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 ISDN Appropriate ISDN support is enabled. 28 ISOL CPU Isolation is enabled. 29 JOY Appropriate joystick support is enabled. 30 KGDB Kernel debugger support is enabled. 31 KVM Kernel Virtual Machine support is enabled. 32 LIBATA Libata driver is enabled 33 LOONGARCH LoongArch architecture is enabled. 34 LOOP Loopback device support is enabled. 35 LP Printer support is enabled. 36 M68k M68k architecture is enabled. 37 These options have more detailed description inside of 38 Documentation/arch/m68k/kernel-options.rst. 39 MDA MDA console support is enabled. 40 MIPS MIPS architecture is enabled. 41 MOUSE Appropriate mouse support is enabled. 42 MSI Message Signaled Interrupts (PCI). 43 MTD MTD (Memory Technology Device) support is enabled. 44 NET Appropriate network support is enabled. 45 NFS Appropriate NFS support is enabled. 46 NUMA NUMA support is enabled. 47 OF Devicetree is enabled. 48 PARISC The PA-RISC architecture is enabled. 49 PCI PCI bus support is enabled. 50 PCIE PCI Express support is enabled. 51 PCMCIA The PCMCIA subsystem is enabled. 52 PNP Plug & Play support is enabled. 53 PPC PowerPC architecture is enabled. 54 PPT Parallel port support is enabled. 55 PS2 Appropriate PS/2 support is enabled. 56 PV_OPS A paravirtualized kernel is enabled. 57 RAM RAM disk support is enabled. 58 RDT Intel Resource Director Technology. 59 RISCV RISCV architecture is enabled. 60 S390 S390 architecture is enabled. 61 SCSI Appropriate SCSI support is enabled. 62 A lot of drivers have their options described inside 63 the Documentation/scsi/ sub-directory. 64 SDW SoundWire support is enabled. 65 SECURITY Different security models are enabled. 66 SELINUX SELinux support is enabled. 67 SERIAL Serial support is enabled. 68 SH SuperH architecture is enabled. 69 SMP The kernel is an SMP kernel. 70 SPARC Sparc architecture is enabled. 71 SUSPEND System suspend states are enabled. 72 SWSUSP Software suspend (hibernation) is enabled. 73 TPM TPM drivers are enabled. 74 UMS USB Mass Storage support is enabled. 75 USB USB support is enabled. 76 NVME NVMe support is enabled 77 USBHID USB Human Interface Device support is enabled. 78 V4L Video For Linux support is enabled. 79 VGA The VGA console has been enabled. 80 VMMIO Driver for memory mapped virtio devices is enabled. 81 VT Virtual terminal support is enabled. 82 WDT Watchdog support is enabled. 83 X86-32 X86-32, aka i386 architecture is enabled. 84 X86-64 X86-64 architecture is enabled. 85 X86 Either 32-bit or 64-bit x86 (same as X86-32+X86-64) 86 X86_UV SGI UV support is enabled. 87 XEN Xen support is enabled 88 XTENSA xtensa architecture is enabled. 89 90In addition, the following text indicates that the option 91 92 BOOT Is a boot loader parameter. 93 BUGS= Relates to possible processor bugs on the said processor. 94 KNL Is a kernel start-up parameter. 95 96 97Kernel parameters 98 99 accept_memory= [MM] 100 Format: { eager | lazy } 101 default: lazy 102 By default, unaccepted memory is accepted lazily to 103 avoid prolonged boot times. The lazy option will add 104 some runtime overhead until all memory is eventually 105 accepted. In most cases the overhead is negligible. 106 For some workloads or for debugging purposes 107 accept_memory=eager can be used to accept all memory 108 at once during boot. 109 110 acpi= [HW,ACPI,X86,ARM64,RISCV64,EARLY] 111 Advanced Configuration and Power Interface 112 Format: { force | on | off | strict | noirq | rsdt | 113 copy_dsdt | nospcr } 114 force -- enable ACPI if default was off 115 on -- enable ACPI but allow fallback to DT [arm64,riscv64] 116 off -- disable ACPI if default was on 117 noirq -- do not use ACPI for IRQ routing 118 strict -- Be less tolerant of platforms that are not 119 strictly ACPI specification compliant. 120 rsdt -- prefer RSDT over (default) XSDT 121 copy_dsdt -- copy DSDT to memory 122 nocmcff -- Disable firmware first mode for corrected 123 errors. This disables parsing the HEST CMC error 124 source to check if firmware has set the FF flag. This 125 may result in duplicate corrected error reports. 126 nospcr -- disable console in ACPI SPCR table as 127 default _serial_ console on ARM64 128 spcr -- enable console in ACPI SPCR table as 129 default _serial_ console on x86 130 For ARM64, ONLY "acpi=off", "acpi=on", "acpi=force" or 131 "acpi=nospcr" are available 132 For RISCV64, ONLY "acpi=off", "acpi=on" or "acpi=force" 133 are available 134 135 See also Documentation/power/runtime_pm.rst, pci=noacpi 136 137 acpi_apic_instance= [ACPI,IOAPIC,EARLY] 138 Format: <int> 139 2: use 2nd APIC table, if available 140 1,0: use 1st APIC table 141 default: 0 142 143 acpi_backlight= [HW,ACPI] 144 { vendor | video | native | none } 145 If set to vendor, prefer vendor-specific driver 146 (e.g. thinkpad_acpi, sony_acpi, etc.) instead 147 of the ACPI video.ko driver. 148 If set to video, use the ACPI video.ko driver. 149 If set to native, use the device's native backlight mode. 150 If set to none, disable the ACPI backlight interface. 151 152 acpi_force_32bit_fadt_addr [ACPI,EARLY] 153 force FADT to use 32 bit addresses rather than the 154 64 bit X_* addresses. Some firmware have broken 64 155 bit addresses for force ACPI ignore these and use 156 the older legacy 32 bit addresses. 157 158 acpica_no_return_repair [HW, ACPI] 159 Disable AML predefined validation mechanism 160 This mechanism can repair the evaluation result to make 161 the return objects more ACPI specification compliant. 162 This option is useful for developers to identify the 163 root cause of an AML interpreter issue when the issue 164 has something to do with the repair mechanism. 165 166 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG] 167 acpi.debug_level= [HW,ACPI,ACPI_DEBUG] 168 Format: <int> 169 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI 170 debug output. Bits in debug_layer correspond to a 171 _COMPONENT in an ACPI source file, e.g., 172 #define _COMPONENT ACPI_EVENTS 173 Bits in debug_level correspond to a level in 174 ACPI_DEBUG_PRINT statements, e.g., 175 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ... 176 The debug_level mask defaults to "info". See 177 Documentation/firmware-guide/acpi/debug.rst for more information about 178 debug layers and levels. 179 180 Enable processor driver info messages: 181 acpi.debug_layer=0x20000000 182 Enable AML "Debug" output, i.e., stores to the Debug 183 object while interpreting AML: 184 acpi.debug_layer=0xffffffff acpi.debug_level=0x2 185 Enable all messages related to ACPI hardware: 186 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff 187 188 Some values produce so much output that the system is 189 unusable. The "log_buf_len" parameter may be useful 190 if you need to capture more output. 191 192 acpi.poweroff_on_fatal= [ACPI] 193 {0 | 1} 194 Causes the system to poweroff when the ACPI bytecode signals 195 a fatal error. The default value of this setting is 1. 196 Overriding this value should only be done for diagnosing 197 ACPI firmware problems, as the system might behave erratically 198 after having encountered a fatal ACPI error. 199 200 acpi_enforce_resources= [ACPI] 201 { strict | lax | no } 202 Check for resource conflicts between native drivers 203 and ACPI OperationRegions (SystemIO and SystemMemory 204 only). IO ports and memory declared in ACPI might be 205 used by the ACPI subsystem in arbitrary AML code and 206 can interfere with legacy drivers. 207 strict (default): access to resources claimed by ACPI 208 is denied; legacy drivers trying to access reserved 209 resources will fail to bind to device using them. 210 lax: access to resources claimed by ACPI is allowed; 211 legacy drivers trying to access reserved resources 212 will bind successfully but a warning message is logged. 213 no: ACPI OperationRegions are not marked as reserved, 214 no further checks are performed. 215 216 acpi_force_table_verification [HW,ACPI,EARLY] 217 Enable table checksum verification during early stage. 218 By default, this is disabled due to x86 early mapping 219 size limitation. 220 221 acpi_irq_balance [HW,ACPI] 222 ACPI will balance active IRQs 223 default in APIC mode 224 225 acpi_irq_nobalance [HW,ACPI] 226 ACPI will not move active IRQs (default) 227 default in PIC mode 228 229 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 230 Format: <irq>,<irq>... 231 232 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 233 use by PCI 234 Format: <irq>,<irq>... 235 236 acpi_mask_gpe= [HW,ACPI] 237 Due to the existence of _Lxx/_Exx, some GPEs triggered 238 by unsupported hardware/firmware features can result in 239 GPE floodings that cannot be automatically disabled by 240 the GPE dispatcher. 241 This facility can be used to prevent such uncontrolled 242 GPE floodings. 243 Format: <byte> or <bitmap-list> 244 245 acpi_no_auto_serialize [HW,ACPI] 246 Disable auto-serialization of AML methods 247 AML control methods that contain the opcodes to create 248 named objects will be marked as "Serialized" by the 249 auto-serialization feature. 250 This feature is enabled by default. 251 This option allows to turn off the feature. 252 253 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 254 kernels. 255 256 acpi_no_static_ssdt [HW,ACPI,EARLY] 257 Disable installation of static SSDTs at early boot time 258 By default, SSDTs contained in the RSDT/XSDT will be 259 installed automatically and they will appear under 260 /sys/firmware/acpi/tables. 261 This option turns off this feature. 262 Note that specifying this option does not affect 263 dynamic table installation which will install SSDT 264 tables to /sys/firmware/acpi/tables/dynamic. 265 266 acpi_no_watchdog [HW,ACPI,WDT] 267 Ignore the ACPI-based watchdog interface (WDAT) and let 268 a native driver control the watchdog device instead. 269 270 acpi_rsdp= [ACPI,EFI,KEXEC,EARLY] 271 Pass the RSDP address to the kernel, mostly used 272 on machines running EFI runtime service to boot the 273 second kernel for kdump. 274 275 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 276 Format: To spoof as Windows 98: ="Microsoft Windows" 277 278 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 279 of 2 which is mandated by ACPI 6) as the supported ACPI 280 specification revision (when using this switch, it may 281 be necessary to carry out a cold reboot _twice_ in a 282 row to make it take effect on the platform firmware). 283 284 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 285 acpi_osi="string1" # add string1 286 acpi_osi="!string2" # remove string2 287 acpi_osi=!* # remove all strings 288 acpi_osi=! # disable all built-in OS vendor 289 strings 290 acpi_osi=!! # enable all built-in OS vendor 291 strings 292 acpi_osi= # disable all strings 293 294 'acpi_osi=!' can be used in combination with single or 295 multiple 'acpi_osi="string1"' to support specific OS 296 vendor string(s). Note that such command can only 297 affect the default state of the OS vendor strings, thus 298 it cannot affect the default state of the feature group 299 strings and the current state of the OS vendor strings, 300 specifying it multiple times through kernel command line 301 is meaningless. This command is useful when one do not 302 care about the state of the feature group strings which 303 should be controlled by the OSPM. 304 Examples: 305 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 306 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 307 can make '_OSI("Windows 2000")' TRUE. 308 309 'acpi_osi=' cannot be used in combination with other 310 'acpi_osi=' command lines, the _OSI method will not 311 exist in the ACPI namespace. NOTE that such command can 312 only affect the _OSI support state, thus specifying it 313 multiple times through kernel command line is also 314 meaningless. 315 Examples: 316 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 317 FALSE. 318 319 'acpi_osi=!*' can be used in combination with single or 320 multiple 'acpi_osi="string1"' to support specific 321 string(s). Note that such command can affect the 322 current state of both the OS vendor strings and the 323 feature group strings, thus specifying it multiple times 324 through kernel command line is meaningful. But it may 325 still not able to affect the final state of a string if 326 there are quirks related to this string. This command 327 is useful when one want to control the state of the 328 feature group strings to debug BIOS issues related to 329 the OSPM features. 330 Examples: 331 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 332 '_OSI("Module Device")' FALSE. 333 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 334 '_OSI("Module Device")' TRUE. 335 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 336 equivalent to 337 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 338 and 339 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 340 they all will make '_OSI("Windows 2000")' TRUE. 341 342 acpi_pm_good [X86] 343 Override the pmtimer bug detection: force the kernel 344 to assume that this machine's pmtimer latches its value 345 and always returns good values. 346 347 acpi_sci= [HW,ACPI,EARLY] ACPI System Control Interrupt trigger mode 348 Format: { level | edge | high | low } 349 350 acpi_skip_timer_override [HW,ACPI,EARLY] 351 Recognize and ignore IRQ0/pin2 Interrupt Override. 352 For broken nForce2 BIOS resulting in XT-PIC timer. 353 354 acpi_sleep= [HW,ACPI] Sleep options 355 Format: { s3_bios, s3_mode, s3_beep, s4_hwsig, 356 s4_nohwsig, old_ordering, nonvs, 357 sci_force_enable, nobl } 358 See Documentation/power/video.rst for information on 359 s3_bios and s3_mode. 360 s3_beep is for debugging; it makes the PC's speaker beep 361 as soon as the kernel's real-mode entry point is called. 362 s4_hwsig causes the kernel to check the ACPI hardware 363 signature during resume from hibernation, and gracefully 364 refuse to resume if it has changed. This complies with 365 the ACPI specification but not with reality, since 366 Windows does not do this and many laptops do change it 367 on docking. So the default behaviour is to allow resume 368 and simply warn when the signature changes, unless the 369 s4_hwsig option is enabled. 370 s4_nohwsig prevents ACPI hardware signature from being 371 used (or even warned about) during resume. 372 old_ordering causes the ACPI 1.0 ordering of the _PTS 373 control method, with respect to putting devices into 374 low power states, to be enforced (the ACPI 2.0 ordering 375 of _PTS is used by default). 376 nonvs prevents the kernel from saving/restoring the 377 ACPI NVS memory during suspend/hibernation and resume. 378 sci_force_enable causes the kernel to set SCI_EN directly 379 on resume from S1/S3 (which is against the ACPI spec, 380 but some broken systems don't work without it). 381 nobl causes the internal blacklist of systems known to 382 behave incorrectly in some ways with respect to system 383 suspend and resume to be ignored (use wisely). 384 385 acpi_use_timer_override [HW,ACPI,EARLY] 386 Use timer override. For some broken Nvidia NF5 boards 387 that require a timer override, but don't have HPET 388 389 add_efi_memmap [EFI,X86,EARLY] Include EFI memory map in 390 kernel's map of available physical RAM. 391 392 agp= [AGP] 393 { off | try_unsupported } 394 off: disable AGP support 395 try_unsupported: try to drive unsupported chipsets 396 (may crash computer or cause data corruption) 397 398 ALSA [HW,ALSA] 399 See Documentation/sound/alsa-configuration.rst 400 401 alignment= [KNL,ARM] 402 Allow the default userspace alignment fault handler 403 behaviour to be specified. Bit 0 enables warnings, 404 bit 1 enables fixups, and bit 2 sends a segfault. 405 406 align_va_addr= [X86-64] 407 Align virtual addresses by clearing slice [14:12] when 408 allocating a VMA at process creation time. This option 409 gives you up to 3% performance improvement on AMD F15h 410 machines (where it is enabled by default) for a 411 CPU-intensive style benchmark, and it can vary highly in 412 a microbenchmark depending on workload and compiler. 413 414 32: only for 32-bit processes 415 64: only for 64-bit processes 416 on: enable for both 32- and 64-bit processes 417 off: disable for both 32- and 64-bit processes 418 419 alloc_snapshot [FTRACE] 420 Allocate the ftrace snapshot buffer on boot up when the 421 main buffer is allocated. This is handy if debugging 422 and you need to use tracing_snapshot() on boot up, and 423 do not want to use tracing_snapshot_alloc() as it needs 424 to be done where GFP_KERNEL allocations are allowed. 425 426 allow_mismatched_32bit_el0 [ARM64,EARLY] 427 Allow execve() of 32-bit applications and setting of the 428 PER_LINUX32 personality on systems where only a strict 429 subset of the CPUs support 32-bit EL0. When this 430 parameter is present, the set of CPUs supporting 32-bit 431 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0 432 and hot-unplug operations may be restricted. 433 434 See Documentation/arch/arm64/asymmetric-32bit.rst for more 435 information. 436 437 amd_iommu= [HW,X86-64] 438 Pass parameters to the AMD IOMMU driver in the system. 439 Possible values are: 440 fullflush - Deprecated, equivalent to iommu.strict=1 441 off - do not initialize any AMD IOMMU found in 442 the system 443 force_isolation - Force device isolation for all 444 devices. The IOMMU driver is not 445 allowed anymore to lift isolation 446 requirements as needed. This option 447 does not override iommu=pt 448 force_enable - Force enable the IOMMU on platforms known 449 to be buggy with IOMMU enabled. Use this 450 option with care. 451 pgtbl_v1 - Use v1 page table for DMA-API (Default). 452 pgtbl_v2 - Use v2 page table for DMA-API. 453 irtcachedis - Disable Interrupt Remapping Table (IRT) caching. 454 nohugepages - Limit page-sizes used for v1 page-tables 455 to 4 KiB. 456 v2_pgsizes_only - Limit page-sizes used for v1 page-tables 457 to 4KiB/2Mib/1GiB. 458 459 460 amd_iommu_dump= [HW,X86-64] 461 Enable AMD IOMMU driver option to dump the ACPI table 462 for AMD IOMMU. With this option enabled, AMD IOMMU 463 driver will print ACPI tables for AMD IOMMU during 464 IOMMU initialization. 465 466 amd_iommu_intr= [HW,X86-64] 467 Specifies one of the following AMD IOMMU interrupt 468 remapping modes: 469 legacy - Use legacy interrupt remapping mode. 470 vapic - Use virtual APIC mode, which allows IOMMU 471 to inject interrupts directly into guest. 472 This mode requires kvm-amd.avic=1. 473 (Default when IOMMU HW support is present.) 474 475 amd_pstate= [X86,EARLY] 476 disable 477 Do not enable amd_pstate as the default 478 scaling driver for the supported processors 479 passive 480 Use amd_pstate with passive mode as a scaling driver. 481 In this mode autonomous selection is disabled. 482 Driver requests a desired performance level and platform 483 tries to match the same performance level if it is 484 satisfied by guaranteed performance level. 485 active 486 Use amd_pstate_epp driver instance as the scaling driver, 487 driver provides a hint to the hardware if software wants 488 to bias toward performance (0x0) or energy efficiency (0xff) 489 to the CPPC firmware. then CPPC power algorithm will 490 calculate the runtime workload and adjust the realtime cores 491 frequency. 492 guided 493 Activate guided autonomous mode. Driver requests minimum and 494 maximum performance level and the platform autonomously 495 selects a performance level in this range and appropriate 496 to the current workload. 497 498 amd_prefcore= 499 [X86] 500 disable 501 Disable amd-pstate preferred core. 502 503 amd_dynamic_epp= 504 [X86] 505 disable 506 Disable amd-pstate dynamic EPP. 507 enable 508 Enable amd-pstate dynamic EPP. 509 510 amijoy.map= [HW,JOY] Amiga joystick support 511 Map of devices attached to JOY0DAT and JOY1DAT 512 Format: <a>,<b> 513 See also Documentation/input/joydev/joystick.rst 514 515 analog.map= [HW,JOY] Analog joystick and gamepad support 516 Specifies type or capabilities of an analog joystick 517 connected to one of 16 gameports 518 Format: <type1>,<type2>,..<type16> 519 520 apc= [HW,SPARC] 521 Power management functions (SPARCstation-4/5 + deriv.) 522 Format: noidle 523 Disable APC CPU standby support. SPARCstation-Fox does 524 not play well with APC CPU idle - disable it if you have 525 APC and your system crashes randomly. 526 527 apic [APIC,X86-64] Use IO-APIC. Default. 528 529 apic= [APIC,X86,EARLY] Advanced Programmable Interrupt Controller 530 Change the output verbosity while booting 531 Format: { quiet (default) | verbose | debug } 532 Change the amount of debugging information output 533 when initialising the APIC and IO-APIC components. 534 535 apic_extnmi= [APIC,X86,EARLY] External NMI delivery setting 536 Format: { bsp (default) | all | none } 537 bsp: External NMI is delivered only to CPU 0 538 all: External NMIs are broadcast to all CPUs as a 539 backup of CPU 0 540 none: External NMI is masked for all CPUs. This is 541 useful so that a dump capture kernel won't be 542 shot down by NMI 543 544 apicpmtimer Do APIC timer calibration using the pmtimer. Implies 545 apicmaintimer. Useful when your PIT timer is totally 546 broken. 547 548 autoconf= [IPV6] 549 See Documentation/networking/ipv6.rst. 550 551 apm= [APM] Advanced Power Management 552 See header of arch/x86/kernel/apm_32.c. 553 554 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 555 Format: { "0" | "1" } 556 See security/apparmor/Kconfig help text 557 0 -- disable. 558 1 -- enable. 559 Default value is set via kernel config option. 560 561 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards 562 Format: <io>,<irq>,<nodeID> 563 564 arm64.no32bit_el0 [ARM64] Unconditionally disable the execution of 565 32 bit applications. 566 567 arm64.nobti [ARM64] Unconditionally disable Branch Target 568 Identification support 569 570 arm64.nogcs [ARM64] Unconditionally disable Guarded Control Stack 571 support 572 573 arm64.nomops [ARM64] Unconditionally disable Memory Copy and Memory 574 Set instructions support 575 576 arm64.nompam [ARM64] Unconditionally disable Memory Partitioning And 577 Monitoring support 578 579 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension 580 support 581 582 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 583 support 584 585 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix 586 Extension support 587 588 arm64.nosve [ARM64] Unconditionally disable Scalable Vector 589 Extension support 590 591 ataflop= [HW,M68k] 592 593 atarimouse= [HW,MOUSE] Atari Mouse 594 595 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 596 EzKey and similar keyboards 597 598 atkbd.reset= [HW] Reset keyboard during initialization 599 600 atkbd.set= [HW] Select keyboard code set 601 Format: <int> (2 = AT (default), 3 = PS/2) 602 603 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 604 keyboards 605 606 atkbd.softraw= [HW] Choose between synthetic and real raw mode 607 Format: <bool> (0 = real, 1 = synthetic (default)) 608 609 atkbd.softrepeat= [HW] 610 Use software keyboard repeat 611 612 audit= [KNL] Enable the audit sub-system 613 Format: { "0" | "1" | "off" | "on" } 614 0 | off - kernel audit is disabled and can not be 615 enabled until the next reboot 616 unset - kernel audit is initialized but disabled and 617 will be fully enabled by the userspace auditd. 618 1 | on - kernel audit is initialized and partially 619 enabled, storing at most audit_backlog_limit 620 messages in RAM until it is fully enabled by the 621 userspace auditd. 622 Default: unset 623 624 audit_backlog_limit= [KNL] Set the audit queue size limit. 625 Format: <int> (must be >=0) 626 Default: 64 627 628 bau= [X86_UV] Enable the BAU on SGI UV. The default 629 behavior is to disable the BAU (i.e. bau=0). 630 Format: { "0" | "1" } 631 0 - Disable the BAU. 632 1 - Enable the BAU. 633 unset - Disable the BAU. 634 635 bdev_allow_write_mounted= 636 Format: <bool> 637 Control the ability to open a mounted block device 638 for writing, i.e., allow / disallow writes that bypass 639 the FS. This was implemented as a means to prevent 640 fuzzers from crashing the kernel by overwriting the 641 metadata underneath a mounted FS without its awareness. 642 This also prevents destructive formatting of mounted 643 filesystems by naive storage tooling that don't use 644 O_EXCL. Default is Y and can be changed through the 645 Kconfig option CONFIG_BLK_DEV_WRITE_MOUNTED. 646 647 bert_disable [ACPI] 648 Disable BERT OS support on buggy BIOSes. 649 650 bgrt_disable [ACPI,X86,EARLY] 651 Disable BGRT to avoid flickering OEM logo. 652 653 blkdevparts= Manual partition parsing of block device(s) for 654 embedded devices based on command line input. 655 See Documentation/block/cmdline-partition.rst 656 657 boot_delay= [KNL,EARLY] 658 Milliseconds to delay each printk during boot. 659 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled, 660 and you may also have to specify "lpj=". Boot_delay 661 values larger than 10 seconds (10000) are assumed 662 erroneous and ignored. 663 Format: integer 664 665 bootconfig [KNL,EARLY] 666 Extended command line options can be added to an initrd 667 and this will cause the kernel to look for it. 668 669 See Documentation/admin-guide/bootconfig.rst 670 671 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 672 bttv.radio= Most important insmod options are available as 673 kernel args too. 674 bttv.pll= See Documentation/admin-guide/media/bttv.rst 675 bttv.tuner= 676 677 bulk_remove=off [PPC] This parameter disables the use of the pSeries 678 firmware feature for flushing multiple hpte entries 679 at a time. 680 681 c101= [NET] Moxa C101 synchronous serial card 682 683 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 684 Sometimes CPU hardware bugs make them report the cache 685 size incorrectly. The kernel will attempt work arounds 686 to fix known problems, but for some CPUs it is not 687 possible to determine what the correct size should be. 688 This option provides an override for these situations. 689 690 carrier_timeout= 691 [NET] Specifies amount of time (in seconds) that 692 the kernel should wait for a network carrier. By default 693 it waits 120 seconds. 694 695 ca_keys= [KEYS] This parameter identifies a specific key(s) on 696 the system trusted keyring to be used for certificate 697 trust validation. 698 format: { id:<keyid> | builtin } 699 700 cca= [MIPS,EARLY] Override the kernel pages' cache coherency 701 algorithm. Accepted values range from 0 to 7 702 inclusive. See arch/mips/include/asm/pgtable-bits.h 703 for platform specific values (SB1, Loongson3 and 704 others). 705 706 ccw_timeout_log [S390] 707 See Documentation/arch/s390/common_io.rst for details. 708 709 cfi= [X86-64] Set Control Flow Integrity checking features 710 when CONFIG_FINEIBT is enabled. 711 Format: feature[,feature...] 712 Default: auto 713 714 auto: Use FineIBT if IBT available, otherwise kCFI. 715 Under FineIBT, enable "paranoid" mode when 716 FRED is not available. 717 off: Turn off CFI checking. 718 kcfi: Use kCFI (disable FineIBT). 719 fineibt: Use FineIBT (even if IBT not available). 720 norand: Do not re-randomize CFI hashes. 721 paranoid: Add caller hash checking under FineIBT. 722 bhi: Enable register poisoning to stop speculation 723 across FineIBT. (Disabled by default.) 724 warn: Do not enforce CFI checking: warn only. 725 debug: Report CFI initialization details. 726 727 cgroup_disable= [KNL] Disable a particular controller or optional feature 728 Format: {name of the controller(s) or feature(s) to disable} 729 The effects of cgroup_disable=foo are: 730 - foo isn't auto-mounted if you mount all cgroups in 731 a single hierarchy 732 - foo isn't visible as an individually mountable 733 subsystem 734 - if foo is an optional feature then the feature is 735 disabled and corresponding cgroup files are not 736 created 737 {Currently only "memory" controller deal with this and 738 cut the overhead, others just disable the usage. So 739 only cgroup_disable=memory is actually worthy} 740 Specifying "pressure" disables per-cgroup pressure 741 stall information accounting feature 742 743 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 744 Format: { { controller | "all" | "named" } 745 [,{ controller | "all" | "named" }...] } 746 Like cgroup_disable, but only applies to cgroup v1; 747 the blacklisted controllers remain available in cgroup2. 748 "all" blacklists all controllers and "named" disables 749 named mounts. Specifying both "all" and "named" disables 750 all v1 hierarchies. 751 752 cgroup_v1_proc= [KNL] Show also missing controllers in /proc/cgroups 753 Format: { "true" | "false" } 754 /proc/cgroups lists only v1 controllers by default. 755 This compatibility option enables listing also v2 756 controllers (whose v1 code is not compiled!), so that 757 semi-legacy software can check this file to decide 758 about usage of v2 (sic) controllers. 759 760 cgroup_favordynmods= [KNL] Enable or Disable favordynmods. 761 Format: { "true" | "false" } 762 Defaults to the value of CONFIG_CGROUP_FAVOR_DYNMODS. 763 764 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 765 Format: <string> 766 nosocket -- Disable socket memory accounting. 767 nokmem -- Disable kernel memory accounting. 768 nobpf -- Disable BPF memory accounting. 769 770 check_pages= [MM,EARLY] Enable sanity checking of pages after 771 allocations / before freeing. This adds checks to catch 772 double-frees, use-after-frees, and other sources of 773 page corruption by inspecting page internals (flags, 774 mapcount/refcount, memcg_data, etc.). 775 Format: { "0" | "1" } 776 Default: 0 (1 if CONFIG_DEBUG_VM is set) 777 778 checkreqprot= [SELINUX] Set initial checkreqprot flag value. 779 Format: { "0" | "1" } 780 See security/selinux/Kconfig help text. 781 0 -- check protection applied by kernel (includes 782 any implied execute protection). 783 1 -- check protection requested by application. 784 Default value is set via a kernel config option. 785 Value can be changed at runtime via 786 /sys/fs/selinux/checkreqprot. 787 Setting checkreqprot to 1 is deprecated. 788 789 cio_ignore= [S390] 790 See Documentation/arch/s390/common_io.rst for details. 791 792 clk_ignore_unused 793 [CLK] 794 Prevents the clock framework from automatically gating 795 clocks that have not been explicitly enabled by a Linux 796 device driver but are enabled in hardware at reset or 797 by the bootloader/firmware. Note that this does not 798 force such clocks to be always-on nor does it reserve 799 those clocks in any way. This parameter is useful for 800 debug and development, but should not be needed on a 801 platform with proper driver support. For more 802 information, see Documentation/driver-api/clk.rst. 803 804 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 805 [Deprecated] 806 Forces specified clocksource (if available) to be used 807 when calculating gettimeofday(). If specified 808 clocksource is not available, it defaults to PIT. 809 Format: { pit | tsc | cyclone | pmtmr } 810 811 clocksource= Override the default clocksource 812 Format: <string> 813 Override the default clocksource and use the clocksource 814 with the name specified. 815 Some clocksource names to choose from, depending on 816 the platform: 817 [all] jiffies (this is the base, fallback clocksource) 818 [ACPI] acpi_pm 819 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 820 pxa_timer,timer3,32k_counter,timer0_1 821 [X86-32] pit,hpet,tsc; 822 scx200_hrt on Geode; cyclone on IBM x440 823 [MIPS] MIPS 824 [PARISC] cr16 825 [S390] tod 826 [SH] SuperH 827 [SPARC64] tick 828 [X86-64] hpet,tsc 829 830 clocksource.arm_arch_timer.evtstrm= 831 [ARM,ARM64,EARLY] 832 Format: <bool> 833 Enable/disable the eventstream feature of the ARM 834 architected timer so that code using WFE-based polling 835 loops can be debugged more effectively on production 836 systems. 837 838 clocksource.verify_n_cpus= [KNL] 839 Limit the number of CPUs checked for clocksources 840 marked with CLOCK_SOURCE_VERIFY_PERCPU that 841 are marked unstable due to excessive skew. 842 A negative value says to check all CPUs, while 843 zero says not to check any. Values larger than 844 nr_cpu_ids are silently truncated to nr_cpu_ids. 845 The actual CPUs are chosen randomly, with 846 no replacement if the same CPU is chosen twice. 847 848 clocksource-wdtest.holdoff= [KNL] 849 Set the time in seconds that the clocksource 850 watchdog test waits before commencing its tests. 851 Defaults to zero when built as a module and to 852 10 seconds when built into the kernel. 853 854 cma=nn[MG]@[start[MG][-end[MG]]] 855 [KNL,CMA,EARLY] 856 Sets the size of kernel global memory area for 857 contiguous memory allocations and optionally the 858 placement constraint by the physical address range of 859 memory allocations. A value of 0 disables CMA 860 altogether. For more information, see 861 kernel/dma/contiguous.c 862 863 cma_pernuma=nn[MG] 864 [KNL,CMA,EARLY] 865 Sets the size of kernel per-numa memory area for 866 contiguous memory allocations. A value of 0 disables 867 per-numa CMA altogether. And If this option is not 868 specified, the default value is 0. 869 With per-numa CMA enabled, DMA users on node nid will 870 first try to allocate buffer from the pernuma area 871 which is located in node nid, if the allocation fails, 872 they will fallback to the global default memory area. 873 874 numa_cma=<node>:nn[MG][,<node>:nn[MG]] 875 [KNL,CMA,EARLY] 876 Sets the size of kernel numa memory area for 877 contiguous memory allocations. It will reserve CMA 878 area for the specified node. 879 880 With numa CMA enabled, DMA users on node nid will 881 first try to allocate buffer from the numa area 882 which is located in node nid, if the allocation fails, 883 they will fallback to the global default memory area. 884 885 cmo_free_hint= [PPC] Format: { yes | no } 886 Specify whether pages are marked as being inactive 887 when they are freed. This is used in CMO environments 888 to determine OS memory pressure for page stealing by 889 a hypervisor. 