to the shared banks, i.e. you might get 3754 duplicated error logs. 3755 3756 dont_log_ce 3757 don't make logs for corrected errors. All events 3758 reported as corrected are silently cleared by OS. This 3759 option will be useful if you have no interest in any 3760 of corrected errors. 3761 3762 ignore_ce 3763 disable features for corrected errors, e.g. 3764 polling timer and CMCI. All events reported as 3765 corrected are not cleared by OS and remained in its 3766 error banks. 3767 3768 Usually this disablement is not recommended, however 3769 if there is an agent checking/clearing corrected 3770 errors (e.g. BIOS or hardware monitoring 3771 applications), conflicting with OS's error handling, 3772 and you cannot deactivate the agent, then this option 3773 will be a help. 3774 3775 no_lmce 3776 do not opt-in to Local MCE delivery. Use legacy method 3777 to broadcast MCEs. 3778 3779 bootlog 3780 enable logging of machine checks left over from 3781 booting. Disabled by default on AMD Fam10h and older 3782 because some BIOS leave bogus ones. 3783 3784 If your BIOS doesn't do that it's a good idea to 3785 enable though to make sure you log even machine check 3786 events that result in a reboot. On Intel systems it is 3787 enabled by default. 3788 3789 nobootlog 3790 disable boot machine check logging. 3791 3792 monarchtimeout (number) 3793 sets the time in us to wait for other CPUs on machine 3794 checks. 0 to disable. 3795 3796 bios_cmci_threshold 3797 don't overwrite the bios-set CMCI threshold. This boot 3798 option prevents Linux from overwriting the CMCI 3799 threshold set by the bios. Without this option, Linux 3800 always sets the CMCI threshold to 1. Enabling this may 3801 make memory predictive failure analysis less effective 3802 if the bios sets thresholds for memory errors since we 3803 will not see details for all errors. 3804 3805 recovery 3806 force-enable recoverable machine check code paths 3807 3808 Everything else is in sysfs now. 3809 3810 3811 md= [HW] RAID subsystems devices and level 3812 See Documentation/admin-guide/md.rst. 3813 3814 mds= [X86,INTEL,EARLY] 3815 Control mitigation for the Micro-architectural Data 3816 Sampling (MDS) vulnerability. 3817 3818 Certain CPUs are vulnerable to an exploit against CPU 3819 internal buffers which can forward information to a 3820 disclosure gadget under certain conditions. 3821 3822 In vulnerable processors, the speculatively 3823 forwarded data can be used in a cache side channel 3824 attack, to access data to which the attacker does 3825 not have direct access. 3826 3827 This parameter controls the MDS mitigation. The 3828 options are: 3829 3830 full - Enable MDS mitigation on vulnerable CPUs 3831 full,nosmt - Enable MDS mitigation and disable 3832 SMT on vulnerable CPUs 3833 off - Unconditionally disable MDS mitigation 3834 3835 On TAA-affected machines, mds=off can be prevented by 3836 an active TAA mitigation as both vulnerabilities are 3837 mitigated with the same mechanism so in order to disable 3838 this mitigation, you need to specify tsx_async_abort=off 3839 too. 3840 3841 Not specifying this option is equivalent to 3842 mds=full. 3843 3844 For details see: Documentation/admin-guide/hw-vuln/mds.rst 3845 3846 mem=nn[KMG] [HEXAGON,EARLY] Set the memory size. 3847 Must be specified, otherwise memory size will be 0. 3848 3849 mem=nn[KMG] [KNL,BOOT,EARLY] Force usage of a specific amount 3850 of memory Amount of memory to be used in cases 3851 as follows: 3852 3853 1 for test; 3854 2 when the kernel is not able to see the whole system memory; 3855 3 memory that lies after 'mem=' boundary is excluded from 3856 the hypervisor, then assigned to KVM guests. 3857 4 to limit the memory available for kdump kernel. 3858 3859 [ARC,MICROBLAZE] - the limit applies only to low memory, 3860 high memory is not affected. 3861 3862 [ARM64] - only limits memory covered by the linear 3863 mapping. The NOMAP regions are not affected. 3864 3865 [X86] Work as limiting max address. Use together 3866 with memmap= to avoid physical address space collisions. 3867 Without memmap= PCI devices could be placed at addresses 3868 belonging to unused RAM. 3869 3870 Note that this only takes effects during boot time since 3871 in above case 3, memory may need be hot added after boot 3872 if system memory of hypervisor is not sufficient. 3873 3874 mem=nn[KMG]@ss[KMG] 3875 [ARM,MIPS,EARLY] - override the memory layout 3876 reported by firmware. 3877 Define a memory region of size nn[KMG] starting at 3878 ss[KMG]. 3879 Multiple different regions can be specified with 3880 multiple mem= parameters on the command line. 3881 3882 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 3883 memory. 3884 3885 memblock=debug [KNL,EARLY] Enable memblock debug messages. 3886 3887 memchunk=nn[KMG] 3888 [KNL,SH] Allow user to override the default size for 3889 per-device physically contiguous DMA buffers. 3890 3891 memhp_default_state=online/offline/online_kernel/online_movable 3892 [KNL] Set the initial state for the memory hotplug 3893 onlining policy. If not specified, the default value is 3894 set according to the 3895 CONFIG_MHP_DEFAULT_ONLINE_TYPE kernel config 3896 options. 3897 See Documentation/admin-guide/mm/memory-hotplug.rst. 3898 3899 memmap=exactmap [KNL,X86,EARLY] Enable setting of an exact 3900 E820 memory map, as specified by the user. 3901 Such memmap=exactmap lines can be constructed based on 3902 BIOS output or other requirements. See the memmap=nn@ss 3903 option description. 3904 3905 memmap=nn[KMG]@ss[KMG] 3906 [KNL, X86,MIPS,XTENSA,EARLY] Force usage of a specific region of memory. 3907 Region of memory to be used is from ss to ss+nn. 3908 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 3909 which limits max address to nn[KMG]. 3910 Multiple different regions can be specified, 3911 comma delimited. 3912 Example: 3913 memmap=100M@2G,100M#3G,1G!1024G 3914 3915 memmap=nn[KMG]#ss[KMG] 3916 [KNL,ACPI,EARLY] Mark specific memory as ACPI data. 3917 Region of memory to be marked is from ss to ss+nn. 3918 3919 memmap=nn[KMG]$ss[KMG] 3920 [KNL,ACPI,EARLY] Mark specific memory as reserved. 3921 Region of memory to be reserved is from ss to ss+nn. 3922 Example: Exclude memory from 0x18690000-0x1869ffff 3923 memmap=64K$0x18690000 3924 or 3925 memmap=0x10000$0x18690000 3926 Some bootloaders may need an escape character before '$', 3927 like Grub2, otherwise '$' and the following number 3928 will be eaten. 3929 3930 memmap=nn[KMG]!ss[KMG,EARLY] 3931 [KNL,X86] Mark specific memory as protected. 3932 Region of memory to be used, from ss to ss+nn. 3933 The memory region may be marked as e820 type 12 (0xc) 3934 and is NVDIMM or ADR memory. 3935 3936 memmap=<size>%<offset>-<oldtype>+<newtype> 3937 [KNL,ACPI,EARLY] Convert memory within the specified region 3938 from <oldtype> to <newtype>. If "-<oldtype>" is left 3939 out, the whole region will be marked as <newtype>, 3940 even if previously unavailable. If "+<newtype>" is left 3941 out, matching memory will be removed. Types are 3942 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 3943 3 = ACPI, 12 = PRAM. 3944 3945 memory_corruption_check=0/1 [X86,EARLY] 3946 Some BIOSes seem to corrupt the first 64k of 3947 memory when doing things like suspend/resume. 3948 Setting this option will scan the memory 3949 looking for corruption. Enabling this will 3950 both detect corruption and prevent the kernel 3951 from using the memory being corrupted. 3952 However, it's intended as a diagnostic tool; if 3953 repeatable BIOS-originated corruption always 3954 affects the same memory, you can use memmap= 3955 to prevent the kernel from using that memory. 3956 3957 memory_corruption_check_size=size [X86,EARLY] 3958 By default it checks for corruption in the low 3959 64k, making this memory unavailable for normal 3960 use. Use this parameter to scan for 3961 corruption in more or less memory. 3962 3963 memory_corruption_check_period=seconds [X86,EARLY] 3964 By default it checks for corruption every 60 3965 seconds. Use this parameter to check at some 3966 other rate. 0 disables periodic checking. 3967 3968 memory_hotplug.memmap_on_memory 3969 [KNL,X86,ARM] Boolean flag to enable this feature. 3970 Format: {on | off (default)} 3971 When enabled, runtime hotplugged memory will 3972 allocate its internal metadata (struct pages, 3973 those vmemmap pages cannot be optimized even 3974 if hugetlb_free_vmemmap is enabled) from the 3975 hotadded memory which will allow to hotadd a 3976 lot of memory without requiring additional 3977 memory to do so. 3978 This feature is disabled by default because it 3979 has some implication on large (e.g. GB) 3980 allocations in some configurations (e.g. small 3981 memory blocks). 3982 The state of the flag can be read in 3983 /sys/module/memory_hotplug/parameters/memmap_on_memory. 3984 Note that even when enabled, there are a few cases where 3985 the feature is not effective. 3986 3987 memtest= [KNL,X86,ARM,M68K,PPC,RISCV,EARLY] Enable memtest 3988 Format: <integer> 3989 default : 0 <disable> 3990 Specifies the number of memtest passes to be 3991 performed. Each pass selects another test 3992 pattern from a given set of patterns. Memtest 3993 fills the memory with this pattern, validates 3994 memory contents and reserves bad memory 3995 regions that are detected. 3996 3997 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 3998 Valid arguments: on, off 3999 Default: off 4000 mem_encrypt=on: Activate SME 4001 mem_encrypt=off: Do not activate SME 4002 4003 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst 4004 for details on when memory encryption can be activated. 4005 4006 mem_sleep_default= [SUSPEND] Default system suspend mode: 4007 s2idle - Suspend-To-Idle 4008 shallow - Power-On Suspend or equivalent (if supported) 4009 deep - Suspend-To-RAM or equivalent (if supported) 4010 See Documentation/admin-guide/pm/sleep-states.rst. 4011 4012 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 4013 the BIOS has incorrectly applied a workaround. TinyBIOS 4014 version 0.98 is known to be affected, 0.99 fixes the 4015 problem by letting the user disable the workaround. 4016 4017 mga= [HW,DRM] 4018 4019 microcode= [X86] Control the behavior of the microcode loader. 4020 Available options, comma separated: 4021 4022 base_rev=X - with <X> with format: <u32> 4023 Set the base microcode revision of each thread when in 4024 debug mode. 4025 4026 dis_ucode_ldr: disable the microcode loader 4027 4028 force_minrev: 4029 Enable or disable the microcode minimal revision 4030 enforcement for the runtime microcode loader. 4031 4032 mini2440= [ARM,HW,KNL] 4033 Format:[0..2][b][c][t] 4034 Default: "0tb" 4035 MINI2440 configuration specification: 4036 0 - The attached screen is the 3.5" TFT 4037 1 - The attached screen is the 7" TFT 4038 2 - The VGA Shield is attached (1024x768) 4039 Leaving out the screen size parameter will not load 4040 the TFT driver, and the framebuffer will be left 4041 unconfigured. 4042 b - Enable backlight. The TFT backlight pin will be 4043 linked to the kernel VESA blanking code and a GPIO 4044 LED. This parameter is not necessary when using the 4045 VGA shield. 4046 c - Enable the s3c camera interface. 4047 t - Reserved for enabling touchscreen support. The 4048 touchscreen support is not enabled in the mainstream 4049 kernel as of 2.6.30, a preliminary port can be found 4050 in the "bleeding edge" mini2440 support kernel at 4051 https://repo.or.cz/w/linux-2.6/mini2440.git 4052 4053 mitigations= 4054 [X86,PPC,S390,ARM64,EARLY] Control optional mitigations for 4055 CPU vulnerabilities. This is a set of curated, 4056 arch-independent options, each of which is an 4057 aggregation of existing arch-specific options. 4058 4059 Note, "mitigations" is supported if and only if the 4060 kernel was built with CPU_MITIGATIONS=y. 4061 4062 off 4063 Disable all optional CPU mitigations. This 4064 improves system performance, but it may also 4065 expose users to several CPU vulnerabilities. 4066 Equivalent to: if nokaslr then kpti=0 [ARM64] 4067 gather_data_sampling=off [X86] 4068 indirect_target_selection=off [X86] 4069 kvm.nx_huge_pages=off [X86] 4070 l1tf=off [X86] 4071 mds=off [X86] 4072 mmio_stale_data=off [X86] 4073 no_entry_flush [PPC] 4074 no_uaccess_flush [PPC] 4075 nobp=0 [S390] 4076 nopti [X86,PPC] 4077 nospectre_bhb [ARM64] 4078 nospectre_v1 [X86,PPC] 4079 nospectre_v2 [X86,PPC,S390,ARM64] 4080 reg_file_data_sampling=off [X86] 4081 retbleed=off [X86] 4082 spec_rstack_overflow=off [X86] 4083 spec_store_bypass_disable=off [X86,PPC] 4084 spectre_bhi=off [X86] 4085 spectre_v2_user=off [X86] 4086 srbds=off [X86,INTEL] 4087 ssbd=force-off [ARM64] 4088 tsa=off [X86,AMD] 4089 tsx_async_abort=off [X86] 4090 vmscape=off [X86] 4091 4092 Exceptions: 4093 This does not have any effect on 4094 kvm.nx_huge_pages when 4095 kvm.nx_huge_pages=force. 4096 4097 auto (default) 4098 Mitigate all CPU vulnerabilities, but leave SMT 4099 enabled, even if it's vulnerable. This is for 4100 users who don't want to be surprised by SMT 4101 getting disabled across kernel upgrades, or who 4102 have other ways of avoiding SMT-based attacks. 4103 Equivalent to: (default behavior) 4104 4105 auto,nosmt 4106 Mitigate all CPU vulnerabilities, disabling SMT 4107 if needed. This is for users who always want to 4108 be fully mitigated, even if it means losing SMT. 4109 Equivalent to: l1tf=flush,nosmt [X86] 4110 mds=full,nosmt [X86] 4111 tsx_async_abort=full,nosmt [X86] 4112 mmio_stale_data=full,nosmt [X86] 4113 retbleed=auto,nosmt [X86] 4114 4115 [X86] After one of the above options, additionally 4116 supports attack-vector based controls as documented in 4117 Documentation/admin-guide/hw-vuln/attack_vector_controls.rst 4118 4119 mminit_loglevel= 4120 [KNL,EARLY] When CONFIG_DEBUG_MEMORY_INIT is set, this 4121 parameter allows control of the logging verbosity for 4122 the additional memory initialisation checks. A value 4123 of 0 disables mminit logging and a level of 4 will 4124 log everything. Information is printed at KERN_DEBUG 4125 so loglevel=8 may also need to be specified. 4126 4127 mmio_stale_data= 4128 [X86,INTEL,EARLY] Control mitigation for the Processor 4129 MMIO Stale Data vulnerabilities. 4130 4131 Processor MMIO Stale Data is a class of 4132 vulnerabilities that may expose data after an MMIO 4133 operation. Exposed data could originate or end in 4134 the same CPU buffers as affected by MDS and TAA. 4135 Therefore, similar to MDS and TAA, the mitigation 4136 is to clear the affected CPU buffers. 4137 4138 This parameter controls the mitigation. The 4139 options are: 4140 4141 full - Enable mitigation on vulnerable CPUs 4142 4143 full,nosmt - Enable mitigation and disable SMT on 4144 vulnerable CPUs. 4145 4146 off - Unconditionally disable mitigation 4147 4148 On MDS or TAA affected machines, 4149 mmio_stale_data=off can be prevented by an active 4150 MDS or TAA mitigation as these vulnerabilities are 4151 mitigated with the same mechanism so in order to 4152 disable this mitigation, you need to specify 4153 mds=off and tsx_async_abort=off too. 4154 4155 Not specifying this option is equivalent to 4156 mmio_stale_data=full. 4157 4158 For details see: 4159 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst 4160 4161 <module>.async_probe[=<bool>] [KNL] 4162 If no <bool> value is specified or if the value 4163 specified is not a valid <bool>, enable asynchronous 4164 probe on this module. Otherwise, enable/disable 4165 asynchronous probe on this module as indicated by the 4166 <bool> value. See also: module.async_probe 4167 4168 module.async_probe=<bool> 4169 [KNL] When set to true, modules will use async probing 4170 by default. To enable/disable async probing for a 4171 specific module, use the module specific control that 4172 is documented under <module>.async_probe. When both 4173 module.async_probe and <module>.async_probe are 4174 specified, <module>.async_probe takes precedence for 4175 the specific module. 4176 4177 module.enable_dups_trace 4178 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set, 4179 this means that duplicate request_module() calls will 4180 trigger a WARN_ON() instead of a pr_warn(). Note that 4181 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s 4182 will always be issued and this option does nothing. 4183 module.sig_enforce 4184 [KNL] When CONFIG_MODULE_SIG is set, this means that 4185 modules without (valid) signatures will fail to load. 4186 Note that if CONFIG_MODULE_SIG_FORCE is set, that 4187 is always true, so this option does nothing. 4188 4189 module_blacklist= [KNL] Do not load a comma-separated list of 4190 modules. Useful for debugging problem modules. 4191 4192 mousedev.tap_time= 4193 [MOUSE] Maximum time between finger touching and 4194 leaving touchpad surface for touch to be considered 4195 a tap and be reported as a left button click (for 4196 touchpads working in absolute mode only). 4197 Format: <msecs> 4198 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 4199 reporting absolute coordinates, such as tablets 4200 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 4201 reporting absolute coordinates, such as tablets 4202 4203 movablecore= [KNL,X86,PPC,EARLY] 4204 Format: nn[KMGTPE] | nn% 4205 This parameter is the complement to kernelcore=, it 4206 specifies the amount of memory used for migratable 4207 allocations. If both kernelcore and movablecore is 4208 specified, then kernelcore will be at *least* the 4209 specified value but may be more. If movablecore on its 4210 own is specified, the administrator must be careful 4211 that the amount of memory usable for all allocations 4212 is not too small. 4213 4214 movable_node [KNL,EARLY] Boot-time switch to make hotplugable memory 4215 NUMA nodes to be movable. This means that the memory 4216 of such nodes will be usable only for movable 4217 allocations which rules out almost all kernel 4218 allocations. Use with caution! 4219 4220 MTD_Partition= [MTD] 4221 Format: <name>,<region-number>,<size>,<offset> 4222 4223 MTD_Region= [MTD] Format: 4224 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 4225 4226 mtdparts= [MTD] 4227 See drivers/mtd/parsers/cmdlinepart.c 4228 4229 mtouchusb.raw_coordinates= 4230 [HW] Make the MicroTouch USB driver use raw coordinates 4231 ('y', default) or cooked coordinates ('n') 4232 4233 mtrr=debug [X86,EARLY] 4234 Enable printing debug information related to MTRR 4235 registers at boot time. 4236 4237 mtrr_chunk_size=nn[KMG,X86,EARLY] 4238 used for mtrr cleanup. It is largest continuous chunk 4239 that could hold holes aka. UC entries. 4240 4241 mtrr_gran_size=nn[KMG,X86,EARLY] 4242 Used for mtrr cleanup. It is granularity of mtrr block. 4243 Default is 1. 4244 Large value could prevent small alignment from 4245 using up MTRRs. 4246 4247 mtrr_spare_reg_nr=n [X86,EARLY] 4248 Format: <integer> 4249 Range: 0,7 : spare reg number 4250 Default : 1 4251 Used for mtrr cleanup. It is spare mtrr entries number. 4252 Set to 2 or more if your graphical card needs more. 4253 4254 multitce=off [PPC] This parameter disables the use of the pSeries 4255 firmware feature for updating multiple TCE entries 4256 at a time. 4257 4258 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 4259 4260 netdev= [NET] NE2000 ISA network devices parameters 4261 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 4262 4263 netpoll.carrier_timeout= 4264 [NET] Specifies amount of time (in seconds) that 4265 netpoll should wait for a carrier. By default netpoll 4266 waits 4 seconds. 4267 4268 nf_conntrack.acct= 4269 [NETFILTER] Enable connection tracking flow accounting 4270 0 to disable accounting 4271 1 to enable accounting 4272 Default value is 0. 4273 4274 nfs.cache_getent= 4275 [NFS] sets the pathname to the program which is used 4276 to update the NFS client cache entries. 4277 4278 nfs.cache_getent_timeout= 4279 [NFS] sets the timeout after which an attempt to 4280 update a cache entry is deemed to have failed. 4281 4282 nfs.callback_nr_threads= 4283 [NFSv4] set the total number of threads that the 4284 NFS client will assign to service NFSv4 callback 4285 requests. 4286 4287 nfs.callback_tcpport= 4288 [NFS] set the TCP port on which the NFSv4 callback 4289 channel should listen. 4290 4291 nfs.delay_retrans= 4292 [NFS] specifies the number of times the NFSv4 client 4293 retries the request before returning an EAGAIN error, 4294 after a reply of NFS4ERR_DELAY from the server. 4295 Only applies if the softerr mount option is enabled, 4296 and the specified value is >= 0. 4297 4298 nfs.idmap_cache_timeout= 4299 [NFS] set the maximum lifetime for idmapper cache 4300 entries. 4301 4302 nfs.max_session_cb_slots= 4303 [NFSv4.1] Sets the maximum number of session 4304 slots the client will assign to the callback 4305 channel. This determines the maximum number of 4306 callbacks the client will process in parallel for 4307 a particular server. 4308 4309 nfs.max_session_slots= 4310 [NFSv4.1] Sets the maximum number of session slots 4311 the client will attempt to negotiate with the server. 4312 This limits the number of simultaneous RPC requests 4313 that the client can send to the NFSv4.1 server. 4314 Note that there is little point in setting this 4315 value higher than the max_tcp_slot_table_limit. 4316 4317 nfs.nfs4_disable_idmapping= 4318 [NFSv4] When set to the default of '1', this option 4319 ensures that both the RPC level authentication 4320 scheme and the NFS level operations agree to use 4321 numeric uids/gids if the mount is using the 4322 'sec=sys' security flavour. In effect it is 4323 disabling idmapping, which can make migration from 4324 legacy NFSv2/v3 systems to NFSv4 easier. 4325 Servers that do not support this mode of operation 4326 will be autodetected by the client, and it will fall 4327 back to using the idmapper. 4328 To turn off this behaviour, set the value to '0'. 4329 4330 nfs.nfs4_unique_id= 4331 [NFS4] Specify an additional fixed unique ident- 4332 ification string that NFSv4 clients can insert into 4333 their nfs_client_id4 string. This is typically a 4334 UUID that is generated at system install time. 4335 4336 nfs.recover_lost_locks= 4337 [NFSv4] Attempt to recover locks that were lost due 4338 to a lease timeout on the server. Please note that 4339 doing this risks data corruption, since there are 4340 no guarantees that the file will remain unchanged 4341 after the locks are lost. 4342 If you want to enable the kernel legacy behaviour of 4343 attempting to recover these locks, then set this 4344 parameter to '1'. 4345 The default parameter value of '0' causes the kernel 4346 not to attempt recovery of lost locks. 4347 4348 nfs.send_implementation_id= 4349 [NFSv4.1] Send client implementation identification 4350 information in exchange_id requests. 4351 If zero, no implementation identification information 4352 will be sent. 4353 The default is to send the implementation identification 4354 information. 4355 4356 nfs4.layoutstats_timer= 4357 [NFSv4.2] Change the rate at which the kernel sends 4358 layoutstats to the pNFS metadata server. 4359 4360 Setting this to value to 0 causes the kernel to use 4361 whatever value is the default set by the layout 4362 driver. A non-zero value sets the minimum interval 4363 in seconds between layoutstats transmissions. 4364 4365 nfsd.inter_copy_offload_enable= 4366 [NFSv4.2] When set to 1, the server will support 4367 server-to-server copies for which this server is 4368 the destination of the copy. 4369 4370 nfsd.nfs4_disable_idmapping= 4371 [NFSv4] When set to the default of '1', the NFSv4 4372 server will return only numeric uids and gids to 4373 clients using auth_sys, and will accept numeric uids 4374 and gids from such clients. This is intended to ease 4375 migration from NFSv2/v3. 