890 Default: yes 891 892 coherent_pool=nn[KMG] [ARM,KNL,EARLY] 893 Sets the size of memory pool for coherent, atomic dma 894 allocations, by default set to 256K. 895 896 com20020= [HW,NET] ARCnet - COM20020 chipset 897 Format: 898 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 899 900 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 901 Format: <io>[,<irq>] 902 903 com90xx= [HW,NET] 904 ARCnet - COM90xx chipset (memory-mapped buffers) 905 Format: <io>[,<irq>[,<memstart>]] 906 907 condev= [HW,S390] console device 908 conmode= 909 910 con3215_drop= [S390,EARLY] 3215 console drop mode. 911 Format: y|n|Y|N|1|0 912 When set to true, drop data on the 3215 console when 913 the console buffer is full. In this case the 914 operator using a 3270 terminal emulator (for example 915 x3270) does not have to enter the clear key for the 916 console output to advance and the kernel to continue. 917 This leads to a much faster boot time when a 3270 918 terminal emulator is active. If no 3270 terminal 919 emulator is used, this parameter has no effect. 920 921 console= [KNL] Output console device and options. 922 923 tty<n> Use the virtual console device <n>. 924 925 ttyS<n>[,options] 926 ttyUSB0[,options] 927 Use the specified serial port. The options are of 928 the form "bbbbpnf", where "bbbb" is the baud rate, 929 "p" is parity ("n", "o", or "e"), "n" is number of 930 bits, and "f" is flow control ("r" for RTS or 931 omit it). Default is "9600n8". 932 933 See Documentation/admin-guide/serial-console.rst for more 934 information. See 935 Documentation/networking/netconsole.rst for an 936 alternative. 937 938 <DEVNAME>:<n>.<n>[,options] 939 Use the specified serial port on the serial core bus. 940 The addressing uses DEVNAME of the physical serial port 941 device, followed by the serial core controller instance, 942 and the serial port instance. The options are the same 943 as documented for the ttyS addressing above. 944 945 The mapping of the serial ports to the tty instances 946 can be viewed with: 947 948 $ ls -d /sys/bus/serial-base/devices/*:*.*/tty/* 949 /sys/bus/serial-base/devices/00:04:0.0/tty/ttyS0 950 951 In the above example, the console can be addressed with 952 console=00:04:0.0. Note that a console addressed this 953 way will only get added when the related device driver 954 is ready. The use of an earlycon parameter in addition to 955 the console may be desired for console output early on. 956 957 uart[8250],io,<addr>[,options] 958 uart[8250],mmio,<addr>[,options] 959 uart[8250],mmio16,<addr>[,options] 960 uart[8250],mmio32,<addr>[,options] 961 uart[8250],0x<addr>[,options] 962 Start an early, polled-mode console on the 8250/16550 963 UART at the specified I/O port or MMIO address, 964 switching to the matching ttyS device later. 965 MMIO inter-register address stride is either 8-bit 966 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 967 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 968 to be equivalent to 'mmio'. 'options' are specified in 969 the same format described for ttyS above; if unspecified, 970 the h/w is not re-initialized. 971 972 hvc<n> Use the hypervisor console device <n>. This is for 973 both Xen and PowerPC hypervisors. 974 975 { null | "" } 976 Use to disable console output, i.e., to have kernel 977 console messages discarded. 978 This must be the only console= parameter used on the 979 kernel command line. 980 981 If the device connected to the port is not a TTY but a braille 982 device, prepend "brl," before the device type, for instance 983 console=brl,ttyS0 984 For now, only VisioBraille is supported. 985 986 console_msg_format= 987 [KNL] Change console messages format 988 default 989 By default we print messages on consoles in 990 "[time stamp] text\n" format (time stamp may not be 991 printed, depending on CONFIG_PRINTK_TIME or 992 `printk_time' param). 993 syslog 994 Switch to syslog format: "<%u>[time stamp] text\n" 995 IOW, each message will have a facility and loglevel 996 prefix. The format is similar to one used by syslog() 997 syscall, or to executing "dmesg -S --raw" or to reading 998 from /proc/kmsg. 999 1000 consoleblank= [KNL] The console blank (screen saver) timeout in 1001 seconds. A value of 0 disables the blank timer. 1002 Defaults to 0. 1003 1004 coredump_filter= 1005 [KNL] Change the default value for 1006 /proc/<pid>/coredump_filter. 1007 See also Documentation/filesystems/proc.rst. 1008 1009 coresight_cpu_debug.enable 1010 [ARM,ARM64] 1011 Format: <bool> 1012 Enable/disable the CPU sampling based debugging. 1013 0: default value, disable debugging 1014 1: enable debugging at boot time 1015 1016 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 1017 Format: 1018 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 1019 1020 cpuidle.off=1 [CPU_IDLE] 1021 disable the cpuidle sub-system 1022 1023 cpuidle.governor= 1024 [CPU_IDLE] Name of the cpuidle governor to use. 1025 1026 cpufreq.off=1 [CPU_FREQ] 1027 disable the cpufreq sub-system 1028 1029 cpufreq.default_governor= 1030 [CPU_FREQ] Name of the default cpufreq governor or 1031 policy to use. This governor must be registered in the 1032 kernel before the cpufreq driver probes. 1033 1034 cpu_init_udelay=N 1035 [X86,EARLY] Delay for N microsec between assert and de-assert 1036 of APIC INIT to start processors. This delay occurs 1037 on every CPU online, such as boot, and resume from suspend. 1038 Default: 10000 1039 1040 cpuhp.parallel= 1041 [SMP] Enable/disable parallel bringup of secondary CPUs 1042 Format: <bool> 1043 Default is enabled if CONFIG_HOTPLUG_PARALLEL=y. Otherwise 1044 the parameter has no effect. 1045 1046 crash_kexec_post_notifiers 1047 Only jump to kdump kernel after running the panic 1048 notifiers and dumping kmsg. This option increases 1049 the risks of a kdump failure, since some panic 1050 notifiers can make the crashed kernel more unstable. 1051 In configurations where kdump may not be reliable, 1052 running the panic notifiers could allow collecting 1053 more data on dmesg, like stack traces from other CPUS 1054 or extra data dumped by panic_print. Note that some 1055 configurations enable this option unconditionally, 1056 like Hyper-V, PowerPC (fadump) and AMD SEV-SNP. 1057 1058 crashkernel=size[KMG][@offset[KMG]] 1059 [KNL,EARLY] Using kexec, Linux can switch to a 'crash kernel' 1060 upon panic. This parameter reserves the physical 1061 memory region [offset, offset + size] for that kernel 1062 image. If '@offset' is omitted, then a suitable offset 1063 is selected automatically. 1064 [KNL, X86-64, ARM64, RISCV, LoongArch] Select a region 1065 under 4G first, and fall back to reserve region above 1066 4G when '@offset' hasn't been specified. 1067 See Documentation/admin-guide/kdump/kdump.rst for further details. 1068 1069 crashkernel=range1:size1[,range2:size2,...][@offset] 1070 [KNL] Same as above, but depends on the memory 1071 in the running system. The syntax of range is 1072 start-[end] where start and end are both 1073 a memory unit (amount[KMG]). See also 1074 Documentation/admin-guide/kdump/kdump.rst for an example. 1075 1076 crashkernel=size[KMG],high 1077 [KNL, X86-64, ARM64, RISCV, LoongArch] range could be 1078 above 4G. 1079 Allow kernel to allocate physical memory region from top, 1080 so could be above 4G if system have more than 4G ram 1081 installed. Otherwise memory region will be allocated 1082 below 4G, if available. 1083 It will be ignored if crashkernel=X is specified. 1084 crashkernel=size[KMG],low 1085 [KNL, X86-64, ARM64, RISCV, LoongArch] range under 4G. 1086 When crashkernel=X,high is passed, kernel could allocate 1087 physical memory region above 4G, that cause second kernel 1088 crash on system that require some amount of low memory, 1089 e.g. swiotlb requires at least 64M+32K low memory, also 1090 enough extra low memory is needed to make sure DMA buffers 1091 for 32-bit devices won't run out. Kernel would try to allocate 1092 default size of memory below 4G automatically. The default 1093 size is platform dependent. 1094 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB) 1095 --> arm64: 128MiB 1096 --> riscv: 128MiB 1097 --> loongarch: 128MiB 1098 This one lets the user specify own low range under 4G 1099 for second kernel instead. 1100 0: to disable low allocation. 1101 It will be ignored when crashkernel=X,high is not used 1102 or memory reserved is below 4G. 1103 crashkernel=size[KMG],cma 1104 [KNL, X86, ppc] Reserve additional crash kernel memory from 1105 CMA. This reservation is usable by the first system's 1106 userspace memory and kernel movable allocations (memory 1107 balloon, zswap). Pages allocated from this memory range 1108 will not be included in the vmcore so this should not 1109 be used if dumping of userspace memory is intended and 1110 it has to be expected that some movable kernel pages 1111 may be missing from the dump. 1112 1113 A standard crashkernel reservation, as described above, 1114 is still needed to hold the crash kernel and initrd. 1115 1116 This option increases the risk of a kdump failure: DMA 1117 transfers configured by the first kernel may end up 1118 corrupting the second kernel's memory. 1119 1120 This reservation method is intended for systems that 1121 can't afford to sacrifice enough memory for standard 1122 crashkernel reservation and where less reliable and 1123 possibly incomplete kdump is preferable to no kdump at 1124 all. 1125 1126 cryptomgr.notests 1127 [KNL] Disable crypto self-tests 1128 1129 cs89x0_dma= [HW,NET] 1130 Format: <dma> 1131 1132 cs89x0_media= [HW,NET] 1133 Format: { rj45 | aui | bnc } 1134 1135 csdlock_debug= [KNL] Enable or disable debug add-ons of cross-CPU 1136 function call handling. When switched on, 1137 additional debug data is printed to the console 1138 in case a hanging CPU is detected, and that 1139 CPU is pinged again in order to try to resolve 1140 the hang situation. The default value of this 1141 option depends on the CSD_LOCK_WAIT_DEBUG_DEFAULT 1142 Kconfig option. 1143 1144 dasd= [HW,NET] 1145 See header of drivers/s390/block/dasd_devmap.c. 1146 1147 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 1148 (one device per port) 1149 Format: <port#>,<type> 1150 See also Documentation/input/devices/joystick-parport.rst 1151 1152 debug [KNL,EARLY] Enable kernel debugging (events log level). 1153 1154 debug_boot_weak_hash 1155 [KNL,EARLY] Enable printing [hashed] pointers early in the 1156 boot sequence. If enabled, we use a weak hash instead 1157 of siphash to hash pointers. Use this option if you are 1158 seeing instances of '(___ptrval___)') and need to see a 1159 value (hashed pointer) instead. Cryptographically 1160 insecure, please do not use on production kernels. 1161 1162 debug_locks_verbose= 1163 [KNL] verbose locking self-tests 1164 Format: <int> 1165 Print debugging info while doing the locking API 1166 self-tests. 1167 Bitmask for the various LOCKTYPE_ tests. Defaults to 0 1168 (no extra messages), setting it to -1 (all bits set) 1169 will print _a_lot_ more information - normally only 1170 useful to lockdep developers. 1171 1172 debug_objects [KNL,EARLY] Enable object debugging 1173 1174 debug_guardpage_minorder= 1175 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this 1176 parameter allows control of the order of pages that will 1177 be intentionally kept free (and hence protected) by the 1178 buddy allocator. Bigger value increase the probability 1179 of catching random memory corruption, but reduce the 1180 amount of memory for normal system use. The maximum 1181 possible value is MAX_PAGE_ORDER/2. Setting this 1182 parameter to 1 or 2 should be enough to identify most 1183 random memory corruption problems caused by bugs in 1184 kernel or driver code when a CPU writes to (or reads 1185 from) a random memory location. Note that there exists 1186 a class of memory corruptions problems caused by buggy 1187 H/W or F/W or by drivers badly programming DMA 1188 (basically when memory is written at bus level and the 1189 CPU MMU is bypassed) which are not detectable by 1190 CONFIG_DEBUG_PAGEALLOC, hence this option will not 1191 help tracking down these problems. 1192 1193 debug_pagealloc= 1194 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 1195 enables the feature at boot time. By default, it is 1196 disabled and the system will work mostly the same as a 1197 kernel built without CONFIG_DEBUG_PAGEALLOC. 1198 Note: to get most of debug_pagealloc error reports, it's 1199 useful to also enable the page_owner functionality. 1200 on: enable the feature 1201 1202 debugfs= [KNL,EARLY] This parameter enables what is exposed to 1203 userspace and debugfs internal clients. 1204 Format: { on, off } 1205 on: All functions are enabled. 1206 off: Filesystem is not registered and clients 1207 get a -EPERM as result when trying to register files 1208 or directories within debugfs. 1209 This is equivalent of the runtime functionality if 1210 debugfs was not enabled in the kernel at all. 1211 Default value is set in build-time with a kernel configuration. 1212 1213 debugpat [X86] Enable PAT debugging 1214 1215 default_hugepagesz= 1216 [HW] The size of the default HugeTLB page. This is 1217 the size represented by the legacy /proc/ hugepages 1218 APIs. In addition, this is the default hugetlb size 1219 used for shmget(), mmap() and mounting hugetlbfs 1220 filesystems. If not specified, defaults to the 1221 architecture's default huge page size. Huge page 1222 sizes are architecture dependent. See also 1223 Documentation/admin-guide/mm/hugetlbpage.rst. 1224 Format: size[KMG] 1225 1226 deferred_probe_timeout= 1227 [KNL] Debugging option to set a timeout in seconds for 1228 deferred probe to give up waiting on dependencies to 1229 probe. Only specific dependencies (subsystems or 1230 drivers) that have opted in will be ignored. A timeout 1231 of 0 will timeout at the end of initcalls. If the time 1232 out hasn't expired, it'll be restarted by each 1233 successful driver registration. This option will also 1234 dump out devices still on the deferred probe list after 1235 retrying. 1236 1237 delayacct [KNL] Enable per-task delay accounting 1238 1239 dell_smm_hwmon.ignore_dmi= 1240 [HW] Continue probing hardware even if DMI data 1241 indicates that the driver is running on unsupported 1242 hardware. 1243 1244 dell_smm_hwmon.force= 1245 [HW] Activate driver even if SMM BIOS signature does 1246 not match list of supported models and enable otherwise 1247 blacklisted features. 1248 1249 dell_smm_hwmon.power_status= 1250 [HW] Report power status in /proc/i8k 1251 (disabled by default). 1252 1253 dell_smm_hwmon.restricted= 1254 [HW] Allow controlling fans only if SYS_ADMIN 1255 capability is set. 1256 1257 dell_smm_hwmon.fan_mult= 1258 [HW] Factor to multiply fan speed with. 1259 1260 dell_smm_hwmon.fan_max= 1261 [HW] Maximum configurable fan speed. 1262 1263 dfltcc= [HW,S390] 1264 Format: { on | off | def_only | inf_only | always } 1265 on: s390 zlib hardware support for compression on 1266 level 1 and decompression (default) 1267 off: No s390 zlib hardware support 1268 def_only: s390 zlib hardware support for deflate 1269 only (compression on level 1) 1270 inf_only: s390 zlib hardware support for inflate 1271 only (decompression) 1272 always: Same as 'on' but ignores the selected compression 1273 level always using hardware support (used for debugging) 1274 1275 dhash_entries= [KNL] 1276 Set number of hash buckets for dentry cache. 1277 1278 disable_1tb_segments [PPC,EARLY] 1279 Disables the use of 1TB hash page table segments. This 1280 causes the kernel to fall back to 256MB segments which 1281 can be useful when debugging issues that require an SLB 1282 miss to occur. 1283 1284 disable= [IPV6] 1285 See Documentation/networking/ipv6.rst. 1286 1287 disable_radix [PPC,EARLY] 1288 Disable RADIX MMU mode on POWER9 1289 1290 disable_tlbie [PPC] 1291 Disable TLBIE instruction. Currently does not work 1292 with KVM, with HASH MMU, or with coherent accelerators. 1293 1294 disable_ddw [PPC/PSERIES,EARLY] 1295 Disable Dynamic DMA Window support. Use this 1296 to workaround buggy firmware. 1297 1298 disable_ipv6= [IPV6] 1299 See Documentation/networking/ipv6.rst. 1300 1301 disable_mtrr_cleanup [X86,EARLY] 1302 The kernel tries to adjust MTRR layout from continuous 1303 to discrete, to make X server driver able to add WB 1304 entry later. This parameter disables that. 1305 1306 disable_mtrr_trim [X86, Intel and AMD only,EARLY] 1307 By default the kernel will trim any uncacheable 1308 memory out of your available memory pool based on 1309 MTRR settings. This parameter disables that behavior, 1310 possibly causing your machine to run very slowly. 1311 1312 disable_timer_pin_1 [X86,EARLY] 1313 Disable PIN 1 of APIC timer 1314 Can be useful to work around chipset bugs. 1315 1316 dis_ucode_ldr [X86] Disable the microcode loader. 1317 1318 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 1319 this option disables the debugging code at boot. 1320 1321 dma_debug_entries=<number> 1322 This option allows to tune the number of preallocated 1323 entries for DMA-API debugging code. One entry is 1324 required per DMA-API allocation. Use this if the 1325 DMA-API debugging code disables itself because the 1326 architectural default is too low. 1327 1328 dma_debug_driver=<driver_name> 1329 With this option the DMA-API debugging driver 1330 filter feature can be enabled at boot time. Just 1331 pass the driver to filter for as the parameter. 1332 The filter can be disabled or changed to another 1333 driver later using sysfs. 1334 1335 reg_file_data_sampling= 1336 [X86] Controls mitigation for Register File Data 1337 Sampling (RFDS) vulnerability. RFDS is a CPU 1338 vulnerability which may allow userspace to infer 1339 kernel data values previously stored in floating point 1340 registers, vector registers, or integer registers. 1341 RFDS only affects Intel Atom processors. 1342 1343 on: Turns ON the mitigation. 1344 off: Turns OFF the mitigation. 1345 1346 This parameter overrides the compile time default set 1347 by CONFIG_MITIGATION_RFDS. Mitigation cannot be 1348 disabled when other VERW based mitigations (like MDS) 1349 are enabled. In order to disable RFDS mitigation all 1350 VERW based mitigations need to be disabled. 1351 1352 For details see: 1353 Documentation/admin-guide/hw-vuln/reg-file-data-sampling.rst 1354 1355 dm_verity.keyring_unsealed= 1356 [KNL] When set to 1, leave the dm-verity keyring 1357 unsealed after initialization so userspace can 1358 provision keys. Once the keyring is restricted 1359 it becomes active and is searched during signature 1360 verification. 1361 1362 driver_async_probe= [KNL] 1363 List of driver names to be probed asynchronously. * 1364 matches with all driver names. If * is specified, the 1365 rest of the listed driver names are those that will NOT 1366 match the *. 1367 Format: <driver_name1>,<driver_name2>... 1368 1369 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1370 Broken monitors, graphic adapters, KVMs and EDIDless 1371 panels may send no or incorrect EDID data sets. 1372 This parameter allows to specify an EDID data sets 1373 in the /lib/firmware directory that are used instead. 1374 An EDID data set will only be used for a particular 1375 connector, if its name and a colon are prepended to 1376 the EDID name. Each connector may use a unique EDID 1377 data set by separating the files with a comma. An EDID 1378 data set with no connector name will be used for 1379 any connectors not explicitly specified. 1380 1381 dscc4.setup= [NET] 1382 1383 dt_cpu_ftrs= [PPC,EARLY] 1384 Format: {"off" | "known"} 1385 Control how the dt_cpu_ftrs device-tree binding is 1386 used for CPU feature discovery and setup (if it 1387 exists). 1388 off: Do not use it, fall back to legacy cpu table. 1389 known: Do not pass through unknown features to guests 1390 or userspace, only those that the kernel is aware of. 1391 1392 dump_apple_properties [X86] 1393 Dump name and content of EFI device properties on 1394 x86 Macs. Useful for driver authors to determine 1395 what data is available or for reverse-engineering. 1396 1397 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1398 <module>.dyndbg[="val"] 1399 Enable debug messages at boot time. See 1400 Documentation/admin-guide/dynamic-debug-howto.rst 1401 for details. 1402 1403 early_ioremap_debug [KNL,EARLY] 1404 Enable debug messages in early_ioremap support. This 1405 is useful for tracking down temporary early mappings 1406 which are not unmapped. 1407 1408 earlycon= [KNL,EARLY] Output early console device and options. 1409 1410 When used with no options, the early console is 1411 determined by stdout-path property in device tree's 1412 chosen node or the ACPI SPCR table if supported by 1413 the platform. 1414 1415 cdns,<addr>[,options] 1416 Start an early, polled-mode console on a Cadence 1417 (xuartps) serial port at the specified address. Only 1418 supported option is baud rate. If baud rate is not 1419 specified, the serial port must already be setup and 1420 configured. 1421 1422 uart[8250],io,<addr>[,options[,uartclk]] 1423 uart[8250],mmio,<addr>[,options[,uartclk]] 1424 uart[8250],mmio32,<addr>[,options[,uartclk]] 1425 uart[8250],mmio32be,<addr>[,options[,uartclk]] 1426 uart[8250],0x<addr>[,options] 1427 Start an early, polled-mode console on the 8250/16550 1428 UART at the specified I/O port or MMIO address. 1429 MMIO inter-register address stride is either 8-bit 1430 (mmio) or 32-bit (mmio32 or mmio32be). 1431 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1432 to be equivalent to 'mmio'. 'options' are specified 1433 in the same format described for "console=ttyS<n>"; if 1434 unspecified, the h/w is not initialized. 'uartclk' is 1435 the uart clock frequency; if unspecified, it is set 1436 to 'BASE_BAUD' * 16. 1437 1438 pl011,<addr> 1439 pl011,mmio32,<addr> 1440 Start an early, polled-mode console on a pl011 serial 1441 port at the specified address. The pl011 serial port 1442 must already be setup and configured. Options are not 1443 yet supported. If 'mmio32' is specified, then only 1444 the driver will use only 32-bit accessors to read/write 1445 the device registers. 1446 1447 liteuart,<addr> 1448 Start an early console on a litex serial port at the 1449 specified address. The serial port must already be 1450 setup and configured. Options are not yet supported. 1451 1452 meson,<addr> 1453 Start an early, polled-mode console on a meson serial 1454 port at the specified address. The serial port must 1455 already be setup and configured. Options are not yet 1456 supported. 1457 1458 msm_serial,<addr> 1459 Start an early, polled-mode console on an msm serial 1460 port at the specified address. The serial port 1461 must already be setup and configured. Options are not 1462 yet supported. 1463 1464 msm_serial_dm,<addr> 1465 Start an early, polled-mode console on an msm serial 1466 dm port at the specified address. The serial port 1467 must already be setup and configured. Options are not 1468 yet supported. 1469 1470 owl,<addr> 1471 Start an early, polled-mode console on a serial port 1472 of an Actions Semi SoC, such as S500 or S900, at the 1473 specified address. The serial port must already be 1474 setup and configured. Options are not yet supported. 1475 1476 rda,<addr> 1477 Start an early, polled-mode console on a serial port 1478 of an RDA Micro SoC, such as RDA8810PL, at the 1479 specified address. The serial port must already be 1480 setup and configured. Options are not yet supported. 1481 1482 sbi 1483 Use RISC-V SBI (Supervisor Binary Interface) for early 1484 console. 1485 1486 smh Use ARM semihosting calls for early console. 1487 1488 s3c2410,<addr> 1489 s3c2412,<addr> 1490 s3c2440,<addr> 1491 s3c6400,<addr> 1492 s5pv210,<addr> 1493 exynos4210,<addr> 1494 Use early console provided by serial driver available 1495 on Samsung SoCs, requires selecting proper type and 1496 a correct base address of the selected UART port. The 1497 serial port must already be setup and configured. 1498 Options are not yet supported. 1499 1500 lantiq,<addr> 1501 Start an early, polled-mode console on a lantiq serial 1502 (lqasc) port at the specified address. The serial port 1503 must already be setup and configured. Options are not 1504 yet supported. 1505 1506 lpuart,<addr> 1507 lpuart32,<addr> 1508 Use early console provided by Freescale LP UART driver 1509 found on Freescale Vybrid and QorIQ LS1021A processors. 1510 A valid base address must be provided, and the serial 1511 port must already be setup and configured. 1512 1513 ec_imx21,<addr> 1514 ec_imx6q,<addr> 1515 Start an early, polled-mode, output-only console on the 1516 Freescale i.MX UART at the specified address. The UART 1517 must already be setup and configured. 1518 1519 ar3700_uart,<addr> 1520 Start an early, polled-mode console on the 1521 Armada 3700 serial port at the specified 1522 address. The serial port must already be setup 1523 and configured. Options are not yet supported. 1524 1525 qcom_geni,<addr> 1526 Start an early, polled-mode console on a Qualcomm 1527 Generic Interface (GENI) based serial port at the 1528 specified address. The serial port must already be 1529 setup and configured. Options are not yet supported. 1530 1531 efifb,[options] 1532 Start an early, unaccelerated console on the EFI 1533 memory mapped framebuffer (if available). On cache 1534 coherent non-x86 systems that use system memory for 1535 the framebuffer, pass the 'ram' option so that it is 1536 mapped with the correct attributes. 1537 1538 linflex,<addr> 1539 Use early console provided by Freescale LINFlexD UART 1540 serial driver for NXP S32V234 SoCs. A valid base 1541 address must be provided, and the serial port must 1542 already be setup and configured. 1543 1544 earlyprintk= [X86,SH,ARM,M68k,S390,UM,EARLY] 1545 earlyprintk=vga 1546 earlyprintk=sclp 1547 earlyprintk=xen 1548 earlyprintk=serial[,ttySn[,baudrate]] 1549 earlyprintk=serial[,0x...[,baudrate]] 1550 earlyprintk=ttySn[,baudrate] 1551 earlyprintk=dbgp[debugController#] 1552 earlyprintk=mmio32,membase[,{nocfg|baudrate}] 1553 earlyprintk=pciserial[,force],bus:device.function[,{nocfg|baudrate}] 1554 earlyprintk=xdbc[xhciController#] 1555 earlyprintk=bios 1556 1557 earlyprintk is useful when the kernel crashes before 1558 the normal console is initialized. It is not enabled by 1559 default because it has some cosmetic problems. 1560 1561 Use "nocfg" to skip UART configuration, assume 1562 BIOS/firmware has configured UART correctly. 1563 1564 Append ",keep" to not disable it when the real console 1565 takes over. 1566 1567 Only one of vga, serial, or usb debug port can 1568 be used at a time. 1569 1570 Currently only ttyS0 and ttyS1 may be specified by 1571 name. Other I/O ports may be explicitly specified 1572 on some architectures (x86 and arm at least) by 1573 replacing ttySn with an I/O port address, like this: 1574 earlyprintk=serial,0x1008,115200 1575 You can find the port for a given device in 1576 /proc/tty/driver/serial: 1577 2: uart:ST16650V2 port:00001008 irq:18 ... 1578 1579 Interaction with the standard serial driver is not 1580 very good. 1581 1582 The VGA output is eventually overwritten by 1583 the real console. 1584 1585 The xen option can only be used in Xen domains. 1586 1587 The sclp output can only be used on s390. 1588 1589 The bios output can only be used on SuperH. 1590 1591 The optional "force" to "pciserial" enables use of a 1592 PCI device even when its classcode is not of the 1593 UART class. 1594 1595 edac_report= [HW,EDAC] Control how to report EDAC event 1596 Format: {"on" | "off" | "force"} 1597 on: enable EDAC to report H/W event. May be overridden 1598 by other higher priority error reporting module. 1599 off: disable H/W event reporting through EDAC. 1600 force: enforce the use of EDAC to report H/W event. 1601 default: on. 1602 1603 edd= [EDD] 1604 Format: {"off" | "on" | "skip[mbr]"} 1605 1606 efi= [EFI,EARLY] 1607 Format: { "debug", "disable_early_pci_dma", 1608 "nochunk", "noruntime", "nosoftreserve", 1609 "novamap", "no_disable_early_pci_dma" } 1610 debug: enable misc debug output. 1611 disable_early_pci_dma: disable the busmaster bit on all 1612 PCI bridges while in the EFI boot stub. 1613 nochunk: disable reading files in "chunks" in the EFI 1614 boot stub, as chunking can cause problems with some 1615 firmware implementations. 1616 noruntime : disable EFI runtime services support 1617 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1618 attribute may cause the kernel to reserve the 1619 memory range for a memory mapping driver to 1620 claim. Specify efi=nosoftreserve to disable this 1621 reservation and treat the memory by its base type 1622 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1623 novamap: do not call SetVirtualAddressMap(). 1624 no_disable_early_pci_dma: Leave the busmaster bit set 1625 on all PCI bridges while in the EFI boot stub 1626 1627 efi_no_storage_paranoia [EFI,X86,EARLY] 1628 Using this parameter you can use more than 50% of 1629 your efi variable storage. Use this parameter only if 1630 you are really sure that your UEFI does sane gc and 1631 fulfills the spec otherwise your board may brick. 1632 1633 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1634 that is to be dynamically loaded by Linux. If there are 1635 multiple variables with the same name but with different 1636 vendor GUIDs, all of them will be loaded. See 1637 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1638 1639 1640 eisa_irq_edge= [PARISC,HW] 1641 See header of drivers/parisc/eisa.c. 1642 1643 ekgdboc= [X86,KGDB,EARLY] Allow early kernel console debugging 1644 Format: ekgdboc=kbd 1645 1646 This is designed to be used in conjunction with 1647 the boot argument: earlyprintk=vga 1648 1649 This parameter works in place of the kgdboc parameter 1650 but can only be used if the backing tty is available 1651 very early in the boot process. For early debugging 1652 via a serial port see kgdboc_earlycon instead. 1653 1654 elanfreq= [X86-32] 1655 See comment before function elanfreq_setup() in 1656 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1657 1658 elfcorehdr=[size[KMG]@]offset[KMG] [PPC,SH,X86,S390,EARLY] 1659 Specifies physical address of start of kernel core 1660 image elf header and optionally the size. Generally 1661 kexec loader will pass this option to capture kernel. 1662 See Documentation/admin-guide/kdump/kdump.rst for details. 1663 1664 enable_mtrr_cleanup [X86,EARLY] 1665 The kernel tries to adjust MTRR layout from continuous 1666 to discrete, to make X server driver able to add WB 1667 entry later. This parameter enables that. 1668 1669 enable_timer_pin_1 [X86] 1670 Enable PIN 1 of APIC timer 1671 Can be useful to work around chipset bugs 1672 (in particular on some ATI chipsets). 1673 The kernel tries to set a reasonable default. 1674 1675 enforcing= [SELINUX] Set initial enforcing status. 1676 Format: {"0" | "1"} 1677 See security/selinux/Kconfig help text. 1678 0 -- permissive (log only, no denials). 1679 1 -- enforcing (deny and log). 1680 Default value is 0. 1681 Value can be changed at runtime via 1682 /sys/fs/selinux/enforce. 1683 1684 erst_disable [ACPI] 1685 Disable Error Record Serialization Table (ERST) 1686 support. 1687 1688 ether= [HW,NET] Ethernet cards parameters 1689 This option is obsoleted by the "netdev=" option, which 1690 has equivalent usage. See its documentation for details. 1691 1692 evm= [EVM] 1693 Format: { "fix" } 1694 Permit 'security.evm' to be updated regardless of 1695 current integrity status. 1696 1697 early_page_ext [KNL,EARLY] Enforces page_ext initialization to earlier 1698 stages so cover more early boot allocations. 1699 Please note that as side effect some optimizations 1700 might be disabled to achieve that (e.g. parallelized 1701 memory initialization is disabled) so the boot process 1702 might take longer, especially on systems with a lot of 1703 memory. Available with CONFIG_PAGE_EXTENSION=y. 1704 1705 failslab= 1706 fail_usercopy= 1707 fail_page_alloc= 1708 fail_skb_realloc= 1709 fail_make_request=[KNL] 1710 General fault injection mechanism. 1711 Format: <interval>,<probability>,<space>,<times> 1712 See also Documentation/fault-injection/. 1713 1714 fb_tunnels= [NET] 1715 Format: { initns | none } 1716 See Documentation/admin-guide/sysctl/net.rst for 1717 fb_tunnels_only_for_init_ns 1718 1719 floppy= [HW] 1720 See Documentation/admin-guide/blockdev/floppy.rst. 1721 1722 forcepae [X86-32] 1723 Forcefully enable Physical Address Extension (PAE). 1724 Many Pentium M systems disable PAE but may have a 1725 functionally usable PAE implementation. 1726 Warning: use of this parameter will taint the kernel 1727 and may cause unknown problems. 1728 1729 fred= [X86-64] 1730 Enable/disable Flexible Return and Event Delivery. 1731 Format: { on | off } 1732 on: enable FRED when it's present, the default setting. 1733 off: disable FRED. 1734 1735 ftrace=[tracer] 1736 [FTRACE] will set and start the specified tracer 1737 as early as possible in order to facilitate early 1738 boot debugging. 1739 1740 ftrace_boot_snapshot 1741 [FTRACE] On boot up, a snapshot will be taken of the 1742 ftrace ring buffer that can be read at: 1743 /sys/kernel/tracing/snapshot. 1744 This is useful if you need tracing information from kernel 1745 boot up that is likely to be overridden by user space 1746 start up functionality. 1747 1748 Optionally, the snapshot can also be defined for a tracing 1749 instance that was created by the trace_instance= command 1750 line parameter. 1751 1752 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo 1753 1754 The above will cause the "foo" tracing instance to trigger 1755 a snapshot at the end of boot up. 1756 1757 ftrace_dump_on_oops[=2(orig_cpu) | =<instance>][,<instance> | 1758 ,<instance>=2(orig_cpu)] 1759 [FTRACE] will dump the trace buffers on oops. 1760 If no parameter is passed, ftrace will dump global 1761 buffers of all CPUs, if you pass 2 or orig_cpu, it 1762 will dump only the buffer of the CPU that triggered 1763 the oops, or the specific instance will be dumped if 1764 its name is passed. Multiple instance dump is also 1765 supported, and instances are separated by commas. Each 1766 instance supports only dump on CPU that triggered the 1767 oops by passing 2 or orig_cpu to it. 1768 1769 ftrace_dump_on_oops=foo=orig_cpu 1770 1771 The above will dump only the buffer of "foo" instance 1772 on CPU that triggered the oops. 1773 1774 ftrace_dump_on_oops,foo,bar=orig_cpu 1775 1776 The above will dump global buffer on all CPUs, the 1777 buffer of "foo" instance on all CPUs and the buffer 1778 of "bar" instance on CPU that triggered the oops. 1779 1780 ftrace_filter=[function-list] 1781 [FTRACE] Limit the functions traced by the function 1782 tracer at boot up. function-list is a comma-separated 1783 list of functions. This list can be changed at run 1784 time by the set_ftrace_filter file in the debugfs 1785 tracing directory. 1786 1787 ftrace_notrace=[function-list] 1788 [FTRACE] Do not trace the functions specified in 1789 function-list. This list can be changed at run time 1790 by the set_ftrace_notrace file in the debugfs 1791 tracing directory. 1792 1793 ftrace_graph_filter=[function-list] 1794 [FTRACE] Limit the top level callers functions traced 1795 by the function graph tracer at boot up. 1796 function-list is a comma-separated list of functions 1797 that can be changed at run time by the 1798 set_graph_function file in the debugfs tracing directory. 1799 1800 ftrace_graph_notrace=[function-list] 1801 [FTRACE] Do not trace from the functions specified in 1802 function-list. This list is a comma-separated list of 1803 functions that can be changed at run time by the 1804 set_graph_notrace file in the debugfs tracing directory. 1805 1806 ftrace_graph_max_depth=<uint> 1807 [FTRACE] Used with the function graph tracer. This is 1808 the max depth it will trace into a function. This value 1809 can be changed at run time by the max_graph_depth file 1810 in the tracefs tracing directory. default: 0 (no limit) 1811 1812 fw_devlink= [KNL,EARLY] Create device links between consumer and supplier 1813 devices by scanning the firmware to infer the 1814 consumer/supplier relationships. This feature is 1815 especially useful when drivers are loaded as modules as 1816 it ensures proper ordering of tasks like device probing 1817 (suppliers first, then consumers), supplier boot state 1818 clean up (only after all consumers have probed), 1819 suspend/resume & runtime PM (consumers first, then 1820 suppliers). 1821 Format: { off | permissive | on | rpm } 1822 off -- Don't create device links from firmware info. 1823 permissive -- Create device links from firmware info 1824 but use it only for ordering boot state clean 1825 up (sync_state() calls). 1826 on -- Create device links from firmware info and use it 1827 to enforce probe and suspend/resume ordering. 1828 rpm -- Like "on", but also use to order runtime PM. 1829 1830 fw_devlink.strict=<bool> 1831 [KNL,EARLY] Treat all inferred dependencies as mandatory 1832 dependencies. This only applies for fw_devlink=on|rpm. 1833 Format: <bool> 1834 1835 fw_devlink.sync_state = 1836 [KNL,EARLY] When all devices that could probe have finished 1837 probing, this parameter controls what to do with 1838 devices that haven't yet received their sync_state() 1839 calls. 1840 Format: { strict | timeout } 1841 strict -- Default. Continue waiting on consumers to 1842 probe successfully. 1843 timeout -- Give up waiting on consumers and call 1844 sync_state() on any devices that haven't yet 1845 received their sync_state() calls after 1846 deferred_probe_timeout has expired or by 1847 late_initcall() if !CONFIG_MODULES. 1848 1849 gamecon.map[2|3]= 1850 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1851 support via parallel port (up to 5 devices per port) 1852 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1853 See also Documentation/input/devices/joystick-parport.rst 1854 1855 gamma= [HW,DRM] 1856 1857 gart_fix_e820= [X86-64,EARLY] disable the fix e820 for K8 GART 1858 Format: off | on 1859 default: on 1860 1861 gather_data_sampling= 1862 [X86,INTEL,EARLY] Control the Gather Data Sampling (GDS) 1863 mitigation. 1864 1865 Gather Data Sampling is a hardware vulnerability which 1866 allows unprivileged speculative access to data which was 1867 previously stored in vector registers. 1868 1869 This issue is mitigated by default in updated microcode. 1870 The mitigation may have a performance impact but can be 1871 disabled. On systems without the microcode mitigation 1872 disabling AVX serves as a mitigation. 1873 1874 force: Disable AVX to mitigate systems without 1875 microcode mitigation. No effect if the microcode 1876 mitigation is present. Known to cause crashes in 1877 userspace with buggy AVX enumeration. 1878 1879 off: Disable GDS mitigation. 1880 1881 gbpages [X86] Use GB pages for kernel direct mappings. 1882 1883 gcov_persist= [GCOV] When non-zero (default), profiling data for 1884 kernel modules is saved and remains accessible via 1885 debugfs, even when the module is unloaded/reloaded. 1886 When zero, profiling data is discarded and associated 1887 debugfs files are removed at module unload time. 1888 1889 goldfish [X86] Enable the goldfish android emulator platform. 1890 Don't use this when you are not running on the 1891 android emulator 1892 1893 gpio-mockup.gpio_mockup_ranges 1894 [HW] Sets the ranges of gpiochip of for this device. 1895 Format: <start1>,<end1>,<start2>,<end2>... 1896 gpio-mockup.gpio_mockup_named_lines 1897 [HW] Let the driver know GPIO lines should be named. 1898 1899 gpt [EFI] Forces disk with valid GPT signature but 1900 invalid Protective MBR to be treated as GPT. If the 1901 primary GPT is corrupted, it enables the backup/alternate 1902 GPT to be used instead. 1903 1904 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1905 the "Enable 0" bit of the configuration register. 1906 Format: 0 | 1 1907 Default: 0 1908 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1909 the "Enable 0" bit of the configuration register. 1910 Format: 0 | 1 1911 Default: 0 1912 grcan.select= [HW] Select which physical interface to use. 1913 Format: 0 | 1 1914 Default: 0 1915 grcan.txsize= [HW] Sets the size of the tx buffer. 1916 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1917 Default: 1024 1918 grcan.rxsize= [HW] Sets the size of the rx buffer. 1919 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1920 Default: 1024 1921 1922 hardened_usercopy= 1923 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 1924 hardening is enabled for this boot. Hardened 1925 usercopy checking is used to protect the kernel 1926 from reading or writing beyond known memory 1927 allocation boundaries as a proactive defense 1928 against bounds-checking flaws in the kernel's 1929 copy_to_user()/copy_from_user() interface. 1930 The default is determined by 1931 CONFIG_HARDENED_USERCOPY_DEFAULT_ON. 1932 on Perform hardened usercopy checks. 1933 off Disable hardened usercopy checks. 1934 1935 hardlockup_all_cpu_backtrace= 1936 [KNL] Should the hard-lockup detector generate 1937 backtraces on all cpus. 1938 Format: 0 | 1 1939 1940 hash_pointers= 1941 [KNL,EARLY] 1942 By default, when pointers are printed to the console 1943 or buffers via the %p format string, that pointer is 1944 "hashed", i.e. obscured by hashing the pointer value. 1945 This is a security feature that hides actual kernel 1946 addresses from unprivileged users, but it also makes 1947 debugging the kernel more difficult since unequal 1948 pointers can no longer be compared. The choices are: 1949 Format: { auto | always | never } 1950 Default: auto 1951 1952 auto - Hash pointers unless slab_debug is enabled. 1953 always - Always hash pointers (even if slab_debug is 1954 enabled). 1955 never - Never hash pointers. This option should only 1956 be specified when debugging the kernel. Do 1957 not use on production kernels. The boot 1958 param "no_hash_pointers" is an alias for 1959 this mode. 1960 1961 For controlling hashing dynamically at runtime, 1962 use the "kernel.kptr_restrict" sysctl instead. 1963 1964 hashdist= [KNL,NUMA] Large hashes allocated during boot 1965 are distributed across NUMA nodes. Defaults on 1966 for 64-bit NUMA, off otherwise. 1967 Format: 0 | 1 (for off | on) 1968 1969 hd= [EIDE] (E)IDE hard drive subsystem geometry 1970 Format: <cyl>,<head>,<sect> 1971 1972 hest_disable [ACPI] 1973 Disable Hardware Error Source Table (HEST) support; 1974 corresponding firmware-first mode error processing 1975 logic will be disabled. 1976 1977 hibernate= [HIBERNATION] 1978 noresume Don't check if there's a hibernation image 1979 present during boot. 1980 nocompress Don't compress/decompress hibernation images. 1981 no Disable hibernation and resume. 1982 protect_image Turn on image protection during restoration 1983 (that will set all pages holding image data 1984 during restoration read-only). 1985 1986 hibernate.compressor= [HIBERNATION] Compression algorithm to be 1987 used with hibernation. 1988 Format: { lzo | lz4 } 1989 Default: lzo 1990 1991 lzo: Select LZO compression algorithm to 1992 compress/decompress hibernation image. 1993 1994 lz4: Select LZ4 compression algorithm to 1995 compress/decompress hibernation image. 1996 1997 hibernate.pm_test_delay= 1998 [HIBERNATION] 1999 Sets the number of seconds to remain in a hibernation test 2000 mode before resuming the system (see 2001 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 2002 is set. Default value is 5. 2003 2004 hibernate_compression_threads= 2005 [HIBERNATION] 2006 Set the number of threads used for compressing or decompressing 2007 hibernation images. 2008 2009 Format: <integer> 2010 Default: 3 2011 Minimum: 1 2012 Example: hibernate_compression_threads=4 2013 2014 highmem=nn[KMG] [KNL,BOOT,EARLY] forces the highmem zone to have an exact 2015 size of <nn>. This works even on boxes that have no 2016 highmem otherwise. This also works to reduce highmem 2017 size on bigger boxes. 2018 2019 highres= [KNL] Enable/disable high resolution timer mode. 2020 Valid parameters: "on", "off" 2021 Default: "on" 2022 2023 hlt [BUGS=ARM,SH] 2024 2025 hostname= [KNL,EARLY] Set the hostname (aka UTS nodename). 2026 Format: <string> 2027 This allows setting the system's hostname during early 2028 startup. This sets the name returned by gethostname. 2029 Using this parameter to set the hostname makes it 2030 possible to ensure the hostname is correctly set before 2031 any userspace processes run, avoiding the possibility 2032 that a process may call gethostname before the hostname 2033 has been explicitly set, resulting in the calling 2034 process getting an incorrect result. The string must 2035 not exceed the maximum allowed hostname length (usually 2036 64 characters) and will be truncated otherwise. 2037 2038 hpet= [X86-32,HPET] option to control HPET usage 2039 Format: { enable (default) | disable | force | 2040 verbose } 2041 disable: disable HPET and use PIT instead 2042 force: allow force enabled of undocumented chips (ICH4, 2043 VIA, nVidia) 2044 verbose: show contents of HPET registers during setup 2045 2046 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 2047 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 2048 2049 hugepages= [HW,EARLY] Number of HugeTLB pages to allocate at boot. 2050 If this follows hugepagesz (below), it specifies 2051 the number of pages of hugepagesz to be allocated. 2052 If this is the first HugeTLB parameter on the command 2053 line, it specifies the number of pages to allocate for 2054 the default huge page size. If using node format, the 2055 number of pages to allocate per-node can be specified. 2056 See also Documentation/admin-guide/mm/hugetlbpage.rst. 2057 Format: <integer> or (node format) 2058 <node>:<integer>[,<node>:<integer>] 2059 2060 hugepagesz= 2061 [HW,EARLY] The size of the HugeTLB pages. This is 2062 used in conjunction with hugepages (above) to 2063 allocate huge pages of a specific size at boot. The 2064 pair hugepagesz=X hugepages=Y can be specified once 2065 for each supported huge page size. Huge page sizes 2066 are architecture dependent. See also 2067 Documentation/admin-guide/mm/hugetlbpage.rst. 2068 Format: size[KMG] 2069 2070 hugepage_alloc_threads= 2071 [HW] The number of threads that should be used to 2072 allocate hugepages during boot. This option can be 2073 used to improve system bootup time when allocating 2074 a large amount of huge pages. 2075 The default value is 25% of the available hardware threads. 2076 2077 Note that this parameter only applies to non-gigantic huge pages. 2078 2079 hugetlb_cma= [HW,CMA,EARLY] The size of a CMA area used for allocation 2080 of gigantic hugepages. Or using node format, the size 2081 of a CMA area per node can be specified. 2082 Format: nn[KMGTPE] or (node format) 2083 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]] 2084 2085 Reserve a CMA area of given size and allocate gigantic 2086 hugepages using the CMA allocator. If enabled, the 2087 boot-time allocation of gigantic hugepages is skipped. 2088 2089 hugetlb_cma_only= 2090 [HW,CMA,EARLY] When allocating new HugeTLB pages, only 2091 try to allocate from the CMA areas. 2092 2093 This option does nothing if hugetlb_cma= is not also 2094 specified. 2095 2096 hugetlb_free_vmemmap= 2097 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP 2098 enabled. 2099 Control if HugeTLB Vmemmap Optimization (HVO) is enabled. 2100 Allows heavy hugetlb users to free up some more 2101 memory (7 * PAGE_SIZE for each 2MB hugetlb page). 2102 Format: { on | off (default) } 2103 2104 on: enable HVO 2105 off: disable HVO 2106 2107 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y, 2108 the default is on. 2109 2110 Note that the vmemmap pages may be allocated from the added 2111 memory block itself when memory_hotplug.memmap_on_memory is 2112 enabled, those vmemmap pages cannot be optimized even if this 2113 feature is enabled. Other vmemmap pages not allocated from 2114 the added memory block itself do not be affected. 2115 2116 hung_task_panic= 2117 [KNL] Number of hung tasks to trigger kernel panic. 2118 Format: <int> 2119 2120 When set to a non-zero value, a kernel panic will be triggered if 2121 the number of detected hung tasks reaches this value. 2122 2123 0: don't panic 2124 1: panic immediately on first hung task 2125 N: panic after N hung tasks are detected in a single scan 2126 2127 The default value is controlled by the 2128 CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time option. The value 2129 selected by this boot parameter can be changed later by the 2130 kernel.hung_task_panic sysctl. 2131 2132 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 2133 terminal devices. Valid values: 0..8 2134 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 2135 If specified, z/VM IUCV HVC accepts connections 2136 from listed z/VM user IDs only. 2137 2138 hv_nopvspin [X86,HYPER_V,EARLY] 2139 Disables the paravirt spinlock optimizations 2140 which allow the hypervisor to 'idle' the guest 2141 on lock contention. 2142 2143 hw_protection= [HW] 2144 Format: reboot | shutdown 2145 2146 Hardware protection action taken on critical events like 2147 overtemperature or imminent voltage loss. 2148 2149 i2c_bus= [HW] Override the default board specific I2C bus speed 2150 or register an additional I2C bus that is not 2151 registered from board initialization code. 2152 Format: 2153 <bus_id>,<clkrate> 2154 2155 i2c_touchscreen_props= [HW,ACPI,X86] 2156 Set device-properties for ACPI-enumerated I2C-attached 2157 touchscreen, to e.g. fix coordinates of upside-down 2158 mounted touchscreens. If you need this option please 2159 submit a drivers/platform/x86/touchscreen_dmi.c patch 2160 adding a DMI quirk for this. 2161 2162 Format: 2163 <ACPI_HW_ID>:<prop_name>=<val>[:prop_name=val][:...] 2164 Where <val> is one of: 2165 Omit "=<val>" entirely Set a boolean device-property 2166 Unsigned number Set a u32 device-property 2167 Anything else Set a string device-property 2168 2169 Examples (split over multiple lines): 2170 i2c_touchscreen_props=GDIX1001:touchscreen-inverted-x: 2171 touchscreen-inverted-y 2172 2173 i2c_touchscreen_props=MSSL1680:touchscreen-size-x=1920: 2174 touchscreen-size-y=1080:touchscreen-inverted-y: 2175 firmware-name=gsl1680-vendor-model.fw:silead,home-button 2176 2177 i8042.debug [HW] Toggle i8042 debug mode 2178 i8042.unmask_kbd_data 2179 [HW] Enable printing of interrupt data from the KBD port 2180 (disabled by default, and as a pre-condition 2181 requires that i8042.debug=1 be enabled) 2182 i8042.direct [HW] Put keyboard port into non-translated mode 2183 i8042.dumbkbd [HW] Pretend that controller can only read data from 2184 keyboard and cannot control its state 2185 (Don't attempt to blink the leds) 2186 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 2187 i8042.nokbd [HW] Don't check/create keyboard port 2188 i8042.noloop [HW] Disable the AUX Loopback command while probing 2189 for the AUX port 2190 i8042.nomux [HW] Don't check presence of an active multiplexing 2191 controller 2192 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 2193 controllers 2194 i8042.notimeout [HW] Ignore timeout condition signalled by controller 2195 i8042.reset [HW] Reset the controller during init, cleanup and 2196 suspend-to-ram transitions, only during s2r 2197 transitions, or never reset 2198 Format: { 1 | Y | y | 0 | N | n } 2199 1, Y, y: always reset controller 2200 0, N, n: don't ever reset controller 2201 Default: only on s2r transitions on x86; most other 2202 architectures force reset to be always executed 2203 i8042.unlock [HW] Unlock (ignore) the keylock 2204 i8042.kbdreset [HW] Reset device connected to KBD port 2205 i8042.probe_defer 2206 [HW] Allow deferred probing upon i8042 probe errors 2207 2208 i810= [HW,DRM] 2209 2210 i915.invert_brightness= 2211 [DRM] Invert the sense of the variable that is used to 2212 set the brightness of the panel backlight. Normally a 2213 brightness value of 0 indicates backlight switched off, 2214 and the maximum of the brightness value sets the backlight 2215 to maximum brightness. If this parameter is set to 0 2216 (default) and the machine requires it, or this parameter 2217 is set to 1, a brightness value of 0 sets the backlight 2218 to maximum brightness, and the maximum of the brightness 2219 value switches the backlight off. 2220 -1 -- never invert brightness 2221 0 -- machine default 2222 1 -- force brightness inversion 2223 2224 ia32_emulation= [X86-64] 2225 Format: <bool> 2226 When true, allows loading 32-bit programs and executing 32-bit 2227 syscalls, essentially overriding IA32_EMULATION_DEFAULT_DISABLED at 2228 boot time. When false, unconditionally disables IA32 emulation. 2229 2230 icn= [HW,ISDN] 2231 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 2232 2233 2234 idle= [X86,EARLY] 2235 Format: idle=poll, idle=halt, idle=nomwait 2236 2237 idle=poll: Don't do power saving in the idle loop 2238 using HLT, but poll for rescheduling event. This will 2239 make the CPUs eat a lot more power, but may be useful 2240 to get slightly better performance in multiprocessor 2241 benchmarks. It also makes some profiling using 2242 performance counters more accurate. Please note that 2243 on systems with MONITOR/MWAIT support (like Intel 2244 EM64T CPUs) this option has no performance advantage 2245 over the normal idle loop. It may also interact badly 2246 with hyperthreading. 2247 2248 idle=halt: Halt is forced to be used for CPU idle. 2249 In such case C2/C3 won't be used again. 2250 2251 idle=nomwait: Disable mwait for CPU C-states 2252 2253 idxd.sva= [HW] 2254 Format: <bool> 2255 Allow force disabling of Shared Virtual Memory (SVA) 2256 support for the idxd driver. By default it is set to 2257 true (1). 2258 2259 idxd.tc_override= [HW] 2260 Format: <bool> 2261 Allow override of default traffic class configuration 2262 for the device. By default it is set to false (0). 2263 2264 ieee754= [MIPS] Select IEEE Std 754 conformance mode 2265 Format: { strict | legacy | 2008 | relaxed | emulated } 2266 Default: strict 2267 2268 Choose which programs will be accepted for execution 2269 based on the IEEE 754 NaN encoding(s) supported by 2270 the FPU and the NaN encoding requested with the value 2271 of an ELF file header flag individually set by each 2272 binary. Hardware implementations are permitted to 2273 support either or both of the legacy and the 2008 NaN 2274 encoding mode. 2275 2276 Available settings are as follows: 2277 strict accept binaries that request a NaN encoding 2278 supported by the FPU 2279 legacy only accept legacy-NaN binaries, if supported 2280 by the FPU 2281 2008 only accept 2008-NaN binaries, if supported 2282 by the FPU 2283 relaxed accept any binaries regardless of whether 2284 supported by the FPU 2285 emulated accept any binaries but enable FPU emulator 2286 if binary mode is unsupported by the FPU. 2287 2288 The FPU emulator is always able to support both NaN 2289 encodings, so if no FPU hardware is present or it has 2290 been disabled with 'nofpu', then the settings of 2291 'legacy' and '2008' strap the emulator accordingly, 2292 'relaxed' straps the emulator for both legacy-NaN and 2293 2008-NaN, whereas 'strict' enables legacy-NaN only on 2294 legacy processors and both NaN encodings on MIPS32 or 2295 MIPS64 CPUs. 2296 2297 The setting for ABS.fmt/NEG.fmt instruction execution 2298 mode generally follows that for the NaN encoding, 2299 except where unsupported by hardware. 2300 2301 ignore_loglevel [KNL,EARLY] 2302 Ignore loglevel setting - this will print /all/ 2303 kernel messages to the console. Useful for debugging. 2304 We also add it as printk module parameter, so users 2305 could change it dynamically, usually by 2306 /sys/module/printk/parameters/ignore_loglevel. 2307 2308 ignore_rlimit_data 2309 Ignore RLIMIT_DATA setting for data mappings, 2310 print warning at first misuse. Can be changed via 2311 /sys/module/kernel/parameters/ignore_rlimit_data. 2312 2313 ihash_entries= [KNL] 2314 Set number of hash buckets for inode cache. 2315 2316 ima_appraise= [IMA] appraise integrity measurements 2317 Format: { "off" | "enforce" | "fix" | "log" } 2318 default: "enforce" 2319 2320 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 2321 The builtin appraise policy appraises all files 2322 owned by uid=0. 2323 2324 ima_canonical_fmt [IMA] 2325 Use the canonical format for the binary runtime 2326 measurements, instead of host native format. 2327 2328 ima_hash= [IMA] 2329 Format: { md5 | sha1 | rmd160 | sha256 | sha384 2330 | sha512 | ... } 2331 default: "sha1" 2332 2333 The list of supported hash algorithms is defined 2334 in crypto/hash_info.h. 2335 2336 ima_policy= [IMA] 2337 The builtin policies to load during IMA setup. 2338 Format: "tcb | appraise_tcb | secure_boot | 2339 fail_securely | critical_data" 2340 2341 The "tcb" policy measures all programs exec'd, files 2342 mmap'd for exec, and all files opened with the read 2343 mode bit set by either the effective uid (euid=0) or 2344 uid=0. 2345 2346 The "appraise_tcb" policy appraises the integrity of 2347 all files owned by root. 2348 2349 The "secure_boot" policy appraises the integrity 2350 of files (eg. kexec kernel image, kernel modules, 2351 firmware, policy, etc) based on file signatures. 2352 2353 The "fail_securely" policy forces file signature 2354 verification failure also on privileged mounted 2355 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 2356 flag. 2357 2358 The "critical_data" policy measures kernel integrity 2359 critical data. 2360 2361 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 2362 Load a policy which meets the needs of the Trusted 2363 Computing Base. This means IMA will measure all 2364 programs exec'd, files mmap'd for exec, and all files 2365 opened for read by uid=0. 2366 2367 ima_template= [IMA] 2368 Select one of defined IMA measurements template formats. 2369 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" | 2370 "ima-sigv2" } 2371 Default: "ima-ng" 2372 2373 ima_template_fmt= 2374 [IMA] Define a custom template format. 2375 Format: { "field1|...|fieldN" } 2376 2377 ima= [IMA] Enable or disable IMA 2378 Format: { "off" | "on" } 2379 Default: "on" 2380 Note that disabling IMA is limited to kdump kernel. 2381 2382 indirect_target_selection= [X86,Intel] Mitigation control for Indirect 2383 Target Selection(ITS) bug in Intel CPUs. Updated 2384 microcode is also required for a fix in IBPB. 2385 2386 on: Enable mitigation (default). 2387 off: Disable mitigation. 2388 force: Force the ITS bug and deploy default 2389 mitigation. 2390 vmexit: Only deploy mitigation if CPU is affected by 2391 guest/host isolation part of ITS. 2392 stuff: Deploy RSB-fill mitigation when retpoline is 2393 also deployed. Otherwise, deploy the default 2394 mitigation. 2395 2396 For details see: 2397 Documentation/admin-guide/hw-vuln/indirect-target-selection.rst 2398 2399 init= [KNL] 2400 Format: <full_path> 2401 Run specified binary instead of /sbin/init as init 2402 process. 2403 2404 initcall_debug [KNL] Trace initcalls as they are executed. Useful 2405 for working out where the kernel is dying during 2406 startup. 2407 2408 initcall_blacklist= [KNL] Do not execute a comma-separated list of 2409 initcall functions. Useful for debugging built-in 2410 modules and initcalls. 2411 2412 initramfs_async= [KNL] 2413 Format: <bool> 2414 Default: 1 2415 This parameter controls whether the initramfs 2416 image is unpacked asynchronously, concurrently 2417 with devices being probed and 2418 initialized. This should normally just work, 2419 but as a debugging aid, one can get the 2420 historical behaviour of the initramfs 2421 unpacking being completed before device_ and 2422 late_ initcalls. 2423 2424 initrd= [BOOT,EARLY] Specify the location of the initial ramdisk 2425 2426 initrdmem= [KNL,EARLY] Specify a physical address and size from which to 2427 load the initrd. If an initrd is compiled in or 2428 specified in the bootparams, it takes priority over this 2429 setting. 2430 Format: ss[KMG],nn[KMG] 2431 Default is 0, 0 2432 2433 init_on_alloc= [MM,EARLY] Fill newly allocated pages and heap objects with 2434 zeroes. 2435 Format: 0 | 1 2436 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 2437 2438 init_on_free= [MM,EARLY] Fill freed pages and heap objects with zeroes. 2439 Format: 0 | 1 2440 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 2441 2442 init_pkru= [X86] Specify the default memory protection keys rights 2443 register contents for all processes. 0x55555554 by 2444 default (disallow access to all but pkey 0). Can 2445 override in debugfs after boot. 2446 2447 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 2448 Format: <irq> 2449 2450 int_pln_enable [X86] Enable power limit notification interrupt 2451 2452 integrity_audit=[IMA] 2453 Format: { "0" | "1" } 2454 0 -- basic integrity auditing messages. (Default) 2455 1 -- additional integrity auditing messages. 2456 2457 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 2458 on 2459 Enable intel iommu driver. 2460 off 2461 Disable intel iommu driver. 2462 igfx_off [Default Off] 2463 By default, gfx is mapped as normal device. If a gfx 2464 device has a dedicated DMAR unit, the DMAR unit is 2465 bypassed by not enabling DMAR with this option. In 2466 this case, gfx device will use physical address for 2467 DMA. 2468 strict [Default Off] 2469 Deprecated, equivalent to iommu.strict=1. 2470 sp_off [Default Off] 2471 By default, super page will be supported if Intel IOMMU 2472 has the capability. With this option, super page will 2473 not be supported. 2474 sm_on 2475 Enable the Intel IOMMU scalable mode if the hardware 2476 advertises that it has support for the scalable mode 2477 translation. 2478 sm_off 2479 Disallow use of the Intel IOMMU scalable mode. 2480 tboot_noforce [Default Off] 2481 Do not force the Intel IOMMU enabled under tboot. 2482 By default, tboot will force Intel IOMMU on, which 2483 could harm performance of some high-throughput 2484 devices like 40GBit network cards, even if identity 2485 mapping is enabled. 2486 Note that using this option lowers the security 2487 provided by tboot because it makes the system 2488 vulnerable to DMA attacks. 2489 2490 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 2491 0 disables intel_idle and fall back on acpi_idle. 2492 1 to 9 specify maximum depth of C-state. 2493 2494 intel_pstate= [X86,EARLY] 2495 disable 2496 Do not enable intel_pstate as the default 2497 scaling driver for the supported processors 2498 active 2499 Use intel_pstate driver to bypass the scaling 2500 governors layer of cpufreq and provides it own 2501 algorithms for p-state selection. There are two 2502 P-state selection algorithms provided by 2503 intel_pstate in the active mode: powersave and 2504 performance. The way they both operate depends 2505 on whether or not the hardware managed P-states 2506 (HWP) feature has been enabled in the processor 2507 and possibly on the processor model. 2508 passive 2509 Use intel_pstate as a scaling driver, but configure it 2510 to work with generic cpufreq governors (instead of 2511 enabling its internal governor). This mode cannot be 2512 used along with the hardware-managed P-states (HWP) 2513 feature. 2514 force 2515 Enable intel_pstate on systems that prohibit it by default 2516 in favor of acpi-cpufreq. Forcing the intel_pstate driver 2517 instead of acpi-cpufreq may disable platform features, such 2518 as thermal controls and power capping, that rely on ACPI 2519 P-States information being indicated to OSPM and therefore 2520 should be used with caution. This option does not work with 2521 processors that aren't supported by the intel_pstate driver 2522 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 2523 no_hwp 2524 Do not enable hardware P state control (HWP) 2525 if available. 2526 hwp_only 2527 Only load intel_pstate on systems which support 2528 hardware P state control (HWP) if available. 2529 support_acpi_ppc 2530 Enforce ACPI _PPC performance limits. If the Fixed ACPI 2531 Description Table, specifies preferred power management 2532 profile as "Enterprise Server" or "Performance Server", 2533 then this feature is turned on by default. 2534 per_cpu_perf_limits 2535 Allow per-logical-CPU P-State performance control limits using 2536 cpufreq sysfs interface 2537 no_cas 2538 Do not enable capacity-aware scheduling (CAS) on 2539 hybrid systems 2540 2541 intremap= [X86-64,Intel-IOMMU,EARLY] 2542 on enable Interrupt Remapping (default) 2543 off disable Interrupt Remapping 2544 nosid disable Source ID checking 2545 no_x2apic_optout 2546 BIOS x2APIC opt-out request will be ignored 2547 nopost disable Interrupt Posting 2548 posted_msi 2549 enable MSIs delivered as posted interrupts 2550 2551 iomem= Disable strict checking of access to MMIO memory 2552 strict regions from userspace. 2553 relaxed 2554 2555 iommu= [X86,EARLY] 2556 2557 off 2558 Don't initialize and use any kind of IOMMU. 2559 2560 force 2561 Force the use of the hardware IOMMU even when 2562 it is not actually needed (e.g. because < 3 GB 2563 memory). 2564 2565 noforce 2566 Don't force hardware IOMMU usage when it is not 2567 needed. (default). 2568 2569 biomerge 2570 panic 2571 nopanic 2572 merge 2573 nomerge 2574 2575 soft 2576 Use software bounce buffering (SWIOTLB) (default for 2577 Intel machines). This can be used to prevent the usage 2578 of an available hardware IOMMU. 2579 2580 pt 2581 nopt 2582 nobypass [PPC/POWERNV] 2583 Disable IOMMU bypass, using IOMMU for PCI devices. 2584 2585 AMD Gart HW IOMMU-specific options: 2586 2587 <size> 2588 Set the size of the remapping area in bytes. 2589 2590 allowed 2591 Overwrite iommu off workarounds for specific chipsets 2592 2593 fullflush 2594 Flush IOMMU on each allocation (default). 2595 2596 nofullflush 2597 Don't use IOMMU fullflush. 2598 2599 memaper[=<order>] 2600 Allocate an own aperture over RAM with size 2601 32MB<<order. (default: order=1, i.e. 64MB) 2602 2603 merge 2604 Do scatter-gather (SG) merging. Implies "force" 2605 (experimental). 2606 2607 nomerge 2608 Don't do scatter-gather (SG) merging. 2609 2610 noaperture 2611 Ask the IOMMU not to touch the aperture for AGP. 2612 2613 noagp 2614 Don't initialize the AGP driver and use full aperture. 2615 2616 panic 2617 Always panic when IOMMU overflows. 2618 2619 iommu.forcedac= [ARM64,X86,EARLY] Control IOVA allocation for PCI devices. 2620 Format: { "0" | "1" } 2621 0 - Try to allocate a 32-bit DMA address first, before 2622 falling back to the full range if needed. 2623 1 - Allocate directly from the full usable range, 2624 forcing Dual Address Cycle for PCI cards supporting 2625 greater than 32-bit addressing. 2626 2627 iommu.strict= [ARM64,X86,S390,EARLY] Configure TLB invalidation behaviour 2628 Format: { "0" | "1" } 2629 0 - Lazy mode. 2630 Request that DMA unmap operations use deferred 2631 invalidation of hardware TLBs, for increased 2632 throughput at the cost of reduced device isolation. 2633 Will fall back to strict mode if not supported by 2634 the relevant IOMMU driver. 2635 1 - Strict mode. 2636 DMA unmap operations invalidate IOMMU hardware TLBs 2637 synchronously. 2638 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}. 2639 Note: on x86, strict mode specified via one of the 2640 legacy driver-specific options takes precedence. 2641 2642 iommu.passthrough= 2643 [ARM64,X86,EARLY] Configure DMA to bypass the IOMMU by default. 2644 Format: { "0" | "1" } 2645 0 - Use IOMMU translation for DMA. 2646 1 - Bypass the IOMMU for DMA. 2647 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2648 2649 iommu.debug_pagealloc= 2650 [KNL,EARLY] When CONFIG_IOMMU_DEBUG_PAGEALLOC is set, this 2651 parameter enables the feature at boot time. By default, it 2652 is disabled and the system behaves the same way as a kernel 2653 built without CONFIG_IOMMU_DEBUG_PAGEALLOC. 2654 Format: { "0" | "1" } 2655 0 - Sanitizer disabled. 2656 1 - Sanitizer enabled, expect runtime overhead. 2657 2658 io7= [HW] IO7 for Marvel-based Alpha systems 2659 See comment before marvel_specify_io7 in 2660 arch/alpha/kernel/core_marvel.c. 2661 2662 io_delay= [X86,EARLY] I/O delay method 2663 0x80 2664 Standard port 0x80 based delay 2665 0xed 2666 Alternate port 0xed based delay (needed on some systems) 2667 udelay 2668 Simple two microseconds delay 2669 none 2670 No delay 2671 2672 ip= [IP_PNP] 2673 See Documentation/admin-guide/nfs/nfsroot.rst. 2674 2675 ipcmni_extend [KNL,EARLY] Extend the maximum number of unique System V 2676 IPC identifiers from 32,768 to 16,777,216. 