4376 4377 nfsd.nfsd4_ssc_umount_timeout= 4378 [NFSv4.2] When used as the destination of a 4379 server-to-server copy, knfsd temporarily mounts 4380 the source server. It caches the mount in case 4381 it will be needed again, and discards it if not 4382 used for the number of milliseconds specified by 4383 this parameter. 4384 4385 nfsaddrs= [NFS] Deprecated. Use ip= instead. 4386 See Documentation/admin-guide/nfs/nfsroot.rst. 4387 4388 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 4389 See Documentation/admin-guide/nfs/nfsroot.rst. 4390 4391 nfsrootdebug [NFS] enable nfsroot debugging messages. 4392 See Documentation/admin-guide/nfs/nfsroot.rst. 4393 4394 nmi_backtrace.backtrace_idle [KNL] 4395 Dump stacks even of idle CPUs in response to an 4396 NMI stack-backtrace request. 4397 4398 nmi_debug= [KNL,SH] Specify one or more actions to take 4399 when a NMI is triggered. 4400 Format: [state][,regs][,debounce][,die] 4401 4402 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 4403 Format: [panic,][nopanic,][rNNN,][num] 4404 Valid num: 0 or 1 4405 0 - turn hardlockup detector in nmi_watchdog off 4406 1 - turn hardlockup detector in nmi_watchdog on 4407 rNNN - configure the watchdog with raw perf event 0xNNN 4408 4409 When panic is specified, panic when an NMI watchdog 4410 timeout occurs (or 'nopanic' to not panic on an NMI 4411 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 4412 To disable both hard and soft lockup detectors, 4413 please see 'nowatchdog'. 4414 This is useful when you use a panic=... timeout and 4415 need the box quickly up again. 4416 4417 These settings can be accessed at runtime via 4418 the nmi_watchdog and hardlockup_panic sysctls. 4419 4420 no4lvl [RISCV,EARLY] Disable 4-level and 5-level paging modes. 4421 Forces kernel to use 3-level paging instead. 4422 4423 no5lvl [X86-64,RISCV,EARLY] Disable 5-level paging mode. Forces 4424 kernel to use 4-level paging instead. 4425 4426 noalign [KNL,ARM] 4427 4428 noapic [SMP,APIC,EARLY] Tells the kernel to not make use of any 4429 IOAPICs that may be present in the system. 4430 4431 noapictimer [APIC,X86] Don't set up the APIC timer 4432 4433 noautogroup Disable scheduler automatic task group creation. 4434 4435 nocache [ARM,EARLY] 4436 4437 no_console_suspend 4438 [HW] Never suspend the console 4439 Disable suspending of consoles during suspend and 4440 hibernate operations. Once disabled, debugging 4441 messages can reach various consoles while the rest 4442 of the system is being put to sleep (ie, while 4443 debugging driver suspend/resume hooks). This may 4444 not work reliably with all consoles, but is known 4445 to work with serial and VGA consoles. 4446 To facilitate more flexible debugging, we also add 4447 console_suspend, a printk module parameter to control 4448 it. Users could use console_suspend (usually 4449 /sys/module/printk/parameters/console_suspend) to 4450 turn on/off it dynamically. 4451 4452 no_debug_objects 4453 [KNL,EARLY] Disable object debugging 4454 4455 nodsp [SH] Disable hardware DSP at boot time. 4456 4457 noefi [EFI,EARLY] Disable EFI runtime services support. 4458 4459 no_entry_flush [PPC,EARLY] Don't flush the L1-D cache when entering the kernel. 4460 4461 noexec32 [X86-64] 4462 This affects only 32-bit executables. 4463 noexec32=on: enable non-executable mappings (default) 4464 read doesn't imply executable mappings 4465 noexec32=off: disable non-executable mappings 4466 read implies executable mappings 4467 4468 no_file_caps Tells the kernel not to honor file capabilities. The 4469 only way then for a file to be executed with privilege 4470 is to be setuid root or executed by root. 4471 4472 nofpu [MIPS,SH] Disable hardware FPU at boot time. 4473 4474 nofsgsbase [X86] Disables FSGSBASE instructions. 4475 4476 nofxsr [BUGS=X86-32] Disables x86 floating point extended 4477 register save and restore. The kernel will only save 4478 legacy floating-point registers on task switch. 4479 4480 nogbpages [X86] Do not use GB pages for kernel direct mappings. 4481 4482 no_hash_pointers 4483 [KNL,EARLY] 4484 Alias for "hash_pointers=never". 4485 4486 nohibernate [HIBERNATION] Disable hibernation and resume. 4487 4488 nohlt [ARM,ARM64,MICROBLAZE,MIPS,PPC,RISCV,SH] Forces the kernel to 4489 busy wait in do_idle() and not use the arch_cpu_idle() 4490 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP 4491 to be effective. This is useful on platforms where the 4492 sleep(SH) or wfi(ARM,ARM64) instructions do not work 4493 correctly or when doing power measurements to evaluate 4494 the impact of the sleep instructions. This is also 4495 useful when using JTAG debugger. 4496 4497 nohpet [X86] Don't use the HPET timer. 4498 4499 nohugeiomap [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge I/O mappings. 4500 4501 nohugevmalloc [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge vmalloc mappings. 4502 4503 nohz= [KNL] Boottime enable/disable dynamic ticks 4504 Valid arguments: on, off 4505 Default: on 4506 4507 nohz_full= [KNL,BOOT,SMP,ISOL] 4508 The argument is a cpu list, as described above. 4509 In kernels built with CONFIG_NO_HZ_FULL=y, set 4510 the specified list of CPUs whose tick will be stopped 4511 whenever possible. The boot CPU will be forced outside 4512 the range to maintain the timekeeping. Any CPUs 4513 in this list will have their RCU callbacks offloaded, 4514 just as if they had also been called out in the 4515 rcu_nocbs= boot parameter. 4516 4517 Note that this argument takes precedence over 4518 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 4519 4520 noinitrd [Deprecated,RAM] Tells the kernel not to load any configured 4521 initial RAM disk. Currently this parameter applies to 4522 initrd only, not to initramfs. But it applies to both 4523 in EFI mode. 4524 4525 nointremap [X86-64,Intel-IOMMU,EARLY] Do not enable interrupt 4526 remapping. 4527 [Deprecated - use intremap=off] 4528 4529 noinvpcid [X86,EARLY] Disable the INVPCID cpu feature. 4530 4531 noiotrap [SH] Disables trapped I/O port accesses. 4532 4533 noirqdebug [X86-32] Disables the code which attempts to detect and 4534 disable unhandled interrupt sources. 4535 4536 noisapnp [ISAPNP] Disables ISA PnP code. 4537 4538 nokaslr [KNL,EARLY] 4539 When CONFIG_RANDOMIZE_BASE is set, this disables 4540 kernel and module base offset ASLR (Address Space 4541 Layout Randomization). 4542 4543 no-kvmapf [X86,KVM,EARLY] Disable paravirtualized asynchronous page 4544 fault handling. 4545 4546 no-kvmclock [X86,KVM,EARLY] Disable paravirtualized KVM clock driver 4547 4548 nolapic [X86-32,APIC,EARLY] Do not enable or use the local APIC. 4549 4550 nolapic_timer [X86-32,APIC,EARLY] Do not use the local APIC timer. 4551 4552 nomce [X86-32] Disable Machine Check Exception 4553 4554 nomfgpt [X86-32] Disable Multi-Function General Purpose 4555 Timer usage (for AMD Geode machines). 4556 4557 nomodeset Disable kernel modesetting. Most systems' firmware 4558 sets up a display mode and provides framebuffer memory 4559 for output. With nomodeset, DRM and fbdev drivers will 4560 not load if they could possibly displace the pre- 4561 initialized output. Only the system framebuffer will 4562 be available for use. The respective drivers will not 4563 perform display-mode changes or accelerated rendering. 4564 4565 Useful as error fallback, or for testing and debugging. 4566 4567 nomodule Disable module load 4568 4569 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 4570 shutdown the other cpus. Instead use the REBOOT_VECTOR 4571 irq. 4572 4573 nopat [X86,EARLY] Disable PAT (page attribute table extension of 4574 pagetables) support. 4575 4576 nopcid [X86-64,EARLY] Disable the PCID cpu feature. 4577 4578 nopku [X86] Disable Memory Protection Keys CPU feature found 4579 in some Intel CPUs. 4580 4581 nopti [X86-64,EARLY] 4582 Equivalent to pti=off 4583 4584 nopv= [X86,XEN,KVM,HYPER_V,VMWARE,EARLY] 4585 Disables the PV optimizations forcing the guest to run 4586 as generic guest with no PV drivers. Currently support 4587 XEN HVM, KVM, HYPER_V and VMWARE guest. 4588 4589 nopvspin [X86,XEN,KVM,EARLY] 4590 Disables the qspinlock slow path using PV optimizations 4591 which allow the hypervisor to 'idle' the guest on lock 4592 contention. 4593 4594 norandmaps Don't use address space randomization. Equivalent to 4595 echo 0 > /proc/sys/kernel/randomize_va_space 4596 4597 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 4598 with UP alternatives 4599 4600 noresume [SWSUSP] Disables resume and restores original swap 4601 space. 4602 4603 no-scroll [VGA] Disables scrollback. 4604 This is required for the Braillex ib80-piezo Braille 4605 reader made by F.H. Papenmeier (Germany). 4606 4607 nosgx [X86-64,SGX,EARLY] Disables Intel SGX kernel support. 4608 4609 nosmap [PPC,EARLY] 4610 Disable SMAP (Supervisor Mode Access Prevention) 4611 even if it is supported by processor. 4612 4613 nosmep [PPC64s,EARLY] 4614 Disable SMEP (Supervisor Mode Execution Prevention) 4615 even if it is supported by processor. 4616 4617 nosmp [SMP,EARLY] Tells an SMP kernel to act as a UP kernel, 4618 and disable the IO APIC. legacy for "maxcpus=0". 4619 4620 nosmt [KNL,MIPS,PPC,EARLY] Disable symmetric multithreading (SMT). 4621 Equivalent to smt=1. 4622 4623 [KNL,LOONGARCH,X86,ARM64,PPC,S390] Disable symmetric multithreading (SMT). 4624 nosmt=force: Force disable SMT, cannot be undone 4625 via the sysfs control file. 4626 4627 nosoftlockup [KNL] Disable the soft-lockup detector. 4628 4629 nospec_store_bypass_disable 4630 [HW,EARLY] Disable all mitigations for the Speculative 4631 Store Bypass vulnerability 4632 4633 nospectre_bhb [ARM64,EARLY] Disable all mitigations for Spectre-BHB (branch 4634 history injection) vulnerability. System may allow data leaks 4635 with this option. 4636 4637 nospectre_v1 [X86,PPC,EARLY] Disable mitigations for Spectre Variant 1 4638 (bounds check bypass). With this option data leaks are 4639 possible in the system. 4640 4641 nospectre_v2 [X86,PPC_E500,ARM64,EARLY] Disable all mitigations 4642 for the Spectre variant 2 (indirect branch 4643 prediction) vulnerability. System may allow data 4644 leaks with this option. 4645 4646 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES,RISCV,LOONGARCH,EARLY] 4647 Disable paravirtualized steal time accounting. steal time 4648 is computed, but won't influence scheduler behaviour 4649 4650 nosync [HW,M68K] Disables sync negotiation for all devices. 4651 4652 no_timer_check [X86,APIC] Disables the code which tests for broken 4653 timer IRQ sources, i.e., the IO-APIC timer. This can 4654 work around problems with incorrect timer 4655 initialization on some boards. 4656 4657 no_uaccess_flush 4658 [PPC,EARLY] Don't flush the L1-D cache after accessing user data. 4659 4660 novmcoredd [KNL,KDUMP] 4661 Disable device dump. Device dump allows drivers to 4662 append dump data to vmcore so you can collect driver 4663 specified debug info. Drivers can append the data 4664 without any limit and this data is stored in memory, 4665 so this may cause significant memory stress. Disabling 4666 device dump can help save memory but the driver debug 4667 data will be no longer available. This parameter 4668 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 4669 is set. 4670 4671 no-vmw-sched-clock 4672 [X86,PV_OPS,EARLY] Disable paravirtualized VMware 4673 scheduler clock and use the default one. 4674 4675 nowatchdog [KNL] Disable both lockup detectors, i.e. 4676 soft-lockup and NMI watchdog (hard-lockup). 4677 4678 nowb [ARM,EARLY] 4679 4680 nox2apic [X86-64,APIC,EARLY] Do not enable x2APIC mode. 4681 4682 NOTE: this parameter will be ignored on systems with the 4683 LEGACY_XAPIC_DISABLED bit set in the 4684 IA32_XAPIC_DISABLE_STATUS MSR. 4685 4686 noxsave [BUGS=X86] Disables x86 extended register state save 4687 and restore using xsave. The kernel will fallback to 4688 enabling legacy floating-point and sse state. 4689 4690 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 4691 register states. The kernel will fall back to use 4692 xsave to save the states. By using this parameter, 4693 performance of saving the states is degraded because 4694 xsave doesn't support modified optimization while 4695 xsaveopt supports it on xsaveopt enabled systems. 4696 4697 noxsaves [X86] Disables xsaves and xrstors used in saving and 4698 restoring x86 extended register state in compacted 4699 form of xsave area. The kernel will fall back to use 4700 xsaveopt and xrstor to save and restore the states 4701 in standard form of xsave area. By using this 4702 parameter, xsave area per process might occupy more 4703 memory on xsaves enabled systems. 4704 4705 nr_cpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 4706 could support. nr_cpus=n : n >= 1 limits the kernel to 4707 support 'n' processors. It could be larger than the 4708 number of already plugged CPU during bootup, later in 4709 runtime you can physically add extra cpu until it reaches 4710 n. So during boot up some boot time memory for per-cpu 4711 variables need be pre-allocated for later physical cpu 4712 hot plugging. 4713 4714 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 4715 4716 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86, EARLY] 4717 Disable NUMA, Only set up a single NUMA node 4718 spanning all memory. 4719 4720 numa=fake=<size>[MG] 4721 [KNL, ARM64, RISCV, X86, EARLY] 4722 If given as a memory unit, fills all system RAM with 4723 nodes of size interleaved over physical nodes. 4724 4725 numa=fake=<N> 4726 [KNL, ARM64, RISCV, X86, EARLY] 4727 If given as an integer, fills all system RAM with N 4728 fake nodes interleaved over physical nodes. 4729 4730 numa=fake=<N>U 4731 [KNL, ARM64, RISCV, X86, EARLY] 4732 If given as an integer followed by 'U', it will 4733 divide each physical node into N emulated nodes. 4734 4735 numa=noacpi [X86] Don't parse the SRAT table for NUMA setup 4736 4737 numa=nohmat [X86] Don't parse the HMAT table for NUMA setup, or 4738 soft-reserved memory partitioning. 4739 4740 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic 4741 NUMA balancing. 4742 Allowed values are enable and disable 4743 4744 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 4745 'node', 'default' can be specified 4746 This can be set from sysctl after boot. 4747 See Documentation/admin-guide/sysctl/vm.rst for details. 4748 4749 nvme.quirks= [NVME] A list of quirk entries to augment the built-in 4750 nvme quirk list. List entries are separated by a 4751 '-' character. 4752 Each entry has the form VendorID:ProductID:quirk_names. 4753 The IDs are 4-digits hex numbers and quirk_names is a 4754 list of quirk names separated by commas. A quirk name 4755 can be prefixed by '^', meaning that the specified 4756 quirk must be disabled. 4757 4758 Example: 4759 nvme.quirks=7710:2267:bogus_nid,^identify_cns-9900:7711:broken_msi 4760 4761 ohci1394_dma=early [HW,EARLY] enable debugging via the ohci1394 driver. 4762 See Documentation/core-api/debugging-via-ohci1394.rst for more 4763 info. 4764 4765 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 4766 Rather than timing out after 20 ms if an EC 4767 command is not properly ACKed, override the length 4768 of the timeout. We have interrupts disabled while 4769 waiting for the ACK, so if this is set too high 4770 interrupts *may* be lost! 4771 4772 omap_mux= [OMAP] Override bootloader pin multiplexing. 4773 Format: <mux_mode0.mode_name=value>... 4774 For example, to override I2C bus2: 4775 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 4776 4777 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 4778 4779 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 4780 4781 boundary - index of last SLC block on Flex-OneNAND. 4782 The remaining blocks are configured as MLC blocks. 4783 lock - Configure if Flex-OneNAND boundary should be locked. 4784 Once locked, the boundary cannot be changed. 4785 1 indicates lock status, 0 indicates unlock status. 4786 4787 oops=panic [KNL,EARLY] 4788 Always panic on oopses. Default is to just kill the 4789 process, but there is a small probability of 4790 deadlocking the machine. 4791 This will also cause panics on machine check exceptions. 4792 Useful together with panic=30 to trigger a reboot. 4793 4794 page_alloc.shuffle= 4795 [KNL] Boolean flag to control whether the page allocator 4796 should randomize its free lists. This parameter can be 4797 used to enable/disable page randomization. The state of 4798 the flag can be read from sysfs at: 4799 /sys/module/page_alloc/parameters/shuffle. 4800 This parameter is only available if CONFIG_SHUFFLE_PAGE_ALLOCATOR=y. 4801 4802 page_owner= [KNL,EARLY] Boot-time page_owner enabling option. 4803 Storage of the information about who allocated 4804 each page is disabled in default. With this switch, 4805 we can turn it on. 4806 on: enable the feature 4807 4808 page_poison= [KNL,EARLY] Boot-time parameter changing the state of 4809 poisoning on the buddy allocator, available with 4810 CONFIG_PAGE_POISONING=y. 4811 off: turn off poisoning (default) 4812 on: turn on poisoning 4813 4814 page_reporting.page_reporting_order= 4815 [KNL] Minimal page reporting order 4816 Format: <integer> 4817 Adjust the minimal page reporting order. The page 4818 reporting is disabled when it exceeds MAX_PAGE_ORDER. 4819 4820 panic= [KNL] Kernel behaviour on panic: delay <timeout> 4821 timeout > 0: seconds before rebooting 4822 timeout = 0: wait forever 4823 timeout < 0: reboot immediately 4824 Format: <timeout> 4825 4826 panic_on_taint= [KNL,EARLY] 4827 Bitmask for conditionally calling panic() in add_taint() 4828 Format: <hex>[,nousertaint] 4829 Hexadecimal bitmask representing the set of TAINT flags 4830 that will cause the kernel to panic when add_taint() is 4831 called with any of the flags in this set. 4832 The optional switch "nousertaint" can be utilized to 4833 prevent userspace forced crashes by writing to sysctl 4834 /proc/sys/kernel/tainted any flagset matching with the 4835 bitmask set on panic_on_taint. 4836 See Documentation/admin-guide/tainted-kernels.rst for 4837 extra details on the taint flags that users can pick 4838 to compose the bitmask to assign to panic_on_taint. 4839 4840 panic_on_warn=1 panic() instead of WARN(). Useful to cause kdump 4841 on a WARN(). 4842 4843 panic_force_cpu= 4844 [KNL,SMP] Force panic handling to execute on a specific CPU. 4845 Format: <cpu number> 4846 Some platforms require panic handling to occur on a 4847 specific CPU for the crash kernel to function correctly. 4848 This can be due to firmware limitations, interrupt routing 4849 constraints, or platform-specific requirements where only 4850 a particular CPU can safely enter the crash kernel. 4851 When set, panic() will redirect execution to the specified 4852 CPU before proceeding with the normal panic and kexec flow. 4853 If the target CPU is offline or unavailable, panic proceeds 4854 on the current CPU. 4855 This option should only be used for systems with the above 4856 constraints as it might cause the panic operation to be less reliable. 4857 4858 panic_print= Bitmask for printing system info when panic happens. 4859 User can chose combination of the following bits: 4860 bit 0: print all tasks info 4861 bit 1: print system memory info 4862 bit 2: print timer info 4863 bit 3: print locks info if CONFIG_LOCKDEP is on 4864 bit 4: print ftrace buffer 4865 bit 5: replay all kernel messages on consoles at the end of panic 4866 bit 6: print all CPUs backtrace (if available in the arch) 4867 bit 7: print only tasks in uninterruptible (blocked) state 4868 *Be aware* that this option may print a _lot_ of lines, 4869 so there are risks of losing older messages in the log. 4870 Use this option carefully, maybe worth to setup a 4871 bigger log buffer with "log_buf_len" along with this. 4872 4873 panic_sys_info= A comma separated list of extra information to be dumped 4874 on panic. 4875 Format: val[,val...] 4876 Where @val can be any of the following: 4877 4878 tasks: print all tasks info 4879 mem: print system memory info 4880 timers: print timers info 4881 locks: print locks info if CONFIG_LOCKDEP is on 4882 ftrace: print ftrace buffer 4883 all_bt: print all CPUs backtrace (if available in the arch) 4884 blocked_tasks: print only tasks in uninterruptible (blocked) state 4885 4886 This is a human readable alternative to the 'panic_print' option. 4887 4888 panic_console_replay 4889 When panic happens, replay all kernel messages on 4890 consoles at the end of panic. 4891 4892 parkbd.port= [HW] Parallel port number the keyboard adapter is 4893 connected to, default is 0. 4894 Format: <parport#> 4895 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 4896 0 for XT, 1 for AT (default is AT). 4897 Format: <mode> 4898 4899 parport= [HW,PPT] Specify parallel ports. 0 disables. 4900 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 4901 Use 'auto' to force the driver to use any 4902 IRQ/DMA settings detected (the default is to 4903 ignore detected IRQ/DMA settings because of 4904 possible conflicts). You can specify the base 4905 address, IRQ, and DMA settings; IRQ and DMA 4906 should be numbers, or 'auto' (for using detected 4907 settings on that particular port), or 'nofifo' 4908 (to avoid using a FIFO even if it is detected). 4909 Parallel ports are assigned in the order they 4910 are specified on the command line, starting 4911 with parport0. 4912 4913 parport_init_mode= [HW,PPT] 4914 Configure VIA parallel port to operate in 4915 a specific mode. This is necessary on Pegasos 4916 computer where firmware has no options for setting 4917 up parallel port mode and sets it to spp. 4918 Currently this function knows 686a and 8231 chips. 4919 Format: [spp|ps2|epp|ecp|ecpepp] 4920 4921 pata_legacy.all= [HW,LIBATA] 4922 Format: <int> 4923 Set to non-zero to probe primary and secondary ISA 4924 port ranges on PCI systems where no PCI PATA device 4925 has been found at either range. Disabled by default. 4926 4927 pata_legacy.autospeed= [HW,LIBATA] 4928 Format: <int> 4929 Set to non-zero if a chip is present that snoops speed 4930 changes. Disabled by default. 4931 4932 pata_legacy.ht6560a= [HW,LIBATA] 4933 Format: <int> 4934 Set to 1, 2, or 3 for HT 6560A on the primary channel, 4935 the secondary channel, or both channels respectively. 4936 Disabled by default. 4937 4938 pata_legacy.ht6560b= [HW,LIBATA] 4939 Format: <int> 4940 Set to 1, 2, or 3 for HT 6560B on the primary channel, 4941 the secondary channel, or both channels respectively. 4942 Disabled by default. 4943 4944 pata_legacy.iordy_mask= [HW,LIBATA] 4945 Format: <int> 4946 IORDY enable mask. Set individual bits to allow IORDY 4947 for the respective channel. Bit 0 is for the first 4948 legacy channel handled by this driver, bit 1 is for 4949 the second channel, and so on. The sequence will often 4950 correspond to the primary legacy channel, the secondary 4951 legacy channel, and so on, but the handling of a PCI 4952 bus and the use of other driver options may interfere 4953 with the sequence. By default IORDY is allowed across 4954 all channels. 4955 4956 pata_legacy.opti82c46x= [HW,LIBATA] 4957 Format: <int> 4958 Set to 1, 2, or 3 for Opti 82c611A on the primary 4959 channel, the secondary channel, or both channels 4960 respectively. Disabled by default. 4961 4962 pata_legacy.opti82c611a= [HW,LIBATA] 4963 Format: <int> 4964 Set to 1, 2, or 3 for Opti 82c465MV on the primary 4965 channel, the secondary channel, or both channels 4966 respectively. Disabled by default. 4967 4968 pata_legacy.pio_mask= [HW,LIBATA] 4969 Format: <int> 4970 PIO mode mask for autospeed devices. Set individual 4971 bits to allow the use of the respective PIO modes. 4972 Bit 0 is for mode 0, bit 1 is for mode 1, and so on. 4973 All modes allowed by default. 