2677 2678 ipe.enforce= [IPE] 2679 Format: <bool> 2680 Determine whether IPE starts in permissive (0) or 2681 enforce (1) mode. The default is enforce. 2682 2683 ipe.success_audit= 2684 [IPE] 2685 Format: <bool> 2686 Start IPE with success auditing enabled, emitting 2687 an audit event when a binary is allowed. The default 2688 is 0. 2689 2690 irqaffinity= [SMP] Set the default irq affinity mask 2691 The argument is a cpu list, as described above. 2692 2693 irqchip.gicv2_force_probe= 2694 [ARM,ARM64,EARLY] 2695 Format: <bool> 2696 Force the kernel to look for the second 4kB page 2697 of a GICv2 controller even if the memory range 2698 exposed by the device tree is too small. 2699 2700 irqchip.gicv3_nolpi= 2701 [ARM,ARM64,EARLY] 2702 Force the kernel to ignore the availability of 2703 LPIs (and by consequence ITSs). Intended for system 2704 that use the kernel as a bootloader, and thus want 2705 to let secondary kernels in charge of setting up 2706 LPIs. 2707 2708 irqchip.gicv3_pseudo_nmi= [ARM64,EARLY] 2709 Enables support for pseudo-NMIs in the kernel. This 2710 requires the kernel to be built with 2711 CONFIG_ARM64_PSEUDO_NMI. 2712 2713 irqchip.riscv_imsic_noipi 2714 [RISC-V,EARLY] 2715 Force the kernel to not use IMSIC software injected MSIs 2716 as IPIs. Intended for system where IMSIC is trap-n-emulated, 2717 and thus want to reduce MMIO traps when triggering IPIs 2718 to multiple harts. 2719 2720 irqfixup [HW] 2721 When an interrupt is not handled search all handlers 2722 for it. Intended to get systems with badly broken 2723 firmware running. 2724 2725 irqhandler.duration_warn_us= [KNL] 2726 Warn if an IRQ handler exceeds the specified duration 2727 threshold in microseconds. Useful for identifying 2728 long-running IRQs in the system. 2729 2730 irqpoll [HW] 2731 When an interrupt is not handled search all handlers 2732 for it. Also check all handlers each timer 2733 interrupt. Intended to get systems with badly broken 2734 firmware running. 2735 2736 isapnp= [ISAPNP] 2737 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2738 2739 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2740 [Deprecated - use cpusets instead] 2741 Format: [flag-list,]<cpu-list> 2742 2743 Specify one or more CPUs to isolate from disturbances 2744 specified in the flag list (default: domain): 2745 2746 nohz 2747 Disable the tick when a single task runs as well as 2748 disabling other kernel noises like having RCU callbacks 2749 offloaded. This is equivalent to the nohz_full parameter. 2750 2751 A residual 1Hz tick is offloaded to workqueues, which you 2752 need to affine to housekeeping through the global 2753 workqueue's affinity configured via the 2754 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2755 by using the 'domain' flag described below. 2756 2757 NOTE: by default the global workqueue runs on all CPUs, 2758 so to protect individual CPUs the 'cpumask' file has to 2759 be configured manually after bootup. 2760 2761 domain 2762 Isolate from the general SMP balancing and scheduling 2763 algorithms. Note that performing domain isolation this way 2764 is irreversible: it's not possible to bring back a CPU to 2765 the domains once isolated through isolcpus. It's strongly 2766 advised to use cpusets instead to disable scheduler load 2767 balancing through the "cpuset.sched_load_balance" file. 2768 It offers a much more flexible interface where CPUs can 2769 move in and out of an isolated set anytime. 2770 2771 You can move a process onto or off an "isolated" CPU via 2772 the CPU affinity syscalls or cpuset. 2773 <cpu number> begins at 0 and the maximum value is 2774 "number of CPUs in system - 1". 2775 2776 managed_irq 2777 2778 Isolate from being targeted by managed interrupts 2779 which have an interrupt mask containing isolated 2780 CPUs. The affinity of managed interrupts is 2781 handled by the kernel and cannot be changed via 2782 the /proc/irq/* interfaces. 2783 2784 This isolation is best effort and only effective 2785 if the automatically assigned interrupt mask of a 2786 device queue contains isolated and housekeeping 2787 CPUs. If housekeeping CPUs are online then such 2788 interrupts are directed to the housekeeping CPU 2789 so that IO submitted on the housekeeping CPU 2790 cannot disturb the isolated CPU. 2791 2792 If a queue's affinity mask contains only isolated 2793 CPUs then this parameter has no effect on the 2794 interrupt routing decision, though interrupts are 2795 only delivered when tasks running on those 2796 isolated CPUs submit IO. IO submitted on 2797 housekeeping CPUs has no influence on those 2798 queues. 2799 2800 The format of <cpu-list> is described above. 2801 2802 iucv= [HW,NET] 2803 2804 ivrs_ioapic [HW,X86-64] 2805 Provide an override to the IOAPIC-ID<->DEVICE-ID 2806 mapping provided in the IVRS ACPI table. 2807 By default, PCI segment is 0, and can be omitted. 2808 2809 For example, to map IOAPIC-ID decimal 10 to 2810 PCI segment 0x1 and PCI device 00:14.0, 2811 write the parameter as: 2812 ivrs_ioapic=10@0001:00:14.0 2813 2814 Deprecated formats: 2815 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0 2816 write the parameter as: 2817 ivrs_ioapic[10]=00:14.0 2818 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and 2819 PCI device 00:14.0 write the parameter as: 2820 ivrs_ioapic[10]=0001:00:14.0 2821 2822 ivrs_hpet [HW,X86-64] 2823 Provide an override to the HPET-ID<->DEVICE-ID 2824 mapping provided in the IVRS ACPI table. 2825 By default, PCI segment is 0, and can be omitted. 2826 2827 For example, to map HPET-ID decimal 10 to 2828 PCI segment 0x1 and PCI device 00:14.0, 2829 write the parameter as: 2830 ivrs_hpet=10@0001:00:14.0 2831 2832 Deprecated formats: 2833 * To map HPET-ID decimal 0 to PCI device 00:14.0 2834 write the parameter as: 2835 ivrs_hpet[0]=00:14.0 2836 * To map HPET-ID decimal 10 to PCI segment 0x1 and 2837 PCI device 00:14.0 write the parameter as: 2838 ivrs_ioapic[10]=0001:00:14.0 2839 2840 ivrs_acpihid [HW,X86-64] 2841 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2842 mapping provided in the IVRS ACPI table. 2843 By default, PCI segment is 0, and can be omitted. 2844 2845 For example, to map UART-HID:UID AMD0020:0 to 2846 PCI segment 0x1 and PCI device ID 00:14.5, 2847 write the parameter as: 2848 ivrs_acpihid=AMD0020:0@0001:00:14.5 2849 2850 Deprecated formats: 2851 * To map UART-HID:UID AMD0020:0 to PCI segment is 0, 2852 PCI device ID 00:14.5, write the parameter as: 2853 ivrs_acpihid[00:14.5]=AMD0020:0 2854 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and 2855 PCI device ID 00:14.5, write the parameter as: 2856 ivrs_acpihid[0001:00:14.5]=AMD0020:0 2857 2858 js= [HW,JOY] Analog joystick 2859 See Documentation/input/joydev/joystick.rst. 2860 2861 kasan_multi_shot 2862 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2863 report on every invalid memory access. Without this 2864 parameter KASAN will print report only for the first 2865 invalid access. 2866 2867 keep_bootcon [KNL,EARLY] 2868 Do not unregister boot console at start. This is only 2869 useful for debugging when something happens in the window 2870 between unregistering the boot console and initializing 2871 the real console. 2872 2873 keepinitrd [HW,ARM] See retain_initrd. 2874 2875 kernelcore= [KNL,X86,PPC,EARLY] 2876 Format: nn[KMGTPE] | nn% | "mirror" 2877 This parameter specifies the amount of memory usable by 2878 the kernel for non-movable allocations. The requested 2879 amount is spread evenly throughout all nodes in the 2880 system as ZONE_NORMAL. The remaining memory is used for 2881 movable memory in its own zone, ZONE_MOVABLE. In the 2882 event, a node is too small to have both ZONE_NORMAL and 2883 ZONE_MOVABLE, kernelcore memory will take priority and 2884 other nodes will have a larger ZONE_MOVABLE. 2885 2886 ZONE_MOVABLE is used for the allocation of pages that 2887 may be reclaimed or moved by the page migration 2888 subsystem. Note that allocations like PTEs-from-HighMem 2889 still use the HighMem zone if it exists, and the Normal 2890 zone if it does not. 2891 2892 It is possible to specify the exact amount of memory in 2893 the form of "nn[KMGTPE]", a percentage of total system 2894 memory in the form of "nn%", or "mirror". If "mirror" 2895 option is specified, mirrored (reliable) memory is used 2896 for non-movable allocations and remaining memory is used 2897 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2898 are exclusive, so you cannot specify multiple forms. 2899 2900 kfence.burst= [MM,KFENCE] The number of additional successive 2901 allocations to be attempted through KFENCE for each 2902 sample interval. 2903 Format: <unsigned integer> 2904 Default: 0 2905 2906 kfence.check_on_panic= 2907 [MM,KFENCE] Whether to check all KFENCE-managed objects' 2908 canaries on panic. 2909 Format: <bool> 2910 Default: false 2911 2912 kfence.deferrable= 2913 [MM,KFENCE] Whether to use a deferrable timer to trigger 2914 allocations. This avoids forcing CPU wake-ups if the 2915 system is idle, at the risk of a less predictable 2916 sample interval. 2917 Format: <bool> 2918 Default: CONFIG_KFENCE_DEFERRABLE 2919 2920 kfence.fault= [MM,KFENCE] Controls the behavior when a KFENCE 2921 error is detected. 2922 report - print the error report and continue (default). 2923 oops - print the error report and oops. 2924 panic - print the error report and panic. 2925 2926 kfence.sample_interval= 2927 [MM,KFENCE] KFENCE's sample interval in milliseconds. 2928 Format: <unsigned integer> 2929 0 - Disable KFENCE. 2930 >0 - Enabled KFENCE with given sample interval. 2931 Default: CONFIG_KFENCE_SAMPLE_INTERVAL 2932 2933 kfence.skip_covered_thresh= 2934 [MM,KFENCE] If pool utilization reaches this threshold 2935 (pool usage%), KFENCE limits currently covered 2936 allocations of the same source from further filling 2937 up the pool. 2938 Format: <unsigned integer> 2939 Default: 75 2940 2941 kgdbdbgp= [KGDB,HW,EARLY] kgdb over EHCI usb debug port. 2942 Format: <Controller#>[,poll interval] 2943 The controller # is the number of the ehci usb debug 2944 port as it is probed via PCI. The poll interval is 2945 optional and is the number seconds in between 2946 each poll cycle to the debug port in case you need 2947 the functionality for interrupting the kernel with 2948 gdb or control-c on the dbgp connection. When 2949 not using this parameter you use sysrq-g to break into 2950 the kernel debugger. 2951 2952 kgdboc= [KGDB,HW] kgdb over consoles. 2953 Requires a tty driver that supports console polling, 2954 or a supported polling keyboard driver (non-usb). 2955 Serial only format: <serial_device>[,baud] 2956 keyboard only format: kbd 2957 keyboard and serial format: kbd,<serial_device>[,baud] 2958 Optional Kernel mode setting: 2959 kms, kbd format: kms,kbd 2960 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2961 2962 kgdboc_earlycon= [KGDB,HW,EARLY] 2963 If the boot console provides the ability to read 2964 characters and can work in polling mode, you can use 2965 this parameter to tell kgdb to use it as a backend 2966 until the normal console is registered. Intended to 2967 be used together with the kgdboc parameter which 2968 specifies the normal console to transition to. 2969 2970 The name of the early console should be specified 2971 as the value of this parameter. Note that the name of 2972 the early console might be different than the tty 2973 name passed to kgdboc. It's OK to leave the value 2974 blank and the first boot console that implements 2975 read() will be picked. 2976 2977 kgdbwait [KGDB,EARLY] Stop kernel execution and enter the 2978 kernel debugger at the earliest opportunity. 2979 2980 kho= [KEXEC,EARLY] 2981 Format: { "0" | "1" | "off" | "on" | "y" | "n" } 2982 Enables or disables Kexec HandOver. 2983 "0" | "off" | "n" - kexec handover is disabled 2984 "1" | "on" | "y" - kexec handover is enabled 2985 2986 kho_scratch= [KEXEC,EARLY] 2987 Format: ll[KMG],mm[KMG],nn[KMG] | nn% 2988 Defines the size of the KHO scratch region. The KHO 2989 scratch regions are physically contiguous memory 2990 ranges that can only be used for non-kernel 2991 allocations. That way, even when memory is heavily 2992 fragmented with handed over memory, the kexeced 2993 kernel will always have enough contiguous ranges to 2994 bootstrap itself. 2995 2996 It is possible to specify the exact amount of 2997 memory in the form of "ll[KMG],mm[KMG],nn[KMG]" 2998 where the first parameter defines the size of a low 2999 memory scratch area, the second parameter defines 3000 the size of a global scratch area and the third 3001 parameter defines the size of additional per-node 3002 scratch areas. The form "nn%" defines scale factor 3003 (in percents) of memory that was used during boot. 3004 3005 kmac= [MIPS] Korina ethernet MAC address. 3006 Configure the RouterBoard 532 series on-chip 3007 Ethernet adapter MAC address. 3008 3009 kmemleak= [KNL,EARLY] Boot-time kmemleak enable/disable 3010 Valid arguments: on, off 3011 Default: on 3012 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 3013 the default is off. 3014 3015 kprobe_event=[probe-list] 3016 [FTRACE] Add kprobe events and enable at boot time. 3017 The probe-list is a semicolon delimited list of probe 3018 definitions. Each definition is same as kprobe_events 3019 interface, but the parameters are comma delimited. 3020 For example, to add a kprobe event on vfs_read with 3021 arg1 and arg2, add to the command line; 3022 3023 kprobe_event=p,vfs_read,$arg1,$arg2 3024 3025 See also Documentation/trace/kprobetrace.rst "Kernel 3026 Boot Parameter" section. 3027 3028 kpti= [ARM64,EARLY] Control page table isolation of 3029 user and kernel address spaces. 3030 Default: enabled on cores which need mitigation. 3031 0: force disabled 3032 1: force enabled 3033 3034 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires 3035 CONFIG_KUNIT to be set to be fully enabled. The 3036 default value can be overridden via 3037 KUNIT_DEFAULT_ENABLED. 3038 Default is 1 (enabled) 3039 3040 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 3041 Default is 0 (don't ignore, but inject #GP) 3042 3043 kvm.eager_page_split= 3044 [KVM,X86] Controls whether or not KVM will try to 3045 proactively split all huge pages during dirty logging. 3046 Eager page splitting reduces interruptions to vCPU 3047 execution by eliminating the write-protection faults 3048 and MMU lock contention that would otherwise be 3049 required to split huge pages lazily. 3050 3051 VM workloads that rarely perform writes or that write 3052 only to a small region of VM memory may benefit from 3053 disabling eager page splitting to allow huge pages to 3054 still be used for reads. 3055 3056 The behavior of eager page splitting depends on whether 3057 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If 3058 disabled, all huge pages in a memslot will be eagerly 3059 split when dirty logging is enabled on that memslot. If 3060 enabled, eager page splitting will be performed during 3061 the KVM_CLEAR_DIRTY ioctl, and only for the pages being 3062 cleared. 3063 3064 Eager page splitting is only supported when kvm.tdp_mmu=Y. 3065 3066 Default is Y (on). 3067 3068 kvm.enable_pmu=[KVM,X86] 3069 If enabled, KVM will virtualize PMU functionality based 3070 on the virtual CPU model defined by userspace. This 3071 can be overridden on a per-VM basis via 3072 KVM_CAP_PMU_CAPABILITY. 3073 3074 If disabled, KVM will not virtualize PMU functionality, 3075 e.g. MSRs, PMCs, PMIs, etc., even if userspace defines 3076 a virtual CPU model that contains PMU assets. 3077 3078 Note, KVM's vPMU support implicitly requires running 3079 with an in-kernel local APIC, e.g. to deliver PMIs to 3080 the guest. Running without an in-kernel local APIC is 3081 not supported, though KVM will allow such a combination 3082 (with severely degraded functionality). 3083 3084 See also enable_mediated_pmu. 3085 3086 Default is Y (on). 3087 3088 kvm.enable_virt_at_load=[KVM,ARM64,LOONGARCH,MIPS,RISCV,X86] 3089 If enabled, KVM will enable virtualization in hardware 3090 when KVM is loaded, and disable virtualization when KVM 3091 is unloaded (if KVM is built as a module). 3092 3093 If disabled, KVM will dynamically enable and disable 3094 virtualization on-demand when creating and destroying 3095 VMs, i.e. on the 0=>1 and 1=>0 transitions of the 3096 number of VMs. 3097 3098 Enabling virtualization at module load avoids potential 3099 latency for creation of the 0=>1 VM, as KVM serializes 3100 virtualization enabling across all online CPUs. The 3101 "cost" of enabling virtualization when KVM is loaded, 3102 is that doing so may interfere with using out-of-tree 3103 hypervisors that want to "own" virtualization hardware. 3104 3105 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 3106 Default is false (don't support). 3107 3108 kvm.nx_huge_pages= 3109 [KVM] Controls the software workaround for the 3110 X86_BUG_ITLB_MULTIHIT bug. 3111 force : Always deploy workaround. 3112 off : Never deploy workaround. 3113 auto : Deploy workaround based on the presence of 3114 X86_BUG_ITLB_MULTIHIT. 3115 3116 Default is 'auto'. 3117 3118 If the software workaround is enabled for the host, 3119 guests do need not to enable it for nested guests. 3120 3121 kvm.nx_huge_pages_recovery_ratio= 3122 [KVM] Controls how many 4KiB pages are periodically zapped 3123 back to huge pages. 0 disables the recovery, otherwise if 3124 the value is N KVM will zap 1/Nth of the 4KiB pages every 3125 period (see below). The default is 60. 3126 3127 kvm.nx_huge_pages_recovery_period_ms= 3128 [KVM] Controls the time period at which KVM zaps 4KiB pages 3129 back to huge pages. If the value is a non-zero N, KVM will 3130 zap a portion (see ratio above) of the pages every N msecs. 3131 If the value is 0 (the default), KVM will pick a period based 3132 on the ratio, such that a page is zapped after 1 hour on average. 3133 3134 kvm-{amd,intel}.enable_mediated_pmu=[KVM,AMD,INTEL] 3135 If enabled, KVM will provide a mediated virtual PMU, 3136 instead of the default perf-based virtual PMU (if 3137 kvm.enable_pmu is true and PMU is enumerated via the 3138 virtual CPU model). 3139 3140 With a perf-based vPMU, KVM operates as a user of perf, 3141 i.e. emulates guest PMU counters using perf events. 3142 KVM-created perf events are managed by perf as regular 3143 (guest-only) events, e.g. are scheduled in/out, contend 3144 for hardware resources, etc. Using a perf-based vPMU 3145 allows guest and host usage of the PMU to co-exist, but 3146 incurs non-trivial overhead and can result in silently 3147 dropped guest events (due to resource contention). 3148 3149 With a mediated vPMU, hardware PMU state is context 3150 switched around the world switch to/from the guest. 3151 KVM mediates which events the guest can utilize, but 3152 gives the guest direct access to all other PMU assets 3153 when possible (KVM may intercept some accesses if the 3154 virtual CPU model provides a subset of hardware PMU 3155 functionality). Using a mediated vPMU significantly 3156 reduces PMU virtualization overhead and eliminates lost 3157 guest events, but is mutually exclusive with using perf 3158 to profile KVM guests and adds latency to most VM-Exits 3159 (to context switch PMU state). 3160 3161 Default is N (off). 3162 3163 kvm-amd.nested= [KVM,AMD] Control nested virtualization feature in 3164 KVM/SVM. Default is 1 (enabled). 3165 3166 kvm-amd.npt= [KVM,AMD] Control KVM's use of Nested Page Tables, 3167 a.k.a. Two-Dimensional Page Tables. Default is 1 3168 (enabled). Disable by KVM if hardware lacks support 3169 for NPT. 3170 3171 kvm-amd.ciphertext_hiding_asids= 3172 [KVM,AMD] Ciphertext hiding prevents disallowed accesses 3173 to SNP private memory from reading ciphertext. Instead, 3174 reads will see constant default values (0xff). 3175 3176 If ciphertext hiding is enabled, the joint SEV-ES and 3177 SEV-SNP ASID space is partitioned into separate SEV-ES 3178 and SEV-SNP ASID ranges, with the SEV-SNP range being 3179 [1..max_snp_asid] and the SEV-ES range being 3180 (max_snp_asid..min_sev_asid), where min_sev_asid is 3181 enumerated by CPUID.0x.8000_001F[EDX]. 3182 3183 A non-zero value enables SEV-SNP ciphertext hiding and 3184 adjusts the ASID ranges for SEV-ES and SEV-SNP guests. 3185 KVM caps the number of SEV-SNP ASIDs at the maximum 3186 possible value, e.g. specifying -1u will assign all 3187 joint SEV-ES and SEV-SNP ASIDs to SEV-SNP. Note, 3188 assigning all joint ASIDs to SEV-SNP, i.e. configuring 3189 max_snp_asid == min_sev_asid-1, will effectively make 3190 SEV-ES unusable. 3191 3192 kvm-arm.mode= 3193 [KVM,ARM,EARLY] Select one of KVM/arm64's modes of 3194 operation. 3195 3196 none: Forcefully disable KVM. 3197 3198 nvhe: Standard nVHE-based mode, without support for 3199 protected guests. 3200 3201 protected: Mode with support for guests whose state is 3202 kept private from the host, using VHE or 3203 nVHE depending on HW support. 3204 3205 nested: VHE-based mode with support for nested 3206 virtualization. Requires at least ARMv8.4 3207 hardware (with FEAT_NV2). 3208 3209 Defaults to VHE/nVHE based on hardware support. Setting 3210 mode to "protected" will disable kexec and hibernation 3211 for the host. To force nVHE on VHE hardware, add 3212 "arm64_sw.hvhe=0 id_aa64mmfr1.vh=0" to the 3213 command-line. 3214 "nested" and "protected" are experimental and should be 3215 used with extreme caution. 3216 3217 kvm-arm.vgic_v3_group0_trap= 3218 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-0 3219 system registers 3220 3221 kvm-arm.vgic_v3_group1_trap= 3222 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-1 3223 system registers 3224 3225 kvm-arm.vgic_v3_common_trap= 3226 [KVM,ARM,EARLY] Trap guest accesses to GICv3 common 3227 system registers 3228 3229 kvm-arm.vgic_v4_enable= 3230 [KVM,ARM,EARLY] Allow use of GICv4 for direct 3231 injection of LPIs. 3232 3233 kvm-arm.wfe_trap_policy= 3234 [KVM,ARM] Control when to set WFE instruction trap for 3235 KVM VMs. Traps are allowed but not guaranteed by the 3236 CPU architecture. 3237 3238 trap: set WFE instruction trap 3239 3240 notrap: clear WFE instruction trap 3241 3242 kvm-arm.wfi_trap_policy= 3243 [KVM,ARM] Control when to set WFI instruction trap for 3244 KVM VMs. Traps are allowed but not guaranteed by the 3245 CPU architecture. 3246 3247 trap: set WFI instruction trap 3248 3249 notrap: clear WFI instruction trap 3250 3251 kvm_cma_resv_ratio=n [PPC,EARLY] 3252 Reserves given percentage from system memory area for 3253 contiguous memory allocation for KVM hash pagetable 3254 allocation. 3255 By default it reserves 5% of total system memory. 3256 Format: <integer> 3257 Default: 5 3258 3259 kvm-intel.ept= [KVM,Intel] Control KVM's use of Extended Page Tables, 3260 a.k.a. Two-Dimensional Page Tables. Default is 1 3261 (enabled). Disable by KVM if hardware lacks support 3262 for EPT. 3263 3264 kvm-intel.emulate_invalid_guest_state= 3265 [KVM,Intel] Control whether to emulate invalid guest 3266 state. Ignored if kvm-intel.enable_unrestricted_guest=1, 3267 as guest state is never invalid for unrestricted 3268 guests. This param doesn't apply to nested guests (L2), 3269 as KVM never emulates invalid L2 guest state. 3270 Default is 1 (enabled). 3271 3272 kvm-intel.flexpriority= 3273 [KVM,Intel] Control KVM's use of FlexPriority feature 3274 (TPR shadow). Default is 1 (enabled). Disable by KVM if 3275 hardware lacks support for it. 3276 3277 kvm-intel.nested= 3278 [KVM,Intel] Control nested virtualization feature in 3279 KVM/VMX. Default is 1 (enabled). 3280 3281 kvm-intel.unrestricted_guest= 3282 [KVM,Intel] Control KVM's use of unrestricted guest 3283 feature (virtualized real and unpaged mode). Default 3284 is 1 (enabled). Disable by KVM if EPT is disabled or 3285 hardware lacks support for it. 3286 3287 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 3288 CVE-2018-3620. 3289 3290 Valid arguments: never, cond, always 3291 3292 always: L1D cache flush on every VMENTER. 3293 cond: Flush L1D on VMENTER only when the code between 3294 VMEXIT and VMENTER can leak host memory. 3295 never: Disables the mitigation 3296 3297 Default is cond (do L1 cache flush in specific instances) 3298 3299 kvm-intel.vpid= [KVM,Intel] Control KVM's use of Virtual Processor 3300 Identification feature (tagged TLBs). Default is 1 3301 (enabled). Disable by KVM if hardware lacks support 3302 for it. 3303 3304 l1d_flush= [X86,INTEL,EARLY] 3305 Control mitigation for L1D based snooping vulnerability. 3306 3307 Certain CPUs are vulnerable to an exploit against CPU 3308 internal buffers which can forward information to a 3309 disclosure gadget under certain conditions. 3310 3311 In vulnerable processors, the speculatively 3312 forwarded data can be used in a cache side channel 3313 attack, to access data to which the attacker does 3314 not have direct access. 3315 3316 This parameter controls the mitigation. The 3317 options are: 3318 3319 on - enable the interface for the mitigation 3320 3321 l1tf= [X86,EARLY] Control mitigation of the L1TF vulnerability on 3322 affected CPUs 3323 3324 The kernel PTE inversion protection is unconditionally 3325 enabled and cannot be disabled. 3326 3327 full 3328 Provides all available mitigations for the 3329 L1TF vulnerability. Disables SMT and 3330 enables all mitigations in the 3331 hypervisors, i.e. unconditional L1D flush. 3332 3333 SMT control and L1D flush control via the 3334 sysfs interface is still possible after 3335 boot. Hypervisors will issue a warning 3336 when the first VM is started in a 3337 potentially insecure configuration, 3338 i.e. SMT enabled or L1D flush disabled. 3339 3340 full,force 3341 Same as 'full', but disables SMT and L1D 3342 flush runtime control. Implies the 3343 'nosmt=force' command line option. 3344 (i.e. sysfs control of SMT is disabled.) 3345 3346 flush 3347 Leaves SMT enabled and enables the default 3348 hypervisor mitigation, i.e. conditional 3349 L1D flush. 3350 3351 SMT control and L1D flush control via the 3352 sysfs interface is still possible after 3353 boot. Hypervisors will issue a warning 3354 when the first VM is started in a 3355 potentially insecure configuration, 3356 i.e. SMT enabled or L1D flush disabled. 3357 3358 flush,nosmt 3359 3360 Disables SMT and enables the default 3361 hypervisor mitigation. 3362 3363 SMT control and L1D flush control via the 3364 sysfs interface is still possible after 3365 boot. Hypervisors will issue a warning 3366 when the first VM is started in a 3367 potentially insecure configuration, 3368 i.e. SMT enabled or L1D flush disabled. 3369 3370 flush,nowarn 3371 Same as 'flush', but hypervisors will not 3372 warn when a VM is started in a potentially 3373 insecure configuration. 3374 3375 off 3376 Disables hypervisor mitigations and doesn't 3377 emit any warnings. 3378 It also drops the swap size and available 3379 RAM limit restriction on both hypervisor and 3380 bare metal. 3381 3382 Default is 'flush'. 3383 3384 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 3385 3386 l2cr= [PPC] 3387 3388 l3cr= [PPC] 3389 3390 lapic [X86-32,APIC,EARLY] Enable the local APIC even if BIOS 3391 disabled it. 3392 3393 lapic= [X86,APIC] Do not use TSC deadline 3394 value for LAPIC timer one-shot implementation. Default 3395 back to the programmable timer unit in the LAPIC. 3396 Format: notscdeadline 3397 3398 lapic_timer_c2_ok [X86,APIC,EARLY] trust the local apic timer 3399 in C2 power state. 3400 3401 libata.dma= [LIBATA] DMA control 3402 libata.dma=0 Disable all PATA and SATA DMA 3403 libata.dma=1 PATA and SATA Disk DMA only 3404 libata.dma=2 ATAPI (CDROM) DMA only 3405 libata.dma=4 Compact Flash DMA only 3406 Combinations also work, so libata.dma=3 enables DMA 3407 for disks and CDROMs, but not CFs. 3408 3409 libata.ignore_hpa= [LIBATA] Ignore HPA limit 3410 libata.ignore_hpa=0 keep BIOS limits (default) 3411 libata.ignore_hpa=1 ignore limits, using full disk 3412 3413 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 3414 when set. 3415 Format: <int> 3416 3417 libata.force= [LIBATA] Force configurations. The format is a comma- 3418 separated list of "[ID:]VAL" where ID is PORT[.DEVICE]. 3419 PORT and DEVICE are decimal numbers matching port, link 3420 or device. Basically, it matches the ATA ID string 3421 printed on console by libata. If the whole ID part is 3422 omitted, the last PORT and DEVICE values are used. If 3423 ID hasn't been specified yet, the configuration applies 3424 to all ports, links and devices. 3425 3426 If only DEVICE is omitted, the parameter applies to 3427 the port and all links and devices behind it. DEVICE 3428 number of 0 either selects the first device or the 3429 first fan-out link behind PMP device. It does not 3430 select the host link. DEVICE number of 15 selects the 3431 host link and device attached to it. 3432 3433 The VAL specifies the configuration to force. As long 3434 as there is no ambiguity, shortcut notation is allowed. 3435 For example, both 1.5 and 1.5G would work for 1.5Gbps. 3436 The following configurations can be forced. 3437 3438 * Cable type: 40c, 80c, short40c, unk, ign or sata. 3439 Any ID with matching PORT is used. 3440 3441 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 3442 3443 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 3444 udma[/][16,25,33,44,66,100,133] notation is also 3445 allowed. 3446 3447 * nohrst, nosrst, norst: suppress hard, soft and both 3448 resets. 3449 3450 * rstonce: only attempt one reset during hot-unplug 3451 link recovery. 3452 3453 * [no]dbdelay: Enable or disable the extra 200ms delay 3454 before debouncing a link PHY and device presence 3455 detection. 3456 3457 * [no]ncq: Turn on or off NCQ. 3458 3459 * [no]ncqtrim: Enable or disable queued DSM TRIM. 3460 3461 * [no]ncqati: Enable or disable NCQ trim on ATI chipset. 3462 3463 * [no]trim: Enable or disable (unqueued) TRIM. 3464 3465 * trim_zero: Indicate that TRIM command zeroes data. 3466 3467 * max_trim_128m: Set 128M maximum trim size limit. 3468 3469 * [no]dma: Turn on or off DMA transfers. 3470 3471 * atapi_dmadir: Enable ATAPI DMADIR bridge support. 3472 3473 * atapi_mod16_dma: Enable the use of ATAPI DMA for 3474 commands that are not a multiple of 16 bytes. 3475 3476 * [no]dmalog: Enable or disable the use of the 3477 READ LOG DMA EXT command to access logs. 3478 3479 * [no]iddevlog: Enable or disable access to the 3480 identify device data log. 3481 3482 * [no]logdir: Enable or disable access to the general 3483 purpose log directory. 3484 3485 * max_sec=<sectors>: Set the transfer size limit, in 3486 number of 512-byte sectors, to the value specified in 3487 <sectors>. The value specified in <sectors> has to be 3488 a non-zero positive integer. 3489 3490 * max_sec_128: Set transfer size limit to 128 sectors. 3491 3492 * max_sec_1024: Set or clear transfer size limit to 3493 1024 sectors. 3494 3495 * max_sec_lba48: Set or clear transfer size limit to 3496 65535 sectors. 3497 3498 * external: Mark port as external (hotplug-capable). 3499 3500 * [no]lpm: Enable or disable link power management. 3501 3502 * [no]setxfer: Indicate if transfer speed mode setting 3503 should be skipped. 3504 3505 * [no]fua: Disable or enable FUA (Force Unit Access) 3506 support for devices supporting this feature. 3507 3508 * dump_id: Dump IDENTIFY data. 3509 3510 * disable: Disable this device. 3511 3512 If there are multiple matching configurations changing 3513 the same attribute, the last one is used. 3514 3515 liveupdate= [KNL,EARLY] 3516 Format: <bool> 3517 Enable Live Update Orchestrator (LUO). 3518 Default: off. 3519 3520 lockd.nlm_grace_period=P [NFS] Assign grace period. 3521 Format: <integer> 3522 3523 lockd.nlm_tcpport=N [NFS] Assign TCP port. 3524 Format: <integer> 3525 3526 lockd.nlm_timeout=T [NFS] Assign timeout value. 3527 Format: <integer> 3528 3529 lockd.nlm_udpport=M [NFS] Assign UDP port. 3530 Format: <integer> 3531 3532 lockdown= [SECURITY,EARLY] 3533 { integrity | confidentiality } 3534 Enable the kernel lockdown feature. If set to 3535 integrity, kernel features that allow userland to 3536 modify the running kernel are disabled. If set to 3537 confidentiality, kernel features that allow userland 3538 to extract confidential information from the kernel 3539 are also disabled. 3540 3541 locktorture.acq_writer_lim= [KNL] 3542 Set the time limit in jiffies for a lock 3543 acquisition. Acquisitions exceeding this limit 3544 will result in a splat once they do complete. 3545 3546 locktorture.bind_readers= [KNL] 3547 Specify the list of CPUs to which the readers are 3548 to be bound. 3549 3550 locktorture.bind_writers= [KNL] 3551 Specify the list of CPUs to which the writers are 3552 to be bound. 3553 3554 locktorture.call_rcu_chains= [KNL] 3555 Specify the number of self-propagating call_rcu() 3556 chains to set up. These are used to ensure that 3557 there is a high probability of an RCU grace period 3558 in progress at any given time. Defaults to 0, 3559 which disables these call_rcu() chains. 3560 3561 locktorture.long_hold= [KNL] 3562 Specify the duration in milliseconds for the 3563 occasional long-duration lock hold time. Defaults 3564 to 100 milliseconds. Select 0 to disable. 3565 3566 locktorture.nested_locks= [KNL] 3567 Specify the maximum lock nesting depth that 3568 locktorture is to exercise, up to a limit of 8 3569 (MAX_NESTED_LOCKS). Specify zero to disable. 3570 Note that this parameter is ineffective on types 3571 of locks that do not support nested acquisition. 3572 3573 locktorture.nreaders_stress= [KNL] 3574 Set the number of locking read-acquisition kthreads. 3575 Defaults to being automatically set based on the 3576 number of online CPUs. 3577 3578 locktorture.nwriters_stress= [KNL] 3579 Set the number of locking write-acquisition kthreads. 3580 3581 locktorture.onoff_holdoff= [KNL] 3582 Set time (s) after boot for CPU-hotplug testing. 3583 3584 locktorture.onoff_interval= [KNL] 3585 Set time (s) between CPU-hotplug operations, or 3586 zero to disable CPU-hotplug testing. 3587 3588 locktorture.rt_boost= [KNL] 3589 Do periodic testing of real-time lock priority 3590 boosting. Select 0 to disable, 1 to boost 3591 only rt_mutex, and 2 to boost unconditionally. 3592 Defaults to 2, which might seem to be an 3593 odd choice, but which should be harmless for 3594 non-real-time spinlocks, due to their disabling 3595 of preemption. Note that non-realtime mutexes 3596 disable boosting. 3597 3598 locktorture.rt_boost_factor= [KNL] 3599 Number that determines how often and for how 3600 long priority boosting is exercised. This is 3601 scaled down by the number of writers, so that the 3602 number of boosts per unit time remains roughly 3603 constant as the number of writers increases. 3604 On the other hand, the duration of each boost 3605 increases with the number of writers. 3606 3607 locktorture.shuffle_interval= [KNL] 3608 Set task-shuffle interval (jiffies). Shuffling 3609 tasks allows some CPUs to go into dyntick-idle 3610 mode during the locktorture test. 3611 3612 locktorture.shutdown_secs= [KNL] 3613 Set time (s) after boot system shutdown. This 3614 is useful for hands-off automated testing. 3615 3616 locktorture.stat_interval= [KNL] 3617 Time (s) between statistics printk()s. 3618 3619 locktorture.stutter= [KNL] 3620 Time (s) to stutter testing, for example, 3621 specifying five seconds causes the test to run for 3622 five seconds, wait for five seconds, and so on. 3623 This tests the locking primitive's ability to 3624 transition abruptly to and from idle. 3625 3626 locktorture.torture_type= [KNL] 3627 Specify the locking implementation to test. 3628 3629 locktorture.verbose= [KNL] 3630 Enable additional printk() statements. 3631 3632 locktorture.writer_fifo= [KNL] 3633 Run the write-side locktorture kthreads at 3634 sched_set_fifo() real-time priority. 3635 3636 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 3637 Format: <irq> 3638 3639 loglevel= [KNL,EARLY] 3640 All Kernel Messages with a loglevel smaller than the 3641 console loglevel will be printed to the console. It can 3642 also be changed with klogd or other programs. The 3643 loglevels are defined as follows: 3644 3645 0 (KERN_EMERG) system is unusable 3646 1 (KERN_ALERT) action must be taken immediately 3647 2 (KERN_CRIT) critical conditions 3648 3 (KERN_ERR) error conditions 3649 4 (KERN_WARNING) warning conditions 3650 5 (KERN_NOTICE) normal but significant condition 3651 6 (KERN_INFO) informational 3652 7 (KERN_DEBUG) debug-level messages 3653 3654 log_buf_len=n[KMG] [KNL,EARLY] 3655 Sets the size of the printk ring buffer, in bytes. 3656 n must be a power of two and greater than the 3657 minimal size. The minimal size is defined by 3658 LOG_BUF_SHIFT kernel config parameter. There 3659 is also CONFIG_LOG_CPU_MAX_BUF_SHIFT config 3660 parameter that allows to increase the default size 3661 depending on the number of CPUs. See init/Kconfig 3662 for more details. 3663 3664 logo.nologo [FB] Disables display of the built-in Linux logo. 3665 This may be used to provide more screen space for 3666 kernel log messages and is useful when debugging 3667 kernel boot problems. 3668 3669 lp=0 [LP] Specify parallel ports to use, e.g, 3670 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 3671 lp=reset first parallel port). 'lp=0' disables the 3672 lp=auto printer driver. 'lp=reset' (which can be 3673 specified in addition to the ports) causes 3674 attached printers to be reset. Using 3675 lp=port1,port2,... specifies the parallel ports 3676 to associate lp devices with, starting with 3677 lp0. A port specification may be 'none' to skip 3678 that lp device, or a parport name such as 3679 'parport0'. Specifying 'lp=auto' instead of a 3680 port specification list means that device IDs 3681 from each port should be examined, to see if 3682 an IEEE 1284-compliant printer is attached; if 3683 so, the driver will manage that printer. 3684 See also header of drivers/char/lp.c. 3685 3686 lpj=n [KNL] 3687 Sets loops_per_jiffy to given constant, thus avoiding 3688 time-consuming boot-time autodetection (up to 250 ms per 3689 CPU). 0 enables autodetection (default). To determine 3690 the correct value for your kernel, boot with normal 3691 autodetection and see what value is printed. Note that 3692 on SMP systems the preset will be applied to all CPUs, 3693 which is likely to cause problems if your CPUs need 3694 significantly divergent settings. An incorrect value 3695 will cause delays in the kernel to be wrong, leading to 3696 unpredictable I/O errors and other breakage. Although 3697 unlikely, in the extreme case this might damage your 3698 hardware. 3699 3700 lsm.debug [SECURITY] Enable LSM initialization debugging output. 3701 3702 lsm=lsm1,...,lsmN 3703 [SECURITY] Choose order of LSM initialization. This 3704 overrides CONFIG_LSM, and the "security=" parameter. 3705 3706 machtype= [Loongson] Share the same kernel image file between 3707 different yeeloong laptops. 3708 Example: machtype=lemote-yeeloong-2f-7inch 3709 3710 maxcpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 3711 will bring up during bootup. maxcpus=n : n >= 0 limits 3712 the kernel to bring up 'n' processors. Surely after 3713 bootup you can bring up the other plugged cpu by executing 3714 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 3715 only takes effect during system bootup. 3716 While n=0 is a special case, it is equivalent to "nosmp", 3717 which also disables the IO APIC. 3718 3719 max_loop= [LOOP] The number of loop block devices that get 3720 (loop.max_loop) unconditionally pre-created at init time. The default 3721 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 3722 of statically allocating a predefined number, loop 3723 devices can be requested on-demand with the 3724 /dev/loop-control interface. 3725 3726 mce= [X86-{32,64}] 3727 3728 Please see Documentation/arch/x86/x86_64/machinecheck.rst for sysfs runtime tunables. 3729 3730 off 3731 disable machine check 3732 3733 no_cmci 3734 disable CMCI(Corrected Machine Check Interrupt) that 3735 Intel processor supports. Usually this disablement is 3736 not recommended, but it might be handy if your 3737 hardware is misbehaving. 3738 3739 Note that you'll get more problems without CMCI than 3740 with due to the shared banks, i.e. you might get 3741 duplicated error logs. 3742 3743 dont_log_ce 3744 don't make logs for corrected errors. All events 3745 reported as corrected are silently cleared by OS. This 3746 option will be useful if you have no interest in any 3747 of corrected errors. 3748 3749 ignore_ce 3750 disable features for corrected errors, e.g. 3751 polling timer and CMCI. All events reported as 3752 corrected are not cleared by OS and remained in its 3753 error banks. 3754 3755 Usually this disablement is not recommended, however 3756 if there is an agent checking/clearing corrected 3757 errors (e.g. BIOS or hardware monitoring 3758 applications), conflicting with OS's error handling, 3759 and you cannot deactivate the agent, then this option 3760 will be a help. 3761 3762 no_lmce 3763 do not opt-in to Local MCE delivery. Use legacy method 3764 to broadcast MCEs. 3765 3766 bootlog 3767 enable logging of machine checks left over from 3768 booting. Disabled by default on AMD Fam10h and older 3769 because some BIOS leave bogus ones. 3770 3771 If your BIOS doesn't do that it's a good idea to 3772 enable though to make sure you log even machine check 3773 events that result in a reboot. On Intel systems it is 3774 enabled by default. 3775 3776 nobootlog 3777 disable boot machine check logging. 3778 3779 monarchtimeout (number) 3780 sets the time in us to wait for other CPUs on machine 3781 checks. 0 to disable. 3782 3783 bios_cmci_threshold 3784 don't overwrite the bios-set CMCI threshold. This boot 3785 option prevents Linux from overwriting the CMCI 3786 threshold set by the bios. Without this option, Linux 3787 always sets the CMCI threshold to 1. Enabling this may 3788 make memory predictive failure analysis less effective 3789 if the bios sets thresholds for memory errors since we 3790 will not see details for all errors. 3791 3792 recovery 3793 force-enable recoverable machine check code paths 3794 3795 Everything else is in sysfs now. 3796 3797 3798 md= [HW] RAID subsystems devices and level 3799 See Documentation/admin-guide/md.rst. 3800 3801 mdacon= [MDA] 3802 Format: <first>,<last> 3803 Specifies range of consoles to be captured by the MDA. 3804 3805 mds= [X86,INTEL,EARLY] 3806 Control mitigation for the Micro-architectural Data 3807 Sampling (MDS) vulnerability. 3808 3809 Certain CPUs are vulnerable to an exploit against CPU 3810 internal buffers which can forward information to a 3811 disclosure gadget under certain conditions. 3812 3813 In vulnerable processors, the speculatively 3814 forwarded data can be used in a cache side channel 3815 attack, to access data to which the attacker does 3816 not have direct access. 3817 3818 This parameter controls the MDS mitigation. The 3819 options are: 3820 3821 full - Enable MDS mitigation on vulnerable CPUs 3822 full,nosmt - Enable MDS mitigation and disable 3823 SMT on vulnerable CPUs 3824 off - Unconditionally disable MDS mitigation 3825 3826 On TAA-affected machines, mds=off can be prevented by 3827 an active TAA mitigation as both vulnerabilities are 3828 mitigated with the same mechanism so in order to disable 3829 this mitigation, you need to specify tsx_async_abort=off 3830 too. 3831 3832 Not specifying this option is equivalent to 3833 mds=full. 3834 3835 For details see: Documentation/admin-guide/hw-vuln/mds.rst 3836 3837 mem=nn[KMG] [HEXAGON,EARLY] Set the memory size. 3838 Must be specified, otherwise memory size will be 0. 3839 3840 mem=nn[KMG] [KNL,BOOT,EARLY] Force usage of a specific amount 3841 of memory Amount of memory to be used in cases 3842 as follows: 3843 3844 1 for test; 3845 2 when the kernel is not able to see the whole system memory; 3846 3 memory that lies after 'mem=' boundary is excluded from 3847 the hypervisor, then assigned to KVM guests. 3848 4 to limit the memory available for kdump kernel. 3849 3850 [ARC,MICROBLAZE] - the limit applies only to low memory, 3851 high memory is not affected. 3852 3853 [ARM64] - only limits memory covered by the linear 3854 mapping. The NOMAP regions are not affected. 3855 3856 [X86] Work as limiting max address. Use together 3857 with memmap= to avoid physical address space collisions. 3858 Without memmap= PCI devices could be placed at addresses 3859 belonging to unused RAM. 3860 3861 Note that this only takes effects during boot time since 3862 in above case 3, memory may need be hot added after boot 3863 if system memory of hypervisor is not sufficient. 3864 3865 mem=nn[KMG]@ss[KMG] 3866 [ARM,MIPS,EARLY] - override the memory layout 3867 reported by firmware. 3868 Define a memory region of size nn[KMG] starting at 3869 ss[KMG]. 3870 Multiple different regions can be specified with 3871 multiple mem= parameters on the command line. 3872 3873 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 3874 memory. 3875 3876 memblock=debug [KNL,EARLY] Enable memblock debug messages. 3877 3878 memchunk=nn[KMG] 3879 [KNL,SH] Allow user to override the default size for 3880 per-device physically contiguous DMA buffers. 3881 3882 memhp_default_state=online/offline/online_kernel/online_movable 3883 [KNL] Set the initial state for the memory hotplug 3884 onlining policy. If not specified, the default value is 3885 set according to the 3886 CONFIG_MHP_DEFAULT_ONLINE_TYPE kernel config 3887 options. 3888 See Documentation/admin-guide/mm/memory-hotplug.rst. 3889 3890 memmap=exactmap [KNL,X86,EARLY] Enable setting of an exact 3891 E820 memory map, as specified by the user. 3892 Such memmap=exactmap lines can be constructed based on 3893 BIOS output or other requirements. See the memmap=nn@ss 3894 option description. 3895 3896 memmap=nn[KMG]@ss[KMG] 3897 [KNL, X86,MIPS,XTENSA,EARLY] Force usage of a specific region of memory. 3898 Region of memory to be used is from ss to ss+nn. 3899 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 3900 which limits max address to nn[KMG]. 3901 Multiple different regions can be specified, 3902 comma delimited. 3903 Example: 3904 memmap=100M@2G,100M#3G,1G!1024G 3905 3906 memmap=nn[KMG]#ss[KMG] 3907 [KNL,ACPI,EARLY] Mark specific memory as ACPI data. 3908 Region of memory to be marked is from ss to ss+nn. 3909 3910 memmap=nn[KMG]$ss[KMG] 3911 [KNL,ACPI,EARLY] Mark specific memory as reserved. 3912 Region of memory to be reserved is from ss to ss+nn. 3913 Example: Exclude memory from 0x18690000-0x1869ffff 3914 memmap=64K$0x18690000 3915 or 3916 memmap=0x10000$0x18690000 3917 Some bootloaders may need an escape character before '$', 3918 like Grub2, otherwise '$' and the following number 3919 will be eaten. 3920 3921 memmap=nn[KMG]!ss[KMG,EARLY] 3922 [KNL,X86] Mark specific memory as protected. 3923 Region of memory to be used, from ss to ss+nn. 3924 The memory region may be marked as e820 type 12 (0xc) 3925 and is NVDIMM or ADR memory. 3926 3927 memmap=<size>%<offset>-<oldtype>+<newtype> 3928 [KNL,ACPI,EARLY] Convert memory within the specified region 3929 from <oldtype> to <newtype>. If "-<oldtype>" is left 3930 out, the whole region will be marked as <newtype>, 3931 even if previously unavailable. If "+<newtype>" is left 3932 out, matching memory will be removed. Types are 3933 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 3934 3 = ACPI, 12 = PRAM. 3935 3936 memory_corruption_check=0/1 [X86,EARLY] 3937 Some BIOSes seem to corrupt the first 64k of 3938 memory when doing things like suspend/resume. 3939 Setting this option will scan the memory 3940 looking for corruption. Enabling this will 3941 both detect corruption and prevent the kernel 3942 from using the memory being corrupted. 3943 However, it's intended as a diagnostic tool; if 3944 repeatable BIOS-originated corruption always 3945 affects the same memory, you can use memmap= 3946 to prevent the kernel from using that memory. 3947 3948 memory_corruption_check_size=size [X86,EARLY] 3949 By default it checks for corruption in the low 3950 64k, making this memory unavailable for normal 3951 use. Use this parameter to scan for 3952 corruption in more or less memory. 3953 3954 memory_corruption_check_period=seconds [X86,EARLY] 3955 By default it checks for corruption every 60 3956 seconds. Use this parameter to check at some 3957 other rate. 0 disables periodic checking. 3958 3959 memory_hotplug.memmap_on_memory 3960 [KNL,X86,ARM] Boolean flag to enable this feature. 3961 Format: {on | off (default)} 3962 When enabled, runtime hotplugged memory will 3963 allocate its internal metadata (struct pages, 3964 those vmemmap pages cannot be optimized even 3965 if hugetlb_free_vmemmap is enabled) from the 3966 hotadded memory which will allow to hotadd a 3967 lot of memory without requiring additional 3968 memory to do so. 3969 This feature is disabled by default because it 3970 has some implication on large (e.g. GB) 3971 allocations in some configurations (e.g. small 3972 memory blocks). 3973 The state of the flag can be read in 3974 /sys/module/memory_hotplug/parameters/memmap_on_memory. 3975 Note that even when enabled, there are a few cases where 3976 the feature is not effective. 3977 3978 memtest= [KNL,X86,ARM,M68K,PPC,RISCV,EARLY] Enable memtest 3979 Format: <integer> 3980 default : 0 <disable> 3981 Specifies the number of memtest passes to be 3982 performed. Each pass selects another test 3983 pattern from a given set of patterns. Memtest 3984 fills the memory with this pattern, validates 3985 memory contents and reserves bad memory 3986 regions that are detected. 3987 3988 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 3989 Valid arguments: on, off 3990 Default: off 3991 mem_encrypt=on: Activate SME 3992 mem_encrypt=off: Do not activate SME 3993 3994 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst 3995 for details on when memory encryption can be activated. 3996 3997 mem_sleep_default= [SUSPEND] Default system suspend mode: 3998 s2idle - Suspend-To-Idle 3999 shallow - Power-On Suspend or equivalent (if supported) 4000 deep - Suspend-To-RAM or equivalent (if supported) 4001 See Documentation/admin-guide/pm/sleep-states.rst. 4002 4003 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 4004 the BIOS has incorrectly applied a workaround. TinyBIOS 4005 version 0.98 is known to be affected, 0.99 fixes the 4006 problem by letting the user disable the workaround. 4007 4008 mga= [HW,DRM] 4009 4010 microcode= [X86] Control the behavior of the microcode loader. 4011 Available options, comma separated: 4012 4013 base_rev=X - with <X> with format: <u32> 4014 Set the base microcode revision of each thread when in 4015 debug mode. 4016 4017 dis_ucode_ldr: disable the microcode loader 4018 4019 force_minrev: 4020 Enable or disable the microcode minimal revision 4021 enforcement for the runtime microcode loader. 4022 4023 mini2440= [ARM,HW,KNL] 4024 Format:[0..2][b][c][t] 4025 Default: "0tb" 4026 MINI2440 configuration specification: 4027 0 - The attached screen is the 3.5" TFT 4028 1 - The attached screen is the 7" TFT 4029 2 - The VGA Shield is attached (1024x768) 4030 Leaving out the screen size parameter will not load 4031 the TFT driver, and the framebuffer will be left 4032 unconfigured. 4033 b - Enable backlight. The TFT backlight pin will be 4034 linked to the kernel VESA blanking code and a GPIO 4035 LED. This parameter is not necessary when using the 4036 VGA shield. 4037 c - Enable the s3c camera interface. 4038 t - Reserved for enabling touchscreen support. The 4039 touchscreen support is not enabled in the mainstream 4040 kernel as of 2.6.30, a preliminary port can be found 4041 in the "bleeding edge" mini2440 support kernel at 4042 https://repo.or.cz/w/linux-2.6/mini2440.git 4043 4044 mitigations= 4045 [X86,PPC,S390,ARM64,EARLY] Control optional mitigations for 4046 CPU vulnerabilities. This is a set of curated, 4047 arch-independent options, each of which is an 4048 aggregation of existing arch-specific options. 4049 4050 Note, "mitigations" is supported if and only if the 4051 kernel was built with CPU_MITIGATIONS=y. 4052 4053 off 4054 Disable all optional CPU mitigations. This 4055 improves system performance, but it may also 4056 expose users to several CPU vulnerabilities. 4057 Equivalent to: if nokaslr then kpti=0 [ARM64] 4058 gather_data_sampling=off [X86] 4059 indirect_target_selection=off [X86] 4060 kvm.nx_huge_pages=off [X86] 4061 l1tf=off [X86] 4062 mds=off [X86] 4063 mmio_stale_data=off [X86] 4064 no_entry_flush [PPC] 4065 no_uaccess_flush [PPC] 4066 nobp=0 [S390] 4067 nopti [X86,PPC] 4068 nospectre_bhb [ARM64] 4069 nospectre_v1 [X86,PPC] 4070 nospectre_v2 [X86,PPC,S390,ARM64] 4071 reg_file_data_sampling=off [X86] 4072 retbleed=off [X86] 4073 spec_rstack_overflow=off [X86] 4074 spec_store_bypass_disable=off [X86,PPC] 4075 spectre_bhi=off [X86] 4076 spectre_v2_user=off [X86] 4077 srbds=off [X86,INTEL] 4078 ssbd=force-off [ARM64] 4079 tsa=off [X86,AMD] 4080 tsx_async_abort=off [X86] 4081 vmscape=off [X86] 4082 4083 Exceptions: 4084 This does not have any effect on 4085 kvm.nx_huge_pages when 4086 kvm.nx_huge_pages=force. 4087 4088 auto (default) 4089 Mitigate all CPU vulnerabilities, but leave SMT 4090 enabled, even if it's vulnerable. This is for 4091 users who don't want to be surprised by SMT 4092 getting disabled across kernel upgrades, or who 4093 have other ways of avoiding SMT-based attacks. 4094 Equivalent to: (default behavior) 4095 4096 auto,nosmt 4097 Mitigate all CPU vulnerabilities, disabling SMT 4098 if needed. This is for users who always want to 4099 be fully mitigated, even if it means losing SMT. 4100 Equivalent to: l1tf=flush,nosmt [X86] 4101 mds=full,nosmt [X86] 4102 tsx_async_abort=full,nosmt [X86] 4103 mmio_stale_data=full,nosmt [X86] 4104 retbleed=auto,nosmt [X86] 4105 4106 [X86] After one of the above options, additionally 4107 supports attack-vector based controls as documented in 4108 Documentation/admin-guide/hw-vuln/attack_vector_controls.rst 4109 4110 mminit_loglevel= 4111 [KNL,EARLY] When CONFIG_DEBUG_MEMORY_INIT is set, this 4112 parameter allows control of the logging verbosity for 4113 the additional memory initialisation checks. A value 4114 of 0 disables mminit logging and a level of 4 will 4115 log everything. Information is printed at KERN_DEBUG 4116 so loglevel=8 may also need to be specified. 4117 4118 mmio_stale_data= 4119 [X86,INTEL,EARLY] Control mitigation for the Processor 4120 MMIO Stale Data vulnerabilities. 4121 4122 Processor MMIO Stale Data is a class of 4123 vulnerabilities that may expose data after an MMIO 4124 operation. Exposed data could originate or end in 4125 the same CPU buffers as affected by MDS and TAA. 4126 Therefore, similar to MDS and TAA, the mitigation 4127 is to clear the affected CPU buffers. 4128 4129 This parameter controls the mitigation. The 4130 options are: 4131 4132 full - Enable mitigation on vulnerable CPUs 4133 4134 full,nosmt - Enable mitigation and disable SMT on 4135 vulnerable CPUs. 4136 4137 off - Unconditionally disable mitigation 4138 4139 On MDS or TAA affected machines, 4140 mmio_stale_data=off can be prevented by an active 4141 MDS or TAA mitigation as these vulnerabilities are 4142 mitigated with the same mechanism so in order to 4143 disable this mitigation, you need to specify 4144 mds=off and tsx_async_abort=off too. 4145 4146 Not specifying this option is equivalent to 4147 mmio_stale_data=full. 4148 4149 For details see: 4150 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst 4151 4152 <module>.async_probe[=<bool>] [KNL] 4153 If no <bool> value is specified or if the value 4154 specified is not a valid <bool>, enable asynchronous 4155 probe on this module. Otherwise, enable/disable 4156 asynchronous probe on this module as indicated by the 4157 <bool> value. See also: module.async_probe 4158 4159 module.async_probe=<bool> 4160 [KNL] When set to true, modules will use async probing 4161 by default. To enable/disable async probing for a 4162 specific module, use the module specific control that 4163 is documented under <module>.async_probe. When both 4164 module.async_probe and <module>.async_probe are 4165 specified, <module>.async_probe takes precedence for 4166 the specific module. 4167 4168 module.enable_dups_trace 4169 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set, 4170 this means that duplicate request_module() calls will 4171 trigger a WARN_ON() instead of a pr_warn(). Note that 4172 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s 4173 will always be issued and this option does nothing. 4174 module.sig_enforce 4175 [KNL] When CONFIG_MODULE_SIG is set, this means that 4176 modules without (valid) signatures will fail to load. 4177 Note that if CONFIG_MODULE_SIG_FORCE is set, that 4178 is always true, so this option does nothing. 4179 4180 module_blacklist= [KNL] Do not load a comma-separated list of 4181 modules. Useful for debugging problem modules. 4182 4183 mousedev.tap_time= 4184 [MOUSE] Maximum time between finger touching and 4185 leaving touchpad surface for touch to be considered 4186 a tap and be reported as a left button click (for 4187 touchpads working in absolute mode only). 4188 Format: <msecs> 4189 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 4190 reporting absolute coordinates, such as tablets 4191 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 4192 reporting absolute coordinates, such as tablets 4193 4194 movablecore= [KNL,X86,PPC,EARLY] 4195 Format: nn[KMGTPE] | nn% 4196 This parameter is the complement to kernelcore=, it 4197 specifies the amount of memory used for migratable 4198 allocations. If both kernelcore and movablecore is 4199 specified, then kernelcore will be at *least* the 4200 specified value but may be more. If movablecore on its 4201 own is specified, the administrator must be careful 4202 that the amount of memory usable for all allocations 4203 is not too small. 4204 4205 movable_node [KNL,EARLY] Boot-time switch to make hotplugable memory 4206 NUMA nodes to be movable. This means that the memory 4207 of such nodes will be usable only for movable 4208 allocations which rules out almost all kernel 4209 allocations. Use with caution! 4210 4211 MTD_Partition= [MTD] 4212 Format: <name>,<region-number>,<size>,<offset> 4213 4214 MTD_Region= [MTD] Format: 4215 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 4216 4217 mtdparts= [MTD] 4218 See drivers/mtd/parsers/cmdlinepart.c 4219 4220 mtouchusb.raw_coordinates= 4221 [HW] Make the MicroTouch USB driver use raw coordinates 4222 ('y', default) or cooked coordinates ('n') 4223 4224 mtrr=debug [X86,EARLY] 4225 Enable printing debug information related to MTRR 4226 registers at boot time. 4227 4228 mtrr_chunk_size=nn[KMG,X86,EARLY] 4229 used for mtrr cleanup. It is largest continuous chunk 4230 that could hold holes aka. UC entries. 4231 4232 mtrr_gran_size=nn[KMG,X86,EARLY] 4233 Used for mtrr cleanup. It is granularity of mtrr block. 4234 Default is 1. 4235 Large value could prevent small alignment from 4236 using up MTRRs. 4237 4238 mtrr_spare_reg_nr=n [X86,EARLY] 4239 Format: <integer> 4240 Range: 0,7 : spare reg number 4241 Default : 1 4242 Used for mtrr cleanup. It is spare mtrr entries number. 4243 Set to 2 or more if your graphical card needs more. 4244 4245 multitce=off [PPC] This parameter disables the use of the pSeries 4246 firmware feature for updating multiple TCE entries 4247 at a time. 4248 4249 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 4250 4251 netdev= [NET] Network devices parameters 4252 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 4253 Note that mem_start is often overloaded to mean 4254 something different and driver-specific. 4255 This usage is only documented in each driver source 4256 file if at all. 4257 4258 netpoll.carrier_timeout= 4259 [NET] Specifies amount of time (in seconds) that 4260 netpoll should wait for a carrier. By default netpoll 4261 waits 4 seconds. 4262 4263 nf_conntrack.acct= 4264 [NETFILTER] Enable connection tracking flow accounting 4265 0 to disable accounting 4266 1 to enable accounting 4267 Default value is 0. 4268 4269 nfs.cache_getent= 4270 [NFS] sets the pathname to the program which is used 4271 to update the NFS client cache entries. 4272 4273 nfs.cache_getent_timeout= 4274 [NFS] sets the timeout after which an attempt to 4275 update a cache entry is deemed to have failed. 4276 4277 nfs.callback_nr_threads= 4278 [NFSv4] set the total number of threads that the 4279 NFS client will assign to service NFSv4 callback 4280 requests. 4281 4282 nfs.callback_tcpport= 4283 [NFS] set the TCP port on which the NFSv4 callback 4284 channel should listen. 4285 4286 nfs.delay_retrans= 4287 [NFS] specifies the number of times the NFSv4 client 4288 retries the request before returning an EAGAIN error, 4289 after a reply of NFS4ERR_DELAY from the server. 4290 Only applies if the softerr mount option is enabled, 4291 and the specified value is >= 0. 4292 4293 nfs.enable_ino64= 4294 [NFS] enable 64-bit inode numbers. 4295 If zero, the NFS client will fake up a 32-bit inode 4296 number for the readdir() and stat() syscalls instead 4297 of returning the full 64-bit number. 4298 The default is to return 64-bit inode numbers. 4299 4300 nfs.idmap_cache_timeout= 4301 [NFS] set the maximum lifetime for idmapper cache 4302 entries. 4303 4304 nfs.max_session_cb_slots= 4305 [NFSv4.1] Sets the maximum number of session 4306 slots the client will assign to the callback 4307 channel. This determines the maximum number of 4308 callbacks the client will process in parallel for 4309 a particular server. 4310 4311 nfs.max_session_slots= 4312 [NFSv4.1] Sets the maximum number of session slots 4313 the client will attempt to negotiate with the server. 4314 This limits the number of simultaneous RPC requests 4315 that the client can send to the NFSv4.1 server. 4316 Note that there is little point in setting this 4317 value higher than the max_tcp_slot_table_limit. 4318 4319 nfs.nfs4_disable_idmapping= 4320 [NFSv4] When set to the default of '1', this option 4321 ensures that both the RPC level authentication 4322 scheme and the NFS level operations agree to use 4323 numeric uids/gids if the mount is using the 4324 'sec=sys' security flavour. In effect it is 4325 disabling idmapping, which can make migration from 4326 legacy NFSv2/v3 systems to NFSv4 easier. 4327 Servers that do not support this mode of operation 4328 will be autodetected by the client, and it will fall 4329 back to using the idmapper. 4330 To turn off this behaviour, set the value to '0'. 4331 4332 nfs.nfs4_unique_id= 4333 [NFS4] Specify an additional fixed unique ident- 4334 ification string that NFSv4 clients can insert into 4335 their nfs_client_id4 string. This is typically a 4336 UUID that is generated at system install time. 4337 4338 nfs.recover_lost_locks= 4339 [NFSv4] Attempt to recover locks that were lost due 4340 to a lease timeout on the server. Please note that 4341 doing this risks data corruption, since there are 4342 no guarantees that the file will remain unchanged 4343 after the locks are lost. 4344 If you want to enable the kernel legacy behaviour of 4345 attempting to recover these locks, then set this 4346 parameter to '1'. 4347 The default parameter value of '0' causes the kernel 4348 not to attempt recovery of lost locks. 4349 4350 nfs.send_implementation_id= 4351 [NFSv4.1] Send client implementation identification 4352 information in exchange_id requests. 4353 If zero, no implementation identification information 4354 will be sent. 4355 The default is to send the implementation identification 4356 information. 4357 4358 nfs4.layoutstats_timer= 4359 [NFSv4.2] Change the rate at which the kernel sends 4360 layoutstats to the pNFS metadata server. 4361 4362 Setting this to value to 0 causes the kernel to use 4363 whatever value is the default set by the layout 4364 driver. A non-zero value sets the minimum interval 4365 in seconds between layoutstats transmissions. 4366 4367 nfsd.inter_copy_offload_enable= 4368 [NFSv4.2] When set to 1, the server will support 4369 server-to-server copies for which this server is 4370 the destination of the copy. 4371 4372 nfsd.nfs4_disable_idmapping= 4373 [NFSv4] When set to the default of '1', the NFSv4 4374 server will return only numeric uids and gids to 4375 clients using auth_sys, and will accept numeric uids 4376 and gids from such clients. This is intended to ease 4377 migration from NFSv2/v3. 4378 4379 nfsd.nfsd4_ssc_umount_timeout= 4380 [NFSv4.2] When used as the destination of a 4381 server-to-server copy, knfsd temporarily mounts 4382 the source server. It caches the mount in case 4383 it will be needed again, and discards it if not 4384 used for the number of milliseconds specified by 4385 this parameter. 4386 4387 nfsaddrs= [NFS] Deprecated. Use ip= instead. 4388 See Documentation/admin-guide/nfs/nfsroot.rst. 4389 4390 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 4391 See Documentation/admin-guide/nfs/nfsroot.rst. 4392 4393 nfsrootdebug [NFS] enable nfsroot debugging messages. 4394 See Documentation/admin-guide/nfs/nfsroot.rst. 4395 4396 nmi_backtrace.backtrace_idle [KNL] 4397 Dump stacks even of idle CPUs in response to an 4398 NMI stack-backtrace request. 4399 4400 nmi_debug= [KNL,SH] Specify one or more actions to take 4401 when a NMI is triggered. 4402 Format: [state][,regs][,debounce][,die] 4403 4404 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 4405 Format: [panic,][nopanic,][rNNN,][num] 4406 Valid num: 0 or 1 4407 0 - turn hardlockup detector in nmi_watchdog off 4408 1 - turn hardlockup detector in nmi_watchdog on 4409 rNNN - configure the watchdog with raw perf event 0xNNN 4410 4411 When panic is specified, panic when an NMI watchdog 4412 timeout occurs (or 'nopanic' to not panic on an NMI 4413 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 4414 To disable both hard and soft lockup detectors, 4415 please see 'nowatchdog'. 4416 This is useful when you use a panic=... timeout and 4417 need the box quickly up again. 4418 4419 These settings can be accessed at runtime via 4420 the nmi_watchdog and hardlockup_panic sysctls. 4421 4422 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 4423 emulation library even if a 387 maths coprocessor 4424 is present. 4425 4426 no4lvl [RISCV,EARLY] Disable 4-level and 5-level paging modes. 4427 Forces kernel to use 3-level paging instead. 4428 4429 no5lvl [X86-64,RISCV,EARLY] Disable 5-level paging mode. Forces 4430 kernel to use 4-level paging instead. 4431 4432 noalign [KNL,ARM] 4433 4434 noapic [SMP,APIC,EARLY] Tells the kernel to not make use of any 4435 IOAPICs that may be present in the system. 4436 4437 noapictimer [APIC,X86] Don't set up the APIC timer 4438 4439 noautogroup Disable scheduler automatic task group creation. 4440 4441 nocache [ARM,EARLY] 4442 4443 no_console_suspend 4444 [HW] Never suspend the console 4445 Disable suspending of consoles during suspend and 4446 hibernate operations. Once disabled, debugging 4447 messages can reach various consoles while the rest 4448 of the system is being put to sleep (ie, while 4449 debugging driver suspend/resume hooks). This may 4450 not work reliably with all consoles, but is known 4451 to work with serial and VGA consoles. 4452 To facilitate more flexible debugging, we also add 4453 console_suspend, a printk module parameter to control 4454 it. Users could use console_suspend (usually 4455 /sys/module/printk/parameters/console_suspend) to 4456 turn on/off it dynamically. 4457 4458 no_debug_objects 4459 [KNL,EARLY] Disable object debugging 4460 4461 nodsp [SH] Disable hardware DSP at boot time. 4462 4463 noefi [EFI,EARLY] Disable EFI runtime services support. 4464 4465 no_entry_flush [PPC,EARLY] Don't flush the L1-D cache when entering the kernel. 4466 4467 noexec32 [X86-64] 4468 This affects only 32-bit executables. 4469 noexec32=on: enable non-executable mappings (default) 4470 read doesn't imply executable mappings 4471 noexec32=off: disable non-executable mappings 4472 read implies executable mappings 4473 4474 no_file_caps Tells the kernel not to honor file capabilities. The 4475 only way then for a file to be executed with privilege 4476 is to be setuid root or executed by root. 4477 4478 nofpu [MIPS,SH] Disable hardware FPU at boot time. 4479 4480 nofsgsbase [X86] Disables FSGSBASE instructions. 