4974 4975 pata_legacy.probe_all= [HW,LIBATA] 4976 Format: <int> 4977 Set to non-zero to probe tertiary and further ISA 4978 port ranges on PCI systems. Disabled by default. 4979 4980 pata_legacy.probe_mask= [HW,LIBATA] 4981 Format: <int> 4982 Probe mask for legacy ISA PATA ports. Depending on 4983 platform configuration and the use of other driver 4984 options up to 6 legacy ports are supported: 0x1f0, 4985 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing 4986 of individual ports can be disabled by setting the 4987 corresponding bits in the mask to 1. Bit 0 is for 4988 the first port in the list above (0x1f0), and so on. 4989 By default all supported ports are probed. 4990 4991 pata_legacy.qdi= [HW,LIBATA] 4992 Format: <int> 4993 Set to non-zero to probe QDI controllers. By default 4994 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise. 4995 4996 pata_legacy.winbond= [HW,LIBATA] 4997 Format: <int> 4998 Set to non-zero to probe Winbond controllers. Use 4999 the standard I/O port (0x130) if 1, otherwise the 5000 value given is the I/O port to use (typically 0x1b0). 5001 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE, 5002 0 otherwise. 5003 5004 pata_platform.pio_mask= [HW,LIBATA] 5005 Format: <int> 5006 Supported PIO mode mask. Set individual bits to allow 5007 the use of the respective PIO modes. Bit 0 is for 5008 mode 0, bit 1 is for mode 1, and so on. Mode 0 only 5009 allowed by default. 5010 5011 pause_on_oops=<int> 5012 Halt all CPUs after the first oops has been printed for 5013 the specified number of seconds. This is to be used if 5014 your oopses keep scrolling off the screen. 5015 5016 pci=option[,option...] [PCI,EARLY] various PCI subsystem options. 5017 5018 Some options herein operate on a specific device 5019 or a set of devices (<pci_dev>). These are 5020 specified in one of the following formats: 5021 5022 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 5023 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 5024 5025 Note: the first format specifies a PCI 5026 bus/device/function address which may change 5027 if new hardware is inserted, if motherboard 5028 firmware changes, or due to changes caused 5029 by other kernel parameters. If the 5030 domain is left unspecified, it is 5031 taken to be zero. Optionally, a path 5032 to a device through multiple device/function 5033 addresses can be specified after the base 5034 address (this is more robust against 5035 renumbering issues). The second format 5036 selects devices using IDs from the 5037 configuration space which may match multiple 5038 devices in the system. 5039 5040 earlydump dump PCI config space before the kernel 5041 changes anything 5042 off [X86] don't probe for the PCI bus 5043 bios [X86-32] force use of PCI BIOS, don't access 5044 the hardware directly. Use this if your machine 5045 has a non-standard PCI host bridge. 5046 nobios [X86-32] disallow use of PCI BIOS, only direct 5047 hardware access methods are allowed. Use this 5048 if you experience crashes upon bootup and you 5049 suspect they are caused by the BIOS. 5050 conf1 [X86] Force use of PCI Configuration Access 5051 Mechanism 1 (config address in IO port 0xCF8, 5052 data in IO port 0xCFC, both 32-bit). 5053 conf2 [X86] Force use of PCI Configuration Access 5054 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 5055 the function, IO port 0xCFA, also 8-bit, sets 5056 bus number. The config space is then accessed 5057 through ports 0xC000-0xCFFF). 5058 See http://wiki.osdev.org/PCI for more info 5059 on the configuration access mechanisms. 5060 noaer [PCIE] If the PCIEAER kernel config parameter is 5061 enabled, this kernel boot option can be used to 5062 disable the use of PCIE advanced error reporting. 5063 nodomains [PCI] Disable support for multiple PCI 5064 root domains (aka PCI segments, in ACPI-speak). 5065 nommconf [X86] Disable use of MMCONFIG for PCI 5066 Configuration 5067 check_enable_amd_mmconf [X86] check for and enable 5068 properly configured MMIO access to PCI 5069 config space on AMD family 10h CPU 5070 nomsi [MSI] If the PCI_MSI kernel config parameter is 5071 enabled, this kernel boot option can be used to 5072 disable the use of MSI interrupts system-wide. 5073 noioapicquirk [APIC] Disable all boot interrupt quirks. 5074 Safety option to keep boot IRQs enabled. This 5075 should never be necessary. 5076 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 5077 primary IO-APIC for bridges that cannot disable 5078 boot IRQs. This fixes a source of spurious IRQs 5079 when the system masks IRQs. 5080 noioapicreroute [APIC] Disable workaround that uses the 5081 boot IRQ equivalent of an IRQ that connects to 5082 a chipset where boot IRQs cannot be disabled. 5083 The opposite of ioapicreroute. 5084 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 5085 routing table. These calls are known to be buggy 5086 on several machines and they hang the machine 5087 when used, but on other computers it's the only 5088 way to get the interrupt routing table. Try 5089 this option if the kernel is unable to allocate 5090 IRQs or discover secondary PCI buses on your 5091 motherboard. 5092 rom [X86] Assign address space to expansion ROMs. 5093 Use with caution as certain devices share 5094 address decoders between ROMs and other 5095 resources. 5096 norom [X86] Do not assign address space to 5097 expansion ROMs that do not already have 5098 BIOS assigned address ranges. 5099 nobar [X86] Do not assign address space to the 5100 BARs that weren't assigned by the BIOS. 5101 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 5102 assigned automatically to PCI devices. You can 5103 make the kernel exclude IRQs of your ISA cards 5104 this way. 5105 pirqaddr=0xAAAAA [X86] Specify the physical address 5106 of the PIRQ table (normally generated 5107 by the BIOS) if it is outside the 5108 F0000h-100000h range. 5109 lastbus=N [X86] Scan all buses thru bus #N. Can be 5110 useful if the kernel is unable to find your 5111 secondary buses and you want to tell it 5112 explicitly which ones they are. 5113 assign-busses [X86] Always assign all PCI bus 5114 numbers ourselves, overriding 5115 whatever the firmware may have done. 5116 usepirqmask [X86] Honor the possible IRQ mask stored 5117 in the BIOS $PIR table. This is needed on 5118 some systems with broken BIOSes, notably 5119 some HP Pavilion N5400 and Omnibook XE3 5120 notebooks. This will have no effect if ACPI 5121 IRQ routing is enabled. 5122 noacpi [X86] Do not use ACPI for IRQ routing 5123 or for PCI scanning. 5124 use_crs [X86] Use PCI host bridge window information 5125 from ACPI. On BIOSes from 2008 or later, this 5126 is enabled by default. If you need to use this, 5127 please report a bug. 5128 nocrs [X86] Ignore PCI host bridge windows from ACPI. 5129 If you need to use this, please report a bug. 5130 use_e820 [X86] Use E820 reservations to exclude parts of 5131 PCI host bridge windows. This is a workaround 5132 for BIOS defects in host bridge _CRS methods. 5133 If you need to use this, please report a bug to 5134 <linux-pci@vger.kernel.org>. 5135 no_e820 [X86] Ignore E820 reservations for PCI host 5136 bridge windows. This is the default on modern 5137 hardware. If you need to use this, please report 5138 a bug to <linux-pci@vger.kernel.org>. 5139 routeirq Do IRQ routing for all PCI devices. 5140 This is normally done in pci_enable_device(), 5141 so this option is a temporary workaround 5142 for broken drivers that don't call it. 5143 skip_isa_align [X86] do not align io start addr, so can 5144 handle more pci cards 5145 noearly [X86] Don't do any early type 1 scanning. 5146 This might help on some broken boards which 5147 machine check when some devices' config space 5148 is read. But various workarounds are disabled 5149 and some IOMMU drivers will not work. 5150 bfsort Sort PCI devices into breadth-first order. 5151 This sorting is done to get a device 5152 order compatible with older (<= 2.4) kernels. 5153 nobfsort Don't sort PCI devices into breadth-first order. 5154 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 5155 tuning and use the BIOS-configured MPS defaults. 5156 pcie_bus_safe Set every device's MPS to the largest value 5157 supported by all devices below the root complex. 5158 pcie_bus_perf Set device MPS to the largest allowable MPS 5159 based on its parent bus. Also set MRRS (Max 5160 Read Request Size) to the largest supported 5161 value (no larger than the MPS that the device 5162 or bus can support) for best performance. 5163 pcie_bus_peer2peer Set every device's MPS to 128B, which 5164 every device is guaranteed to support. This 5165 configuration allows peer-to-peer DMA between 5166 any pair of devices, possibly at the cost of 5167 reduced performance. This also guarantees 5168 that hot-added devices will work. 5169 cbiosize=nn[KMG] The fixed amount of bus space which is 5170 reserved for the CardBus bridge's IO window. 5171 The default value is 256 bytes. 5172 cbmemsize=nn[KMG] The fixed amount of bus space which is 5173 reserved for the CardBus bridge's memory 5174 window. The default value is 64 megabytes. 5175 resource_alignment= 5176 Format: 5177 [<order of align>@]<pci_dev>[; ...] 5178 Specifies alignment and device to reassign 5179 aligned memory resources. How to 5180 specify the device is described above. 5181 If <order of align> is not specified, 5182 PAGE_SIZE is used as alignment. 5183 A PCI-PCI bridge can be specified if resource 5184 windows need to be expanded. 5185 To specify the alignment for several 5186 instances of a device, the PCI vendor, 5187 device, subvendor, and subdevice may be 5188 specified, e.g., 12@pci:8086:9c22:103c:198f 5189 for 4096-byte alignment. 5190 ecrc= Enable/disable PCIe ECRC (transaction layer 5191 end-to-end CRC checking). Only effective if 5192 OS has native AER control (either granted by 5193 ACPI _OSC or forced via "pcie_ports=native") 5194 bios: Use BIOS/firmware settings. This is the 5195 the default. 5196 off: Turn ECRC off 5197 on: Turn ECRC on. 5198 hpiosize=nn[KMG] The fixed amount of bus space which is 5199 reserved for hotplug bridge's IO window. 5200 Default size is 256 bytes. 5201 hpmmiosize=nn[KMG] The fixed amount of bus space which is 5202 reserved for hotplug bridge's MMIO window. 5203 Default size is 2 megabytes. 5204 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 5205 reserved for hotplug bridge's MMIO_PREF window. 5206 Default size is 2 megabytes. 5207 hpmemsize=nn[KMG] The fixed amount of bus space which is 5208 reserved for hotplug bridge's MMIO and 5209 MMIO_PREF window. 5210 Default size is 2 megabytes. 5211 hpbussize=nn The minimum amount of additional bus numbers 5212 reserved for buses below a hotplug bridge. 5213 Default is 1. 5214 realloc= Enable/disable reallocating PCI bridge resources 5215 if allocations done by BIOS are too small to 5216 accommodate resources required by all child 5217 devices. 5218 off: Turn realloc off 5219 on: Turn realloc on 5220 realloc same as realloc=on 5221 noari do not use PCIe ARI. 5222 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 5223 do not use PCIe ATS (and IOMMU device IOTLB). 5224 pcie_scan_all Scan all possible PCIe devices. Otherwise we 5225 only look for one device below a PCIe downstream 5226 port. 5227 big_root_window Try to add a big 64bit memory window to the PCIe 5228 root complex on AMD CPUs. Some GFX hardware 5229 can resize a BAR to allow access to all VRAM. 5230 Adding the window is slightly risky (it may 5231 conflict with unreported devices), so this 5232 taints the kernel. 5233 disable_acs_redir=<pci_dev>[; ...] 5234 Specify one or more PCI devices (in the format 5235 specified above) separated by semicolons. 5236 Each device specified will have the PCI ACS 5237 redirect capabilities forced off which will 5238 allow P2P traffic between devices through 5239 bridges without forcing it upstream. Note: 5240 this removes isolation between devices and 5241 may put more devices in an IOMMU group. 5242 config_acs= 5243 Format: 5244 <ACS flags>@<pci_dev>[; ...] 5245 Specify one or more PCI devices (in the format 5246 specified above) optionally prepended with flags 5247 and separated by semicolons. The respective 5248 capabilities will be enabled, disabled or 5249 unchanged based on what is specified in 5250 flags. 5251 5252 ACS Flags is defined as follows: 5253 bit-0 : ACS Source Validation 5254 bit-1 : ACS Translation Blocking 5255 bit-2 : ACS P2P Request Redirect 5256 bit-3 : ACS P2P Completion Redirect 5257 bit-4 : ACS Upstream Forwarding 5258 bit-5 : ACS P2P Egress Control 5259 bit-6 : ACS Direct Translated P2P 5260 Each bit can be marked as: 5261 '0' – force disabled 5262 '1' – force enabled 5263 'x' – unchanged 5264 For example, 5265 pci=config_acs=10x@pci:0:0 5266 would configure all devices that support 5267 ACS to enable P2P Request Redirect, disable 5268 Translation Blocking, and leave Source 5269 Validation unchanged from whatever power-up 5270 or firmware set it to. 5271 5272 Note: this may remove isolation between devices 5273 and may put more devices in an IOMMU group. 5274 force_floating [S390] Force usage of floating interrupts. 5275 nomio [S390] Do not use MIO instructions. 5276 norid [S390] ignore the RID field and force use of 5277 one PCI domain per PCI function 5278 notph [PCIE] If the PCIE_TPH kernel config parameter 5279 is enabled, this kernel boot option can be used 5280 to disable PCIe TLP Processing Hints support 5281 system-wide. 5282 5283 pcie_aspm= [PCIE] Forcibly enable or ignore PCIe Active State Power 5284 Management. 5285 off Don't touch ASPM configuration at all. Leave any 5286 configuration done by firmware unchanged. 5287 force Enable ASPM even on devices that claim not to support it. 5288 WARNING: Forcing ASPM on may cause system lockups. 5289 5290 pcie_ports= [PCIE] PCIe port services handling: 5291 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 5292 even if the platform doesn't give the OS permission to 5293 use them. This may cause conflicts if the platform 5294 also tries to use these services. 5295 dpc-native Use native PCIe service for DPC only. May 5296 cause conflicts if firmware uses AER or DPC. 5297 compat Disable native PCIe services (PME, AER, DPC, PCIe 5298 hotplug). 5299 5300 pcie_port_pm= [PCIE] PCIe port power management handling: 5301 off Disable power management of all PCIe ports 5302 force Forcibly enable power management of all PCIe ports 5303 5304 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 5305 nomsi Do not use MSI for native PCIe PME signaling (this makes 5306 all PCIe root ports use INTx for all services). 5307 5308 pcmv= [HW,PCMCIA] BadgePAD 4 5309 5310 pd_ignore_unused 5311 [PM] 5312 Keep all power-domains already enabled by bootloader on, 5313 even if no driver has claimed them. This is useful 5314 for debug and development, but should not be 5315 needed on a platform with proper driver support. 5316 5317 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 5318 boot time. 5319 Format: { 0 | 1 } 5320 See arch/parisc/kernel/pdc_chassis.c 5321 5322 percpu_alloc= [MM,EARLY] 5323 Select which percpu first chunk allocator to use. 5324 Currently supported values are "embed" and "page". 5325 Archs may support subset or none of the selections. 5326 See comments in mm/percpu.c for details on each 5327 allocator. This parameter is primarily for debugging 5328 and performance comparison. 5329 5330 pirq= [SMP,APIC] Manual mp-table setup 5331 See Documentation/arch/x86/i386/IO-APIC.rst. 5332 5333 plip= [PPT,NET] Parallel port network link 5334 Format: { parport<nr> | timid | 0 } 5335 See also Documentation/admin-guide/parport.rst. 5336 5337 pmtmr= [X86] Manual setup of pmtmr I/O Port. 5338 Override pmtimer IOPort with a hex value. 5339 e.g. pmtmr=0x508 5340 5341 pmu_override= [PPC] Override the PMU. 5342 This option takes over the PMU facility, so it is no 5343 longer usable by perf. Setting this option starts the 5344 PMU counters by setting MMCR0 to 0 (the FC bit is 5345 cleared). If a number is given, then MMCR1 is set to 5346 that number, otherwise (e.g., 'pmu_override=on'), MMCR1 5347 remains 0. 5348 5349 pm_async= [PM] 5350 Format: off 5351 This parameter sets the initial value of the 5352 /sys/power/pm_async sysfs knob at boot time. 5353 If set to "off", disables asynchronous suspend and 5354 resume of devices during system-wide power transitions. 5355 This can be useful on platforms where device 5356 dependencies are not well-defined, or for debugging 5357 power management issues. Asynchronous operations are 5358 enabled by default. 5359 5360 5361 pm_debug_messages [SUSPEND,KNL] 5362 Enable suspend/resume debug messages during boot up. 5363 5364 pnp.debug=1 [PNP] 5365 Enable PNP debug messages (depends on the 5366 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 5367 via /sys/module/pnp/parameters/debug. We always show 5368 current resource usage; turning this on also shows 5369 possible settings and some assignment information. 5370 5371 pnpacpi= [ACPI] 5372 { off } 5373 5374 pnpbios= [ISAPNP] 5375 { on | off | curr | res | no-curr | no-res } 5376 5377 pnp_reserve_irq= 5378 [ISAPNP] Exclude IRQs for the autoconfiguration 5379 5380 pnp_reserve_dma= 5381 [ISAPNP] Exclude DMAs for the autoconfiguration 5382 5383 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 5384 Ranges are in pairs (I/O port base and size). 5385 5386 pnp_reserve_mem= 5387 [ISAPNP] Exclude memory regions for the 5388 autoconfiguration. 5389 Ranges are in pairs (memory base and size). 5390 5391 ports= [IP_VS_FTP] IPVS ftp helper module 5392 Default is 21. 5393 Up to 8 (IP_VS_APP_MAX_PORTS) ports 5394 may be specified. 5395 Format: <port>,<port>.... 5396 5397 possible_cpus= [SMP,S390,X86] 5398 Format: <unsigned int> 5399 Set the number of possible CPUs, overriding the 5400 regular discovery mechanisms (such as ACPI/FW, etc). 5401 5402 powersave=off [PPC] This option disables power saving features. 5403 It specifically disables cpuidle and sets the 5404 platform machine description specific power_save 5405 function to NULL. On Idle the CPU just reduces 5406 execution priority. 5407 5408 ppc_strict_facility_enable 5409 [PPC,ENABLE] This option catches any kernel floating point, 5410 Altivec, VSX and SPE outside of regions specifically 5411 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 5412 There is some performance impact when enabling this. 5413 5414 ppc_tm= [PPC,EARLY] 5415 Format: {"off"} 5416 Disable Hardware Transactional Memory 5417 5418 preempt= [KNL] 5419 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC 5420 none - Limited to cond_resched() calls 5421 voluntary - Limited to cond_resched() and might_sleep() calls 5422 full - Any section that isn't explicitly preempt disabled 5423 can be preempted anytime. Tasks will also yield 5424 contended spinlocks (if the critical section isn't 5425 explicitly preempt disabled beyond the lock itself). 5426 lazy - Scheduler controlled. Similar to full but instead 5427 of preempting the task immediately, the task gets 5428 one HZ tick time to yield itself before the 5429 preemption will be forced. One preemption is when the 5430 task returns to user space. 5431 5432 print-fatal-signals= 5433 [KNL] debug: print fatal signals 5434 5435 If enabled, warn about various signal handling 5436 related application anomalies: too many signals, 5437 too many POSIX.1 timers, fatal signals causing a 5438 coredump - etc. 5439 5440 If you hit the warning due to signal overflow, 5441 you might want to try "ulimit -i unlimited". 5442 5443 default: off. 5444 5445 printk.always_kmsg_dump= 5446 Trigger kmsg_dump for cases other than kernel oops or 5447 panics 5448 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5449 default: disabled 5450 5451 printk.console_no_auto_verbose= 5452 Disable console loglevel raise on oops, panic 5453 or lockdep-detected issues (only if lock debug is on). 5454 With an exception to setups with low baudrate on 5455 serial console, keeping this 0 is a good choice 5456 in order to provide more debug information. 5457 Format: <bool> 5458 default: 0 (auto_verbose is enabled) 5459 5460 printk.debug_non_panic_cpus= 5461 Allows storing messages from non-panic CPUs into 5462 the printk log buffer during panic(). They are 5463 flushed to consoles by the panic-CPU on 5464 a best-effort basis. 5465 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5466 Default: disabled 5467 5468 printk.devkmsg={on,off,ratelimit} 5469 Control writing to /dev/kmsg. 5470 on - unlimited logging to /dev/kmsg from userspace 5471 off - logging to /dev/kmsg disabled 5472 ratelimit - ratelimit the logging 5473 Default: ratelimit 5474 5475 printk.time= Show timing data prefixed to each printk message line 5476 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5477 5478 proc_mem.force_override= [KNL] 5479 Format: {always | ptrace | never} 5480 Traditionally /proc/pid/mem allows memory permissions to be 5481 overridden without restrictions. This option may be set to 5482 restrict that. Can be one of: 5483 - 'always': traditional behavior always allows mem overrides. 5484 - 'ptrace': only allow mem overrides for active ptracers. 5485 - 'never': never allow mem overrides. 5486 If not specified, default is the CONFIG_PROC_MEM_* choice. 5487 5488 processor.max_cstate= [HW,ACPI] 5489 Limit processor to maximum C-state 5490 max_cstate=9 overrides any DMI blacklist limit. 5491 5492 processor.nocst [HW,ACPI] 5493 Ignore the _CST method to determine C-states, 5494 instead using the legacy FADT method 5495 5496 profile= [KNL] Enable kernel profiling via /proc/profile 5497 Format: [<profiletype>,]<number> 5498 Param: <profiletype>: "schedule" or "kvm" 5499 [defaults to kernel profiling] 5500 Param: "schedule" - profile schedule points. 5501 Param: "kvm" - profile VM exits. 5502 Param: <number> - step/bucket size as a power of 2 for 5503 statistical time based profiling. 5504 5505 prot_virt= [S390] enable hosting protected virtual machines 5506 isolated from the hypervisor (if hardware supports 5507 that). If enabled, the default kernel base address 5508 might be overridden even when Kernel Address Space 5509 Layout Randomization is disabled. 5510 Format: <bool> 5511 5512 psi= [KNL] Enable or disable pressure stall information 5513 tracking. 5514 Format: <bool> 5515 5516 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 5517 probe for; one of (bare|imps|exps|lifebook|any). 5518 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 5519 per second. 5520 psmouse.resetafter= [HW,MOUSE] 5521 Try to reset the device after so many bad packets 5522 (0 = never). 5523 psmouse.resolution= 5524 [HW,MOUSE] Set desired mouse resolution, in dpi. 5525 psmouse.smartscroll= 5526 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 5527 0 = disabled, 1 = enabled (default). 5528 5529 pstore.backend= Specify the name of the pstore backend to use 5530 5531 pti= [X86-64] Control Page Table Isolation of user and 5532 kernel address spaces. Disabling this feature 5533 removes hardening, but improves performance of 5534 system calls and interrupts. 5535 5536 on - unconditionally enable 5537 off - unconditionally disable 5538 auto - kernel detects whether your CPU model is 5539 vulnerable to issues that PTI mitigates 5540 5541 Not specifying this option is equivalent to pti=auto. 5542 5543 pty.legacy_count= 5544 [KNL] Number of legacy pty's. Overwrites compiled-in 5545 default number. 5546 5547 quiet [KNL,EARLY] Disable most log messages 5548 5549 r128= [HW,DRM] 5550 5551 radix_hcall_invalidate=on [PPC/PSERIES] 5552 Disable RADIX GTSE feature and use hcall for TLB 5553 invalidate. 