4481 4482 nofxsr [BUGS=X86-32] Disables x86 floating point extended 4483 register save and restore. The kernel will only save 4484 legacy floating-point registers on task switch. 4485 4486 nogbpages [X86] Do not use GB pages for kernel direct mappings. 4487 4488 no_hash_pointers 4489 [KNL,EARLY] 4490 Alias for "hash_pointers=never". 4491 4492 nohibernate [HIBERNATION] Disable hibernation and resume. 4493 4494 nohlt [ARM,ARM64,MICROBLAZE,MIPS,PPC,RISCV,SH] Forces the kernel to 4495 busy wait in do_idle() and not use the arch_cpu_idle() 4496 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP 4497 to be effective. This is useful on platforms where the 4498 sleep(SH) or wfi(ARM,ARM64) instructions do not work 4499 correctly or when doing power measurements to evaluate 4500 the impact of the sleep instructions. This is also 4501 useful when using JTAG debugger. 4502 4503 nohpet [X86] Don't use the HPET timer. 4504 4505 nohugeiomap [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge I/O mappings. 4506 4507 nohugevmalloc [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge vmalloc mappings. 4508 4509 nohz= [KNL] Boottime enable/disable dynamic ticks 4510 Valid arguments: on, off 4511 Default: on 4512 4513 nohz_full= [KNL,BOOT,SMP,ISOL] 4514 The argument is a cpu list, as described above. 4515 In kernels built with CONFIG_NO_HZ_FULL=y, set 4516 the specified list of CPUs whose tick will be stopped 4517 whenever possible. The boot CPU will be forced outside 4518 the range to maintain the timekeeping. Any CPUs 4519 in this list will have their RCU callbacks offloaded, 4520 just as if they had also been called out in the 4521 rcu_nocbs= boot parameter. 4522 4523 Note that this argument takes precedence over 4524 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 4525 4526 noinitrd [Deprecated,RAM] Tells the kernel not to load any configured 4527 initial RAM disk. Currently this parameter applies to 4528 initrd only, not to initramfs. But it applies to both 4529 in EFI mode. 4530 4531 nointremap [X86-64,Intel-IOMMU,EARLY] Do not enable interrupt 4532 remapping. 4533 [Deprecated - use intremap=off] 4534 4535 noinvpcid [X86,EARLY] Disable the INVPCID cpu feature. 4536 4537 noiotrap [SH] Disables trapped I/O port accesses. 4538 4539 noirqdebug [X86-32] Disables the code which attempts to detect and 4540 disable unhandled interrupt sources. 4541 4542 noisapnp [ISAPNP] Disables ISA PnP code. 4543 4544 nokaslr [KNL,EARLY] 4545 When CONFIG_RANDOMIZE_BASE is set, this disables 4546 kernel and module base offset ASLR (Address Space 4547 Layout Randomization). 4548 4549 no-kvmapf [X86,KVM,EARLY] Disable paravirtualized asynchronous page 4550 fault handling. 4551 4552 no-kvmclock [X86,KVM,EARLY] Disable paravirtualized KVM clock driver 4553 4554 nolapic [X86-32,APIC,EARLY] Do not enable or use the local APIC. 4555 4556 nolapic_timer [X86-32,APIC,EARLY] Do not use the local APIC timer. 4557 4558 nomce [X86-32] Disable Machine Check Exception 4559 4560 nomfgpt [X86-32] Disable Multi-Function General Purpose 4561 Timer usage (for AMD Geode machines). 4562 4563 nomodeset Disable kernel modesetting. Most systems' firmware 4564 sets up a display mode and provides framebuffer memory 4565 for output. With nomodeset, DRM and fbdev drivers will 4566 not load if they could possibly displace the pre- 4567 initialized output. Only the system framebuffer will 4568 be available for use. The respective drivers will not 4569 perform display-mode changes or accelerated rendering. 4570 4571 Useful as error fallback, or for testing and debugging. 4572 4573 nomodule Disable module load 4574 4575 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 4576 shutdown the other cpus. Instead use the REBOOT_VECTOR 4577 irq. 4578 4579 nopat [X86,EARLY] Disable PAT (page attribute table extension of 4580 pagetables) support. 4581 4582 nopcid [X86-64,EARLY] Disable the PCID cpu feature. 4583 4584 nopku [X86] Disable Memory Protection Keys CPU feature found 4585 in some Intel CPUs. 4586 4587 nopti [X86-64,EARLY] 4588 Equivalent to pti=off 4589 4590 nopv= [X86,XEN,KVM,HYPER_V,VMWARE,EARLY] 4591 Disables the PV optimizations forcing the guest to run 4592 as generic guest with no PV drivers. Currently support 4593 XEN HVM, KVM, HYPER_V and VMWARE guest. 4594 4595 nopvspin [X86,XEN,KVM,EARLY] 4596 Disables the qspinlock slow path using PV optimizations 4597 which allow the hypervisor to 'idle' the guest on lock 4598 contention. 4599 4600 norandmaps Don't use address space randomization. Equivalent to 4601 echo 0 > /proc/sys/kernel/randomize_va_space 4602 4603 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 4604 with UP alternatives 4605 4606 noresume [SWSUSP] Disables resume and restores original swap 4607 space. 4608 4609 no-scroll [VGA] Disables scrollback. 4610 This is required for the Braillex ib80-piezo Braille 4611 reader made by F.H. Papenmeier (Germany). 4612 4613 nosgx [X86-64,SGX,EARLY] Disables Intel SGX kernel support. 4614 4615 nosmap [PPC,EARLY] 4616 Disable SMAP (Supervisor Mode Access Prevention) 4617 even if it is supported by processor. 4618 4619 nosmep [PPC64s,EARLY] 4620 Disable SMEP (Supervisor Mode Execution Prevention) 4621 even if it is supported by processor. 4622 4623 nosmp [SMP,EARLY] Tells an SMP kernel to act as a UP kernel, 4624 and disable the IO APIC. legacy for "maxcpus=0". 4625 4626 nosmt [KNL,MIPS,PPC,EARLY] Disable symmetric multithreading (SMT). 4627 Equivalent to smt=1. 4628 4629 [KNL,LOONGARCH,X86,PPC,S390] Disable symmetric multithreading (SMT). 4630 nosmt=force: Force disable SMT, cannot be undone 4631 via the sysfs control file. 4632 4633 nosoftlockup [KNL] Disable the soft-lockup detector. 4634 4635 nospec_store_bypass_disable 4636 [HW,EARLY] Disable all mitigations for the Speculative 4637 Store Bypass vulnerability 4638 4639 nospectre_bhb [ARM64,EARLY] Disable all mitigations for Spectre-BHB (branch 4640 history injection) vulnerability. System may allow data leaks 4641 with this option. 4642 4643 nospectre_v1 [X86,PPC,EARLY] Disable mitigations for Spectre Variant 1 4644 (bounds check bypass). With this option data leaks are 4645 possible in the system. 4646 4647 nospectre_v2 [X86,PPC_E500,ARM64,EARLY] Disable all mitigations 4648 for the Spectre variant 2 (indirect branch 4649 prediction) vulnerability. System may allow data 4650 leaks with this option. 4651 4652 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES,RISCV,LOONGARCH,EARLY] 4653 Disable paravirtualized steal time accounting. steal time 4654 is computed, but won't influence scheduler behaviour 4655 4656 nosync [HW,M68K] Disables sync negotiation for all devices. 4657 4658 no_timer_check [X86,APIC] Disables the code which tests for broken 4659 timer IRQ sources, i.e., the IO-APIC timer. This can 4660 work around problems with incorrect timer 4661 initialization on some boards. 4662 4663 no_uaccess_flush 4664 [PPC,EARLY] Don't flush the L1-D cache after accessing user data. 4665 4666 novmcoredd [KNL,KDUMP] 4667 Disable device dump. Device dump allows drivers to 4668 append dump data to vmcore so you can collect driver 4669 specified debug info. Drivers can append the data 4670 without any limit and this data is stored in memory, 4671 so this may cause significant memory stress. Disabling 4672 device dump can help save memory but the driver debug 4673 data will be no longer available. This parameter 4674 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 4675 is set. 4676 4677 no-vmw-sched-clock 4678 [X86,PV_OPS,EARLY] Disable paravirtualized VMware 4679 scheduler clock and use the default one. 4680 4681 nowatchdog [KNL] Disable both lockup detectors, i.e. 4682 soft-lockup and NMI watchdog (hard-lockup). 4683 4684 nowb [ARM,EARLY] 4685 4686 nox2apic [X86-64,APIC,EARLY] Do not enable x2APIC mode. 4687 4688 NOTE: this parameter will be ignored on systems with the 4689 LEGACY_XAPIC_DISABLED bit set in the 4690 IA32_XAPIC_DISABLE_STATUS MSR. 4691 4692 noxsave [BUGS=X86] Disables x86 extended register state save 4693 and restore using xsave. The kernel will fallback to 4694 enabling legacy floating-point and sse state. 4695 4696 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 4697 register states. The kernel will fall back to use 4698 xsave to save the states. By using this parameter, 4699 performance of saving the states is degraded because 4700 xsave doesn't support modified optimization while 4701 xsaveopt supports it on xsaveopt enabled systems. 4702 4703 noxsaves [X86] Disables xsaves and xrstors used in saving and 4704 restoring x86 extended register state in compacted 4705 form of xsave area. The kernel will fall back to use 4706 xsaveopt and xrstor to save and restore the states 4707 in standard form of xsave area. By using this 4708 parameter, xsave area per process might occupy more 4709 memory on xsaves enabled systems. 4710 4711 nr_cpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 4712 could support. nr_cpus=n : n >= 1 limits the kernel to 4713 support 'n' processors. It could be larger than the 4714 number of already plugged CPU during bootup, later in 4715 runtime you can physically add extra cpu until it reaches 4716 n. So during boot up some boot time memory for per-cpu 4717 variables need be pre-allocated for later physical cpu 4718 hot plugging. 4719 4720 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 4721 4722 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86, EARLY] 4723 Disable NUMA, Only set up a single NUMA node 4724 spanning all memory. 4725 4726 numa=fake=<size>[MG] 4727 [KNL, ARM64, RISCV, X86, EARLY] 4728 If given as a memory unit, fills all system RAM with 4729 nodes of size interleaved over physical nodes. 4730 4731 numa=fake=<N> 4732 [KNL, ARM64, RISCV, X86, EARLY] 4733 If given as an integer, fills all system RAM with N 4734 fake nodes interleaved over physical nodes. 4735 4736 numa=fake=<N>U 4737 [KNL, ARM64, RISCV, X86, EARLY] 4738 If given as an integer followed by 'U', it will 4739 divide each physical node into N emulated nodes. 4740 4741 numa=noacpi [X86] Don't parse the SRAT table for NUMA setup 4742 4743 numa=nohmat [X86] Don't parse the HMAT table for NUMA setup, or 4744 soft-reserved memory partitioning. 4745 4746 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic 4747 NUMA balancing. 4748 Allowed values are enable and disable 4749 4750 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 4751 'node', 'default' can be specified 4752 This can be set from sysctl after boot. 4753 See Documentation/admin-guide/sysctl/vm.rst for details. 4754 4755 nvme.quirks= [NVME] A list of quirk entries to augment the built-in 4756 nvme quirk list. List entries are separated by a 4757 '-' character. 4758 Each entry has the form VendorID:ProductID:quirk_names. 4759 The IDs are 4-digits hex numbers and quirk_names is a 4760 list of quirk names separated by commas. A quirk name 4761 can be prefixed by '^', meaning that the specified 4762 quirk must be disabled. 4763 4764 Example: 4765 nvme.quirks=7710:2267:bogus_nid,^identify_cns-9900:7711:broken_msi 4766 4767 ohci1394_dma=early [HW,EARLY] enable debugging via the ohci1394 driver. 4768 See Documentation/core-api/debugging-via-ohci1394.rst for more 4769 info. 4770 4771 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 4772 Rather than timing out after 20 ms if an EC 4773 command is not properly ACKed, override the length 4774 of the timeout. We have interrupts disabled while 4775 waiting for the ACK, so if this is set too high 4776 interrupts *may* be lost! 4777 4778 omap_mux= [OMAP] Override bootloader pin multiplexing. 4779 Format: <mux_mode0.mode_name=value>... 4780 For example, to override I2C bus2: 4781 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 4782 4783 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 4784 4785 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 4786 4787 boundary - index of last SLC block on Flex-OneNAND. 4788 The remaining blocks are configured as MLC blocks. 4789 lock - Configure if Flex-OneNAND boundary should be locked. 4790 Once locked, the boundary cannot be changed. 4791 1 indicates lock status, 0 indicates unlock status. 4792 4793 oops=panic [KNL,EARLY] 4794 Always panic on oopses. Default is to just kill the 4795 process, but there is a small probability of 4796 deadlocking the machine. 4797 This will also cause panics on machine check exceptions. 4798 Useful together with panic=30 to trigger a reboot. 4799 4800 page_alloc.shuffle= 4801 [KNL] Boolean flag to control whether the page allocator 4802 should randomize its free lists. This parameter can be 4803 used to enable/disable page randomization. The state of 4804 the flag can be read from sysfs at: 4805 /sys/module/page_alloc/parameters/shuffle. 4806 This parameter is only available if CONFIG_SHUFFLE_PAGE_ALLOCATOR=y. 4807 4808 page_owner= [KNL,EARLY] Boot-time page_owner enabling option. 4809 Storage of the information about who allocated 4810 each page is disabled in default. With this switch, 4811 we can turn it on. 4812 on: enable the feature 4813 4814 page_poison= [KNL,EARLY] Boot-time parameter changing the state of 4815 poisoning on the buddy allocator, available with 4816 CONFIG_PAGE_POISONING=y. 4817 off: turn off poisoning (default) 4818 on: turn on poisoning 4819 4820 page_reporting.page_reporting_order= 4821 [KNL] Minimal page reporting order 4822 Format: <integer> 4823 Adjust the minimal page reporting order. The page 4824 reporting is disabled when it exceeds MAX_PAGE_ORDER. 4825 4826 panic= [KNL] Kernel behaviour on panic: delay <timeout> 4827 timeout > 0: seconds before rebooting 4828 timeout = 0: wait forever 4829 timeout < 0: reboot immediately 4830 Format: <timeout> 4831 4832 panic_on_taint= [KNL,EARLY] 4833 Bitmask for conditionally calling panic() in add_taint() 4834 Format: <hex>[,nousertaint] 4835 Hexadecimal bitmask representing the set of TAINT flags 4836 that will cause the kernel to panic when add_taint() is 4837 called with any of the flags in this set. 4838 The optional switch "nousertaint" can be utilized to 4839 prevent userspace forced crashes by writing to sysctl 4840 /proc/sys/kernel/tainted any flagset matching with the 4841 bitmask set on panic_on_taint. 4842 See Documentation/admin-guide/tainted-kernels.rst for 4843 extra details on the taint flags that users can pick 4844 to compose the bitmask to assign to panic_on_taint. 4845 4846 panic_on_warn=1 panic() instead of WARN(). Useful to cause kdump 4847 on a WARN(). 4848 4849 panic_force_cpu= 4850 [KNL,SMP] Force panic handling to execute on a specific CPU. 4851 Format: <cpu number> 4852 Some platforms require panic handling to occur on a 4853 specific CPU for the crash kernel to function correctly. 4854 This can be due to firmware limitations, interrupt routing 4855 constraints, or platform-specific requirements where only 4856 a particular CPU can safely enter the crash kernel. 4857 When set, panic() will redirect execution to the specified 4858 CPU before proceeding with the normal panic and kexec flow. 4859 If the target CPU is offline or unavailable, panic proceeds 4860 on the current CPU. 4861 This option should only be used for systems with the above 4862 constraints as it might cause the panic operation to be less reliable. 4863 4864 panic_print= Bitmask for printing system info when panic happens. 4865 User can chose combination of the following bits: 4866 bit 0: print all tasks info 4867 bit 1: print system memory info 4868 bit 2: print timer info 4869 bit 3: print locks info if CONFIG_LOCKDEP is on 4870 bit 4: print ftrace buffer 4871 bit 5: replay all kernel messages on consoles at the end of panic 4872 bit 6: print all CPUs backtrace (if available in the arch) 4873 bit 7: print only tasks in uninterruptible (blocked) state 4874 *Be aware* that this option may print a _lot_ of lines, 4875 so there are risks of losing older messages in the log. 4876 Use this option carefully, maybe worth to setup a 4877 bigger log buffer with "log_buf_len" along with this. 4878 4879 panic_sys_info= A comma separated list of extra information to be dumped 4880 on panic. 4881 Format: val[,val...] 4882 Where @val can be any of the following: 4883 4884 tasks: print all tasks info 4885 mem: print system memory info 4886 timers: print timers info 4887 locks: print locks info if CONFIG_LOCKDEP is on 4888 ftrace: print ftrace buffer 4889 all_bt: print all CPUs backtrace (if available in the arch) 4890 blocked_tasks: print only tasks in uninterruptible (blocked) state 4891 4892 This is a human readable alternative to the 'panic_print' option. 4893 4894 panic_console_replay 4895 When panic happens, replay all kernel messages on 4896 consoles at the end of panic. 4897 4898 parkbd.port= [HW] Parallel port number the keyboard adapter is 4899 connected to, default is 0. 4900 Format: <parport#> 4901 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 4902 0 for XT, 1 for AT (default is AT). 4903 Format: <mode> 4904 4905 parport= [HW,PPT] Specify parallel ports. 0 disables. 4906 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 4907 Use 'auto' to force the driver to use any 4908 IRQ/DMA settings detected (the default is to 4909 ignore detected IRQ/DMA settings because of 4910 possible conflicts). You can specify the base 4911 address, IRQ, and DMA settings; IRQ and DMA 4912 should be numbers, or 'auto' (for using detected 4913 settings on that particular port), or 'nofifo' 4914 (to avoid using a FIFO even if it is detected). 4915 Parallel ports are assigned in the order they 4916 are specified on the command line, starting 4917 with parport0. 4918 4919 parport_init_mode= [HW,PPT] 4920 Configure VIA parallel port to operate in 4921 a specific mode. This is necessary on Pegasos 4922 computer where firmware has no options for setting 4923 up parallel port mode and sets it to spp. 4924 Currently this function knows 686a and 8231 chips. 4925 Format: [spp|ps2|epp|ecp|ecpepp] 4926 4927 pata_legacy.all= [HW,LIBATA] 4928 Format: <int> 4929 Set to non-zero to probe primary and secondary ISA 4930 port ranges on PCI systems where no PCI PATA device 4931 has been found at either range. Disabled by default. 4932 4933 pata_legacy.autospeed= [HW,LIBATA] 4934 Format: <int> 4935 Set to non-zero if a chip is present that snoops speed 4936 changes. Disabled by default. 4937 4938 pata_legacy.ht6560a= [HW,LIBATA] 4939 Format: <int> 4940 Set to 1, 2, or 3 for HT 6560A on the primary channel, 4941 the secondary channel, or both channels respectively. 4942 Disabled by default. 4943 4944 pata_legacy.ht6560b= [HW,LIBATA] 4945 Format: <int> 4946 Set to 1, 2, or 3 for HT 6560B on the primary channel, 4947 the secondary channel, or both channels respectively. 4948 Disabled by default. 4949 4950 pata_legacy.iordy_mask= [HW,LIBATA] 4951 Format: <int> 4952 IORDY enable mask. Set individual bits to allow IORDY 4953 for the respective channel. Bit 0 is for the first 4954 legacy channel handled by this driver, bit 1 is for 4955 the second channel, and so on. The sequence will often 4956 correspond to the primary legacy channel, the secondary 4957 legacy channel, and so on, but the handling of a PCI 4958 bus and the use of other driver options may interfere 4959 with the sequence. By default IORDY is allowed across 4960 all channels. 4961 4962 pata_legacy.opti82c46x= [HW,LIBATA] 4963 Format: <int> 4964 Set to 1, 2, or 3 for Opti 82c611A on the primary 4965 channel, the secondary channel, or both channels 4966 respectively. Disabled by default. 4967 4968 pata_legacy.opti82c611a= [HW,LIBATA] 4969 Format: <int> 4970 Set to 1, 2, or 3 for Opti 82c465MV on the primary 4971 channel, the secondary channel, or both channels 4972 respectively. Disabled by default. 4973 4974 pata_legacy.pio_mask= [HW,LIBATA] 4975 Format: <int> 4976 PIO mode mask for autospeed devices. Set individual 4977 bits to allow the use of the respective PIO modes. 4978 Bit 0 is for mode 0, bit 1 is for mode 1, and so on. 4979 All modes allowed by default. 4980 4981 pata_legacy.probe_all= [HW,LIBATA] 4982 Format: <int> 4983 Set to non-zero to probe tertiary and further ISA 4984 port ranges on PCI systems. Disabled by default. 4985 4986 pata_legacy.probe_mask= [HW,LIBATA] 4987 Format: <int> 4988 Probe mask for legacy ISA PATA ports. Depending on 4989 platform configuration and the use of other driver 4990 options up to 6 legacy ports are supported: 0x1f0, 4991 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing 4992 of individual ports can be disabled by setting the 4993 corresponding bits in the mask to 1. Bit 0 is for 4994 the first port in the list above (0x1f0), and so on. 4995 By default all supported ports are probed. 4996 4997 pata_legacy.qdi= [HW,LIBATA] 4998 Format: <int> 4999 Set to non-zero to probe QDI controllers. By default 5000 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise. 5001 5002 pata_legacy.winbond= [HW,LIBATA] 5003 Format: <int> 5004 Set to non-zero to probe Winbond controllers. Use 5005 the standard I/O port (0x130) if 1, otherwise the 5006 value given is the I/O port to use (typically 0x1b0). 5007 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE, 5008 0 otherwise. 5009 5010 pata_platform.pio_mask= [HW,LIBATA] 5011 Format: <int> 5012 Supported PIO mode mask. Set individual bits to allow 5013 the use of the respective PIO modes. Bit 0 is for 5014 mode 0, bit 1 is for mode 1, and so on. Mode 0 only 5015 allowed by default. 5016 5017 pause_on_oops=<int> 5018 Halt all CPUs after the first oops has been printed for 5019 the specified number of seconds. This is to be used if 5020 your oopses keep scrolling off the screen. 5021 5022 pcbit= [HW,ISDN] 5023 5024 pci=option[,option...] [PCI,EARLY] various PCI subsystem options. 5025 5026 Some options herein operate on a specific device 5027 or a set of devices (<pci_dev>). These are 5028 specified in one of the following formats: 5029 5030 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 5031 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 5032 5033 Note: the first format specifies a PCI 5034 bus/device/function address which may change 5035 if new hardware is inserted, if motherboard 5036 firmware changes, or due to changes caused 5037 by other kernel parameters. If the 5038 domain is left unspecified, it is 5039 taken to be zero. Optionally, a path 5040 to a device through multiple device/function 5041 addresses can be specified after the base 5042 address (this is more robust against 5043 renumbering issues). The second format 5044 selects devices using IDs from the 5045 configuration space which may match multiple 5046 devices in the system. 5047 5048 earlydump dump PCI config space before the kernel 5049 changes anything 5050 off [X86] don't probe for the PCI bus 5051 bios [X86-32] force use of PCI BIOS, don't access 5052 the hardware directly. Use this if your machine 5053 has a non-standard PCI host bridge. 5054 nobios [X86-32] disallow use of PCI BIOS, only direct 5055 hardware access methods are allowed. Use this 5056 if you experience crashes upon bootup and you 5057 suspect they are caused by the BIOS. 5058 conf1 [X86] Force use of PCI Configuration Access 5059 Mechanism 1 (config address in IO port 0xCF8, 5060 data in IO port 0xCFC, both 32-bit). 5061 conf2 [X86] Force use of PCI Configuration Access 5062 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 5063 the function, IO port 0xCFA, also 8-bit, sets 5064 bus number. The config space is then accessed 5065 through ports 0xC000-0xCFFF). 5066 See http://wiki.osdev.org/PCI for more info 5067 on the configuration access mechanisms. 5068 noaer [PCIE] If the PCIEAER kernel config parameter is 5069 enabled, this kernel boot option can be used to 5070 disable the use of PCIE advanced error reporting. 5071 nodomains [PCI] Disable support for multiple PCI 5072 root domains (aka PCI segments, in ACPI-speak). 5073 nommconf [X86] Disable use of MMCONFIG for PCI 5074 Configuration 5075 check_enable_amd_mmconf [X86] check for and enable 5076 properly configured MMIO access to PCI 5077 config space on AMD family 10h CPU 5078 nomsi [MSI] If the PCI_MSI kernel config parameter is 5079 enabled, this kernel boot option can be used to 5080 disable the use of MSI interrupts system-wide. 5081 noioapicquirk [APIC] Disable all boot interrupt quirks. 5082 Safety option to keep boot IRQs enabled. This 5083 should never be necessary. 5084 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 5085 primary IO-APIC for bridges that cannot disable 5086 boot IRQs. This fixes a source of spurious IRQs 5087 when the system masks IRQs. 5088 noioapicreroute [APIC] Disable workaround that uses the 5089 boot IRQ equivalent of an IRQ that connects to 5090 a chipset where boot IRQs cannot be disabled. 5091 The opposite of ioapicreroute. 5092 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 5093 routing table. These calls are known to be buggy 5094 on several machines and they hang the machine 5095 when used, but on other computers it's the only 5096 way to get the interrupt routing table. Try 5097 this option if the kernel is unable to allocate 5098 IRQs or discover secondary PCI buses on your 5099 motherboard. 5100 rom [X86] Assign address space to expansion ROMs. 5101 Use with caution as certain devices share 5102 address decoders between ROMs and other 5103 resources. 5104 norom [X86] Do not assign address space to 5105 expansion ROMs that do not already have 5106 BIOS assigned address ranges. 5107 nobar [X86] Do not assign address space to the 5108 BARs that weren't assigned by the BIOS. 5109 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 5110 assigned automatically to PCI devices. You can 5111 make the kernel exclude IRQs of your ISA cards 5112 this way. 5113 pirqaddr=0xAAAAA [X86] Specify the physical address 5114 of the PIRQ table (normally generated 5115 by the BIOS) if it is outside the 5116 F0000h-100000h range. 5117 lastbus=N [X86] Scan all buses thru bus #N. Can be 5118 useful if the kernel is unable to find your 5119 secondary buses and you want to tell it 5120 explicitly which ones they are. 5121 assign-busses [X86] Always assign all PCI bus 5122 numbers ourselves, overriding 5123 whatever the firmware may have done. 5124 usepirqmask [X86] Honor the possible IRQ mask stored 5125 in the BIOS $PIR table. This is needed on 5126 some systems with broken BIOSes, notably 5127 some HP Pavilion N5400 and Omnibook XE3 5128 notebooks. This will have no effect if ACPI 5129 IRQ routing is enabled. 5130 noacpi [X86] Do not use ACPI for IRQ routing 5131 or for PCI scanning. 5132 use_crs [X86] Use PCI host bridge window information 5133 from ACPI. On BIOSes from 2008 or later, this 5134 is enabled by default. If you need to use this, 5135 please report a bug. 5136 nocrs [X86] Ignore PCI host bridge windows from ACPI. 5137 If you need to use this, please report a bug. 5138 use_e820 [X86] Use E820 reservations to exclude parts of 5139 PCI host bridge windows. This is a workaround 5140 for BIOS defects in host bridge _CRS methods. 5141 If you need to use this, please report a bug to 5142 <linux-pci@vger.kernel.org>. 5143 no_e820 [X86] Ignore E820 reservations for PCI host 5144 bridge windows. This is the default on modern 5145 hardware. If you need to use this, please report 5146 a bug to <linux-pci@vger.kernel.org>. 5147 routeirq Do IRQ routing for all PCI devices. 5148 This is normally done in pci_enable_device(), 5149 so this option is a temporary workaround 5150 for broken drivers that don't call it. 5151 skip_isa_align [X86] do not align io start addr, so can 5152 handle more pci cards 5153 noearly [X86] Don't do any early type 1 scanning. 5154 This might help on some broken boards which 5155 machine check when some devices' config space 5156 is read. But various workarounds are disabled 5157 and some IOMMU drivers will not work. 5158 bfsort Sort PCI devices into breadth-first order. 5159 This sorting is done to get a device 5160 order compatible with older (<= 2.4) kernels. 5161 nobfsort Don't sort PCI devices into breadth-first order. 5162 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 5163 tuning and use the BIOS-configured MPS defaults. 5164 pcie_bus_safe Set every device's MPS to the largest value 5165 supported by all devices below the root complex. 5166 pcie_bus_perf Set device MPS to the largest allowable MPS 5167 based on its parent bus. Also set MRRS (Max 5168 Read Request Size) to the largest supported 5169 value (no larger than the MPS that the device 5170 or bus can support) for best performance. 5171 pcie_bus_peer2peer Set every device's MPS to 128B, which 5172 every device is guaranteed to support. This 5173 configuration allows peer-to-peer DMA between 5174 any pair of devices, possibly at the cost of 5175 reduced performance. This also guarantees 5176 that hot-added devices will work. 5177 cbiosize=nn[KMG] The fixed amount of bus space which is 5178 reserved for the CardBus bridge's IO window. 5179 The default value is 256 bytes. 5180 cbmemsize=nn[KMG] The fixed amount of bus space which is 5181 reserved for the CardBus bridge's memory 5182 window. The default value is 64 megabytes. 5183 resource_alignment= 5184 Format: 5185 [<order of align>@]<pci_dev>[; ...] 5186 Specifies alignment and device to reassign 5187 aligned memory resources. How to 5188 specify the device is described above. 5189 If <order of align> is not specified, 5190 PAGE_SIZE is used as alignment. 5191 A PCI-PCI bridge can be specified if resource 5192 windows need to be expanded. 5193 To specify the alignment for several 5194 instances of a device, the PCI vendor, 5195 device, subvendor, and subdevice may be 5196 specified, e.g., 12@pci:8086:9c22:103c:198f 5197 for 4096-byte alignment. 5198 ecrc= Enable/disable PCIe ECRC (transaction layer 5199 end-to-end CRC checking). Only effective if 5200 OS has native AER control (either granted by 5201 ACPI _OSC or forced via "pcie_ports=native") 5202 bios: Use BIOS/firmware settings. This is the 5203 the default. 5204 off: Turn ECRC off 5205 on: Turn ECRC on. 5206 hpiosize=nn[KMG] The fixed amount of bus space which is 5207 reserved for hotplug bridge's IO window. 5208 Default size is 256 bytes. 5209 hpmmiosize=nn[KMG] The fixed amount of bus space which is 5210 reserved for hotplug bridge's MMIO window. 5211 Default size is 2 megabytes. 5212 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 5213 reserved for hotplug bridge's MMIO_PREF window. 5214 Default size is 2 megabytes. 5215 hpmemsize=nn[KMG] The fixed amount of bus space which is 5216 reserved for hotplug bridge's MMIO and 5217 MMIO_PREF window. 5218 Default size is 2 megabytes. 5219 hpbussize=nn The minimum amount of additional bus numbers 5220 reserved for buses below a hotplug bridge. 5221 Default is 1. 5222 realloc= Enable/disable reallocating PCI bridge resources 5223 if allocations done by BIOS are too small to 5224 accommodate resources required by all child 5225 devices. 5226 off: Turn realloc off 5227 on: Turn realloc on 5228 realloc same as realloc=on 5229 noari do not use PCIe ARI. 5230 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 5231 do not use PCIe ATS (and IOMMU device IOTLB). 5232 pcie_scan_all Scan all possible PCIe devices. Otherwise we 5233 only look for one device below a PCIe downstream 5234 port. 5235 big_root_window Try to add a big 64bit memory window to the PCIe 5236 root complex on AMD CPUs. Some GFX hardware 5237 can resize a BAR to allow access to all VRAM. 5238 Adding the window is slightly risky (it may 5239 conflict with unreported devices), so this 5240 taints the kernel. 5241 disable_acs_redir=<pci_dev>[; ...] 5242 Specify one or more PCI devices (in the format 5243 specified above) separated by semicolons. 5244 Each device specified will have the PCI ACS 5245 redirect capabilities forced off which will 5246 allow P2P traffic between devices through 5247 bridges without forcing it upstream. Note: 5248 this removes isolation between devices and 5249 may put more devices in an IOMMU group. 5250 config_acs= 5251 Format: 5252 <ACS flags>@<pci_dev>[; ...] 5253 Specify one or more PCI devices (in the format 5254 specified above) optionally prepended with flags 5255 and separated by semicolons. The respective 5256 capabilities will be enabled, disabled or 5257 unchanged based on what is specified in 5258 flags. 5259 5260 ACS Flags is defined as follows: 5261 bit-0 : ACS Source Validation 5262 bit-1 : ACS Translation Blocking 5263 bit-2 : ACS P2P Request Redirect 5264 bit-3 : ACS P2P Completion Redirect 5265 bit-4 : ACS Upstream Forwarding 5266 bit-5 : ACS P2P Egress Control 5267 bit-6 : ACS Direct Translated P2P 5268 Each bit can be marked as: 5269 '0' – force disabled 5270 '1' – force enabled 5271 'x' – unchanged 5272 For example, 5273 pci=config_acs=10x@pci:0:0 5274 would configure all devices that support 5275 ACS to enable P2P Request Redirect, disable 5276 Translation Blocking, and leave Source 5277 Validation unchanged from whatever power-up 5278 or firmware set it to. 5279 5280 Note: this may remove isolation between devices 5281 and may put more devices in an IOMMU group. 5282 force_floating [S390] Force usage of floating interrupts. 5283 nomio [S390] Do not use MIO instructions. 5284 norid [S390] ignore the RID field and force use of 5285 one PCI domain per PCI function 5286 notph [PCIE] If the PCIE_TPH kernel config parameter 5287 is enabled, this kernel boot option can be used 5288 to disable PCIe TLP Processing Hints support 5289 system-wide. 5290 5291 pcie_aspm= [PCIE] Forcibly enable or ignore PCIe Active State Power 5292 Management. 5293 off Don't touch ASPM configuration at all. Leave any 5294 configuration done by firmware unchanged. 5295 force Enable ASPM even on devices that claim not to support it. 5296 WARNING: Forcing ASPM on may cause system lockups. 5297 5298 pcie_ports= [PCIE] PCIe port services handling: 5299 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 5300 even if the platform doesn't give the OS permission to 5301 use them. This may cause conflicts if the platform 5302 also tries to use these services. 