5554 5555 raid= [HW,RAID] 5556 See Documentation/admin-guide/md.rst. 5557 5558 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 5559 See Documentation/admin-guide/blockdev/ramdisk.rst. 5560 5561 ramdisk_start= [Deprecated,RAM] RAM disk image start address 5562 5563 random.trust_cpu=off 5564 [KNL,EARLY] Disable trusting the use of the CPU's 5565 random number generator (if available) to 5566 initialize the kernel's RNG. 5567 5568 random.trust_bootloader=off 5569 [KNL,EARLY] Disable trusting the use of the a seed 5570 passed by the bootloader (if available) to 5571 initialize the kernel's RNG. 5572 5573 randomize_kstack_offset= 5574 [KNL,EARLY] Enable or disable kernel stack offset 5575 randomization, which provides roughly 5 bits of 5576 entropy, frustrating memory corruption attacks 5577 that depend on stack address determinism or 5578 cross-syscall address exposures. This is only 5579 available on architectures that have defined 5580 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET. 5581 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5582 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT. 5583 5584 ras=option[,option,...] [KNL] RAS-specific options 5585 5586 cec_disable [X86] 5587 Disable the Correctable Errors Collector, 5588 see CONFIG_RAS_CEC help text. 5589 5590 rcu_nocbs[=cpu-list] 5591 [KNL] The optional argument is a cpu list, 5592 as described above. 5593 5594 In kernels built with CONFIG_RCU_NOCB_CPU=y, 5595 enable the no-callback CPU mode, which prevents 5596 such CPUs' callbacks from being invoked in 5597 softirq context. Invocation of such CPUs' RCU 5598 callbacks will instead be offloaded to "rcuox/N" 5599 kthreads created for that purpose, where "x" is 5600 "p" for RCU-preempt, "s" for RCU-sched, and "g" 5601 for the kthreads that mediate grace periods; and 5602 "N" is the CPU number. This reduces OS jitter on 5603 the offloaded CPUs, which can be useful for HPC 5604 and real-time workloads. It can also improve 5605 energy efficiency for asymmetric multiprocessors. 5606 5607 If a cpulist is passed as an argument, the specified 5608 list of CPUs is set to no-callback mode from boot. 5609 5610 Otherwise, if the '=' sign and the cpulist 5611 arguments are omitted, no CPU will be set to 5612 no-callback mode from boot but the mode may be 5613 toggled at runtime via cpusets. 5614 5615 Note that this argument takes precedence over 5616 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 5617 5618 rcu_nocb_poll [KNL] 5619 Rather than requiring that offloaded CPUs 5620 (specified by rcu_nocbs= above) explicitly 5621 awaken the corresponding "rcuoN" kthreads, 5622 make these kthreads poll for callbacks. 5623 This improves the real-time response for the 5624 offloaded CPUs by relieving them of the need to 5625 wake up the corresponding kthread, but degrades 5626 energy efficiency by requiring that the kthreads 5627 periodically wake up to do the polling. 5628 5629 rcutree.blimit= [KNL] 5630 Set maximum number of finished RCU callbacks to 5631 process in one batch. 5632 5633 rcutree.csd_lock_suppress_rcu_stall= [KNL] 5634 Do only a one-line RCU CPU stall warning when 5635 there is an ongoing too-long CSD-lock wait. 5636 5637 rcutree.do_rcu_barrier= [KNL] 5638 Request a call to rcu_barrier(). This is 5639 throttled so that userspace tests can safely 5640 hammer on the sysfs variable if they so choose. 5641 If triggered before the RCU grace-period machinery 5642 is fully active, this will error out with EAGAIN. 5643 5644 rcutree.dump_tree= [KNL] 5645 Dump the structure of the rcu_node combining tree 5646 out at early boot. This is used for diagnostic 5647 purposes, to verify correct tree setup. 5648 5649 rcutree.gp_cleanup_delay= [KNL] 5650 Set the number of jiffies to delay each step of 5651 RCU grace-period cleanup. 5652 5653 rcutree.gp_init_delay= [KNL] 5654 Set the number of jiffies to delay each step of 5655 RCU grace-period initialization. 5656 5657 rcutree.gp_preinit_delay= [KNL] 5658 Set the number of jiffies to delay each step of 5659 RCU grace-period pre-initialization, that is, 5660 the propagation of recent CPU-hotplug changes up 5661 the rcu_node combining tree. 5662 5663 rcutree.jiffies_till_first_fqs= [KNL] 5664 Set delay from grace-period initialization to 5665 first attempt to force quiescent states. 5666 Units are jiffies, minimum value is zero, 5667 and maximum value is HZ. 5668 5669 rcutree.jiffies_till_next_fqs= [KNL] 5670 Set delay between subsequent attempts to force 5671 quiescent states. Units are jiffies, minimum 5672 value is one, and maximum value is HZ. 5673 5674 rcutree.jiffies_till_sched_qs= [KNL] 5675 Set required age in jiffies for a 5676 given grace period before RCU starts 5677 soliciting quiescent-state help from 5678 rcu_note_context_switch() and cond_resched(). 5679 If not specified, the kernel will calculate 5680 a value based on the most recent settings 5681 of rcutree.jiffies_till_first_fqs 5682 and rcutree.jiffies_till_next_fqs. 5683 This calculated value may be viewed in 5684 rcutree.jiffies_to_sched_qs. Any attempt to set 5685 rcutree.jiffies_to_sched_qs will be cheerfully 5686 overwritten. 5687 5688 rcutree.kthread_prio= [KNL,BOOT] 5689 Set the SCHED_FIFO priority of the RCU per-CPU 5690 kthreads (rcuc/N). This value is also used for 5691 the priority of the RCU boost threads (rcub/N) 5692 and for the RCU grace-period kthreads (rcu_bh, 5693 rcu_preempt, and rcu_sched). If RCU_BOOST is 5694 set, valid values are 1-99 and the default is 1 5695 (the least-favored priority). Otherwise, when 5696 RCU_BOOST is not set, valid values are 0-99 and 5697 the default is zero (non-realtime operation). 5698 When RCU_NOCB_CPU is set, also adjust the 5699 priority of NOCB callback kthreads. 5700 5701 rcutree.nocb_nobypass_lim_per_jiffy= [KNL] 5702 On callback-offloaded (rcu_nocbs) CPUs, 5703 RCU reduces the lock contention that would 5704 otherwise be caused by callback floods through 5705 use of the ->nocb_bypass list. However, in the 5706 common non-flooded case, RCU queues directly to 5707 the main ->cblist in order to avoid the extra 5708 overhead of the ->nocb_bypass list and its lock. 5709 But if there are too many callbacks queued during 5710 a single jiffy, RCU pre-queues the callbacks into 5711 the ->nocb_bypass queue. The definition of "too 5712 many" is supplied by this kernel boot parameter. 5713 5714 rcutree.nohz_full_patience_delay= [KNL] 5715 On callback-offloaded (rcu_nocbs) CPUs, avoid 5716 disturbing RCU unless the grace period has 5717 reached the specified age in milliseconds. 5718 Defaults to zero. Large values will be capped 5719 at five seconds. All values will be rounded down 5720 to the nearest value representable by jiffies. 5721 5722 rcutree.qhimark= [KNL] 5723 Set threshold of queued RCU callbacks beyond which 5724 batch limiting is disabled. 5725 5726 rcutree.qlowmark= [KNL] 5727 Set threshold of queued RCU callbacks below which 5728 batch limiting is re-enabled. 5729 5730 rcutree.qovld= [KNL] 5731 Set threshold of queued RCU callbacks beyond which 5732 RCU's force-quiescent-state scan will aggressively 5733 enlist help from cond_resched() and sched IPIs to 5734 help CPUs more quickly reach quiescent states. 5735 Set to less than zero to make this be set based 5736 on rcutree.qhimark at boot time and to zero to 5737 disable more aggressive help enlistment. 5738 5739 rcutree.rcu_delay_page_cache_fill_msec= [KNL] 5740 Set the page-cache refill delay (in milliseconds) 5741 in response to low-memory conditions. The range 5742 of permitted values is in the range 0:100000. 5743 5744 rcutree.rcu_divisor= [KNL] 5745 Set the shift-right count to use to compute 5746 the callback-invocation batch limit bl from 5747 the number of callbacks queued on this CPU. 5748 The result will be bounded below by the value of 5749 the rcutree.blimit kernel parameter. Every bl 5750 callbacks, the softirq handler will exit in 5751 order to allow the CPU to do other work. 5752 5753 Please note that this callback-invocation batch 5754 limit applies only to non-offloaded callback 5755 invocation. Offloaded callbacks are instead 5756 invoked in the context of an rcuoc kthread, which 5757 scheduler will preempt as it does any other task. 5758 5759 rcutree.rcu_fanout_exact= [KNL] 5760 Disable autobalancing of the rcu_node combining 5761 tree. This is used by rcutorture, and might 5762 possibly be useful for architectures having high 5763 cache-to-cache transfer latencies. 5764 5765 rcutree.rcu_fanout_leaf= [KNL] 5766 Change the number of CPUs assigned to each 5767 leaf rcu_node structure. Useful for very 5768 large systems, which will choose the value 64, 5769 and for NUMA systems with large remote-access 5770 latencies, which will choose a value aligned 5771 with the appropriate hardware boundaries. 5772 5773 rcutree.rcu_min_cached_objs= [KNL] 5774 Minimum number of objects which are cached and 5775 maintained per one CPU. Object size is equal 5776 to PAGE_SIZE. The cache allows to reduce the 5777 pressure to page allocator, also it makes the 5778 whole algorithm to behave better in low memory 5779 condition. 5780 5781 rcutree.rcu_nocb_gp_stride= [KNL] 5782 Set the number of NOCB callback kthreads in 5783 each group, which defaults to the square root 5784 of the number of CPUs. Larger numbers reduce 5785 the wakeup overhead on the global grace-period 5786 kthread, but increases that same overhead on 5787 each group's NOCB grace-period kthread. 5788 5789 rcutree.rcu_kick_kthreads= [KNL] 5790 Cause the grace-period kthread to get an extra 5791 wake_up() if it sleeps three times longer than 5792 it should at force-quiescent-state time. 5793 This wake_up() will be accompanied by a 5794 WARN_ONCE() splat and an ftrace_dump(). 5795 5796 rcutree.rcu_resched_ns= [KNL] 5797 Limit the time spend invoking a batch of RCU 5798 callbacks to the specified number of nanoseconds. 5799 By default, this limit is checked only once 5800 every 32 callbacks in order to limit the pain 5801 inflicted by local_clock() overhead. 5802 5803 rcutree.rcu_unlock_delay= [KNL] 5804 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 5805 this specifies an rcu_read_unlock()-time delay 5806 in microseconds. This defaults to zero. 5807 Larger delays increase the probability of 5808 catching RCU pointer leaks, that is, buggy use 5809 of RCU-protected pointers after the relevant 5810 rcu_read_unlock() has completed. 5811 5812 rcutree.sysrq_rcu= [KNL] 5813 Commandeer a sysrq key to dump out Tree RCU's 5814 rcu_node tree with an eye towards determining 5815 why a new grace period has not yet started. 5816 5817 rcutree.use_softirq= [KNL] 5818 If set to zero, move all RCU_SOFTIRQ processing to 5819 per-CPU rcuc kthreads. Defaults to a non-zero 5820 value, meaning that RCU_SOFTIRQ is used by default. 5821 Specify rcutree.use_softirq=0 to use rcuc kthreads. 5822 5823 But note that CONFIG_PREEMPT_RT=y kernels disable 5824 this kernel boot parameter, forcibly setting it 5825 to zero. 5826 5827 rcutree.enable_rcu_lazy= [KNL] 5828 To save power, batch RCU callbacks and flush after 5829 delay, memory pressure or callback list growing too 5830 big. 5831 5832 rcutree.rcu_normal_wake_from_gp= [KNL] 5833 Reduces a latency of synchronize_rcu() call. This approach 5834 maintains its own track of synchronize_rcu() callers, so it 5835 does not interact with regular callbacks because it does not 5836 use a call_rcu[_hurry]() path. Please note, this is for a 5837 normal grace period. 5838 5839 How to disable it: 5840 5841 echo 0 > /sys/module/rcutree/parameters/rcu_normal_wake_from_gp 5842 or pass a boot parameter "rcutree.rcu_normal_wake_from_gp=0" 5843 5844 Default is 1 if it is not explicitly disabled by the boot parameter 5845 passing 0. 5846 5847 rcuscale.gp_async= [KNL] 5848 Measure performance of asynchronous 5849 grace-period primitives such as call_rcu(). 5850 5851 rcuscale.gp_async_max= [KNL] 5852 Specify the maximum number of outstanding 5853 callbacks per writer thread. When a writer 5854 thread exceeds this limit, it invokes the 5855 corresponding flavor of rcu_barrier() to allow 5856 previously posted callbacks to drain. 5857 5858 rcuscale.gp_exp= [KNL] 5859 Measure performance of expedited synchronous 5860 grace-period primitives. 5861 5862 rcuscale.holdoff= [KNL] 5863 Set test-start holdoff period. The purpose of 5864 this parameter is to delay the start of the 5865 test until boot completes in order to avoid 5866 interference. 5867 5868 rcuscale.kfree_by_call_rcu= [KNL] 5869 In kernels built with CONFIG_RCU_LAZY=y, test 5870 call_rcu() instead of kfree_rcu(). 5871 5872 rcuscale.kfree_mult= [KNL] 5873 Instead of allocating an object of size kfree_obj, 5874 allocate one of kfree_mult * sizeof(kfree_obj). 5875 Defaults to 1. 5876 5877 rcuscale.kfree_rcu_test= [KNL] 5878 Set to measure performance of kfree_rcu() flooding. 5879 5880 rcuscale.kfree_rcu_test_double= [KNL] 5881 Test the double-argument variant of kfree_rcu(). 5882 If this parameter has the same value as 5883 rcuscale.kfree_rcu_test_single, both the single- 5884 and double-argument variants are tested. 5885 5886 rcuscale.kfree_rcu_test_single= [KNL] 5887 Test the single-argument variant of kfree_rcu(). 5888 If this parameter has the same value as 5889 rcuscale.kfree_rcu_test_double, both the single- 5890 and double-argument variants are tested. 5891 5892 rcuscale.kfree_nthreads= [KNL] 5893 The number of threads running loops of kfree_rcu(). 5894 5895 rcuscale.kfree_alloc_num= [KNL] 5896 Number of allocations and frees done in an iteration. 5897 5898 rcuscale.kfree_loops= [KNL] 5899 Number of loops doing rcuscale.kfree_alloc_num number 5900 of allocations and frees. 5901 5902 rcuscale.minruntime= [KNL] 5903 Set the minimum test run time in seconds. This 5904 does not affect the data-collection interval, 5905 but instead allows better measurement of things 5906 like CPU consumption. 5907 5908 rcuscale.nreaders= [KNL] 5909 Set number of RCU readers. The value -1 selects 5910 N, where N is the number of CPUs. A value 5911 "n" less than -1 selects N-n+1, where N is again 5912 the number of CPUs. For example, -2 selects N 5913 (the number of CPUs), -3 selects N+1, and so on. 5914 A value of "n" less than or equal to -N selects 5915 a single reader. 5916 5917 rcuscale.nwriters= [KNL] 5918 Set number of RCU writers. The values operate 5919 the same as for rcuscale.nreaders. 5920 N, where N is the number of CPUs 5921 5922 rcuscale.scale_type= [KNL] 5923 Specify the RCU implementation to test. 5924 5925 rcuscale.shutdown= [KNL] 5926 Shut the system down after performance tests 5927 complete. This is useful for hands-off automated 5928 testing. 5929 5930 rcuscale.verbose= [KNL] 5931 Enable additional printk() statements. 5932 5933 rcuscale.writer_holdoff= [KNL] 5934 Write-side holdoff between grace periods, 5935 in microseconds. The default of zero says 5936 no holdoff. 5937 5938 rcuscale.writer_holdoff_jiffies= [KNL] 5939 Additional write-side holdoff between grace 5940 periods, but in jiffies. The default of zero 5941 says no holdoff. 5942 5943 rcutorture.fqs_duration= [KNL] 5944 Set duration of force_quiescent_state bursts 5945 in microseconds. 5946 5947 rcutorture.fqs_holdoff= [KNL] 5948 Set holdoff time within force_quiescent_state bursts 5949 in microseconds. 5950 5951 rcutorture.fqs_stutter= [KNL] 5952 Set wait time between force_quiescent_state bursts 5953 in seconds. 5954 5955 rcutorture.fwd_progress= [KNL] 5956 Specifies the number of kthreads to be used 5957 for RCU grace-period forward-progress testing 5958 for the types of RCU supporting this notion. 5959 Defaults to 1 kthread, values less than zero or 5960 greater than the number of CPUs cause the number 5961 of CPUs to be used. 5962 5963 rcutorture.fwd_progress_div= [KNL] 5964 Specify the fraction of a CPU-stall-warning 5965 period to do tight-loop forward-progress testing. 5966 5967 rcutorture.fwd_progress_holdoff= [KNL] 5968 Number of seconds to wait between successive 5969 forward-progress tests. 5970 5971 rcutorture.fwd_progress_need_resched= [KNL] 5972 Enclose cond_resched() calls within checks for 5973 need_resched() during tight-loop forward-progress 5974 testing. 5975 5976 rcutorture.gp_cond= [KNL] 5977 Use conditional/asynchronous update-side 5978 normal-grace-period primitives, if available. 5979 5980 rcutorture.gp_cond_exp= [KNL] 5981 Use conditional/asynchronous update-side 5982 expedited-grace-period primitives, if available. 5983 5984 rcutorture.gp_cond_full= [KNL] 5985 Use conditional/asynchronous update-side 5986 normal-grace-period primitives that also take 5987 concurrent expedited grace periods into account, 5988 if available. 5989 5990 rcutorture.gp_cond_exp_full= [KNL] 5991 Use conditional/asynchronous update-side 5992 expedited-grace-period primitives that also take 5993 concurrent normal grace periods into account, 5994 if available. 5995 5996 rcutorture.gp_cond_wi= [KNL] 5997 Nominal wait interval for normal conditional 5998 grace periods (specified by rcutorture's 5999 gp_cond and gp_cond_full module parameters), 6000 in microseconds. The actual wait interval will 6001 be randomly selected to nanosecond granularity up 6002 to this wait interval. Defaults to 16 jiffies, 6003 for example, 16,000 microseconds on a system 6004 with HZ=1000. 6005 6006 rcutorture.gp_cond_wi_exp= [KNL] 6007 Nominal wait interval for expedited conditional 6008 grace periods (specified by rcutorture's 6009 gp_cond_exp and gp_cond_exp_full module 6010 parameters), in microseconds. The actual wait 6011 interval will be randomly selected to nanosecond 6012 granularity up to this wait interval. Defaults to 6013 128 microseconds. 6014 6015 rcutorture.gp_exp= [KNL] 6016 Use expedited update-side primitives, if available. 6017 6018 rcutorture.gp_normal= [KNL] 6019 Use normal (non-expedited) asynchronous 6020 update-side primitives, if available. 6021 6022 rcutorture.gp_poll= [KNL] 6023 Use polled update-side normal-grace-period 6024 primitives, if available. 6025 6026 rcutorture.gp_poll_exp= [KNL] 6027 Use polled update-side expedited-grace-period 6028 primitives, if available. 6029 6030 rcutorture.gp_poll_full= [KNL] 6031 Use polled update-side normal-grace-period 6032 primitives that also take concurrent expedited 6033 grace periods into account, if available. 6034 6035 rcutorture.gp_poll_exp_full= [KNL] 6036 Use polled update-side expedited-grace-period 6037 primitives that also take concurrent normal 6038 grace periods into account, if available. 6039 6040 rcutorture.gp_poll_wi= [KNL] 6041 Nominal wait interval for normal conditional 6042 grace periods (specified by rcutorture's 6043 gp_poll and gp_poll_full module parameters), 6044 in microseconds. The actual wait interval will 6045 be randomly selected to nanosecond granularity up 6046 to this wait interval. Defaults to 16 jiffies, 6047 for example, 16,000 microseconds on a system 6048 with HZ=1000. 6049 6050 rcutorture.gp_poll_wi_exp= [KNL] 6051 Nominal wait interval for expedited conditional 6052 grace periods (specified by rcutorture's 6053 gp_poll_exp and gp_poll_exp_full module 6054 parameters), in microseconds. The actual wait 6055 interval will be randomly selected to nanosecond 6056 granularity up to this wait interval. Defaults to 6057 128 microseconds. 6058 6059 rcutorture.gp_sync= [KNL] 6060 Use normal (non-expedited) synchronous 6061 update-side primitives, if available. If all 6062 of rcutorture.gp_cond=, rcutorture.gp_exp=, 6063 rcutorture.gp_normal=, and rcutorture.gp_sync= 6064 are zero, rcutorture acts as if is interpreted 6065 they are all non-zero. 6066 6067 rcutorture.gpwrap_lag= [KNL] 6068 Enable grace-period wrap lag testing. Setting 6069 to false prevents the gpwrap lag test from 6070 running. Default is true. 6071 6072 rcutorture.gpwrap_lag_gps= [KNL] 6073 Set the value for grace-period wrap lag during 6074 active lag testing periods. This controls how many 6075 grace periods differences we tolerate between 6076 rdp and rnp's gp_seq before setting overflow flag. 6077 The default is always set to 8. 6078 6079 rcutorture.gpwrap_lag_cycle_mins= [KNL] 6080 Set the total cycle duration for gpwrap lag 6081 testing in minutes. This is the total time for 6082 one complete cycle of active and inactive 6083 testing periods. Default is 30 minutes. 6084 6085 rcutorture.gpwrap_lag_active_mins= [KNL] 6086 Set the duration for which gpwrap lag is active 6087 within each cycle, in minutes. During this time, 6088 the grace-period wrap lag will be set to the 6089 value specified by gpwrap_lag_gps. Default is 6090 5 minutes. 6091 6092 rcutorture.irqreader= [KNL] 6093 Run RCU readers from irq handlers, or, more 6094 accurately, from a timer handler. Not all RCU 6095 flavors take kindly to this sort of thing. 6096 6097 rcutorture.leakpointer= [KNL] 6098 Leak an RCU-protected pointer out of the reader. 6099 This can of course result in splats, and is 6100 intended to test the ability of things like 6101 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 6102 such leaks. 6103 6104 rcutorture.n_barrier_cbs= [KNL] 6105 Set callbacks/threads for rcu_barrier() testing. 6106 6107 rcutorture.nfakewriters= [KNL] 6108 Set number of concurrent RCU writers. These just 6109 stress RCU, they don't participate in the actual 6110 test, hence the "fake". 6111 6112 rcutorture.nocbs_nthreads= [KNL] 6113 Set number of RCU callback-offload togglers. 6114 Zero (the default) disables toggling. 6115 6116 rcutorture.nocbs_toggle= [KNL] 6117 Set the delay in milliseconds between successive 6118 callback-offload toggling attempts. 6119 6120 rcutorture.nreaders= [KNL] 6121 Set number of RCU readers. The value -1 selects 6122 N-1, where N is the number of CPUs. A value 6123 "n" less than -1 selects N-n-2, where N is again 6124 the number of CPUs. For example, -2 selects N 6125 (the number of CPUs), -3 selects N+1, and so on. 6126 6127 rcutorture.object_debug= [KNL] 6128 Enable debug-object double-call_rcu() testing. 6129 6130 rcutorture.onoff_holdoff= [KNL] 6131 Set time (s) after boot for CPU-hotplug testing. 6132 6133 rcutorture.onoff_interval= [KNL] 6134 Set time (jiffies) between CPU-hotplug operations, 6135 or zero to disable CPU-hotplug testing. 6136 6137 rcutorture.preempt_duration= [KNL] 6138 Set duration (in milliseconds) of preemptions 6139 by a high-priority FIFO real-time task. Set to 6140 zero (the default) to disable. The CPUs to 6141 preempt are selected randomly from the set that 6142 are online at a given point in time. Races with 6143 CPUs going offline are ignored, with that attempt 6144 at preemption skipped. 6145 6146 rcutorture.preempt_interval= [KNL] 6147 Set interval (in milliseconds, defaulting to one 6148 second) between preemptions by a high-priority 6149 FIFO real-time task. This delay is mediated 6150 by an hrtimer and is further fuzzed to avoid 6151 inadvertent synchronizations. 6152 6153 rcutorture.read_exit_burst= [KNL] 6154 The number of times in a given read-then-exit 6155 episode that a set of read-then-exit kthreads 6156 is spawned. 6157 6158 rcutorture.read_exit_delay= [KNL] 6159 The delay, in seconds, between successive 6160 read-then-exit testing episodes. 6161 6162 rcutorture.reader_flavor= [KNL] 6163 A bit mask indicating which readers to use. 6164 If there is more than one bit set, the readers 6165 are entered from low-order bit up, and are 6166 exited in the opposite order. For SRCU, the 6167 0x1 bit is normal readers, 0x2 NMI-safe readers, 6168 and 0x4 light-weight readers. 6169 6170 rcutorture.shuffle_interval= [KNL] 6171 Set task-shuffle interval (s). Shuffling tasks 6172 allows some CPUs to go into dyntick-idle mode 6173 during the rcutorture test. 6174 6175 rcutorture.shutdown_secs= [KNL] 6176 Set time (s) after boot system shutdown. This 6177 is useful for hands-off automated testing. 6178 6179 rcutorture.stall_cpu= [KNL] 6180 Duration of CPU stall (s) to test RCU CPU stall 6181 warnings, zero to disable. 6182 6183 rcutorture.stall_cpu_block= [KNL] 6184 Sleep while stalling if set. This will result 6185 in warnings from preemptible RCU in addition to 6186 any other stall-related activity. Note that 6187 in kernels built with CONFIG_PREEMPTION=n and 6188 CONFIG_PREEMPT_COUNT=y, this parameter will 6189 cause the CPU to pass through a quiescent state. 6190 Given CONFIG_PREEMPTION=n, this will suppress 6191 RCU CPU stall warnings, but will instead result 6192 in scheduling-while-atomic splats. 6193 6194 Use of this module parameter results in splats. 6195 6196 6197 rcutorture.stall_cpu_holdoff= [KNL] 6198 Time to wait (s) after boot before inducing stall. 6199 6200 rcutorture.stall_cpu_irqsoff= [KNL] 6201 Disable interrupts while stalling if set, but only 6202 on the first stall in the set. 6203 6204 rcutorture.stall_cpu_repeat= [KNL] 6205 Number of times to repeat the stall sequence, 6206 so that rcutorture.stall_cpu_repeat=3 will result 6207 in four stall sequences. 6208 6209 rcutorture.stall_gp_kthread= [KNL] 6210 Duration (s) of forced sleep within RCU 6211 grace-period kthread to test RCU CPU stall 6212 warnings, zero to disable. If both stall_cpu 6213 and stall_gp_kthread are specified, the 6214 kthread is starved first, then the CPU. 6215 6216 rcutorture.stat_interval= [KNL] 6217 Time (s) between statistics printk()s. 6218 6219 rcutorture.stutter= [KNL] 6220 Time (s) to stutter testing, for example, specifying 6221 five seconds causes the test to run for five seconds, 6222 wait for five seconds, and so on. This tests RCU's 6223 ability to transition abruptly to and from idle. 6224 6225 rcutorture.test_boost= [KNL] 6226 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 6227 "Maybe" means test if the RCU implementation 6228 under test support RCU priority boosting. 6229 6230 rcutorture.test_boost_duration= [KNL] 6231 Duration (s) of each individual boost test. 6232 6233 rcutorture.test_boost_holdoff= [KNL] 6234 Holdoff time (s) from start of test to the start 6235 of RCU priority-boost testing. Defaults to zero, 6236 that is, no holdoff. 6237 6238 rcutorture.test_boost_interval= [KNL] 6239 Interval (s) between each boost test. 6240 6241 rcutorture.test_no_idle_hz= [KNL] 6242 Test RCU's dyntick-idle handling. See also the 6243 rcutorture.shuffle_interval parameter. 6244 6245 rcutorture.torture_type= [KNL] 6246 Specify the RCU implementation to test. 6247 6248 rcutorture.verbose= [KNL] 6249 Enable additional printk() statements. 6250 6251 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 6252 Dump ftrace buffer after reporting RCU CPU 6253 stall warning. 6254 6255 rcupdate.rcu_cpu_stall_notifiers= [KNL] 6256 Provide RCU CPU stall notifiers, but see the 6257 warnings in the RCU_CPU_STALL_NOTIFIER Kconfig 6258 option's help text. TL;DR: You almost certainly 6259 do not want rcupdate.rcu_cpu_stall_notifiers. 6260 6261 rcupdate.rcu_cpu_stall_suppress= [KNL] 6262 Suppress RCU CPU stall warning messages. 6263 6264 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 6265 Suppress RCU CPU stall warning messages and 6266 rcutorture writer stall warnings that occur 6267 during early boot, that is, during the time 6268 before the init task is spawned. 6269 6270 rcupdate.rcu_cpu_stall_timeout= [KNL] 6271 Set timeout for RCU CPU stall warning messages. 6272 The value is in seconds and the maximum allowed 6273 value is 300 seconds. 6274 6275 rcupdate.rcu_exp_cpu_stall_timeout= [KNL] 6276 Set timeout for expedited RCU CPU stall warning 6277 messages. The value is in milliseconds 6278 and the maximum allowed value is 21000 6279 milliseconds. Please note that this value is 6280 adjusted to an arch timer tick resolution. 6281 Setting this to zero causes the value from 6282 rcupdate.rcu_cpu_stall_timeout to be used (after 6283 conversion from seconds to milliseconds). 6284 6285 rcupdate.rcu_cpu_stall_cputime= [KNL] 6286 Provide statistics on the cputime and count of 6287 interrupts and tasks during the sampling period. For 6288 multiple continuous RCU stalls, all sampling periods 6289 begin at half of the first RCU stall timeout. 6290 6291 rcupdate.rcu_exp_stall_task_details= [KNL] 6292 Print stack dumps of any tasks blocking the 6293 current expedited RCU grace period during an 6294 expedited RCU CPU stall warning. 6295 6296 rcupdate.rcu_expedited= [KNL] 6297 Use expedited grace-period primitives, for 6298 example, synchronize_rcu_expedited() instead 6299 of synchronize_rcu(). This reduces latency, 6300 but can increase CPU utilization, degrade 6301 real-time latency, and degrade energy efficiency. 6302 No effect on CONFIG_TINY_RCU kernels. 6303 6304 rcupdate.rcu_normal= [KNL] 6305 Use only normal grace-period primitives, 6306 for example, synchronize_rcu() instead of 6307 synchronize_rcu_expedited(). This improves 6308 real-time latency, CPU utilization, and 6309 energy efficiency, but can expose users to 6310 increased grace-period latency. This parameter 6311 overrides rcupdate.rcu_expedited. No effect on 6312 CONFIG_TINY_RCU kernels. 6313 6314 rcupdate.rcu_normal_after_boot= [KNL] 6315 Once boot has completed (that is, after 6316 rcu_end_inkernel_boot() has been invoked), use 6317 only normal grace-period primitives. No effect 6318 on CONFIG_TINY_RCU kernels. 6319 6320 But note that CONFIG_PREEMPT_RT=y kernels enables 6321 this kernel boot parameter, forcibly setting 6322 it to the value one, that is, converting any 6323 post-boot attempt at an expedited RCU grace 6324 period to instead use normal non-expedited 6325 grace-period processing. 6326 6327 rcupdate.rcu_task_collapse_lim= [KNL] 6328 Set the maximum number of callbacks present 6329 at the beginning of a grace period that allows 6330 the RCU Tasks flavors to collapse back to using 6331 a single callback queue. This switching only 6332 occurs when rcupdate.rcu_task_enqueue_lim is 6333 set to the default value of -1. 6334 6335 rcupdate.rcu_task_contend_lim= [KNL] 6336 Set the minimum number of callback-queuing-time 6337 lock-contention events per jiffy required to 6338 cause the RCU Tasks flavors to switch to per-CPU 6339 callback queuing. This switching only occurs 6340 when rcupdate.rcu_task_enqueue_lim is set to 6341 the default value of -1. 6342 6343 rcupdate.rcu_task_enqueue_lim= [KNL] 6344 Set the number of callback queues to use for the 6345 RCU Tasks family of RCU flavors. The default 6346 of -1 allows this to be automatically (and 6347 dynamically) adjusted. This parameter is intended 6348 for use in testing. 6349 6350 rcupdate.rcu_task_lazy_lim= [KNL] 6351 Number of callbacks on a given CPU that will 6352 cancel laziness on that CPU. Use -1 to disable 6353 cancellation of laziness, but be advised that 6354 doing so increases the danger of OOM due to 6355 callback flooding. 6356 6357 rcupdate.rcu_task_stall_info= [KNL] 6358 Set initial timeout in jiffies for RCU task stall 6359 informational messages, which give some indication 6360 of the problem for those not patient enough to 6361 wait for ten minutes. Informational messages are 6362 only printed prior to the stall-warning message 6363 for a given grace period. Disable with a value 6364 less than or equal to zero. Defaults to ten 6365 seconds. A change in value does not take effect 6366 until the beginning of the next grace period. 6367 6368 rcupdate.rcu_task_stall_info_mult= [KNL] 6369 Multiplier for time interval between successive 6370 RCU task stall informational messages for a given 6371 RCU tasks grace period. This value is clamped 6372 to one through ten, inclusive. It defaults to 6373 the value three, so that the first informational 6374 message is printed 10 seconds into the grace 6375 period, the second at 40 seconds, the third at 6376 160 seconds, and then the stall warning at 600 6377 seconds would prevent a fourth at 640 seconds. 6378 6379 rcupdate.rcu_task_stall_timeout= [KNL] 6380 Set timeout in jiffies for RCU task stall 6381 warning messages. Disable with a value less 6382 than or equal to zero. Defaults to ten minutes. 6383 A change in value does not take effect until 6384 the beginning of the next grace period. 6385 6386 rcupdate.rcu_tasks_lazy_ms= [KNL] 6387 Set timeout in milliseconds RCU Tasks asynchronous 6388 callback batching for call_rcu_tasks(). 6389 A negative value will take the default. A value 6390 of zero will disable batching. Batching is 6391 always disabled for synchronize_rcu_tasks(). 6392 6393 rcupdate.rcu_self_test= [KNL] 6394 Run the RCU early boot self tests 6395 6396 rdinit= [KNL] 6397 Format: <full_path> 6398 Run specified binary instead of /init from the ramdisk, 6399 used for early userspace startup. See initrd. 6400 6401 rdrand= [X86,EARLY] 6402 force - Override the decision by the kernel to hide the 6403 advertisement of RDRAND support (this affects 6404 certain AMD processors because of buggy BIOS 6405 support, specifically around the suspend/resume 6406 path). 6407 6408 rdt= [HW,X86,RDT] 6409 Turn on/off individual RDT features. List is: 6410 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 6411 mba, smba, bmec, abmc, sdciae, energy[:guid], 6412 perf[:guid]. 6413 E.g. to turn on cmt and turn off mba use: 6414 rdt=cmt,!mba 6415 To turn off all energy telemetry monitoring and ensure that 6416 perf telemetry monitoring associated with guid 0x12345 6417 is enabled use: 6418 rdt=!energy,perf:0x12345 6419 6420 reboot= [KNL] 6421 Format (x86 or x86_64): 6422 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \ 6423 [[,]s[mp]####] \ 6424 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 6425 [[,]f[orce]] 6426 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 6427 (prefix with 'panic_' to set mode for panic 6428 reboot only), 6429 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 6430 reboot_force is either force or not specified, 6431 reboot_cpu is s[mp]#### with #### being the processor 6432 to be used for rebooting. 6433 6434 acpi 6435 Use the ACPI RESET_REG in the FADT. If ACPI is not 6436 configured or the ACPI reset does not work, the reboot 6437 path attempts the reset using the keyboard controller. 6438 6439 bios 6440 Use the CPU reboot vector for warm reset 6441 6442 cold 6443 Set the cold reboot flag 6444 6445 default 6446 There are some built-in platform specific "quirks" 6447 - you may see: "reboot: <name> series board detected. 6448 Selecting <type> for reboots." In the case where you 6449 think the quirk is in error (e.g. you have newer BIOS, 6450 or newer board) using this option will ignore the 6451 built-in quirk table, and use the generic default 6452 reboot actions. 6453 6454 efi 6455 Use efi reset_system runtime service. If EFI is not 6456 configured or the EFI reset does not work, the reboot 6457 path attempts the reset using the keyboard controller. 6458 6459 force 6460 Don't stop other CPUs on reboot. This can make reboot 6461 more reliable in some cases. 6462 6463 kbd 6464 Use the keyboard controller. cold reset (default) 6465 6466 pci 6467 Use a write to the PCI config space register 0xcf9 to 6468 trigger reboot. 6469 6470 triple 6471 Force a triple fault (init) 6472 6473 warm 6474 Don't set the cold reboot flag 6475 6476 Using warm reset will be much faster especially on big 6477 memory systems because the BIOS will not go through 6478 the memory check. Disadvantage is that not all 6479 hardware will be completely reinitialized on reboot so 6480 there may be boot problems on some systems. 6481 6482 6483 refscale.holdoff= [KNL] 6484 Set test-start holdoff period. The purpose of 6485 this parameter is to delay the start of the 6486 test until boot completes in order to avoid 6487 interference. 6488 6489 refscale.lookup_instances= [KNL] 6490 Number of data elements to use for the forms of 6491 SLAB_TYPESAFE_BY_RCU testing. A negative number 6492 is negated and multiplied by nr_cpu_ids, while 6493 zero specifies nr_cpu_ids. 6494 6495 refscale.loops= [KNL] 6496 Set the number of loops over the synchronization 6497 primitive under test. Increasing this number 6498 reduces noise due to loop start/end overhead, 6499 but the default has already reduced the per-pass 6500 noise to a handful of picoseconds on ca. 2020 6501 x86 laptops. 6502 6503 refscale.nreaders= [KNL] 6504 Set number of readers. The default value of -1 6505 selects N, where N is roughly 75% of the number 6506 of CPUs. A value of zero is an interesting choice. 6507 6508 refscale.nruns= [KNL] 6509 Set number of runs, each of which is dumped onto 6510 the console log. 6511 6512 refscale.readdelay= [KNL] 6513 Set the read-side critical-section duration, 6514 measured in microseconds. 6515 6516 refscale.scale_type= [KNL] 6517 Specify the read-protection implementation to test. 6518 6519 refscale.shutdown= [KNL] 6520 Shut down the system at the end of the performance 6521 test. This defaults to 1 (shut it down) when 6522 refscale is built into the kernel and to 0 (leave 6523 it running) when refscale is built as a module. 6524 6525 refscale.verbose= [KNL] 6526 Enable additional printk() statements. 6527 6528 refscale.verbose_batched= [KNL] 6529 Batch the additional printk() statements. If zero 6530 (the default) or negative, print everything. Otherwise, 6531 print every Nth verbose statement, where N is the value 6532 specified. 6533 6534 regulator_ignore_unused 6535 [REGULATOR] 6536 Prevents regulator framework from disabling regulators 6537 that are unused, due no driver claiming them. This may 6538 be useful for debug and development, but should not be 6539 needed on a platform with proper driver support. 6540 6541 relax_domain_level= 6542 [KNL, SMP] Set scheduler's default relax_domain_level. 6543 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 6544 6545 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 6546 Format: <base1>,<size1>[,<base2>,<size2>,...] 6547 Reserve I/O ports or memory so the kernel won't use 6548 them. If <base> is less than 0x10000, the region 6549 is assumed to be I/O ports; otherwise it is memory. 6550 6551 reserve_mem= [RAM] 6552 Format: nn[KMG]:<align>:<label> 6553 Reserve physical memory and label it with a name that 6554 other subsystems can use to access it. This is typically 6555 used for systems that do not wipe the RAM, and this command 6556 line will try to reserve the same physical memory on 6557 soft reboots. Note, it is not guaranteed to be the same 6558 location. For example, if anything about the system changes 6559 or if booting a different kernel. It can also fail if KASLR 6560 places the kernel at the location of where the RAM reservation 6561 was from a previous boot, the new reservation will be at a 6562 different location. 6563 Any subsystem using this feature must add a way to verify 6564 that the contents of the physical memory is from a previous 6565 boot, as there may be cases where the memory will not be 6566 located at the same location. 6567 6568 The format is size:align:label for example, to request 6569 12 megabytes of 4096 alignment for ramoops: 6570 6571 reserve_mem=12M:4096:oops ramoops.mem_name=oops 6572 6573 reservetop= [X86-32,EARLY] 6574 Format: nn[KMG] 6575 Reserves a hole at the top of the kernel virtual 6576 address space. 6577 6578 reset_devices [KNL] Force drivers to reset the underlying device 6579 during initialization. 6580 6581 resume= [SWSUSP] 6582 Specify the partition device for software suspend 6583 Format: 6584 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 6585 6586 resume_offset= [SWSUSP] 6587 Specify the offset from the beginning of the partition 6588 given by "resume=" at which the swap header is located, 6589 in <PAGE_SIZE> units (needed only for swap files). 6590 See Documentation/power/swsusp-and-swap-files.rst 6591 6592 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 6593 read the resume files 6594 6595 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 6596 Useful for devices that are detected asynchronously 6597 (e.g. USB and MMC devices). 6598 6599 retain_initrd [RAM] Keep initrd memory after extraction. After boot, it will 6600 be accessible via /sys/firmware/initrd. 6601 6602 retbleed= [X86] Control mitigation of RETBleed (Arbitrary 6603 Speculative Code Execution with Return Instructions) 6604 vulnerability. 6605 6606 AMD-based UNRET and IBPB mitigations alone do not stop 6607 sibling threads from influencing the predictions of other 6608 sibling threads. For that reason, STIBP is used on pro- 6609 cessors that support it, and mitigate SMT on processors 6610 that don't. 6611 6612 off - no mitigation 6613 auto - automatically select a mitigation 6614 auto,nosmt - automatically select a mitigation, 6615 disabling SMT if necessary for 6616 the full mitigation (only on Zen1 6617 and older without STIBP). 6618 ibpb - On AMD, mitigate short speculation 6619 windows on basic block boundaries too. 6620 Safe, highest perf impact. It also 6621 enables STIBP if present. Not suitable 6622 on Intel. 6623 ibpb,nosmt - Like "ibpb" above but will disable SMT 6624 when STIBP is not available. This is 6625 the alternative for systems which do not 6626 have STIBP. 6627 unret - Force enable untrained return thunks, 6628 only effective on AMD f15h-f17h based 6629 systems. 6630 unret,nosmt - Like unret, but will disable SMT when STIBP 6631 is not available. This is the alternative for 6632 systems which do not have STIBP. 6633 6634 Selecting 'auto' will choose a mitigation method at run 6635 time according to the CPU. 6636 6637 Not specifying this option is equivalent to retbleed=auto. 6638 6639 rfkill.default_state= 6640 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 6641 etc. communication is blocked by default. 6642 1 Unblocked. 6643 6644 rfkill.master_switch_mode= 6645 0 The "airplane mode" button does nothing. 6646 1 The "airplane mode" button toggles between everything 6647 blocked and the previous configuration. 6648 2 The "airplane mode" button toggles between everything 6649 blocked and everything unblocked. 6650 6651 ring3mwait=disable 6652 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 6653 CPUs. 6654 6655 riscv_isa_fallback [RISCV,EARLY] 6656 When CONFIG_RISCV_ISA_FALLBACK is not enabled, permit 6657 falling back to detecting extension support by parsing 6658 "riscv,isa" property on devicetree systems when the 6659 replacement properties are not found. See the Kconfig 6660 entry for RISCV_ISA_FALLBACK. 6661 6662 riscv_nousercfi= 6663 all Disable user CFI ABI to userspace even if cpu extension 6664 are available. 6665 bcfi Disable user backward CFI ABI to userspace even if 6666 the shadow stack extension is available. 6667 fcfi Disable user forward CFI ABI to userspace even if the 6668 landing pad extension is available. 6669 6670 ro [KNL] Mount root device read-only on boot 6671 6672 rodata= [KNL,EARLY] 6673 on Mark read-only kernel memory as read-only (default). 6674 off Leave read-only kernel memory writable for debugging. 6675 noalias Mark read-only kernel memory as read-only but retain 6676 writable aliases in the direct map for regions outside 6677 of the kernel image. [arm64] 6678 6679 rockchip.usb_uart 6680 [EARLY] 6681 Enable the uart passthrough on the designated usb port 6682 on Rockchip SoCs. When active, the signals of the 6683 debug-uart get routed to the D+ and D- pins of the usb 6684 port and the regular usb controller gets disabled. 6685 6686 root= [KNL] Root filesystem 6687 Usually this is a block device specifier of some kind, 6688 see the early_lookup_bdev comment in 6689 block/early-lookup.c for details. 6690 Alternatively this can be "ram" for the legacy initial 6691 ramdisk, "nfs" and "cifs" for root on a network file 6692 system, or "mtd" and "ubi" for mounting from raw flash. 6693 6694 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 6695 mount the root filesystem 6696 6697 rootflags= [KNL] Set root filesystem mount option string 6698 6699 rseq_slice_ext= [KNL] RSEQ based time slice extension 6700 Format: boolean 6701 Control enablement of RSEQ based time slice extension. 6702 Default is 'on'. 6703 6704 initramfs_options= [KNL] 6705 Specify mount options for the initramfs mount. 6706 6707 rootfstype= [KNL] Set root filesystem type 6708 6709 rootwait [KNL] Wait (indefinitely) for root device to show up. 6710 Useful for devices that are detected asynchronously 6711 (e.g. USB and MMC devices). 6712 6713 rootwait= [KNL] Maximum time (in seconds) to wait for root device 6714 to show up before attempting to mount the root 6715 filesystem. 6716 6717 rproc_mem=nn[KMG][@address] 6718 [KNL,ARM,CMA] Remoteproc physical memory block. 6719 Memory area to be used by remote processor image, 6720 managed by CMA. 6721 6722 rseq_debug= [KNL] Enable or disable restartable sequence 6723 debug mode. Defaults to CONFIG_RSEQ_DEBUG_DEFAULT_ENABLE. 6724 Format: <bool> 6725 6726 rt_group_sched= [KNL] Enable or disable SCHED_RR/FIFO group scheduling 6727 when CONFIG_RT_GROUP_SCHED=y. Defaults to 6728 !CONFIG_RT_GROUP_SCHED_DEFAULT_DISABLED. 6729 Format: <bool> 6730 6731 rw [KNL] Mount root device read-write on boot 6732 6733 S [KNL] Run init in single mode 6734 6735 s390_iommu= [HW,S390] 6736 Set s390 IOTLB flushing mode 6737 strict 6738 With strict flushing every unmap operation will result 6739 in an IOTLB flush. Default is lazy flushing before 6740 reuse, which is faster. Deprecated, equivalent to 6741 iommu.strict=1. 6742 6743 s390_iommu_aperture= [KNL,S390] 6744 Specifies the size of the per device DMA address space 6745 accessible through the DMA and IOMMU APIs as a decimal 6746 factor of the size of main memory. 6747 The default is 1 meaning that one can concurrently use 6748 as many DMA addresses as physical memory is installed, 6749 if supported by hardware, and thus map all of memory 6750 once. With a value of 2 one can map all of memory twice 6751 and so on. As a special case a factor of 0 imposes no 6752 restrictions other than those given by hardware at the 6753 cost of significant additional memory use for tables. 6754 6755 6756 sched_proxy_exec= [KNL] 6757 Enables or disables "proxy execution" style 6758 solution to mutex-based priority inversion. 6759 Format: <bool> 6760 6761 sched_verbose [KNL,EARLY] Enables verbose scheduler debug messages. 6762 6763 schedstats= [KNL,X86] Enable or disable scheduled statistics. 6764 Allowed values are enable and disable. This feature 6765 incurs a small amount of overhead in the scheduler 6766 but is useful for debugging and performance tuning. 6767 6768 sched_thermal_decay_shift= 6769 [Deprecated] 6770 [KNL, SMP] Set a decay shift for scheduler thermal 6771 pressure signal. Thermal pressure signal follows the 6772 default decay period of other scheduler pelt 6773 signals(usually 32 ms but configurable). Setting 6774 sched_thermal_decay_shift will left shift the decay 6775 period for the thermal pressure signal by the shift 6776 value. 6777 i.e. with the default pelt decay period of 32 ms 6778 sched_thermal_decay_shift thermal pressure decay pr 6779 1 64 ms 6780 2 128 ms 6781 and so on. 6782 Format: integer between 0 and 10 6783 Default is 0. 6784 6785 scftorture.holdoff= [KNL] 6786 Number of seconds to hold off before starting 6787 test. Defaults to zero for module insertion and 6788 to 10 seconds for built-in smp_call_function() 6789 tests. 6790 6791 scftorture.longwait= [KNL] 6792 Request ridiculously long waits randomly selected 6793 up to the chosen limit in seconds. Zero (the 6794 default) disables this feature. Please note 6795 that requesting even small non-zero numbers of 6796 seconds can result in RCU CPU stall warnings, 6797 softlockup complaints, and so on. 6798 6799 scftorture.nthreads= [KNL] 6800 Number of kthreads to spawn to invoke the 6801 smp_call_function() family of functions. 