5303 dpc-native Use native PCIe service for DPC only. May 5304 cause conflicts if firmware uses AER or DPC. 5305 compat Disable native PCIe services (PME, AER, DPC, PCIe 5306 hotplug). 5307 5308 pcie_port_pm= [PCIE] PCIe port power management handling: 5309 off Disable power management of all PCIe ports 5310 force Forcibly enable power management of all PCIe ports 5311 5312 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 5313 nomsi Do not use MSI for native PCIe PME signaling (this makes 5314 all PCIe root ports use INTx for all services). 5315 5316 pcmv= [HW,PCMCIA] BadgePAD 4 5317 5318 pd_ignore_unused 5319 [PM] 5320 Keep all power-domains already enabled by bootloader on, 5321 even if no driver has claimed them. This is useful 5322 for debug and development, but should not be 5323 needed on a platform with proper driver support. 5324 5325 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 5326 boot time. 5327 Format: { 0 | 1 } 5328 See arch/parisc/kernel/pdc_chassis.c 5329 5330 percpu_alloc= [MM,EARLY] 5331 Select which percpu first chunk allocator to use. 5332 Currently supported values are "embed" and "page". 5333 Archs may support subset or none of the selections. 5334 See comments in mm/percpu.c for details on each 5335 allocator. This parameter is primarily for debugging 5336 and performance comparison. 5337 5338 pirq= [SMP,APIC] Manual mp-table setup 5339 See Documentation/arch/x86/i386/IO-APIC.rst. 5340 5341 plip= [PPT,NET] Parallel port network link 5342 Format: { parport<nr> | timid | 0 } 5343 See also Documentation/admin-guide/parport.rst. 5344 5345 pmtmr= [X86] Manual setup of pmtmr I/O Port. 5346 Override pmtimer IOPort with a hex value. 5347 e.g. pmtmr=0x508 5348 5349 pmu_override= [PPC] Override the PMU. 5350 This option takes over the PMU facility, so it is no 5351 longer usable by perf. Setting this option starts the 5352 PMU counters by setting MMCR0 to 0 (the FC bit is 5353 cleared). If a number is given, then MMCR1 is set to 5354 that number, otherwise (e.g., 'pmu_override=on'), MMCR1 5355 remains 0. 5356 5357 pm_async= [PM] 5358 Format: off 5359 This parameter sets the initial value of the 5360 /sys/power/pm_async sysfs knob at boot time. 5361 If set to "off", disables asynchronous suspend and 5362 resume of devices during system-wide power transitions. 5363 This can be useful on platforms where device 5364 dependencies are not well-defined, or for debugging 5365 power management issues. Asynchronous operations are 5366 enabled by default. 5367 5368 5369 pm_debug_messages [SUSPEND,KNL] 5370 Enable suspend/resume debug messages during boot up. 5371 5372 pnp.debug=1 [PNP] 5373 Enable PNP debug messages (depends on the 5374 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 5375 via /sys/module/pnp/parameters/debug. We always show 5376 current resource usage; turning this on also shows 5377 possible settings and some assignment information. 5378 5379 pnpacpi= [ACPI] 5380 { off } 5381 5382 pnpbios= [ISAPNP] 5383 { on | off | curr | res | no-curr | no-res } 5384 5385 pnp_reserve_irq= 5386 [ISAPNP] Exclude IRQs for the autoconfiguration 5387 5388 pnp_reserve_dma= 5389 [ISAPNP] Exclude DMAs for the autoconfiguration 5390 5391 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 5392 Ranges are in pairs (I/O port base and size). 5393 5394 pnp_reserve_mem= 5395 [ISAPNP] Exclude memory regions for the 5396 autoconfiguration. 5397 Ranges are in pairs (memory base and size). 5398 5399 ports= [IP_VS_FTP] IPVS ftp helper module 5400 Default is 21. 5401 Up to 8 (IP_VS_APP_MAX_PORTS) ports 5402 may be specified. 5403 Format: <port>,<port>.... 5404 5405 possible_cpus= [SMP,S390,X86] 5406 Format: <unsigned int> 5407 Set the number of possible CPUs, overriding the 5408 regular discovery mechanisms (such as ACPI/FW, etc). 5409 5410 powersave=off [PPC] This option disables power saving features. 5411 It specifically disables cpuidle and sets the 5412 platform machine description specific power_save 5413 function to NULL. On Idle the CPU just reduces 5414 execution priority. 5415 5416 ppc_strict_facility_enable 5417 [PPC,ENABLE] This option catches any kernel floating point, 5418 Altivec, VSX and SPE outside of regions specifically 5419 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 5420 There is some performance impact when enabling this. 5421 5422 ppc_tm= [PPC,EARLY] 5423 Format: {"off"} 5424 Disable Hardware Transactional Memory 5425 5426 preempt= [KNL] 5427 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC 5428 none - Limited to cond_resched() calls 5429 voluntary - Limited to cond_resched() and might_sleep() calls 5430 full - Any section that isn't explicitly preempt disabled 5431 can be preempted anytime. Tasks will also yield 5432 contended spinlocks (if the critical section isn't 5433 explicitly preempt disabled beyond the lock itself). 5434 lazy - Scheduler controlled. Similar to full but instead 5435 of preempting the task immediately, the task gets 5436 one HZ tick time to yield itself before the 5437 preemption will be forced. One preemption is when the 5438 task returns to user space. 5439 5440 print-fatal-signals= 5441 [KNL] debug: print fatal signals 5442 5443 If enabled, warn about various signal handling 5444 related application anomalies: too many signals, 5445 too many POSIX.1 timers, fatal signals causing a 5446 coredump - etc. 5447 5448 If you hit the warning due to signal overflow, 5449 you might want to try "ulimit -i unlimited". 5450 5451 default: off. 5452 5453 printk.always_kmsg_dump= 5454 Trigger kmsg_dump for cases other than kernel oops or 5455 panics 5456 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5457 default: disabled 5458 5459 printk.console_no_auto_verbose= 5460 Disable console loglevel raise on oops, panic 5461 or lockdep-detected issues (only if lock debug is on). 5462 With an exception to setups with low baudrate on 5463 serial console, keeping this 0 is a good choice 5464 in order to provide more debug information. 5465 Format: <bool> 5466 default: 0 (auto_verbose is enabled) 5467 5468 printk.debug_non_panic_cpus= 5469 Allows storing messages from non-panic CPUs into 5470 the printk log buffer during panic(). They are 5471 flushed to consoles by the panic-CPU on 5472 a best-effort basis. 5473 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5474 Default: disabled 5475 5476 printk.devkmsg={on,off,ratelimit} 5477 Control writing to /dev/kmsg. 5478 on - unlimited logging to /dev/kmsg from userspace 5479 off - logging to /dev/kmsg disabled 5480 ratelimit - ratelimit the logging 5481 Default: ratelimit 5482 5483 printk.time= Show timing data prefixed to each printk message line 5484 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5485 5486 proc_mem.force_override= [KNL] 5487 Format: {always | ptrace | never} 5488 Traditionally /proc/pid/mem allows memory permissions to be 5489 overridden without restrictions. This option may be set to 5490 restrict that. Can be one of: 5491 - 'always': traditional behavior always allows mem overrides. 5492 - 'ptrace': only allow mem overrides for active ptracers. 5493 - 'never': never allow mem overrides. 5494 If not specified, default is the CONFIG_PROC_MEM_* choice. 5495 5496 processor.max_cstate= [HW,ACPI] 5497 Limit processor to maximum C-state 5498 max_cstate=9 overrides any DMI blacklist limit. 5499 5500 processor.nocst [HW,ACPI] 5501 Ignore the _CST method to determine C-states, 5502 instead using the legacy FADT method 5503 5504 profile= [KNL] Enable kernel profiling via /proc/profile 5505 Format: [<profiletype>,]<number> 5506 Param: <profiletype>: "schedule" or "kvm" 5507 [defaults to kernel profiling] 5508 Param: "schedule" - profile schedule points. 5509 Param: "kvm" - profile VM exits. 5510 Param: <number> - step/bucket size as a power of 2 for 5511 statistical time based profiling. 5512 5513 prot_virt= [S390] enable hosting protected virtual machines 5514 isolated from the hypervisor (if hardware supports 5515 that). If enabled, the default kernel base address 5516 might be overridden even when Kernel Address Space 5517 Layout Randomization is disabled. 5518 Format: <bool> 5519 5520 psi= [KNL] Enable or disable pressure stall information 5521 tracking. 5522 Format: <bool> 5523 5524 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 5525 probe for; one of (bare|imps|exps|lifebook|any). 5526 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 5527 per second. 5528 psmouse.resetafter= [HW,MOUSE] 5529 Try to reset the device after so many bad packets 5530 (0 = never). 5531 psmouse.resolution= 5532 [HW,MOUSE] Set desired mouse resolution, in dpi. 5533 psmouse.smartscroll= 5534 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 5535 0 = disabled, 1 = enabled (default). 5536 5537 pstore.backend= Specify the name of the pstore backend to use 5538 5539 pti= [X86-64] Control Page Table Isolation of user and 5540 kernel address spaces. Disabling this feature 5541 removes hardening, but improves performance of 5542 system calls and interrupts. 5543 5544 on - unconditionally enable 5545 off - unconditionally disable 5546 auto - kernel detects whether your CPU model is 5547 vulnerable to issues that PTI mitigates 5548 5549 Not specifying this option is equivalent to pti=auto. 5550 5551 pty.legacy_count= 5552 [KNL] Number of legacy pty's. Overwrites compiled-in 5553 default number. 5554 5555 quiet [KNL,EARLY] Disable most log messages 5556 5557 r128= [HW,DRM] 5558 5559 radix_hcall_invalidate=on [PPC/PSERIES] 5560 Disable RADIX GTSE feature and use hcall for TLB 5561 invalidate. 5562 5563 raid= [HW,RAID] 5564 See Documentation/admin-guide/md.rst. 5565 5566 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 5567 See Documentation/admin-guide/blockdev/ramdisk.rst. 5568 5569 ramdisk_start= [Deprecated,RAM] RAM disk image start address 5570 5571 random.trust_cpu=off 5572 [KNL,EARLY] Disable trusting the use of the CPU's 5573 random number generator (if available) to 5574 initialize the kernel's RNG. 5575 5576 random.trust_bootloader=off 5577 [KNL,EARLY] Disable trusting the use of the a seed 5578 passed by the bootloader (if available) to 5579 initialize the kernel's RNG. 5580 5581 randomize_kstack_offset= 5582 [KNL,EARLY] Enable or disable kernel stack offset 5583 randomization, which provides roughly 5 bits of 5584 entropy, frustrating memory corruption attacks 5585 that depend on stack address determinism or 5586 cross-syscall address exposures. This is only 5587 available on architectures that have defined 5588 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET. 5589 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5590 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT. 5591 5592 ras=option[,option,...] [KNL] RAS-specific options 5593 5594 cec_disable [X86] 5595 Disable the Correctable Errors Collector, 5596 see CONFIG_RAS_CEC help text. 5597 5598 rcu_nocbs[=cpu-list] 5599 [KNL] The optional argument is a cpu list, 5600 as described above. 5601 5602 In kernels built with CONFIG_RCU_NOCB_CPU=y, 5603 enable the no-callback CPU mode, which prevents 5604 such CPUs' callbacks from being invoked in 5605 softirq context. Invocation of such CPUs' RCU 5606 callbacks will instead be offloaded to "rcuox/N" 5607 kthreads created for that purpose, where "x" is 5608 "p" for RCU-preempt, "s" for RCU-sched, and "g" 5609 for the kthreads that mediate grace periods; and 5610 "N" is the CPU number. This reduces OS jitter on 5611 the offloaded CPUs, which can be useful for HPC 5612 and real-time workloads. It can also improve 5613 energy efficiency for asymmetric multiprocessors. 5614 5615 If a cpulist is passed as an argument, the specified 5616 list of CPUs is set to no-callback mode from boot. 5617 5618 Otherwise, if the '=' sign and the cpulist 5619 arguments are omitted, no CPU will be set to 5620 no-callback mode from boot but the mode may be 5621 toggled at runtime via cpusets. 5622 5623 Note that this argument takes precedence over 5624 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 5625 5626 rcu_nocb_poll [KNL] 5627 Rather than requiring that offloaded CPUs 5628 (specified by rcu_nocbs= above) explicitly 5629 awaken the corresponding "rcuoN" kthreads, 5630 make these kthreads poll for callbacks. 5631 This improves the real-time response for the 5632 offloaded CPUs by relieving them of the need to 5633 wake up the corresponding kthread, but degrades 5634 energy efficiency by requiring that the kthreads 5635 periodically wake up to do the polling. 5636 5637 rcutree.blimit= [KNL] 5638 Set maximum number of finished RCU callbacks to 5639 process in one batch. 5640 5641 rcutree.csd_lock_suppress_rcu_stall= [KNL] 5642 Do only a one-line RCU CPU stall warning when 5643 there is an ongoing too-long CSD-lock wait. 5644 5645 rcutree.do_rcu_barrier= [KNL] 5646 Request a call to rcu_barrier(). This is 5647 throttled so that userspace tests can safely 5648 hammer on the sysfs variable if they so choose. 5649 If triggered before the RCU grace-period machinery 5650 is fully active, this will error out with EAGAIN. 5651 5652 rcutree.dump_tree= [KNL] 5653 Dump the structure of the rcu_node combining tree 5654 out at early boot. This is used for diagnostic 5655 purposes, to verify correct tree setup. 5656 5657 rcutree.gp_cleanup_delay= [KNL] 5658 Set the number of jiffies to delay each step of 5659 RCU grace-period cleanup. 5660 5661 rcutree.gp_init_delay= [KNL] 5662 Set the number of jiffies to delay each step of 5663 RCU grace-period initialization. 5664 5665 rcutree.gp_preinit_delay= [KNL] 5666 Set the number of jiffies to delay each step of 5667 RCU grace-period pre-initialization, that is, 5668 the propagation of recent CPU-hotplug changes up 5669 the rcu_node combining tree. 5670 5671 rcutree.jiffies_till_first_fqs= [KNL] 5672 Set delay from grace-period initialization to 5673 first attempt to force quiescent states. 5674 Units are jiffies, minimum value is zero, 5675 and maximum value is HZ. 5676 5677 rcutree.jiffies_till_next_fqs= [KNL] 5678 Set delay between subsequent attempts to force 5679 quiescent states. Units are jiffies, minimum 5680 value is one, and maximum value is HZ. 5681 5682 rcutree.jiffies_till_sched_qs= [KNL] 5683 Set required age in jiffies for a 5684 given grace period before RCU starts 5685 soliciting quiescent-state help from 5686 rcu_note_context_switch() and cond_resched(). 5687 If not specified, the kernel will calculate 5688 a value based on the most recent settings 5689 of rcutree.jiffies_till_first_fqs 5690 and rcutree.jiffies_till_next_fqs. 5691 This calculated value may be viewed in 5692 rcutree.jiffies_to_sched_qs. Any attempt to set 5693 rcutree.jiffies_to_sched_qs will be cheerfully 5694 overwritten. 5695 5696 rcutree.kthread_prio= [KNL,BOOT] 5697 Set the SCHED_FIFO priority of the RCU per-CPU 5698 kthreads (rcuc/N). This value is also used for 5699 the priority of the RCU boost threads (rcub/N) 5700 and for the RCU grace-period kthreads (rcu_bh, 5701 rcu_preempt, and rcu_sched). If RCU_BOOST is 5702 set, valid values are 1-99 and the default is 1 5703 (the least-favored priority). Otherwise, when 5704 RCU_BOOST is not set, valid values are 0-99 and 5705 the default is zero (non-realtime operation). 5706 When RCU_NOCB_CPU is set, also adjust the 5707 priority of NOCB callback kthreads. 5708 5709 rcutree.nocb_nobypass_lim_per_jiffy= [KNL] 5710 On callback-offloaded (rcu_nocbs) CPUs, 5711 RCU reduces the lock contention that would 5712 otherwise be caused by callback floods through 5713 use of the ->nocb_bypass list. However, in the 5714 common non-flooded case, RCU queues directly to 5715 the main ->cblist in order to avoid the extra 5716 overhead of the ->nocb_bypass list and its lock. 5717 But if there are too many callbacks queued during 5718 a single jiffy, RCU pre-queues the callbacks into 5719 the ->nocb_bypass queue. The definition of "too 5720 many" is supplied by this kernel boot parameter. 5721 5722 rcutree.nohz_full_patience_delay= [KNL] 5723 On callback-offloaded (rcu_nocbs) CPUs, avoid 5724 disturbing RCU unless the grace period has 5725 reached the specified age in milliseconds. 5726 Defaults to zero. Large values will be capped 5727 at five seconds. All values will be rounded down 5728 to the nearest value representable by jiffies. 5729 5730 rcutree.qhimark= [KNL] 5731 Set threshold of queued RCU callbacks beyond which 5732 batch limiting is disabled. 5733 5734 rcutree.qlowmark= [KNL] 5735 Set threshold of queued RCU callbacks below which 5736 batch limiting is re-enabled. 5737 5738 rcutree.qovld= [KNL] 5739 Set threshold of queued RCU callbacks beyond which 5740 RCU's force-quiescent-state scan will aggressively 5741 enlist help from cond_resched() and sched IPIs to 5742 help CPUs more quickly reach quiescent states. 5743 Set to less than zero to make this be set based 5744 on rcutree.qhimark at boot time and to zero to 5745 disable more aggressive help enlistment. 5746 5747 rcutree.rcu_delay_page_cache_fill_msec= [KNL] 5748 Set the page-cache refill delay (in milliseconds) 5749 in response to low-memory conditions. The range 5750 of permitted values is in the range 0:100000. 5751 5752 rcutree.rcu_divisor= [KNL] 5753 Set the shift-right count to use to compute 5754 the callback-invocation batch limit bl from 5755 the number of callbacks queued on this CPU. 5756 The result will be bounded below by the value of 5757 the rcutree.blimit kernel parameter. Every bl 5758 callbacks, the softirq handler will exit in 5759 order to allow the CPU to do other work. 5760 5761 Please note that this callback-invocation batch 5762 limit applies only to non-offloaded callback 5763 invocation. Offloaded callbacks are instead 5764 invoked in the context of an rcuoc kthread, which 5765 scheduler will preempt as it does any other task. 5766 5767 rcutree.rcu_fanout_exact= [KNL] 5768 Disable autobalancing of the rcu_node combining 5769 tree. This is used by rcutorture, and might 5770 possibly be useful for architectures having high 5771 cache-to-cache transfer latencies. 5772 5773 rcutree.rcu_fanout_leaf= [KNL] 5774 Change the number of CPUs assigned to each 5775 leaf rcu_node structure. Useful for very 5776 large systems, which will choose the value 64, 5777 and for NUMA systems with large remote-access 5778 latencies, which will choose a value aligned 5779 with the appropriate hardware boundaries. 5780 5781 rcutree.rcu_min_cached_objs= [KNL] 5782 Minimum number of objects which are cached and 5783 maintained per one CPU. Object size is equal 5784 to PAGE_SIZE. The cache allows to reduce the 5785 pressure to page allocator, also it makes the 5786 whole algorithm to behave better in low memory 5787 condition. 5788 5789 rcutree.rcu_nocb_gp_stride= [KNL] 5790 Set the number of NOCB callback kthreads in 5791 each group, which defaults to the square root 5792 of the number of CPUs. Larger numbers reduce 5793 the wakeup overhead on the global grace-period 5794 kthread, but increases that same overhead on 5795 each group's NOCB grace-period kthread. 5796 5797 rcutree.rcu_kick_kthreads= [KNL] 5798 Cause the grace-period kthread to get an extra 5799 wake_up() if it sleeps three times longer than 5800 it should at force-quiescent-state time. 5801 This wake_up() will be accompanied by a 5802 WARN_ONCE() splat and an ftrace_dump(). 5803 5804 rcutree.rcu_resched_ns= [KNL] 5805 Limit the time spend invoking a batch of RCU 5806 callbacks to the specified number of nanoseconds. 5807 By default, this limit is checked only once 5808 every 32 callbacks in order to limit the pain 5809 inflicted by local_clock() overhead. 5810 5811 rcutree.rcu_unlock_delay= [KNL] 5812 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 5813 this specifies an rcu_read_unlock()-time delay 5814 in microseconds. This defaults to zero. 5815 Larger delays increase the probability of 5816 catching RCU pointer leaks, that is, buggy use 5817 of RCU-protected pointers after the relevant 5818 rcu_read_unlock() has completed. 5819 5820 rcutree.sysrq_rcu= [KNL] 5821 Commandeer a sysrq key to dump out Tree RCU's 5822 rcu_node tree with an eye towards determining 5823 why a new grace period has not yet started. 5824 5825 rcutree.use_softirq= [KNL] 5826 If set to zero, move all RCU_SOFTIRQ processing to 5827 per-CPU rcuc kthreads. Defaults to a non-zero 5828 value, meaning that RCU_SOFTIRQ is used by default. 5829 Specify rcutree.use_softirq=0 to use rcuc kthreads. 5830 5831 But note that CONFIG_PREEMPT_RT=y kernels disable 5832 this kernel boot parameter, forcibly setting it 5833 to zero. 5834 5835 rcutree.enable_rcu_lazy= [KNL] 5836 To save power, batch RCU callbacks and flush after 5837 delay, memory pressure or callback list growing too 5838 big. 5839 5840 rcutree.rcu_normal_wake_from_gp= [KNL] 5841 Reduces a latency of synchronize_rcu() call. This approach 5842 maintains its own track of synchronize_rcu() callers, so it 5843 does not interact with regular callbacks because it does not 5844 use a call_rcu[_hurry]() path. Please note, this is for a 5845 normal grace period. 5846 5847 How to enable it: 5848 5849 echo 1 > /sys/module/rcutree/parameters/rcu_normal_wake_from_gp 5850 or pass a boot parameter "rcutree.rcu_normal_wake_from_gp=1" 5851 5852 Default is 1 if num_possible_cpus() <= 16 and it is not explicitly 5853 disabled by the boot parameter passing 0. 5854 5855 rcuscale.gp_async= [KNL] 5856 Measure performance of asynchronous 5857 grace-period primitives such as call_rcu(). 5858 5859 rcuscale.gp_async_max= [KNL] 5860 Specify the maximum number of outstanding 5861 callbacks per writer thread. When a writer 5862 thread exceeds this limit, it invokes the 5863 corresponding flavor of rcu_barrier() to allow 5864 previously posted callbacks to drain. 5865 5866 rcuscale.gp_exp= [KNL] 5867 Measure performance of expedited synchronous 5868 grace-period primitives. 5869 5870 rcuscale.holdoff= [KNL] 5871 Set test-start holdoff period. The purpose of 5872 this parameter is to delay the start of the 5873 test until boot completes in order to avoid 5874 interference. 5875 5876 rcuscale.kfree_by_call_rcu= [KNL] 5877 In kernels built with CONFIG_RCU_LAZY=y, test 5878 call_rcu() instead of kfree_rcu(). 5879 5880 rcuscale.kfree_mult= [KNL] 5881 Instead of allocating an object of size kfree_obj, 5882 allocate one of kfree_mult * sizeof(kfree_obj). 5883 Defaults to 1. 5884 5885 rcuscale.kfree_rcu_test= [KNL] 5886 Set to measure performance of kfree_rcu() flooding. 5887 5888 rcuscale.kfree_rcu_test_double= [KNL] 5889 Test the double-argument variant of kfree_rcu(). 5890 If this parameter has the same value as 5891 rcuscale.kfree_rcu_test_single, both the single- 5892 and double-argument variants are tested. 5893 5894 rcuscale.kfree_rcu_test_single= [KNL] 5895 Test the single-argument variant of kfree_rcu(). 5896 If this parameter has the same value as 5897 rcuscale.kfree_rcu_test_double, both the single- 5898 and double-argument variants are tested. 5899 5900 rcuscale.kfree_nthreads= [KNL] 5901 The number of threads running loops of kfree_rcu(). 5902 5903 rcuscale.kfree_alloc_num= [KNL] 5904 Number of allocations and frees done in an iteration. 5905 5906 rcuscale.kfree_loops= [KNL] 5907 Number of loops doing rcuscale.kfree_alloc_num number 5908 of allocations and frees. 5909 5910 rcuscale.minruntime= [KNL] 5911 Set the minimum test run time in seconds. This 5912 does not affect the data-collection interval, 5913 but instead allows better measurement of things 5914 like CPU consumption. 5915 5916 rcuscale.nreaders= [KNL] 5917 Set number of RCU readers. The value -1 selects 5918 N, where N is the number of CPUs. A value 5919 "n" less than -1 selects N-n+1, where N is again 5920 the number of CPUs. For example, -2 selects N 5921 (the number of CPUs), -3 selects N+1, and so on. 5922 A value of "n" less than or equal to -N selects 5923 a single reader. 5924 5925 rcuscale.nwriters= [KNL] 5926 Set number of RCU writers. The values operate 5927 the same as for rcuscale.nreaders. 5928 N, where N is the number of CPUs 5929 5930 rcuscale.scale_type= [KNL] 5931 Specify the RCU implementation to test. 5932 5933 rcuscale.shutdown= [KNL] 5934 Shut the system down after performance tests 5935 complete. This is useful for hands-off automated 5936 testing. 5937 5938 rcuscale.verbose= [KNL] 5939 Enable additional printk() statements. 5940 5941 rcuscale.writer_holdoff= [KNL] 5942 Write-side holdoff between grace periods, 5943 in microseconds. The default of zero says 5944 no holdoff. 5945 5946 rcuscale.writer_holdoff_jiffies= [KNL] 5947 Additional write-side holdoff between grace 5948 periods, but in jiffies. The default of zero 5949 says no holdoff. 5950 5951 rcutorture.fqs_duration= [KNL] 5952 Set duration of force_quiescent_state bursts 5953 in microseconds. 5954 5955 rcutorture.fqs_holdoff= [KNL] 5956 Set holdoff time within force_quiescent_state bursts 5957 in microseconds. 5958 5959 rcutorture.fqs_stutter= [KNL] 5960 Set wait time between force_quiescent_state bursts 5961 in seconds. 5962 5963 rcutorture.fwd_progress= [KNL] 5964 Specifies the number of kthreads to be used 5965 for RCU grace-period forward-progress testing 5966 for the types of RCU supporting this notion. 5967 Defaults to 1 kthread, values less than zero or 5968 greater than the number of CPUs cause the number 5969 of CPUs to be used. 5970 5971 rcutorture.fwd_progress_div= [KNL] 5972 Specify the fraction of a CPU-stall-warning 5973 period to do tight-loop forward-progress testing. 5974 5975 rcutorture.fwd_progress_holdoff= [KNL] 5976 Number of seconds to wait between successive 5977 forward-progress tests. 5978 5979 rcutorture.fwd_progress_need_resched= [KNL] 5980 Enclose cond_resched() calls within checks for 5981 need_resched() during tight-loop forward-progress 5982 testing. 5983 5984 rcutorture.gp_cond= [KNL] 5985 Use conditional/asynchronous update-side 5986 normal-grace-period primitives, if available. 5987 5988 rcutorture.gp_cond_exp= [KNL] 5989 Use conditional/asynchronous update-side 5990 expedited-grace-period primitives, if available. 5991 5992 rcutorture.gp_cond_full= [KNL] 5993 Use conditional/asynchronous update-side 5994 normal-grace-period primitives that also take 5995 concurrent expedited grace periods into account, 5996 if available. 5997 5998 rcutorture.gp_cond_exp_full= [KNL] 5999 Use conditional/asynchronous update-side 6000 expedited-grace-period primitives that also take 6001 concurrent normal grace periods into account, 6002 if available. 6003 6004 rcutorture.gp_cond_wi= [KNL] 6005 Nominal wait interval for normal conditional 6006 grace periods (specified by rcutorture's 6007 gp_cond and gp_cond_full module parameters), 6008 in microseconds. The actual wait interval will 6009 be randomly selected to nanosecond granularity up 6010 to this wait interval. Defaults to 16 jiffies, 6011 for example, 16,000 microseconds on a system 6012 with HZ=1000. 6013 6014 rcutorture.gp_cond_wi_exp= [KNL] 6015 Nominal wait interval for expedited conditional 6016 grace periods (specified by rcutorture's 6017 gp_cond_exp and gp_cond_exp_full module 6018 parameters), in microseconds. The actual wait 6019 interval will be randomly selected to nanosecond 6020 granularity up to this wait interval. Defaults to 6021 128 microseconds. 6022 6023 rcutorture.gp_exp= [KNL] 6024 Use expedited update-side primitives, if available. 6025 6026 rcutorture.gp_normal= [KNL] 6027 Use normal (non-expedited) asynchronous 6028 update-side primitives, if available. 6029 6030 rcutorture.gp_poll= [KNL] 6031 Use polled update-side normal-grace-period 6032 primitives, if available. 6033 6034 rcutorture.gp_poll_exp= [KNL] 6035 Use polled update-side expedited-grace-period 6036 primitives, if available. 6037 6038 rcutorture.gp_poll_full= [KNL] 6039 Use polled update-side normal-grace-period 6040 primitives that also take concurrent expedited 6041 grace periods into account, if available. 6042 6043 rcutorture.gp_poll_exp_full= [KNL] 6044 Use polled update-side expedited-grace-period 6045 primitives that also take concurrent normal 6046 grace periods into account, if available. 6047 6048 rcutorture.gp_poll_wi= [KNL] 6049 Nominal wait interval for normal conditional 6050 grace periods (specified by rcutorture's 6051 gp_poll and gp_poll_full module parameters), 6052 in microseconds. The actual wait interval will 6053 be randomly selected to nanosecond granularity up 6054 to this wait interval. Defaults to 16 jiffies, 6055 for example, 16,000 microseconds on a system 6056 with HZ=1000. 6057 6058 rcutorture.gp_poll_wi_exp= [KNL] 6059 Nominal wait interval for expedited conditional 6060 grace periods (specified by rcutorture's 6061 gp_poll_exp and gp_poll_exp_full module 6062 parameters), in microseconds. The actual wait 6063 interval will be randomly selected to nanosecond 6064 granularity up to this wait interval. Defaults to 6065 128 microseconds. 6066 6067 rcutorture.gp_sync= [KNL] 6068 Use normal (non-expedited) synchronous 6069 update-side primitives, if available. If all 6070 of rcutorture.gp_cond=, rcutorture.gp_exp=, 6071 rcutorture.gp_normal=, and rcutorture.gp_sync= 6072 are zero, rcutorture acts as if is interpreted 6073 they are all non-zero. 6074 6075 rcutorture.gpwrap_lag= [KNL] 6076 Enable grace-period wrap lag testing. Setting 6077 to false prevents the gpwrap lag test from 6078 running. Default is true. 6079 6080 rcutorture.gpwrap_lag_gps= [KNL] 6081 Set the value for grace-period wrap lag during 6082 active lag testing periods. This controls how many 6083 grace periods differences we tolerate between 6084 rdp and rnp's gp_seq before setting overflow flag. 6085 The default is always set to 8. 6086 6087 rcutorture.gpwrap_lag_cycle_mins= [KNL] 6088 Set the total cycle duration for gpwrap lag 6089 testing in minutes. This is the total time for 6090 one complete cycle of active and inactive 6091 testing periods. Default is 30 minutes. 6092 6093 rcutorture.gpwrap_lag_active_mins= [KNL] 6094 Set the duration for which gpwrap lag is active 6095 within each cycle, in minutes. During this time, 6096 the grace-period wrap lag will be set to the 6097 value specified by gpwrap_lag_gps. Default is 6098 5 minutes. 6099 6100 rcutorture.irqreader= [KNL] 6101 Run RCU readers from irq handlers, or, more 6102 accurately, from a timer handler. Not all RCU 6103 flavors take kindly to this sort of thing. 6104 6105 rcutorture.leakpointer= [KNL] 6106 Leak an RCU-protected pointer out of the reader. 6107 This can of course result in splats, and is 6108 intended to test the ability of things like 6109 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 6110 such leaks. 6111 6112 rcutorture.n_barrier_cbs= [KNL] 6113 Set callbacks/threads for rcu_barrier() testing. 6114 6115 rcutorture.nfakewriters= [KNL] 6116 Set number of concurrent RCU writers. These just 6117 stress RCU, they don't participate in the actual 6118 test, hence the "fake". 6119 6120 rcutorture.nocbs_nthreads= [KNL] 6121 Set number of RCU callback-offload togglers. 6122 Zero (the default) disables toggling. 6123 6124 rcutorture.nocbs_toggle= [KNL] 6125 Set the delay in milliseconds between successive 6126 callback-offload toggling attempts. 6127 6128 rcutorture.nreaders= [KNL] 6129 Set number of RCU readers. The value -1 selects 6130 N-1, where N is the number of CPUs. A value 6131 "n" less than -1 selects N-n-2, where N is again 6132 the number of CPUs. For example, -2 selects N 6133 (the number of CPUs), -3 selects N+1, and so on. 6134 6135 rcutorture.object_debug= [KNL] 6136 Enable debug-object double-call_rcu() testing. 6137 6138 rcutorture.onoff_holdoff= [KNL] 6139 Set time (s) after boot for CPU-hotplug testing. 6140 6141 rcutorture.onoff_interval= [KNL] 6142 Set time (jiffies) between CPU-hotplug operations, 6143 or zero to disable CPU-hotplug testing. 6144 6145 rcutorture.preempt_duration= [KNL] 6146 Set duration (in milliseconds) of preemptions 6147 by a high-priority FIFO real-time task. Set to 6148 zero (the default) to disable. The CPUs to 6149 preempt are selected randomly from the set that 6150 are online at a given point in time. Races with 6151 CPUs going offline are ignored, with that attempt 6152 at preemption skipped. 6153 6154 rcutorture.preempt_interval= [KNL] 6155 Set interval (in milliseconds, defaulting to one 6156 second) between preemptions by a high-priority 6157 FIFO real-time task. This delay is mediated 6158 by an hrtimer and is further fuzzed to avoid 6159 inadvertent synchronizations. 6160 6161 rcutorture.read_exit_burst= [KNL] 6162 The number of times in a given read-then-exit 6163 episode that a set of read-then-exit kthreads 6164 is spawned. 6165 6166 rcutorture.read_exit_delay= [KNL] 6167 The delay, in seconds, between successive 6168 read-then-exit testing episodes. 6169 6170 rcutorture.reader_flavor= [KNL] 6171 A bit mask indicating which readers to use. 6172 If there is more than one bit set, the readers 6173 are entered from low-order bit up, and are 6174 exited in the opposite order. For SRCU, the 6175 0x1 bit is normal readers, 0x2 NMI-safe readers, 6176 and 0x4 light-weight readers. 6177 6178 rcutorture.shuffle_interval= [KNL] 6179 Set task-shuffle interval (s). Shuffling tasks 6180 allows some CPUs to go into dyntick-idle mode 6181 during the rcutorture test. 6182 6183 rcutorture.shutdown_secs= [KNL] 6184 Set time (s) after boot system shutdown. This 6185 is useful for hands-off automated testing. 6186 6187 rcutorture.stall_cpu= [KNL] 6188 Duration of CPU stall (s) to test RCU CPU stall 6189 warnings, zero to disable. 6190 6191 rcutorture.stall_cpu_block= [KNL] 6192 Sleep while stalling if set. This will result 6193 in warnings from preemptible RCU in addition to 6194 any other stall-related activity. Note that 6195 in kernels built with CONFIG_PREEMPTION=n and 6196 CONFIG_PREEMPT_COUNT=y, this parameter will 6197 cause the CPU to pass through a quiescent state. 6198 Given CONFIG_PREEMPTION=n, this will suppress 6199 RCU CPU stall warnings, but will instead result 6200 in scheduling-while-atomic splats. 6201 6202 Use of this module parameter results in splats. 6203 6204 6205 rcutorture.stall_cpu_holdoff= [KNL] 6206 Time to wait (s) after boot before inducing stall. 