6802 The default of -1 specifies a number of kthreads 6803 equal to the number of CPUs. 6804 6805 scftorture.onoff_holdoff= [KNL] 6806 Number seconds to wait after the start of the 6807 test before initiating CPU-hotplug operations. 6808 6809 scftorture.onoff_interval= [KNL] 6810 Number seconds to wait between successive 6811 CPU-hotplug operations. Specifying zero (which 6812 is the default) disables CPU-hotplug operations. 6813 6814 scftorture.shutdown_secs= [KNL] 6815 The number of seconds following the start of the 6816 test after which to shut down the system. The 6817 default of zero avoids shutting down the system. 6818 Non-zero values are useful for automated tests. 6819 6820 scftorture.stat_interval= [KNL] 6821 The number of seconds between outputting the 6822 current test statistics to the console. A value 6823 of zero disables statistics output. 6824 6825 scftorture.stutter_cpus= [KNL] 6826 The number of jiffies to wait between each change 6827 to the set of CPUs under test. 6828 6829 scftorture.use_cpus_read_lock= [KNL] 6830 Use use_cpus_read_lock() instead of the default 6831 preempt_disable() to disable CPU hotplug 6832 while invoking one of the smp_call_function*() 6833 functions. 6834 6835 scftorture.verbose= [KNL] 6836 Enable additional printk() statements. 6837 6838 scftorture.weight_single= [KNL] 6839 The probability weighting to use for the 6840 smp_call_function_single() function with a zero 6841 "wait" parameter. A value of -1 selects the 6842 default if all other weights are -1. However, 6843 if at least one weight has some other value, a 6844 value of -1 will instead select a weight of zero. 6845 6846 scftorture.weight_single_wait= [KNL] 6847 The probability weighting to use for the 6848 smp_call_function_single() function with a 6849 non-zero "wait" parameter. See weight_single. 6850 6851 scftorture.weight_many= [KNL] 6852 The probability weighting to use for the 6853 smp_call_function_many() function with a zero 6854 "wait" parameter. See weight_single. 6855 Note well that setting a high probability for 6856 this weighting can place serious IPI load 6857 on the system. 6858 6859 scftorture.weight_many_wait= [KNL] 6860 The probability weighting to use for the 6861 smp_call_function_many() function with a 6862 non-zero "wait" parameter. See weight_single 6863 and weight_many. 6864 6865 scftorture.weight_all= [KNL] 6866 The probability weighting to use for the 6867 smp_call_function_all() function with a zero 6868 "wait" parameter. See weight_single and 6869 weight_many. 6870 6871 scftorture.weight_all_wait= [KNL] 6872 The probability weighting to use for the 6873 smp_call_function_all() function with a 6874 non-zero "wait" parameter. See weight_single 6875 and weight_many. 6876 6877 sdw_mclk_divider=[SDW] 6878 Specify the MCLK divider for Intel SoundWire buses in 6879 case the BIOS does not provide the clock rate properly. 6880 6881 skew_tick= [KNL,EARLY] Offset the periodic timer tick per cpu to mitigate 6882 xtime_lock contention on larger systems, and/or RCU lock 6883 contention on all systems with CONFIG_MAXSMP set. 6884 Format: { "0" | "1" } 6885 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1") 6886 1 -- enable. 6887 Note: increases power consumption, thus should only be 6888 enabled if running jitter sensitive (HPC/RT) workloads. 6889 6890 security= [SECURITY] Choose a legacy "major" security module to 6891 enable at boot. This has been deprecated by the 6892 "lsm=" parameter. 6893 6894 selinux= [SELINUX] Disable or enable SELinux at boot time. 6895 Format: { "0" | "1" } 6896 See security/selinux/Kconfig help text. 6897 0 -- disable. 6898 1 -- enable. 6899 Default value is 1. 6900 6901 serialnumber [BUGS=X86-32] 6902 6903 sev=option[,option...] [X86-64] 6904 6905 debug 6906 Enable debug messages. 6907 6908 nosnp 6909 Do not enable SEV-SNP (applies to host/hypervisor 6910 only). Setting 'nosnp' avoids the RMP check overhead 6911 in memory accesses when users do not want to run 6912 SEV-SNP guests. 6913 6914 shapers= [NET] 6915 Maximal number of shapers. 6916 6917 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 6918 Limit apic dumping. The parameter defines the maximal 6919 number of local apics being dumped. Also it is possible 6920 to set it to "all" by meaning -- no limit here. 6921 Format: { 1 (default) | 2 | ... | all }. 6922 The parameter valid if only apic=debug or 6923 apic=verbose is specified. 6924 Example: apic=debug show_lapic=all 6925 6926 slab_debug[=options[,slabs][;[options[,slabs]]...]] [MM] 6927 Enabling slab_debug allows one to determine the 6928 culprit if slab objects become corrupted. Enabling 6929 slab_debug can create guard zones around objects and 6930 may poison objects when not in use. Also tracks the 6931 last alloc / free. For more information see 6932 Documentation/admin-guide/mm/slab.rst. 6933 (slub_debug legacy name also accepted for now) 6934 6935 Using this option implies the "no_hash_pointers" 6936 option which can be undone by adding the 6937 "hash_pointers=always" option. 6938 6939 slab_max_order= [MM] 6940 Determines the maximum allowed order for slabs. 6941 A high setting may cause OOMs due to memory 6942 fragmentation. For more information see 6943 Documentation/admin-guide/mm/slab.rst. 6944 (slub_max_order legacy name also accepted for now) 6945 6946 slab_merge [MM] 6947 Enable merging of slabs with similar size when the 6948 kernel is built without CONFIG_SLAB_MERGE_DEFAULT. 6949 (slub_merge legacy name also accepted for now) 6950 6951 slab_min_objects= [MM] 6952 The minimum number of objects per slab. SLUB will 6953 increase the slab order up to slab_max_order to 6954 generate a sufficiently large slab able to contain 6955 the number of objects indicated. The higher the number 6956 of objects the smaller the overhead of tracking slabs 6957 and the less frequently locks need to be acquired. 6958 For more information see 6959 Documentation/admin-guide/mm/slab.rst. 6960 (slub_min_objects legacy name also accepted for now) 6961 6962 slab_min_order= [MM] 6963 Determines the minimum page order for slabs. Must be 6964 lower or equal to slab_max_order. For more information see 6965 Documentation/admin-guide/mm/slab.rst. 6966 (slub_min_order legacy name also accepted for now) 6967 6968 slab_nomerge [MM] 6969 Disable merging of slabs with similar size. May be 6970 necessary if there is some reason to distinguish 6971 allocs to different slabs, especially in hardened 6972 environments where the risk of heap overflows and 6973 layout control by attackers can usually be 6974 frustrated by disabling merging. This will reduce 6975 most of the exposure of a heap attack to a single 6976 cache (risks via metadata attacks are mostly 6977 unchanged). Debug options disable merging on their 6978 own. 6979 For more information see 6980 Documentation/admin-guide/mm/slab.rst. 6981 (slub_nomerge legacy name also accepted for now) 6982 6983 slab_strict_numa [MM] 6984 Support memory policies on a per object level 6985 in the slab allocator. The default is for memory 6986 policies to be applied at the folio level when 6987 a new folio is needed or a partial folio is 6988 retrieved from the lists. Increases overhead 6989 in the slab fastpaths but gains more accurate 6990 NUMA kernel object placement which helps with slow 6991 interconnects in NUMA systems. 6992 6993 slram= [HW,MTD] 6994 6995 smart2= [HW] 6996 Format: <io1>[,<io2>[,...,<io8>]] 6997 6998 smp.csd_lock_timeout= [KNL] 6999 Specify the period of time in milliseconds 7000 that smp_call_function() and friends will wait 7001 for a CPU to release the CSD lock. This is 7002 useful when diagnosing bugs involving CPUs 7003 disabling interrupts for extended periods 7004 of time. Defaults to 5,000 milliseconds, and 7005 setting a value of zero disables this feature. 7006 This feature may be more efficiently disabled 7007 using the csdlock_debug- kernel parameter. 7008 7009 smp.panic_on_ipistall= [KNL] 7010 If a csd_lock_timeout extends for more than 7011 the specified number of milliseconds, panic the 7012 system. By default, let CSD-lock acquisition 7013 take as long as they take. Specifying 300,000 7014 for this value provides a 5-minute timeout. 7015 7016 smt= [KNL,MIPS,S390,EARLY] Set the maximum number of threads 7017 (logical CPUs) to use per physical CPU on systems 7018 capable of symmetric multithreading (SMT). Will 7019 be capped to the actual hardware limit. 7020 Format: <integer> 7021 Default: -1 (no limit) 7022 7023 softlockup_panic= 7024 [KNL] Should the soft-lockup detector generate panics. 7025 Format: <int> 7026 7027 A value of non-zero instructs the soft-lockup detector 7028 to panic the machine when a soft-lockup duration exceeds 7029 N thresholds. It is also controlled by the kernel.softlockup_panic 7030 sysctl and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 7031 respective build-time switch to that functionality. 7032 7033 softlockup_all_cpu_backtrace= 7034 [KNL] Should the soft-lockup detector generate 7035 backtraces on all cpus. 7036 Format: 0 | 1 7037 7038 sonypi.*= [HW] Sony Programmable I/O Control Device driver 7039 See Documentation/admin-guide/laptops/sonypi.rst 7040 7041 spectre_bhi= [X86] Control mitigation of Branch History Injection 7042 (BHI) vulnerability. This setting affects the 7043 deployment of the HW BHI control and the SW BHB 7044 clearing sequence. 7045 7046 on - (default) Enable the HW or SW mitigation as 7047 needed. This protects the kernel from 7048 both syscalls and VMs. 7049 vmexit - On systems which don't have the HW mitigation 7050 available, enable the SW mitigation on vmexit 7051 ONLY. On such systems, the host kernel is 7052 protected from VM-originated BHI attacks, but 7053 may still be vulnerable to syscall attacks. 7054 off - Disable the mitigation. 7055 7056 spectre_v2= [X86,EARLY] Control mitigation of Spectre variant 2 7057 (indirect branch speculation) vulnerability. 7058 The default operation protects the kernel from 7059 user space attacks. 7060 7061 on - unconditionally enable, implies 7062 spectre_v2_user=on 7063 off - unconditionally disable, implies 7064 spectre_v2_user=off 7065 auto - kernel detects whether your CPU model is 7066 vulnerable 7067 7068 Selecting 'on' will, and 'auto' may, choose a 7069 mitigation method at run time according to the 7070 CPU, the available microcode, the setting of the 7071 CONFIG_MITIGATION_RETPOLINE configuration option, 7072 and the compiler with which the kernel was built. 7073 7074 Selecting 'on' will also enable the mitigation 7075 against user space to user space task attacks. 7076 Selecting specific mitigation does not force enable 7077 user mitigations. 7078 7079 Selecting 'off' will disable both the kernel and 7080 the user space protections. 7081 7082 Specific mitigations can also be selected manually: 7083 7084 retpoline - replace indirect branches 7085 retpoline,generic - Retpolines 7086 retpoline,lfence - LFENCE; indirect branch 7087 retpoline,amd - alias for retpoline,lfence 7088 eibrs - Enhanced/Auto IBRS 7089 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines 7090 eibrs,lfence - Enhanced/Auto IBRS + LFENCE 7091 ibrs - use IBRS to protect kernel 7092 7093 Not specifying this option is equivalent to 7094 spectre_v2=auto. 7095 7096 spectre_v2_user= 7097 [X86] Control mitigation of Spectre variant 2 7098 (indirect branch speculation) vulnerability between 7099 user space tasks 7100 7101 on - Unconditionally enable mitigations. Is 7102 enforced by spectre_v2=on 7103 7104 off - Unconditionally disable mitigations. Is 7105 enforced by spectre_v2=off 7106 7107 prctl - Indirect branch speculation is enabled, 7108 but mitigation can be enabled via prctl 7109 per thread. The mitigation control state 7110 is inherited on fork. 7111 7112 prctl,ibpb 7113 - Like "prctl" above, but only STIBP is 7114 controlled per thread. IBPB is issued 7115 always when switching between different user 7116 space processes. 7117 7118 seccomp 7119 - Same as "prctl" above, but all seccomp 7120 threads will enable the mitigation unless 7121 they explicitly opt out. 7122 7123 seccomp,ibpb 7124 - Like "seccomp" above, but only STIBP is 7125 controlled per thread. IBPB is issued 7126 always when switching between different 7127 user space processes. 7128 7129 auto - Kernel selects the mitigation depending on 7130 the available CPU features and vulnerability. 7131 7132 Default mitigation: "prctl" 7133 7134 Not specifying this option is equivalent to 7135 spectre_v2_user=auto. 7136 7137 spec_rstack_overflow= 7138 [X86,EARLY] Control RAS overflow mitigation on AMD Zen CPUs 7139 7140 off - Disable mitigation 7141 microcode - Enable microcode mitigation only 7142 safe-ret - Enable sw-only safe RET mitigation (default) 7143 ibpb - Enable mitigation by issuing IBPB on 7144 kernel entry 7145 ibpb-vmexit - Issue IBPB only on VMEXIT 7146 (cloud-specific mitigation) 7147 7148 spec_store_bypass_disable= 7149 [HW,EARLY] Control Speculative Store Bypass (SSB) Disable mitigation 7150 (Speculative Store Bypass vulnerability) 7151 7152 Certain CPUs are vulnerable to an exploit against a 7153 a common industry wide performance optimization known 7154 as "Speculative Store Bypass" in which recent stores 7155 to the same memory location may not be observed by 7156 later loads during speculative execution. The idea 7157 is that such stores are unlikely and that they can 7158 be detected prior to instruction retirement at the 7159 end of a particular speculation execution window. 7160 7161 In vulnerable processors, the speculatively forwarded 7162 store can be used in a cache side channel attack, for 7163 example to read memory to which the attacker does not 7164 directly have access (e.g. inside sandboxed code). 7165 7166 This parameter controls whether the Speculative Store 7167 Bypass optimization is used. 7168 7169 On x86 the options are: 7170 7171 on - Unconditionally disable Speculative Store Bypass 7172 off - Unconditionally enable Speculative Store Bypass 7173 auto - Kernel detects whether the CPU model contains an 7174 implementation of Speculative Store Bypass and 7175 picks the most appropriate mitigation. If the 7176 CPU is not vulnerable, "off" is selected. If the 7177 CPU is vulnerable the default mitigation is 7178 architecture and Kconfig dependent. See below. 7179 prctl - Control Speculative Store Bypass per thread 7180 via prctl. Speculative Store Bypass is enabled 7181 for a process by default. The state of the control 7182 is inherited on fork. 7183 seccomp - Same as "prctl" above, but all seccomp threads 7184 will disable SSB unless they explicitly opt out. 7185 7186 Default mitigations: 7187 X86: "prctl" 7188 7189 On powerpc the options are: 7190 7191 on,auto - On Power8 and Power9 insert a store-forwarding 7192 barrier on kernel entry and exit. On Power7 7193 perform a software flush on kernel entry and 7194 exit. 7195 off - No action. 7196 7197 Not specifying this option is equivalent to 7198 spec_store_bypass_disable=auto. 7199 7200 split_llc= 7201 [X86,EARLY] Split the LLC N-ways 7202 7203 When set, the LLC is split this many ways by matching 7204 'core_id % n'. This is setup before SMP bringup and 7205 used during SMP bringup before it knows the full 7206 topology. If your core count doesn't nicely divide by 7207 the number given, you get to keep the pieces. 7208 7209 This is mostly a debug feature to emulate multiple LLCs 7210 on hardware that only have a single LLC. 7211 7212 split_lock_detect= 7213 [X86] Enable split lock detection or bus lock detection 7214 7215 When enabled (and if hardware support is present), atomic 7216 instructions that access data across cache line 7217 boundaries will result in an alignment check exception 7218 for split lock detection or a debug exception for 7219 bus lock detection. 7220 7221 off - not enabled 7222 7223 warn - the kernel will emit rate-limited warnings 7224 about applications triggering the #AC 7225 exception or the #DB exception. This mode is 7226 the default on CPUs that support split lock 7227 detection or bus lock detection. Default 7228 behavior is by #AC if both features are 7229 enabled in hardware. 7230 7231 fatal - the kernel will send SIGBUS to applications 7232 that trigger the #AC exception or the #DB 7233 exception. Default behavior is by #AC if 7234 both features are enabled in hardware. 7235 7236 ratelimit:N - 7237 Set system wide rate limit to N bus locks 7238 per second for bus lock detection. 7239 0 < N <= 1000. 7240 7241 N/A for split lock detection. 7242 7243 7244 If an #AC exception is hit in the kernel or in 7245 firmware (i.e. not while executing in user mode) 7246 the kernel will oops in either "warn" or "fatal" 7247 mode. 7248 7249 #DB exception for bus lock is triggered only when 7250 CPL > 0. 7251 7252 srbds= [X86,INTEL,EARLY] 7253 Control the Special Register Buffer Data Sampling 7254 (SRBDS) mitigation. 7255 7256 Certain CPUs are vulnerable to an MDS-like 7257 exploit which can leak bits from the random 7258 number generator. 7259 7260 By default, this issue is mitigated by 7261 microcode. However, the microcode fix can cause 7262 the RDRAND and RDSEED instructions to become 7263 much slower. Among other effects, this will 7264 result in reduced throughput from /dev/urandom. 7265 7266 The microcode mitigation can be disabled with 7267 the following option: 7268 7269 off: Disable mitigation and remove 7270 performance impact to RDRAND and RDSEED 7271 7272 srcutree.big_cpu_lim [KNL] 7273 Specifies the number of CPUs constituting a 7274 large system, such that srcu_struct structures 7275 should immediately allocate an srcu_node array. 7276 This kernel-boot parameter defaults to 128, 7277 but takes effect only when the low-order four 7278 bits of srcutree.convert_to_big is equal to 3 7279 (decide at boot). 7280 7281 srcutree.convert_to_big [KNL] 7282 Specifies under what conditions an SRCU tree 7283 srcu_struct structure will be converted to big 7284 form, that is, with an rcu_node tree: 7285 7286 0: Never. 7287 1: At init_srcu_struct() time. 7288 2: When rcutorture decides to. 7289 3: Decide at boot time (default). 7290 0x1X: Above plus if high contention. 7291 7292 Either way, the srcu_node tree will be sized based 7293 on the actual runtime number of CPUs (nr_cpu_ids) 7294 instead of the compile-time CONFIG_NR_CPUS. 7295 7296 srcutree.counter_wrap_check [KNL] 7297 Specifies how frequently to check for 7298 grace-period sequence counter wrap for the 7299 srcu_data structure's ->srcu_gp_seq_needed field. 7300 The greater the number of bits set in this kernel 7301 parameter, the less frequently counter wrap will 7302 be checked for. Note that the bottom two bits 7303 are ignored. 7304 7305 srcutree.exp_holdoff [KNL] 7306 Specifies how many nanoseconds must elapse 7307 since the end of the last SRCU grace period for 7308 a given srcu_struct until the next normal SRCU 7309 grace period will be considered for automatic 7310 expediting. Set to zero to disable automatic 7311 expediting. 7312 7313 srcutree.srcu_max_nodelay [KNL] 7314 Specifies the number of no-delay instances 7315 per jiffy for which the SRCU grace period 7316 worker thread will be rescheduled with zero 7317 delay. Beyond this limit, worker thread will 7318 be rescheduled with a sleep delay of one jiffy. 7319 7320 srcutree.srcu_max_nodelay_phase [KNL] 7321 Specifies the per-grace-period phase, number of 7322 non-sleeping polls of readers. Beyond this limit, 7323 grace period worker thread will be rescheduled 7324 with a sleep delay of one jiffy, between each 7325 rescan of the readers, for a grace period phase. 7326 7327 srcutree.srcu_retry_check_delay [KNL] 7328 Specifies number of microseconds of non-sleeping 7329 delay between each non-sleeping poll of readers. 7330 7331 srcutree.small_contention_lim [KNL] 7332 Specifies the number of update-side contention 7333 events per jiffy will be tolerated before 7334 initiating a conversion of an srcu_struct 7335 structure to big form. Note that the value of 7336 srcutree.convert_to_big must have the 0x10 bit 7337 set for contention-based conversions to occur. 7338 7339 ssbd= [ARM64,HW,EARLY] 7340 Speculative Store Bypass Disable control 7341 7342 On CPUs that are vulnerable to the Speculative 7343 Store Bypass vulnerability and offer a 7344 firmware based mitigation, this parameter 7345 indicates how the mitigation should be used: 7346 7347 force-on: Unconditionally enable mitigation for 7348 for both kernel and userspace 7349 force-off: Unconditionally disable mitigation for 7350 for both kernel and userspace 7351 kernel: Always enable mitigation in the 7352 kernel, and offer a prctl interface 7353 to allow userspace to register its 7354 interest in being mitigated too. 7355 7356 stack_guard_gap= [MM] 7357 override the default stack gap protection. The value 7358 is in page units and it defines how many pages prior 7359 to (for stacks growing down) resp. after (for stacks 7360 growing up) the main stack are reserved for no other 7361 mapping. Default value is 256 pages. 7362 7363 stack_depot_disable= [KNL,EARLY] 7364 Setting this to true through kernel command line will 7365 disable the stack depot thereby saving the static memory 7366 consumed by the stack hash table. By default this is set 7367 to false. 7368 7369 stack_depot_max_pools= [KNL,EARLY] 7370 Specify the maximum number of pools to use for storing 7371 stack traces. Pools are allocated on-demand up to this 7372 limit. Default value is 8191 pools. 7373 7374 stacktrace [FTRACE] 7375 Enable the stack tracer on boot up. 7376 7377 stacktrace_filter=[function-list] 7378 [FTRACE] Limit the functions that the stack tracer 7379 will trace at boot up. function-list is a comma-separated 7380 list of functions. This list can be changed at run 7381 time by the stack_trace_filter file in the debugfs 7382 tracing directory. Note, this enables stack tracing 7383 and the stacktrace above is not needed. 7384 7385 sti= [PARISC,HW] 7386 Format: <num> 7387 Set the STI (builtin display/keyboard on the HP-PARISC 7388 machines) console (graphic card) which should be used 7389 as the initial boot-console. 7390 See also comment in drivers/video/sticore.c. 7391 7392 sti_font= [HW] 7393 See comment in drivers/video/sticore.c. 7394 7395 stifb= [HW] 7396 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 7397 7398 strict_sas_size= 7399 [X86] 7400 Format: <bool> 7401 Enable or disable strict sigaltstack size checks 7402 against the required signal frame size which 7403 depends on the supported FPU features. This can 7404 be used to filter out binaries which have 7405 not yet been made aware of AT_MINSIGSTKSZ. 7406 7407 stress_hpt [PPC,EARLY] 7408 Limits the number of kernel HPT entries in the hash 7409 page table to increase the rate of hash page table 7410 faults on kernel addresses. 7411 7412 stress_slb [PPC,EARLY] 7413 Limits the number of kernel SLB entries, and flushes 7414 them frequently to increase the rate of SLB faults 7415 on kernel addresses. 7416 7417 no_slb_preload [PPC,EARLY] 7418 Disables slb preloading for userspace. 7419 7420 sunrpc.min_resvport= 7421 sunrpc.max_resvport= 7422 [NFS,SUNRPC] 7423 SunRPC servers often require that client requests 7424 originate from a privileged port (i.e. a port in the 7425 range 0 < portnr < 1024). 7426 An administrator who wishes to reserve some of these 7427 ports for other uses may adjust the range that the 7428 kernel's sunrpc client considers to be privileged 7429 using these two parameters to set the minimum and 7430 maximum port values. 7431 7432 sunrpc.svc_rpc_per_connection_limit= 7433 [NFS,SUNRPC] 7434 Limit the number of requests that the server will 7435 process in parallel from a single connection. 7436 The default value is 0 (no limit). 7437 7438 sunrpc.pool_mode= 7439 [NFS] 7440 Control how the NFS server code allocates CPUs to 7441 service thread pools. Depending on how many NICs 7442 you have and where their interrupts are bound, this 7443 option will affect which CPUs will do NFS serving. 7444 Note: this parameter cannot be changed while the 7445 NFS server is running. 7446 7447 auto the server chooses an appropriate mode 7448 automatically using heuristics 7449 global a single global pool contains all CPUs 7450 percpu one pool for each CPU 7451 pernode one pool for each NUMA node (equivalent 7452 to global on non-NUMA machines) 7453 7454 sunrpc.tcp_slot_table_entries= 7455 sunrpc.udp_slot_table_entries= 7456 [NFS,SUNRPC] 7457 Sets the upper limit on the number of simultaneous 7458 RPC calls that can be sent from the client to a 7459 server. Increasing these values may allow you to 7460 improve throughput, but will also increase the 7461 amount of memory reserved for use by the client. 7462 7463 suspend.pm_test_delay= 7464 [SUSPEND] 7465 Sets the number of seconds to remain in a suspend test 7466 mode before resuming the system (see 7467 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 7468 is set. Default value is 5. 7469 7470 svm= [PPC] 7471 Format: { on | off | y | n | 1 | 0 } 7472 This parameter controls use of the Protected 7473 Execution Facility on pSeries. 7474 7475 swiotlb= [ARM,PPC,MIPS,X86,S390,EARLY] 7476 Format: { <int> [,<int>] | force | noforce } 7477 <int> -- Number of I/O TLB slabs 7478 <int> -- Second integer after comma. Number of swiotlb 7479 areas with their own lock. Will be rounded up 7480 to a power of 2. 7481 force -- force using of bounce buffers even if they 7482 wouldn't be automatically used by the kernel 7483 noforce -- Never use bounce buffers (for debugging) 7484 7485 switches= [HW,M68k,EARLY] 7486 7487 sysctl.*= [KNL] 7488 Set a sysctl parameter, right before loading the init 7489 process, as if the value was written to the respective 7490 /proc/sys/... file. Both '.' and '/' are recognized as 7491 separators. Unrecognized parameters and invalid values 7492 are reported in the kernel log. Sysctls registered 7493 later by a loaded module cannot be set this way. 7494 Example: sysctl.vm.swappiness=40 7495 7496 sysrq_always_enabled 7497 [KNL] 7498 Ignore sysrq setting - this boot parameter will 7499 neutralize any effect of /proc/sys/kernel/sysrq. 7500 Useful for debugging. 7501 7502 tcpmhash_entries= [KNL,NET] 7503 Set the number of tcp_metrics_hash slots. 7504 Default value is 8192 or 16384 depending on total 7505 ram pages. This is used to specify the TCP metrics 7506 cache size. See Documentation/networking/ip-sysctl.rst 7507 "tcp_no_metrics_save" section for more details. 7508 7509 tdfx= [HW,DRM] 7510 7511 test_suspend= [SUSPEND] 7512 Format: { "mem" | "standby" | "freeze" }[,N] 7513 Specify "mem" (for Suspend-to-RAM) or "standby" (for 7514 standby suspend) or "freeze" (for suspend type freeze) 7515 as the system sleep state during system startup with 7516 the optional capability to repeat N number of times. 7517 The system is woken from this state using a 7518 wakeup-capable RTC alarm. 7519 7520 thash_entries= [KNL,NET] 7521 Set number of hash buckets for TCP connection 7522 7523 thermal.act= [HW,ACPI] 7524 -1: disable all active trip points in all thermal zones 7525 <degrees C>: override all lowest active trip points 7526 7527 thermal.crt= [HW,ACPI] 7528 -1: disable all critical trip points in all thermal zones 7529 <degrees C>: override all critical trip points 7530 7531 thermal.off= [HW,ACPI] 7532 1: disable ACPI thermal control 7533 7534 thermal.psv= [HW,ACPI] 7535 -1: disable all passive trip points 7536 <degrees C>: override all passive trip points to this 7537 value 7538 7539 thermal.tzp= [HW,ACPI] 7540 Specify global default ACPI thermal zone polling rate 7541 <deci-seconds>: poll all this frequency 7542 0: no polling (default) 7543 7544 thp_anon= [KNL] 7545 Format: <size>[KMG],<size>[KMG]:<state>;<size>[KMG]-<size>[KMG]:<state> 7546 state is one of "always", "madvise", "never" or "inherit". 7547 Control the default behavior of the system with respect 7548 to anonymous transparent hugepages. 7549 Can be used multiple times for multiple anon THP sizes. 7550 See Documentation/admin-guide/mm/transhuge.rst for more 7551 details. 7552 7553 threadirqs [KNL,EARLY] 7554 Force threading of all interrupt handlers except those 7555 marked explicitly IRQF_NO_THREAD. 7556 7557 thp_shmem= [KNL] 7558 Format: <size>[KMG],<size>[KMG]:<policy>;<size>[KMG]-<size>[KMG]:<policy> 7559 Control the default policy of each hugepage size for the 7560 internal shmem mount. <policy> is one of policies available 7561 for the shmem mount ("always", "inherit", "never", "within_size", 7562 and "advise"). 7563 It can be used multiple times for multiple shmem THP sizes. 7564 See Documentation/admin-guide/mm/transhuge.rst for more 7565 details. 7566 7567 topology= [S390,EARLY] 7568 Format: {off | on} 7569 Specify if the kernel should make use of the cpu 7570 topology information if the hardware supports this. 7571 The scheduler will make use of this information and 7572 e.g. base its process migration decisions on it. 7573 Default is on. 7574 7575 torture.disable_onoff_at_boot= [KNL] 7576 Prevent the CPU-hotplug component of torturing 7577 until after init has spawned. 7578 7579 torture.ftrace_dump_at_shutdown= [KNL] 7580 Dump the ftrace buffer at torture-test shutdown, 7581 even if there were no errors. This can be a 7582 very costly operation when many torture tests 7583 are running concurrently, especially on systems 7584 with rotating-rust storage. 7585 7586 torture.verbose_sleep_frequency= [KNL] 7587 Specifies how many verbose printk()s should be 7588 emitted between each sleep. The default of zero 7589 disables verbose-printk() sleeping. 7590 7591 torture.verbose_sleep_duration= [KNL] 7592 Duration of each verbose-printk() sleep in jiffies. 7593 7594 tpm.disable_pcr_integrity= [HW,TPM] 7595 Do not protect PCR registers from unintended physical 7596 access, or interposers in the bus by the means of 7597 having an integrity protected session wrapped around 7598 TPM2_PCR_Extend command. Consider this in a situation 7599 where TPM is heavily utilized by IMA, thus protection 7600 causing a major performance hit, and the space where 7601 machines are deployed is by other means guarded. 7602 7603 tpm_crb_ffa.busy_timeout_ms= [ARM64,TPM] 7604 Maximum time in milliseconds to retry sending a message 7605 to the TPM service before giving up. This parameter controls 7606 how long the system will continue retrying when the TPM 7607 service is busy. 7608 Format: <unsigned int> 7609 Default: 2000 (2 seconds) 7610 7611 tpm_suspend_pcr=[HW,TPM] 7612 Format: integer pcr id 7613 Specify that at suspend time, the tpm driver 7614 should extend the specified pcr with zeros, 7615 as a workaround for some chips which fail to 7616 flush the last written pcr on TPM_SaveState. 7617 This will guarantee that all the other pcrs 7618 are saved. 7619 7620 tpm_tis.interrupts= [HW,TPM] 7621 Enable interrupts for the MMIO based physical layer 7622 for the FIFO interface. By default it is set to false 7623 (0). For more information about TPM hardware interfaces 7624 defined by Trusted Computing Group (TCG) see 7625 https://trustedcomputinggroup.org/resource/pc-client-platform-tpm-profile-ptp-specification/ 7626 7627 tp_printk [FTRACE] 7628 Have the tracepoints sent to printk as well as the 7629 tracing ring buffer. This is useful for early boot up 7630 where the system hangs or reboots and does not give the 7631 option for reading the tracing buffer or performing a 7632 ftrace_dump_on_oops. 7633 7634 To turn off having tracepoints sent to printk, 7635 echo 0 > /proc/sys/kernel/tracepoint_printk 7636 Note, echoing 1 into this file without the 7637 tp_printk kernel cmdline option has no effect. 7638 7639 The tp_printk_stop_on_boot (see below) can also be used 7640 to stop the printing of events to console at 7641 late_initcall_sync. 7642 7643 ** CAUTION ** 7644 7645 Having tracepoints sent to printk() and activating high 7646 frequency tracepoints such as irq or sched, can cause 7647 the system to live lock. 7648 7649 tp_printk_stop_on_boot [FTRACE] 7650 When tp_printk (above) is set, it can cause a lot of noise 7651 on the console. It may be useful to only include the 7652 printing of events during boot up, as user space may 7653 make the system inoperable. 7654 7655 This command line option will stop the printing of events 7656 to console at the late_initcall_sync() time frame. 7657 7658 trace_buf_size=nn[KMG] 7659 [FTRACE] will set tracing buffer size on each cpu. 7660 7661 trace_clock= [FTRACE] Set the clock used for tracing events 7662 at boot up. 7663 local - Use the per CPU time stamp counter 7664 (converted into nanoseconds). Fast, but 7665 depending on the architecture, may not be 7666 in sync between CPUs. 7667 global - Event time stamps are synchronized across 7668 CPUs. May be slower than the local clock, 7669 but better for some race conditions. 7670 counter - Simple counting of events (1, 2, ..) 7671 note, some counts may be skipped due to the 7672 infrastructure grabbing the clock more than 7673 once per event. 7674 uptime - Use jiffies as the time stamp. 7675 perf - Use the same clock that perf uses. 7676 mono - Use ktime_get_mono_fast_ns() for time stamps. 7677 mono_raw - Use ktime_get_raw_fast_ns() for time 7678 stamps. 7679 boot - Use ktime_get_boot_fast_ns() for time stamps. 7680 Architectures may add more clocks. See 7681 Documentation/trace/ftrace.rst for more details. 7682 7683 trace_event=[event-list] 7684 [FTRACE] Set and start specified trace events in order 7685 to facilitate early boot debugging. The event-list is a 7686 comma-separated list of trace events to enable. See 7687 also Documentation/trace/events.rst 7688 7689 To enable modules, use :mod: keyword: 7690 7691 trace_event=:mod:<module> 7692 7693 The value before :mod: will only enable specific events 7694 that are part of the module. See the above mentioned 7695 document for more information. 7696 7697 trace_instance=[instance-info] 7698 [FTRACE] Create a ring buffer instance early in boot up. 7699 This will be listed in: 7700 7701 /sys/kernel/tracing/instances 7702 7703 Events can be enabled at the time the instance is created 7704 via: 7705 7706 trace_instance=<name>,<system1>:<event1>,<system2>:<event2> 7707 7708 Note, the "<system*>:" portion is optional if the event is 7709 unique. 7710 7711 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall 7712 7713 will enable the "sched_switch" event (note, the "sched:" is optional, and 7714 the same thing would happen if it was left off). The irq_handler_entry 7715 event, and all events under the "initcall" system. 7716 7717 Flags can be added to the instance to modify its behavior when it is 7718 created. The flags are separated by '^'. 7719 7720 The available flags are: 7721 7722 traceoff - Have the tracing instance tracing disabled after it is created. 7723 traceprintk - Have trace_printk() write into this trace instance 7724 (note, "printk" and "trace_printk" can also be used) 7725 7726 trace_instance=foo^traceoff^traceprintk,sched,irq 7727 7728 The flags must come before the defined events. 7729 7730 If memory has been reserved (see memmap for x86), the instance 7731 can use that memory: 7732 7733 memmap=12M$0x284500000 trace_instance=boot_map@0x284500000:12M 7734 7735 The above will create a "boot_map" instance that uses the physical 7736 memory at 0x284500000 that is 12Megs. The per CPU buffers of that 7737 instance will be split up accordingly. 7738 7739 Alternatively, the memory can be reserved by the reserve_mem option: 7740 7741 reserve_mem=12M:4096:trace trace_instance=boot_map@trace 7742 7743 This will reserve 12 megabytes at boot up with a 4096 byte alignment 7744 and place the ring buffer in this memory. Note that due to KASLR, the 7745 memory may not be the same location each time, which will not preserve 7746 the buffer content. 7747 7748 Also note that the layout of the ring buffer data may change between 7749 kernel versions where the validator will fail and reset the ring buffer 7750 if the layout is not the same as the previous kernel. 7751 7752 If the ring buffer is used for persistent bootups and has events enabled, 7753 it is recommend to disable tracing so that events from a previous boot do not 7754 mix with events of the current boot (unless you are debugging a random crash 7755 at boot up). 7756 7757 reserve_mem=12M:4096:trace trace_instance=boot_map^traceoff^traceprintk@trace,sched,irq 7758 7759 Note, saving the trace buffer across reboots does require that the system 7760 is set up to not wipe memory. For instance, CONFIG_RESET_ATTACK_MITIGATION 7761 can force a memory reset on boot which will clear any trace that was stored. 7762 This is just one of many ways that can clear memory. Make sure your system 7763 keeps the content of memory across reboots before relying on this option. 7764 7765 NB: Both the mapped address and size must be page aligned for the architecture. 7766 7767 See also Documentation/trace/debugging.rst 7768 7769 7770 trace_options=[option-list] 7771 [FTRACE] Enable or disable tracer options at boot. 7772 The option-list is a comma delimited list of options 7773 that can be enabled or disabled just as if you were 7774 to echo the option name into 7775 7776 /sys/kernel/tracing/trace_options 7777 7778 For example, to enable stacktrace option (to dump the 7779 stack trace of each event), add to the command line: 7780 7781 trace_options=stacktrace 7782 7783 See also Documentation/trace/ftrace.rst "trace options" 7784 section. 7785 7786 trace_trigger=[trigger-list] 7787 [FTRACE] Add an event trigger on specific events. 7788 Set a trigger on top of a specific event, with an optional 7789 filter. 7790 7791 The format is "trace_trigger=<event>.<trigger>[ if <filter>],..." 7792 Where more than one trigger may be specified that are comma delimited. 7793 7794 For example: 7795 7796 trace_trigger="sched_switch.stacktrace if prev_state == 2" 7797 7798 The above will enable the "stacktrace" trigger on the "sched_switch" 7799 event but only trigger it if the "prev_state" of the "sched_switch" 7800 event is "2" (TASK_UNINTERRUPTIBLE). 7801 7802 See also "Event triggers" in Documentation/trace/events.rst 7803 7804 7805 traceoff_after_boot 7806 [FTRACE] Sometimes tracing is used to debug issues 7807 during the boot process. Since the trace buffer has a 7808 limited amount of storage, it may be prudent to 7809 disable tracing after the boot is finished, otherwise 7810 the critical information may be overwritten. With this 7811 option, the main tracing buffer will be turned off at 7812 the end of the boot process. 7813 7814 traceoff_on_warning 7815 [FTRACE] enable this option to disable tracing when a 7816 warning is hit. This turns off "tracing_on". Tracing can 7817 be enabled again by echoing '1' into the "tracing_on" 7818 file located in /sys/kernel/tracing/ 7819 7820 This option is useful, as it disables the trace before 7821 the WARNING dump is called, which prevents the trace to 7822 be filled with content caused by the warning output. 7823 7824 This option can also be set at run time via the sysctl 7825 option: kernel/traceoff_on_warning 7826 7827 transparent_hugepage= 7828 [KNL] 7829 Format: [always|madvise|never] 7830 Can be used to control the default behavior of the system 7831 with respect to transparent hugepages. 7832 See Documentation/admin-guide/mm/transhuge.rst 7833 for more details. 7834 7835 transparent_hugepage_shmem= [KNL] 7836 Format: [always|within_size|advise|never|deny|force] 7837 Can be used to control the hugepage allocation policy for 7838 the internal shmem mount. 7839 See Documentation/admin-guide/mm/transhuge.rst 7840 for more details. 7841 7842 transparent_hugepage_tmpfs= [KNL] 7843 Format: [always|within_size|advise|never] 7844 Can be used to control the default hugepage allocation policy 7845 for the tmpfs mount. 7846 See Documentation/admin-guide/mm/transhuge.rst 7847 for more details. 7848 7849 trusted.source= [KEYS] 7850 Format: <string> 7851 This parameter identifies the trust source as a backend 7852 for trusted keys implementation. Supported trust 7853 sources: 7854 - "tpm" 7855 - "tee" 7856 - "caam" 7857 - "dcp" 7858 - "pkwm" 7859 If not specified then it defaults to iterating through 7860 the trust source list starting with TPM and assigns the 7861 first trust source as a backend which is initialized 7862 successfully during iteration. 7863 7864 trusted.debug= [KEYS] 7865 Format: <bool> 7866 Enable trusted keys debug traces at runtime when 7867 CONFIG_TRUSTED_KEYS_DEBUG=y. 7868 7869 To make the traces visible after enabling the option, 7870 use trusted.dyndbg='+p' as needed. By convention, 7871 the subsystem uses pr_debug() for these traces. 7872 7873 SAFETY: The traces can leak sensitive data, so be 7874 cautious before enabling this. They remain inactive 7875 unless this parameter is set this option to a true 7876 value. 7877 7878 Default: false 7879 7880 trusted.rng= [KEYS] 7881 Format: <string> 7882 The RNG used to generate key material for trusted keys. 7883 Can be one of: 7884 - "kernel" 7885 - the same value as trusted.source: "tpm" or "tee" 7886 - "default" 7887 If not specified, "default" is used. In this case, 7888 the RNG's choice is left to each individual trust source. 7889 7890 trusted.dcp_use_otp_key 7891 This is intended to be used in combination with 7892 trusted.source=dcp and will select the DCP OTP key 7893 instead of the DCP UNIQUE key blob encryption. 7894 7895 trusted.dcp_skip_zk_test 7896 This is intended to be used in combination with 7897 trusted.source=dcp and will disable the check if the 7898 blob key is all zeros. This is helpful for situations where 7899 having this key zero'ed is acceptable. E.g. in testing 7900 scenarios. 7901 7902 tsa= [X86] Control mitigation for Transient Scheduler 7903 Attacks on AMD CPUs. Search the following in your 7904 favourite search engine for more details: 7905 7906 "Technical guidance for mitigating transient scheduler 7907 attacks". 7908 7909 off - disable the mitigation 7910 on - enable the mitigation (default) 7911 user - mitigate only user/kernel transitions 7912 vm - mitigate only guest/host transitions 7913 7914 7915 tsc= Disable clocksource stability checks for TSC. 7916 Format: <string> 7917 [x86] reliable: mark tsc clocksource as reliable, this 7918 disables clocksource verification at runtime, as well 7919 as the stability checks done at bootup. Used to enable 7920 high-resolution timer mode on older hardware, and in 7921 virtualized environment. 7922 [x86] noirqtime: Do not use TSC to do irq accounting. 7923 Used to run time disable IRQ_TIME_ACCOUNTING on any 7924 platforms where RDTSC is slow and this accounting 7925 can add overhead. 7926 [x86] unstable: mark the TSC clocksource as unstable, this 7927 marks the TSC unconditionally unstable at bootup and 7928 avoids any further wobbles once the TSC watchdog notices. 7929 [x86] nowatchdog: disable clocksource watchdog. Used 7930 in situations with strict latency requirements (where 7931 interruptions from clocksource watchdog are not 7932 acceptable). 7933 [x86] recalibrate: force recalibration against a HW timer 7934 (HPET or PM timer) on systems whose TSC frequency was 7935 obtained from HW or FW using either an MSR or CPUID(0x15). 7936 Warn if the difference is more than 500 ppm. 7937 [x86] watchdog: Enforce the clocksource watchdog on TSC 7938 7939 tsc_early_khz= [X86,EARLY] Skip early TSC calibration and use the given 7940 value instead. Useful when the early TSC frequency discovery 7941 procedure is not reliable, such as on overclocked systems 7942 with CPUID.16h support and partial CPUID.15h support. 7943 Format: <unsigned int> 7944 7945 tsx= [X86] Control Transactional Synchronization 7946 Extensions (TSX) feature in Intel processors that 7947 support TSX control. 7948 7949 This parameter controls the TSX feature. The options are: 7950 7951 on - Enable TSX on the system. Although there are 7952 mitigations for all known security vulnerabilities, 7953 TSX has been known to be an accelerator for 7954 several previous speculation-related CVEs, and 7955 so there may be unknown security risks associated 7956 with leaving it enabled. 7957 7958 off - Disable TSX on the system. (Note that this 7959 option takes effect only on newer CPUs which are 7960 not vulnerable to MDS, i.e., have 7961 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 7962 the new IA32_TSX_CTRL MSR through a microcode 7963 update. This new MSR allows for the reliable 7964 deactivation of the TSX functionality.) 7965 7966 auto - Disable TSX if X86_BUG_TAA is present, 7967 otherwise enable TSX on the system. 7968 7969 Not specifying this option is equivalent to tsx=off. 7970 7971 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 7972 for more details. 7973 7974 tsx_async_abort= [X86,INTEL,EARLY] Control mitigation for the TSX Async 7975 Abort (TAA) vulnerability. 7976 7977 Similar to Micro-architectural Data Sampling (MDS) 7978 certain CPUs that support Transactional 7979 Synchronization Extensions (TSX) are vulnerable to an 7980 exploit against CPU internal buffers which can forward 7981 information to a disclosure gadget under certain 7982 conditions. 7983 7984 In vulnerable processors, the speculatively forwarded 7985 data can be used in a cache side channel attack, to 7986 access data to which the attacker does not have direct 7987 access. 7988 7989 This parameter controls the TAA mitigation. The 7990 options are: 7991 7992 full - Enable TAA mitigation on vulnerable CPUs 7993 if TSX is enabled. 7994 7995 full,nosmt - Enable TAA mitigation and disable SMT on 7996 vulnerable CPUs. If TSX is disabled, SMT 7997 is not disabled because CPU is not 7998 vulnerable to cross-thread TAA attacks. 7999 off - Unconditionally disable TAA mitigation 8000 8001 On MDS-affected machines, tsx_async_abort=off can be 8002 prevented by an active MDS mitigation as both vulnerabilities 8003 are mitigated with the same mechanism so in order to disable 8004 this mitigation, you need to specify mds=off too. 8005 8006 Not specifying this option is equivalent to 8007 tsx_async_abort=full. On CPUs which are MDS affected 8008 and deploy MDS mitigation, TAA mitigation is not 8009 required and doesn't provide any additional 8010 mitigation. 8011 8012 For details see: 8013 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 8014 8015 turbografx.map[2|3]= [HW,JOY] 8016 TurboGraFX parallel port interface 8017 Format: 8018 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 8019 See also Documentation/input/devices/joystick-parport.rst 8020 8021 udbg-immortal [PPC] When debugging early kernel crashes that 8022 happen after console_init() and before a proper 8023 console driver takes over, this boot options might 8024 help "seeing" what's going on. 8025 8026 uhash_entries= [KNL,NET] 8027 Set number of hash buckets for UDP/UDP-Lite connections 8028 8029 uhci-hcd.ignore_oc= 8030 [USB] Ignore overcurrent events (default N). 8031 Some badly-designed motherboards generate lots of 8032 bogus events, for ports that aren't wired to 8033 anything. Set this parameter to avoid log spamming. 8034 Note that genuine overcurrent events won't be 8035 reported either. 8036 8037 unaligned_scalar_speed= 8038 [RISCV] 8039 Format: {slow | fast | unsupported} 8040 Allow skipping scalar unaligned access speed tests. This 8041 is useful for testing alternative code paths and to skip 8042 the tests in environments where they run too slowly. All 8043 CPUs must have the same scalar unaligned access speed. 8044 8045 unaligned_vector_speed= 8046 [RISCV] 8047 Format: {slow | fast | unsupported} 8048 Allow skipping vector unaligned access speed tests. This 8049 is useful for testing alternative code paths and to skip 8050 the tests in environments where they run too slowly. All 8051 CPUs must have the same vector unaligned access speed. 8052 8053 unknown_nmi_panic 8054 [X86] Cause panic on unknown NMI. 8055 8056 unwind_debug [X86-64,EARLY] 8057 Enable unwinder debug output. This can be 8058 useful for debugging certain unwinder error 8059 conditions, including corrupt stacks and 8060 bad/missing unwinder metadata. 8061 8062 usbcore.authorized_default= 8063 [USB] Default USB device authorization: 8064 (default -1 = authorized (same as 1), 8065 0 = not authorized, 1 = authorized, 2 = authorized 8066 if device connected to internal port) 8067 8068 usbcore.autosuspend= 8069 [USB] The autosuspend time delay (in seconds) used 8070 for newly-detected USB devices (default 2). This 8071 is the time required before an idle device will be 8072 autosuspended. Devices for which the delay is set 8073 to a negative value won't be autosuspended at all. 8074 8075 usbcore.usbfs_snoop= 8076 [USB] Set to log all usbfs traffic (default 0 = off). 8077 8078 usbcore.usbfs_snoop_max= 8079 [USB] Maximum number of bytes to snoop in each URB 8080 (default = 65536). 8081 8082 usbcore.blinkenlights= 8083 [USB] Set to cycle leds on hubs (default 0 = off). 8084 8085 usbcore.old_scheme_first= 8086 [USB] Start with the old device initialization 8087 scheme (default 0 = off). 8088 8089 usbcore.usbfs_memory_mb= 8090 [USB] Memory limit (in MB) for buffers allocated by 8091 usbfs (default = 16, 0 = max = 2047). 8092 8093 usbcore.use_both_schemes= 8094 [USB] Try the other device initialization scheme 8095 if the first one fails (default 1 = enabled). 8096 8097 usbcore.initial_descriptor_timeout= 8098 [USB] Specifies timeout for the initial 64-byte 8099 USB_REQ_GET_DESCRIPTOR request in milliseconds 8100 (default 5000 = 5.0 seconds). 8101 8102 usbcore.nousb [USB] Disable the USB subsystem 8103 8104 usbcore.quirks= 8105 [USB] A list of quirk entries to augment the built-in 8106 usb core quirk list. List entries are separated by 8107 commas. Each entry has the form 8108 VendorID:ProductID:Flags. The IDs are 4-digit hex 8109 numbers and Flags is a set of letters. Each letter 8110 will change the built-in quirk; setting it if it is 8111 clear and clearing it if it is set. The letters have 8112 the following meanings: 8113 a = USB_QUIRK_STRING_FETCH_255 (string 8114 descriptors must not be fetched using 8115 a 255-byte read); 8116 b = USB_QUIRK_RESET_RESUME (device can't resume 8117 correctly so reset it instead); 8118 c = USB_QUIRK_NO_SET_INTF (device can't handle 8119 Set-Interface requests); 8120 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 8121 handle its Configuration or Interface 8122 strings); 8123 e = USB_QUIRK_RESET (device can't be reset 8124 (e.g morph devices), don't use reset); 8125 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 8126 more interface descriptions than the 8127 bNumInterfaces count, and can't handle 8128 talking to these interfaces); 8129 g = USB_QUIRK_DELAY_INIT (device needs a pause 8130 during initialization, after we read 8131 the device descriptor); 8132 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 8133 high speed and super speed interrupt 8134 endpoints, the USB 2.0 and USB 3.0 spec 8135 require the interval in microframes (1 8136 microframe = 125 microseconds) to be 8137 calculated as interval = 2 ^ 8138 (bInterval-1). 8139 Devices with this quirk report their 8140 bInterval as the result of this 8141 calculation instead of the exponent 8142 variable used in the calculation); 8143 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 8144 handle device_qualifier descriptor 8145 requests); 8146 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 8147 generates spurious wakeup, ignore 8148 remote wakeup capability); 8149 k = USB_QUIRK_NO_LPM (device can't handle Link 8150 Power Management); 8151 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 8152 (Device reports its bInterval as linear 8153 frames instead of the USB 2.0 8154 calculation); 8155 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 8156 to be disconnected before suspend to 8157 prevent spurious wakeup); 8158 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 8159 pause after every control message); 8160 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 8161 delay after resetting its port); 8162 p = USB_QUIRK_SHORT_SET_ADDRESS_REQ_TIMEOUT 8163 (Reduce timeout of the SET_ADDRESS 8164 request from 5000 ms to 500 ms); 8165 q = USB_QUIRK_FORCE_ONE_CONFIG (Device 8166 claims zero configurations, 8167 forcing to 1); 8168 Example: quirks=0781:5580:bk,0a5c:5834:gij 8169 8170 usbhid.mousepoll= 8171 [USBHID] The interval which mice are to be polled at. 8172 8173 usbhid.jspoll= 8174 [USBHID] The interval which joysticks are to be polled at. 8175 8176 usbhid.kbpoll= 8177 [USBHID] The interval which keyboards are to be polled at. 8178 8179 usb-storage.delay_use= 8180 [UMS] The delay in seconds before a new device is 8181 scanned for Logical Units (default 1). 8182 Optionally the delay in milliseconds if the value has 8183 suffix with "ms". 8184 Example: delay_use=2567ms 8185 8186 usb-storage.quirks= 8187 [UMS] A list of quirks entries to supplement or 8188 override the built-in unusual_devs list. List 8189 entries are separated by commas. Each entry has 8190 the form VID:PID:Flags where VID and PID are Vendor 8191 and Product ID values (4-digit hex numbers) and 8192 Flags is a set of characters, each corresponding 8193 to a common usb-storage quirk flag as follows: 8194 a = SANE_SENSE (collect more than 18 bytes 8195 of sense data, not on uas); 8196 b = BAD_SENSE (don't collect more than 18 8197 bytes of sense data, not on uas); 8198 c = FIX_CAPACITY (decrease the reported 8199 device capacity by one sector); 8200 d = NO_READ_DISC_INFO (don't use 8201 READ_DISC_INFO command, not on uas); 8202 e = NO_READ_CAPACITY_16 (don't use 8203 READ_CAPACITY_16 command); 8204 f = NO_REPORT_OPCODES (don't use report opcodes 8205 command, uas only); 8206 g = MAX_SECTORS_240 (don't transfer more than 8207 240 sectors at a time, uas only); 8208 h = CAPACITY_HEURISTICS (decrease the 8209 reported device capacity by one 8210 sector if the number is odd); 8211 i = IGNORE_DEVICE (don't bind to this 8212 device); 8213 j = NO_REPORT_LUNS (don't use report luns 8214 command, uas only); 8215 k = NO_SAME (do not use WRITE_SAME, uas only) 8216 l = NOT_LOCKABLE (don't try to lock and 8217 unlock ejectable media, not on uas); 8218 m = MAX_SECTORS_64 (don't transfer more 8219 than 64 sectors = 32 KB at a time, 8220 not on uas); 8221 n = INITIAL_READ10 (force a retry of the 8222 initial READ(10) command, not on uas); 8223 o = CAPACITY_OK (accept the capacity 8224 reported by the device, not on uas); 8225 p = WRITE_CACHE (the device cache is ON 8226 by default, not on uas); 8227 r = IGNORE_RESIDUE (the device reports 8228 bogus residue values, not on uas); 8229 s = SINGLE_LUN (the device has only one 8230 Logical Unit); 8231 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 8232 commands, uas only); 8233 u = IGNORE_UAS (don't bind to the uas driver); 8234 w = NO_WP_DETECT (don't test whether the 8235 medium is write-protected). 8236 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 8237 even if the device claims no cache, 8238 not on uas) 8239 Example: quirks=0419:aaf5:rl,0421:0433:rc 8240 8241 user_debug= [KNL,ARM] 8242 Format: <int> 8243 See arch/arm/Kconfig.debug help text. 8244 1 - undefined instruction events 8245 2 - system calls 8246 4 - invalid data aborts 8247 8 - SIGSEGV faults 8248 16 - SIGBUS faults 8249 Example: user_debug=31 8250 8251 vdso= [X86,SH,SPARC] 8252 On X86_32, this is an alias for vdso32=. Otherwise: 8253 8254 vdso=1: enable VDSO (the default) 8255 vdso=0: disable VDSO mapping 8256 8257 vdso32= [X86] Control the 32-bit vDSO 8258 vdso32=1: enable 32-bit VDSO 8259 vdso32=0 or vdso32=2: disable 32-bit VDSO 8260 8261 See the help text for CONFIG_COMPAT_VDSO for more 8262 details. If CONFIG_COMPAT_VDSO is set, the default is 8263 vdso32=0; otherwise, the default is vdso32=1. 8264 8265 For compatibility with older kernels, vdso32=2 is an 8266 alias for vdso32=0. 8267 8268 Try vdso32=0 if you encounter an error that says: 8269 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 8270 8271 video= [FB,EARLY] Frame buffer configuration 8272 See Documentation/fb/modedb.rst. 8273 8274 video.brightness_switch_enabled= [ACPI] 8275 Format: [0|1] 8276 If set to 1, on receiving an ACPI notify event 8277 generated by hotkey, video driver will adjust brightness 8278 level and then send out the event to user space through 8279 the allocated input device. If set to 0, video driver 8280 will only send out the event without touching backlight 8281 brightness level. 8282 default: 1 8283 8284 virtio_mmio.device= 8285 [VMMIO] Memory mapped virtio (platform) device. 8286 8287 <size>@<baseaddr>:<irq>[:<id>] 8288 where: 8289 <size> := size (can use standard suffixes 8290 like K, M and G) 8291 <baseaddr> := physical base address 8292 <irq> := interrupt number (as passed to 8293 request_irq()) 8294 <id> := (optional) platform device id 8295 example: 8296 virtio_mmio.device=1K@0x100b0000:48:7 8297 8298 Can be used multiple times for multiple devices. 8299 8300 vga= [BOOT,X86-32] Select a particular video mode 8301 See Documentation/arch/x86/boot.rst and 8302 Documentation/admin-guide/svga.rst. 8303 Use vga=ask for menu. 8304 This is actually a boot loader parameter; the value is 8305 passed to the kernel using a special protocol. 8306 8307 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 8308 May slow down system boot speed, especially when 8309 enabled on systems with a large amount of memory. 8310 All options are enabled by default, and this 8311 interface is meant to allow for selectively 8312 enabling or disabling specific virtual memory 8313 debugging features. 8314 8315 Available options are: 8316 P Enable page structure init time poisoning 8317 - Disable all of the above options 8318 8319 vmalloc=nn[KMG] [KNL,BOOT,EARLY] Forces the vmalloc area to have an 8320 exact size of <nn>. This can be used to increase 8321 the minimum size (128MB on x86, arm32 platforms). 8322 It can also be used to decrease the size and leave more room 8323 for directly mapped kernel RAM. Note that this parameter does 8324 not exist on many other platforms (including arm64, alpha, 8325 loongarch, arc, csky, hexagon, microblaze, mips, nios2, openrisc, 8326 parisc, m64k, powerpc, riscv, sh, um, xtensa, s390, sparc). 8327 8328 vmcp_cma=nn[MG] [KNL,S390,EARLY] 8329 Sets the memory size reserved for contiguous memory 8330 allocations for the vmcp device driver. 8331 8332 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 8333 Format: <command> 8334 8335 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 8336 Format: <command> 8337 8338 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 8339 Format: <command> 8340 8341 vmscape= [X86] Controls mitigation for VMscape attacks. 8342 VMscape attacks can leak information from a userspace 8343 hypervisor to a guest via speculative side-channels. 8344 8345 off - disable the mitigation 8346 ibpb - use Indirect Branch Prediction Barrier 8347 (IBPB) mitigation (default) 8348 force - force vulnerability detection even on 8349 unaffected processors 8350 8351 vsyscall= [X86-64,EARLY] 8352 Controls the behavior of vsyscalls (i.e. calls to 8353 fixed addresses of 0xffffffffff600x00 from legacy 8354 code). Most statically-linked binaries and older 8355 versions of glibc use these calls. Because these 8356 functions are at fixed addresses, they make nice 8357 targets for exploits that can control RIP. 8358 8359 emulate Vsyscalls turn into traps and are emulated 8360 reasonably safely. The vsyscall page is 8361 readable. This disables the Linear 8362 Address Space Separation (LASS) security 8363 feature and makes the system less secure. 8364 8365 xonly [default] Vsyscalls turn into traps and are 8366 emulated reasonably safely. The vsyscall 8367 page is not readable. 8368 8369 none Vsyscalls don't work at all. This makes 8370 them quite hard to use for exploits but 8371 might break your system. 8372 8373 vt.color= [VT] Default text color. 8374 Format: 0xYX, X = foreground, Y = background. 8375 Default: 0x07 = light gray on black. 8376 8377 vt.cur_default= [VT] Default cursor shape. 8378 Format: 0xCCBBAA, where AA, BB, and CC are the same as 8379 the parameters of the <Esc>[?A;B;Cc escape sequence; 8380 see vga-softcursor.rst. Default: 2 = underline. 8381 8382 vt.default_blu= [VT] 8383 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 8384 Change the default blue palette of the console. 8385 This is a 16-member array composed of values 8386 ranging from 0-255. 8387 8388 vt.default_grn= [VT] 8389 Format: <green0>,<green1>,<green2>,...,<green15> 8390 Change the default green palette of the console. 8391 This is a 16-member array composed of values 8392 ranging from 0-255. 8393 8394 vt.default_red= [VT] 8395 Format: <red0>,<red1>,<red2>,...,<red15> 8396 Change the default red palette of the console. 8397 This is a 16-member array composed of values 8398 ranging from 0-255. 8399 8400 vt.default_utf8= 8401 [VT] 8402 Format=<0|1> 8403 Set system-wide default UTF-8 mode for all tty's. 8404 Default is 1, i.e. UTF-8 mode is enabled for all 8405 newly opened terminals. 8406 8407 vt.global_cursor_default= 8408 [VT] 8409 Format=<-1|0|1> 8410 Set system-wide default for whether a cursor 8411 is shown on new VTs. Default is -1, 8412 i.e. cursors will be created by default unless 8413 overridden by individual drivers. 0 will hide 8414 cursors, 1 will display them. 8415 8416 vt.italic= [VT] Default color for italic text; 0-15. 8417 Default: 2 = green. 8418 8419 vt.underline= [VT] Default color for underlined text; 0-15. 8420 Default: 3 = cyan. 8421 8422 watchdog timers [HW,WDT] For information on watchdog timers, 8423 see Documentation/watchdog/watchdog-parameters.rst 8424 or other driver-specific files in the 8425 Documentation/watchdog/ directory. 8426 8427 watchdog_thresh= 8428 [KNL] 8429 Set the hard lockup detector stall duration 8430 threshold in seconds. The soft lockup detector 8431 threshold is set to twice the value. A value of 0 8432 disables both lockup detectors. Default is 10 8433 seconds. 8434 8435 workqueue.unbound_cpus= 8436 [KNL,SMP] Specify to constrain one or some CPUs 8437 to use in unbound workqueues. 8438 Format: <cpu-list> 8439 By default, all online CPUs are available for 8440 unbound workqueues. 8441 8442 workqueue.watchdog_thresh= 8443 If CONFIG_WQ_WATCHDOG is configured, workqueue can 8444 warn stall conditions and dump internal state to 8445 help debugging. 0 disables workqueue stall 8446 detection; otherwise, it's the stall threshold 8447 duration in seconds. The default value is 30 and 8448 it can be updated at runtime by writing to the 8449 corresponding sysfs file. 8450 8451 workqueue.panic_on_stall=<uint> 8452 Panic when workqueue stall is detected by 8453 CONFIG_WQ_WATCHDOG. It sets the number times of the 8454 stall to trigger panic. 8455 8456 The default is set by CONFIG_BOOTPARAM_WQ_STALL_PANIC, 8457 which is 0 (disabled) if not configured. 8458 8459 workqueue.panic_on_stall_time=<uint> 8460 Panic when a workqueue stall has been continuous for 8461 the specified number of seconds. Unlike panic_on_stall 8462 which counts accumulated stall events, this triggers 8463 based on the duration of a single continuous stall. 8464 8465 The default is 0, which disables the time-based panic. 8466 8467 workqueue.cpu_intensive_thresh_us= 8468 Per-cpu work items which run for longer than this 8469 threshold are automatically considered CPU intensive 8470 and excluded from concurrency management to prevent 8471 them from noticeably delaying other per-cpu work 8472 items. Default is 10000 (10ms). 8473 8474 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8475 will report the work functions which violate this 8476 threshold repeatedly. They are likely good 8477 candidates for using WQ_UNBOUND workqueues instead. 8478 8479 workqueue.cpu_intensive_warning_thresh=<uint> 8480 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8481 will report the work functions which violate the 8482 intensive_threshold_us repeatedly. In order to prevent 8483 spurious warnings, start printing only after a work 8484 function has violated this threshold number of times. 8485 8486 The default is 4 times. 0 disables the warning. 8487 8488 workqueue.power_efficient 8489 Per-cpu workqueues are generally preferred because 8490 they show better performance thanks to cache 8491 locality; unfortunately, per-cpu workqueues tend to 8492 be more power hungry than unbound workqueues. 8493 8494 Enabling this makes the per-cpu workqueues which 8495 were observed to contribute significantly to power 8496 consumption unbound, leading to measurably lower 8497 power usage at the cost of small performance 8498 overhead. 8499 8500 The default value of this parameter is determined by 8501 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 8502 8503 workqueue.default_affinity_scope= 8504 Select the default affinity scope to use for unbound 8505 workqueues. Can be one of "cpu", "smt", "cache", 8506 "cache_shard", "numa" and "system". Default is 8507 "cache_shard". For more 8508 information, see the Affinity Scopes section in 8509 Documentation/core-api/workqueue.rst. 8510 8511 This can be changed after boot by writing to the 8512 matching /sys/module/workqueue/parameters file. All 8513 workqueues with the "default" affinity scope will be 8514 updated accordingly. 8515 8516 workqueue.debug_force_rr_cpu 8517 Workqueue used to implicitly guarantee that work 8518 items queued without explicit CPU specified are put 8519 on the local CPU. This guarantee is no longer true 8520 and while local CPU is still preferred work items 8521 may be put on foreign CPUs. This debug option 8522 forces round-robin CPU selection to flush out 8523 usages which depend on the now broken guarantee. 8524 When enabled, memory and cache locality will be 8525 impacted. 8526 8527 writecombine= [LOONGARCH,EARLY] Control the MAT (Memory Access 8528 Type) of ioremap_wc(). 8529 8530 on - Enable writecombine, use WUC for ioremap_wc() 8531 off - Disable writecombine, use SUC for ioremap_wc() 8532 8533 x2apic_phys [X86-64,APIC,EARLY] Use x2apic physical mode instead of 8534 default x2apic cluster mode on platforms 8535 supporting x2apic. 8536 8537 xen_512gb_limit [KNL,X86-64,XEN] 8538 Restricts the kernel running paravirtualized under Xen 8539 to use only up to 512 GB of RAM. The reason to do so is 8540 crash analysis tools and Xen tools for doing domain 8541 save/restore/migration must be enabled to handle larger 8542 domains. 8543 8544 xen_console_io [XEN,EARLY] 8545 Boolean option to enable/disable the usage of the Xen 8546 console_io hypercalls to read and write to the console. 8547 Mostly useful for debugging and development. 8548 8549 xen_emul_unplug= [HW,X86,XEN,EARLY] 8550 Unplug Xen emulated devices 8551 Format: [unplug0,][unplug1] 8552 ide-disks -- unplug primary master IDE devices 8553 aux-ide-disks -- unplug non-primary-master IDE devices 8554 nics -- unplug network devices 8555 all -- unplug all emulated devices (NICs and IDE disks) 8556 unnecessary -- unplugging emulated devices is 8557 unnecessary even if the host did not respond to 8558 the unplug protocol 8559 never -- do not unplug even if version check succeeds 8560 8561 xen_legacy_crash [X86,XEN,EARLY] 8562 Crash from Xen panic notifier, without executing late 8563 panic() code such as dumping handler. 8564 8565 xen_mc_debug [X86,XEN,EARLY] 8566 Enable multicall debugging when running as a Xen PV guest. 8567 Enabling this feature will reduce performance a little 8568 bit, so it should only be enabled for obtaining extended 8569 debug data in case of multicall errors. 8570 8571 xen_msr_safe= [X86,XEN,EARLY] 8572 Format: <bool> 8573 Select whether to always use non-faulting (safe) MSR 8574 access functions when running as Xen PV guest. The 8575 default value is controlled by CONFIG_XEN_PV_MSR_SAFE. 8576 8577 xen_nopv [X86] 8578 Disables the PV optimizations forcing the HVM guest to 8579 run as generic HVM guest with no PV drivers. 8580 This option is obsoleted by the "nopv" option, which 8581 has equivalent effect for XEN platform. 8582 8583 xen_no_vector_callback 8584 [KNL,X86,XEN,EARLY] Disable the vector callback for Xen 8585 event channel interrupts. 8586 8587 xen_scrub_pages= [XEN] 8588 Boolean option to control scrubbing pages before giving them back 8589 to Xen, for use by other domains. Can be also changed at runtime 8590 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 8591 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 8592 8593 xen_timer_slop= [X86-64,XEN,EARLY] 8594 Set the timer slop (in nanoseconds) for the virtual Xen 8595 timers (default is 100000). This adjusts the minimum 8596 delta of virtualized Xen timers, where lower values 8597 improve timer resolution at the expense of processing 8598 more timer interrupts. 8599 8600 xen.balloon_boot_timeout= [XEN] 8601 The time (in seconds) to wait before giving up to boot 8602 in case initial ballooning fails to free enough memory. 8603 Applies only when running as HVM or PVH guest and 8604 started with less memory configured than allowed at 8605 max. Default is 180. 8606 8607 xen.event_eoi_delay= [XEN] 8608 How long to delay EOI handling in case of event 8609 storms (jiffies). Default is 10. 8610 8611 xen.event_loop_timeout= [XEN] 8612 After which time (jiffies) the event handling loop 8613 should start to delay EOI handling. Default is 2. 8614 8615 xen.fifo_events= [XEN] 8616 Boolean parameter to disable using fifo event handling 8617 even if available. Normally fifo event handling is 8618 preferred over the 2-level event handling, as it is 8619 fairer and the number of possible event channels is 8620 much higher. Default is on (use fifo events). 8621 8622 xirc2ps_cs= [NET,PCMCIA] 8623 Format: 8624 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 8625 8626 xive= [PPC] 8627 By default on POWER9 and above, the kernel will 8628 natively use the XIVE interrupt controller. This option 8629 allows the fallback firmware mode to be used: 8630 8631 off Fallback to firmware control of XIVE interrupt 8632 controller on both pseries and powernv 8633 platforms. Only useful on POWER9 and above. 8634 8635 xive.store-eoi=off [PPC] 8636 By default on POWER10 and above, the kernel will use 8637 stores for EOI handling when the XIVE interrupt mode 8638 is active. This option allows the XIVE driver to use 8639 loads instead, as on POWER9. 8640 8641 xhci-hcd.quirks [USB,KNL] 8642 A hex value specifying bitmask with supplemental xhci 8643 host controller quirks. Meaning of each bit can be 8644 consulted in header drivers/usb/host/xhci.h. 8645 8646 xmon [PPC,EARLY] 8647 Format: { early | on | rw | ro | off } 8648 Controls if xmon debugger is enabled. Default is off. 8649 Passing only "xmon" is equivalent to "xmon=early". 8650 early Call xmon as early as possible on boot; xmon 8651 debugger is called from setup_arch(). 8652 on xmon debugger hooks will be installed so xmon 8653 is only called on a kernel crash. Default mode, 8654 i.e. either "ro" or "rw" mode, is controlled 8655 with CONFIG_XMON_DEFAULT_RO_MODE. 8656 rw xmon debugger hooks will be installed so xmon 8657 is called only on a kernel crash, mode is write, 8658 meaning SPR registers, memory and, other data 8659 can be written using xmon commands. 8660 ro same as "rw" option above but SPR registers, 8661 memory, and other data can't be written using 8662 xmon commands. 8663 off xmon is disabled. 8664