6207 6208 rcutorture.stall_cpu_irqsoff= [KNL] 6209 Disable interrupts while stalling if set, but only 6210 on the first stall in the set. 6211 6212 rcutorture.stall_cpu_repeat= [KNL] 6213 Number of times to repeat the stall sequence, 6214 so that rcutorture.stall_cpu_repeat=3 will result 6215 in four stall sequences. 6216 6217 rcutorture.stall_gp_kthread= [KNL] 6218 Duration (s) of forced sleep within RCU 6219 grace-period kthread to test RCU CPU stall 6220 warnings, zero to disable. If both stall_cpu 6221 and stall_gp_kthread are specified, the 6222 kthread is starved first, then the CPU. 6223 6224 rcutorture.stat_interval= [KNL] 6225 Time (s) between statistics printk()s. 6226 6227 rcutorture.stutter= [KNL] 6228 Time (s) to stutter testing, for example, specifying 6229 five seconds causes the test to run for five seconds, 6230 wait for five seconds, and so on. This tests RCU's 6231 ability to transition abruptly to and from idle. 6232 6233 rcutorture.test_boost= [KNL] 6234 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 6235 "Maybe" means test if the RCU implementation 6236 under test support RCU priority boosting. 6237 6238 rcutorture.test_boost_duration= [KNL] 6239 Duration (s) of each individual boost test. 6240 6241 rcutorture.test_boost_holdoff= [KNL] 6242 Holdoff time (s) from start of test to the start 6243 of RCU priority-boost testing. Defaults to zero, 6244 that is, no holdoff. 6245 6246 rcutorture.test_boost_interval= [KNL] 6247 Interval (s) between each boost test. 6248 6249 rcutorture.test_no_idle_hz= [KNL] 6250 Test RCU's dyntick-idle handling. See also the 6251 rcutorture.shuffle_interval parameter. 6252 6253 rcutorture.torture_type= [KNL] 6254 Specify the RCU implementation to test. 6255 6256 rcutorture.verbose= [KNL] 6257 Enable additional printk() statements. 6258 6259 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 6260 Dump ftrace buffer after reporting RCU CPU 6261 stall warning. 6262 6263 rcupdate.rcu_cpu_stall_notifiers= [KNL] 6264 Provide RCU CPU stall notifiers, but see the 6265 warnings in the RCU_CPU_STALL_NOTIFIER Kconfig 6266 option's help text. TL;DR: You almost certainly 6267 do not want rcupdate.rcu_cpu_stall_notifiers. 6268 6269 rcupdate.rcu_cpu_stall_suppress= [KNL] 6270 Suppress RCU CPU stall warning messages. 6271 6272 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 6273 Suppress RCU CPU stall warning messages and 6274 rcutorture writer stall warnings that occur 6275 during early boot, that is, during the time 6276 before the init task is spawned. 6277 6278 rcupdate.rcu_cpu_stall_timeout= [KNL] 6279 Set timeout for RCU CPU stall warning messages. 6280 The value is in seconds and the maximum allowed 6281 value is 300 seconds. 6282 6283 rcupdate.rcu_exp_cpu_stall_timeout= [KNL] 6284 Set timeout for expedited RCU CPU stall warning 6285 messages. The value is in milliseconds 6286 and the maximum allowed value is 21000 6287 milliseconds. Please note that this value is 6288 adjusted to an arch timer tick resolution. 6289 Setting this to zero causes the value from 6290 rcupdate.rcu_cpu_stall_timeout to be used (after 6291 conversion from seconds to milliseconds). 6292 6293 rcupdate.rcu_cpu_stall_cputime= [KNL] 6294 Provide statistics on the cputime and count of 6295 interrupts and tasks during the sampling period. For 6296 multiple continuous RCU stalls, all sampling periods 6297 begin at half of the first RCU stall timeout. 6298 6299 rcupdate.rcu_exp_stall_task_details= [KNL] 6300 Print stack dumps of any tasks blocking the 6301 current expedited RCU grace period during an 6302 expedited RCU CPU stall warning. 6303 6304 rcupdate.rcu_expedited= [KNL] 6305 Use expedited grace-period primitives, for 6306 example, synchronize_rcu_expedited() instead 6307 of synchronize_rcu(). This reduces latency, 6308 but can increase CPU utilization, degrade 6309 real-time latency, and degrade energy efficiency. 6310 No effect on CONFIG_TINY_RCU kernels. 6311 6312 rcupdate.rcu_normal= [KNL] 6313 Use only normal grace-period primitives, 6314 for example, synchronize_rcu() instead of 6315 synchronize_rcu_expedited(). This improves 6316 real-time latency, CPU utilization, and 6317 energy efficiency, but can expose users to 6318 increased grace-period latency. This parameter 6319 overrides rcupdate.rcu_expedited. No effect on 6320 CONFIG_TINY_RCU kernels. 6321 6322 rcupdate.rcu_normal_after_boot= [KNL] 6323 Once boot has completed (that is, after 6324 rcu_end_inkernel_boot() has been invoked), use 6325 only normal grace-period primitives. No effect 6326 on CONFIG_TINY_RCU kernels. 6327 6328 But note that CONFIG_PREEMPT_RT=y kernels enables 6329 this kernel boot parameter, forcibly setting 6330 it to the value one, that is, converting any 6331 post-boot attempt at an expedited RCU grace 6332 period to instead use normal non-expedited 6333 grace-period processing. 6334 6335 rcupdate.rcu_task_collapse_lim= [KNL] 6336 Set the maximum number of callbacks present 6337 at the beginning of a grace period that allows 6338 the RCU Tasks flavors to collapse back to using 6339 a single callback queue. This switching only 6340 occurs when rcupdate.rcu_task_enqueue_lim is 6341 set to the default value of -1. 6342 6343 rcupdate.rcu_task_contend_lim= [KNL] 6344 Set the minimum number of callback-queuing-time 6345 lock-contention events per jiffy required to 6346 cause the RCU Tasks flavors to switch to per-CPU 6347 callback queuing. This switching only occurs 6348 when rcupdate.rcu_task_enqueue_lim is set to 6349 the default value of -1. 6350 6351 rcupdate.rcu_task_enqueue_lim= [KNL] 6352 Set the number of callback queues to use for the 6353 RCU Tasks family of RCU flavors. The default 6354 of -1 allows this to be automatically (and 6355 dynamically) adjusted. This parameter is intended 6356 for use in testing. 6357 6358 rcupdate.rcu_task_lazy_lim= [KNL] 6359 Number of callbacks on a given CPU that will 6360 cancel laziness on that CPU. Use -1 to disable 6361 cancellation of laziness, but be advised that 6362 doing so increases the danger of OOM due to 6363 callback flooding. 6364 6365 rcupdate.rcu_task_stall_info= [KNL] 6366 Set initial timeout in jiffies for RCU task stall 6367 informational messages, which give some indication 6368 of the problem for those not patient enough to 6369 wait for ten minutes. Informational messages are 6370 only printed prior to the stall-warning message 6371 for a given grace period. Disable with a value 6372 less than or equal to zero. Defaults to ten 6373 seconds. A change in value does not take effect 6374 until the beginning of the next grace period. 6375 6376 rcupdate.rcu_task_stall_info_mult= [KNL] 6377 Multiplier for time interval between successive 6378 RCU task stall informational messages for a given 6379 RCU tasks grace period. This value is clamped 6380 to one through ten, inclusive. It defaults to 6381 the value three, so that the first informational 6382 message is printed 10 seconds into the grace 6383 period, the second at 40 seconds, the third at 6384 160 seconds, and then the stall warning at 600 6385 seconds would prevent a fourth at 640 seconds. 6386 6387 rcupdate.rcu_task_stall_timeout= [KNL] 6388 Set timeout in jiffies for RCU task stall 6389 warning messages. Disable with a value less 6390 than or equal to zero. Defaults to ten minutes. 6391 A change in value does not take effect until 6392 the beginning of the next grace period. 6393 6394 rcupdate.rcu_tasks_lazy_ms= [KNL] 6395 Set timeout in milliseconds RCU Tasks asynchronous 6396 callback batching for call_rcu_tasks(). 6397 A negative value will take the default. A value 6398 of zero will disable batching. Batching is 6399 always disabled for synchronize_rcu_tasks(). 6400 6401 rcupdate.rcu_self_test= [KNL] 6402 Run the RCU early boot self tests 6403 6404 rdinit= [KNL] 6405 Format: <full_path> 6406 Run specified binary instead of /init from the ramdisk, 6407 used for early userspace startup. See initrd. 6408 6409 rdrand= [X86,EARLY] 6410 force - Override the decision by the kernel to hide the 6411 advertisement of RDRAND support (this affects 6412 certain AMD processors because of buggy BIOS 6413 support, specifically around the suspend/resume 6414 path). 6415 6416 rdt= [HW,X86,RDT] 6417 Turn on/off individual RDT features. List is: 6418 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 6419 mba, smba, bmec, abmc, sdciae, energy[:guid], 6420 perf[:guid]. 6421 E.g. to turn on cmt and turn off mba use: 6422 rdt=cmt,!mba 6423 To turn off all energy telemetry monitoring and ensure that 6424 perf telemetry monitoring associated with guid 0x12345 6425 is enabled use: 6426 rdt=!energy,perf:0x12345 6427 6428 reboot= [KNL] 6429 Format (x86 or x86_64): 6430 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \ 6431 [[,]s[mp]#### \ 6432 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 6433 [[,]f[orce] 6434 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 6435 (prefix with 'panic_' to set mode for panic 6436 reboot only), 6437 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 6438 reboot_force is either force or not specified, 6439 reboot_cpu is s[mp]#### with #### being the processor 6440 to be used for rebooting. 6441 6442 acpi 6443 Use the ACPI RESET_REG in the FADT. If ACPI is not 6444 configured or the ACPI reset does not work, the reboot 6445 path attempts the reset using the keyboard controller. 6446 6447 bios 6448 Use the CPU reboot vector for warm reset 6449 6450 cold 6451 Set the cold reboot flag 6452 6453 default 6454 There are some built-in platform specific "quirks" 6455 - you may see: "reboot: <name> series board detected. 6456 Selecting <type> for reboots." In the case where you 6457 think the quirk is in error (e.g. you have newer BIOS, 6458 or newer board) using this option will ignore the 6459 built-in quirk table, and use the generic default 6460 reboot actions. 6461 6462 efi 6463 Use efi reset_system runtime service. If EFI is not 6464 configured or the EFI reset does not work, the reboot 6465 path attempts the reset using the keyboard controller. 6466 6467 force 6468 Don't stop other CPUs on reboot. This can make reboot 6469 more reliable in some cases. 6470 6471 kbd 6472 Use the keyboard controller. cold reset (default) 6473 6474 pci 6475 Use a write to the PCI config space register 0xcf9 to 6476 trigger reboot. 6477 6478 triple 6479 Force a triple fault (init) 6480 6481 warm 6482 Don't set the cold reboot flag 6483 6484 Using warm reset will be much faster especially on big 6485 memory systems because the BIOS will not go through 6486 the memory check. Disadvantage is that not all 6487 hardware will be completely reinitialized on reboot so 6488 there may be boot problems on some systems. 6489 6490 6491 refscale.holdoff= [KNL] 6492 Set test-start holdoff period. The purpose of 6493 this parameter is to delay the start of the 6494 test until boot completes in order to avoid 6495 interference. 6496 6497 refscale.lookup_instances= [KNL] 6498 Number of data elements to use for the forms of 6499 SLAB_TYPESAFE_BY_RCU testing. A negative number 6500 is negated and multiplied by nr_cpu_ids, while 6501 zero specifies nr_cpu_ids. 6502 6503 refscale.loops= [KNL] 6504 Set the number of loops over the synchronization 6505 primitive under test. Increasing this number 6506 reduces noise due to loop start/end overhead, 6507 but the default has already reduced the per-pass 6508 noise to a handful of picoseconds on ca. 2020 6509 x86 laptops. 6510 6511 refscale.nreaders= [KNL] 6512 Set number of readers. The default value of -1 6513 selects N, where N is roughly 75% of the number 6514 of CPUs. A value of zero is an interesting choice. 6515 6516 refscale.nruns= [KNL] 6517 Set number of runs, each of which is dumped onto 6518 the console log. 6519 6520 refscale.readdelay= [KNL] 6521 Set the read-side critical-section duration, 6522 measured in microseconds. 6523 6524 refscale.scale_type= [KNL] 6525 Specify the read-protection implementation to test. 6526 6527 refscale.shutdown= [KNL] 6528 Shut down the system at the end of the performance 6529 test. This defaults to 1 (shut it down) when 6530 refscale is built into the kernel and to 0 (leave 6531 it running) when refscale is built as a module. 6532 6533 refscale.verbose= [KNL] 6534 Enable additional printk() statements. 6535 6536 refscale.verbose_batched= [KNL] 6537 Batch the additional printk() statements. If zero 6538 (the default) or negative, print everything. Otherwise, 6539 print every Nth verbose statement, where N is the value 6540 specified. 6541 6542 regulator_ignore_unused 6543 [REGULATOR] 6544 Prevents regulator framework from disabling regulators 6545 that are unused, due no driver claiming them. This may 6546 be useful for debug and development, but should not be 6547 needed on a platform with proper driver support. 6548 6549 relax_domain_level= 6550 [KNL, SMP] Set scheduler's default relax_domain_level. 6551 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 6552 6553 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 6554 Format: <base1>,<size1>[,<base2>,<size2>,...] 6555 Reserve I/O ports or memory so the kernel won't use 6556 them. If <base> is less than 0x10000, the region 6557 is assumed to be I/O ports; otherwise it is memory. 6558 6559 reserve_mem= [RAM] 6560 Format: nn[KMG]:<align>:<label> 6561 Reserve physical memory and label it with a name that 6562 other subsystems can use to access it. This is typically 6563 used for systems that do not wipe the RAM, and this command 6564 line will try to reserve the same physical memory on 6565 soft reboots. Note, it is not guaranteed to be the same 6566 location. For example, if anything about the system changes 6567 or if booting a different kernel. It can also fail if KASLR 6568 places the kernel at the location of where the RAM reservation 6569 was from a previous boot, the new reservation will be at a 6570 different location. 6571 Any subsystem using this feature must add a way to verify 6572 that the contents of the physical memory is from a previous 6573 boot, as there may be cases where the memory will not be 6574 located at the same location. 6575 6576 The format is size:align:label for example, to request 6577 12 megabytes of 4096 alignment for ramoops: 6578 6579 reserve_mem=12M:4096:oops ramoops.mem_name=oops 6580 6581 reservetop= [X86-32,EARLY] 6582 Format: nn[KMG] 6583 Reserves a hole at the top of the kernel virtual 6584 address space. 6585 6586 reset_devices [KNL] Force drivers to reset the underlying device 6587 during initialization. 6588 6589 resume= [SWSUSP] 6590 Specify the partition device for software suspend 6591 Format: 6592 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 6593 6594 resume_offset= [SWSUSP] 6595 Specify the offset from the beginning of the partition 6596 given by "resume=" at which the swap header is located, 6597 in <PAGE_SIZE> units (needed only for swap files). 6598 See Documentation/power/swsusp-and-swap-files.rst 6599 6600 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 6601 read the resume files 6602 6603 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 6604 Useful for devices that are detected asynchronously 6605 (e.g. USB and MMC devices). 6606 6607 retain_initrd [RAM] Keep initrd memory after extraction. After boot, it will 6608 be accessible via /sys/firmware/initrd. 6609 6610 retbleed= [X86] Control mitigation of RETBleed (Arbitrary 6611 Speculative Code Execution with Return Instructions) 6612 vulnerability. 6613 6614 AMD-based UNRET and IBPB mitigations alone do not stop 6615 sibling threads from influencing the predictions of other 6616 sibling threads. For that reason, STIBP is used on pro- 6617 cessors that support it, and mitigate SMT on processors 6618 that don't. 6619 6620 off - no mitigation 6621 auto - automatically select a mitigation 6622 auto,nosmt - automatically select a mitigation, 6623 disabling SMT if necessary for 6624 the full mitigation (only on Zen1 6625 and older without STIBP). 6626 ibpb - On AMD, mitigate short speculation 6627 windows on basic block boundaries too. 6628 Safe, highest perf impact. It also 6629 enables STIBP if present. Not suitable 6630 on Intel. 6631 ibpb,nosmt - Like "ibpb" above but will disable SMT 6632 when STIBP is not available. This is 6633 the alternative for systems which do not 6634 have STIBP. 6635 unret - Force enable untrained return thunks, 6636 only effective on AMD f15h-f17h based 6637 systems. 6638 unret,nosmt - Like unret, but will disable SMT when STIBP 6639 is not available. This is the alternative for 6640 systems which do not have STIBP. 6641 6642 Selecting 'auto' will choose a mitigation method at run 6643 time according to the CPU. 6644 6645 Not specifying this option is equivalent to retbleed=auto. 6646 6647 rfkill.default_state= 6648 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 6649 etc. communication is blocked by default. 6650 1 Unblocked. 6651 6652 rfkill.master_switch_mode= 6653 0 The "airplane mode" button does nothing. 6654 1 The "airplane mode" button toggles between everything 6655 blocked and the previous configuration. 6656 2 The "airplane mode" button toggles between everything 6657 blocked and everything unblocked. 6658 6659 ring3mwait=disable 6660 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 6661 CPUs. 6662 6663 riscv_isa_fallback [RISCV,EARLY] 6664 When CONFIG_RISCV_ISA_FALLBACK is not enabled, permit 6665 falling back to detecting extension support by parsing 6666 "riscv,isa" property on devicetree systems when the 6667 replacement properties are not found. See the Kconfig 6668 entry for RISCV_ISA_FALLBACK. 6669 6670 riscv_nousercfi= 6671 all Disable user CFI ABI to userspace even if cpu extension 6672 are available. 6673 bcfi Disable user backward CFI ABI to userspace even if 6674 the shadow stack extension is available. 6675 fcfi Disable user forward CFI ABI to userspace even if the 6676 landing pad extension is available. 6677 6678 ro [KNL] Mount root device read-only on boot 6679 6680 rodata= [KNL,EARLY] 6681 on Mark read-only kernel memory as read-only (default). 6682 off Leave read-only kernel memory writable for debugging. 6683 noalias Mark read-only kernel memory as read-only but retain 6684 writable aliases in the direct map for regions outside 6685 of the kernel image. [arm64] 6686 6687 rockchip.usb_uart 6688 [EARLY] 6689 Enable the uart passthrough on the designated usb port 6690 on Rockchip SoCs. When active, the signals of the 6691 debug-uart get routed to the D+ and D- pins of the usb 6692 port and the regular usb controller gets disabled. 6693 6694 root= [KNL] Root filesystem 6695 Usually this is a block device specifier of some kind, 6696 see the early_lookup_bdev comment in 6697 block/early-lookup.c for details. 6698 Alternatively this can be "ram" for the legacy initial 6699 ramdisk, "nfs" and "cifs" for root on a network file 6700 system, or "mtd" and "ubi" for mounting from raw flash. 6701 6702 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 6703 mount the root filesystem 6704 6705 rootflags= [KNL] Set root filesystem mount option string 6706 6707 rseq_slice_ext= [KNL] RSEQ based time slice extension 6708 Format: boolean 6709 Control enablement of RSEQ based time slice extension. 6710 Default is 'on'. 6711 6712 initramfs_options= [KNL] 6713 Specify mount options for the initramfs mount. 6714 6715 rootfstype= [KNL] Set root filesystem type 6716 6717 rootwait [KNL] Wait (indefinitely) for root device to show up. 6718 Useful for devices that are detected asynchronously 6719 (e.g. USB and MMC devices). 6720 6721 rootwait= [KNL] Maximum time (in seconds) to wait for root device 6722 to show up before attempting to mount the root 6723 filesystem. 6724 6725 rproc_mem=nn[KMG][@address] 6726 [KNL,ARM,CMA] Remoteproc physical memory block. 6727 Memory area to be used by remote processor image, 6728 managed by CMA. 6729 6730 rseq_debug= [KNL] Enable or disable restartable sequence 6731 debug mode. Defaults to CONFIG_RSEQ_DEBUG_DEFAULT_ENABLE. 6732 Format: <bool> 6733 6734 rt_group_sched= [KNL] Enable or disable SCHED_RR/FIFO group scheduling 6735 when CONFIG_RT_GROUP_SCHED=y. Defaults to 6736 !CONFIG_RT_GROUP_SCHED_DEFAULT_DISABLED. 6737 Format: <bool> 6738 6739 rw [KNL] Mount root device read-write on boot 6740 6741 S [KNL] Run init in single mode 6742 6743 s390_iommu= [HW,S390] 6744 Set s390 IOTLB flushing mode 6745 strict 6746 With strict flushing every unmap operation will result 6747 in an IOTLB flush. Default is lazy flushing before 6748 reuse, which is faster. Deprecated, equivalent to 6749 iommu.strict=1. 6750 6751 s390_iommu_aperture= [KNL,S390] 6752 Specifies the size of the per device DMA address space 6753 accessible through the DMA and IOMMU APIs as a decimal 6754 factor of the size of main memory. 6755 The default is 1 meaning that one can concurrently use 6756 as many DMA addresses as physical memory is installed, 6757 if supported by hardware, and thus map all of memory 6758 once. With a value of 2 one can map all of memory twice 6759 and so on. As a special case a factor of 0 imposes no 6760 restrictions other than those given by hardware at the 6761 cost of significant additional memory use for tables. 6762 6763 sa1100ir [NET] 6764 See drivers/net/irda/sa1100_ir.c. 6765 6766 sched_proxy_exec= [KNL] 6767 Enables or disables "proxy execution" style 6768 solution to mutex-based priority inversion. 6769 Format: <bool> 6770 6771 sched_verbose [KNL,EARLY] Enables verbose scheduler debug messages. 6772 6773 schedstats= [KNL,X86] Enable or disable scheduled statistics. 6774 Allowed values are enable and disable. This feature 6775 incurs a small amount of overhead in the scheduler 6776 but is useful for debugging and performance tuning. 6777 6778 sched_thermal_decay_shift= 6779 [Deprecated] 6780 [KNL, SMP] Set a decay shift for scheduler thermal 6781 pressure signal. Thermal pressure signal follows the 6782 default decay period of other scheduler pelt 6783 signals(usually 32 ms but configurable). Setting 6784 sched_thermal_decay_shift will left shift the decay 6785 period for the thermal pressure signal by the shift 6786 value. 6787 i.e. with the default pelt decay period of 32 ms 6788 sched_thermal_decay_shift thermal pressure decay pr 6789 1 64 ms 6790 2 128 ms 6791 and so on. 6792 Format: integer between 0 and 10 6793 Default is 0. 6794 6795 scftorture.holdoff= [KNL] 6796 Number of seconds to hold off before starting 6797 test. Defaults to zero for module insertion and 6798 to 10 seconds for built-in smp_call_function() 6799 tests. 6800 6801 scftorture.longwait= [KNL] 6802 Request ridiculously long waits randomly selected 6803 up to the chosen limit in seconds. Zero (the 6804 default) disables this feature. Please note 6805 that requesting even small non-zero numbers of 6806 seconds can result in RCU CPU stall warnings, 6807 softlockup complaints, and so on. 6808 6809 scftorture.nthreads= [KNL] 6810 Number of kthreads to spawn to invoke the 6811 smp_call_function() family of functions. 6812 The default of -1 specifies a number of kthreads 6813 equal to the number of CPUs. 6814 6815 scftorture.onoff_holdoff= [KNL] 6816 Number seconds to wait after the start of the 6817 test before initiating CPU-hotplug operations. 6818 6819 scftorture.onoff_interval= [KNL] 6820 Number seconds to wait between successive 6821 CPU-hotplug operations. Specifying zero (which 6822 is the default) disables CPU-hotplug operations. 6823 6824 scftorture.shutdown_secs= [KNL] 6825 The number of seconds following the start of the 6826 test after which to shut down the system. The 6827 default of zero avoids shutting down the system. 6828 Non-zero values are useful for automated tests. 6829 6830 scftorture.stat_interval= [KNL] 6831 The number of seconds between outputting the 6832 current test statistics to the console. A value 6833 of zero disables statistics output. 6834 6835 scftorture.stutter_cpus= [KNL] 6836 The number of jiffies to wait between each change 6837 to the set of CPUs under test. 6838 6839 scftorture.use_cpus_read_lock= [KNL] 6840 Use use_cpus_read_lock() instead of the default 6841 preempt_disable() to disable CPU hotplug 6842 while invoking one of the smp_call_function*() 6843 functions. 6844 6845 scftorture.verbose= [KNL] 6846 Enable additional printk() statements. 6847 6848 scftorture.weight_single= [KNL] 6849 The probability weighting to use for the 6850 smp_call_function_single() function with a zero 6851 "wait" parameter. A value of -1 selects the 6852 default if all other weights are -1. However, 6853 if at least one weight has some other value, a 6854 value of -1 will instead select a weight of zero. 6855 6856 scftorture.weight_single_wait= [KNL] 6857 The probability weighting to use for the 6858 smp_call_function_single() function with a 6859 non-zero "wait" parameter. See weight_single. 6860 6861 scftorture.weight_many= [KNL] 6862 The probability weighting to use for the 6863 smp_call_function_many() function with a zero 6864 "wait" parameter. See weight_single. 6865 Note well that setting a high probability for 6866 this weighting can place serious IPI load 6867 on the system. 6868 6869 scftorture.weight_many_wait= [KNL] 6870 The probability weighting to use for the 6871 smp_call_function_many() function with a 6872 non-zero "wait" parameter. See weight_single 6873 and weight_many. 6874 6875 scftorture.weight_all= [KNL] 6876 The probability weighting to use for the 6877 smp_call_function_all() function with a zero 6878 "wait" parameter. See weight_single and 6879 weight_many. 6880 6881 scftorture.weight_all_wait= [KNL] 6882 The probability weighting to use for the 6883 smp_call_function_all() function with a 6884 non-zero "wait" parameter. See weight_single 6885 and weight_many. 6886 6887 sdw_mclk_divider=[SDW] 6888 Specify the MCLK divider for Intel SoundWire buses in 6889 case the BIOS does not provide the clock rate properly. 6890 6891 skew_tick= [KNL,EARLY] Offset the periodic timer tick per cpu to mitigate 6892 xtime_lock contention on larger systems, and/or RCU lock 6893 contention on all systems with CONFIG_MAXSMP set. 6894 Format: { "0" | "1" } 6895 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 6896 1 -- enable. 6897 Note: increases power consumption, thus should only be 6898 enabled if running jitter sensitive (HPC/RT) workloads. 6899 6900 security= [SECURITY] Choose a legacy "major" security module to 6901 enable at boot. This has been deprecated by the 6902 "lsm=" parameter. 6903 6904 selinux= [SELINUX] Disable or enable SELinux at boot time. 6905 Format: { "0" | "1" } 6906 See security/selinux/Kconfig help text. 6907 0 -- disable. 6908 1 -- enable. 6909 Default value is 1. 6910 6911 serialnumber [BUGS=X86-32] 6912 6913 sev=option[,option...] [X86-64] 6914 6915 debug 6916 Enable debug messages. 6917 6918 nosnp 6919 Do not enable SEV-SNP (applies to host/hypervisor 6920 only). Setting 'nosnp' avoids the RMP check overhead 6921 in memory accesses when users do not want to run 6922 SEV-SNP guests. 6923 6924 shapers= [NET] 6925 Maximal number of shapers. 6926 6927 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 6928 Limit apic dumping. The parameter defines the maximal 6929 number of local apics being dumped. Also it is possible 6930 to set it to "all" by meaning -- no limit here. 6931 Format: { 1 (default) | 2 | ... | all }. 6932 The parameter valid if only apic=debug or 6933 apic=verbose is specified. 6934 Example: apic=debug show_lapic=all 6935 6936 slab_debug[=options[,slabs][;[options[,slabs]]...] [MM] 6937 Enabling slab_debug allows one to determine the 6938 culprit if slab objects become corrupted. Enabling 6939 slab_debug can create guard zones around objects and 6940 may poison objects when not in use. Also tracks the 6941 last alloc / free. For more information see 6942 Documentation/admin-guide/mm/slab.rst. 6943 (slub_debug legacy name also accepted for now) 6944 6945 Using this option implies the "no_hash_pointers" 6946 option which can be undone by adding the 6947 "hash_pointers=always" option. 6948 6949 slab_max_order= [MM] 6950 Determines the maximum allowed order for slabs. 6951 A high setting may cause OOMs due to memory 6952 fragmentation. For more information see 6953 Documentation/admin-guide/mm/slab.rst. 6954 (slub_max_order legacy name also accepted for now) 6955 6956 slab_merge [MM] 6957 Enable merging of slabs with similar size when the 6958 kernel is built without CONFIG_SLAB_MERGE_DEFAULT. 6959 (slub_merge legacy name also accepted for now) 6960 6961 slab_min_objects= [MM] 6962 The minimum number of objects per slab. SLUB will 6963 increase the slab order up to slab_max_order to 6964 generate a sufficiently large slab able to contain 6965 the number of objects indicated. The higher the number 6966 of objects the smaller the overhead of tracking slabs 6967 and the less frequently locks need to be acquired. 6968 For more information see 6969 Documentation/admin-guide/mm/slab.rst. 6970 (slub_min_objects legacy name also accepted for now) 6971 6972 slab_min_order= [MM] 6973 Determines the minimum page order for slabs. Must be 6974 lower or equal to slab_max_order. For more information see 6975 Documentation/admin-guide/mm/slab.rst. 6976 (slub_min_order legacy name also accepted for now) 6977 6978 slab_nomerge [MM] 6979 Disable merging of slabs with similar size. May be 6980 necessary if there is some reason to distinguish 6981 allocs to different slabs, especially in hardened 6982 environments where the risk of heap overflows and 6983 layout control by attackers can usually be 6984 frustrated by disabling merging. This will reduce 6985 most of the exposure of a heap attack to a single 6986 cache (risks via metadata attacks are mostly 6987 unchanged). Debug options disable merging on their 6988 own. 6989 For more information see 6990 Documentation/admin-guide/mm/slab.rst. 6991 (slub_nomerge legacy name also accepted for now) 6992 6993 slab_strict_numa [MM] 6994 Support memory policies on a per object level 6995 in the slab allocator. The default is for memory 6996 policies to be applied at the folio level when 6997 a new folio is needed or a partial folio is 6998 retrieved from the lists. Increases overhead 6999 in the slab fastpaths but gains more accurate 7000 NUMA kernel object placement which helps with slow 7001 interconnects in NUMA systems. 7002 7003 slram= [HW,MTD] 7004 7005 smart2= [HW] 7006 Format: <io1>[,<io2>[,...,<io8>]] 7007 7008 smp.csd_lock_timeout= [KNL] 7009 Specify the period of time in milliseconds 7010 that smp_call_function() and friends will wait 7011 for a CPU to release the CSD lock. This is 7012 useful when diagnosing bugs involving CPUs 7013 disabling interrupts for extended periods 7014 of time. Defaults to 5,000 milliseconds, and 7015 setting a value of zero disables this feature. 7016 This feature may be more efficiently disabled 7017 using the csdlock_debug- kernel parameter. 7018 7019 smp.panic_on_ipistall= [KNL] 7020 If a csd_lock_timeout extends for more than 7021 the specified number of milliseconds, panic the 7022 system. By default, let CSD-lock acquisition 7023 take as long as they take. Specifying 300,000 7024 for this value provides a 5-minute timeout. 7025 7026 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 7027 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 7028 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 7029 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 7030 smsc-ircc2.ircc_irq= [HW] IRQ line 7031 smsc-ircc2.ircc_dma= [HW] DMA channel 7032 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 7033 0: Toshiba Satellite 1800 (GP data pin select) 7034 1: Fast pin select (default) 7035 2: ATC IRMode 7036 7037 smt= [KNL,MIPS,S390,EARLY] Set the maximum number of threads 7038 (logical CPUs) to use per physical CPU on systems 7039 capable of symmetric multithreading (SMT). Will 7040 be capped to the actual hardware limit. 7041 Format: <integer> 7042 Default: -1 (no limit) 7043 7044 softlockup_panic= 7045 [KNL] Should the soft-lockup detector generate panics. 7046 Format: <int> 7047 7048 A value of non-zero instructs the soft-lockup detector 7049 to panic the machine when a soft-lockup duration exceeds 7050 N thresholds. It is also controlled by the kernel.softlockup_panic 7051 sysctl and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 7052 respective build-time switch to that functionality. 7053 7054 softlockup_all_cpu_backtrace= 7055 [KNL] Should the soft-lockup detector generate 7056 backtraces on all cpus. 7057 Format: 0 | 1 7058 7059 sonypi.*= [HW] Sony Programmable I/O Control Device driver 7060 See Documentation/admin-guide/laptops/sonypi.rst 7061 7062 spectre_bhi= [X86] Control mitigation of Branch History Injection 7063 (BHI) vulnerability. This setting affects the 7064 deployment of the HW BHI control and the SW BHB 7065 clearing sequence. 7066 7067 on - (default) Enable the HW or SW mitigation as 7068 needed. This protects the kernel from 7069 both syscalls and VMs. 7070 vmexit - On systems which don't have the HW mitigation 7071 available, enable the SW mitigation on vmexit 7072 ONLY. On such systems, the host kernel is 7073 protected from VM-originated BHI attacks, but 7074 may still be vulnerable to syscall attacks. 7075 off - Disable the mitigation. 7076 7077 spectre_v2= [X86,EARLY] Control mitigation of Spectre variant 2 7078 (indirect branch speculation) vulnerability. 7079 The default operation protects the kernel from 7080 user space attacks. 7081 7082 on - unconditionally enable, implies 7083 spectre_v2_user=on 7084 off - unconditionally disable, implies 7085 spectre_v2_user=off 7086 auto - kernel detects whether your CPU model is 7087 vulnerable 7088 7089 Selecting 'on' will, and 'auto' may, choose a 7090 mitigation method at run time according to the 7091 CPU, the available microcode, the setting of the 7092 CONFIG_MITIGATION_RETPOLINE configuration option, 7093 and the compiler with which the kernel was built. 7094 7095 Selecting 'on' will also enable the mitigation 7096 against user space to user space task attacks. 7097 Selecting specific mitigation does not force enable 7098 user mitigations. 7099 7100 Selecting 'off' will disable both the kernel and 7101 the user space protections. 7102 7103 Specific mitigations can also be selected manually: 7104 7105 retpoline - replace indirect branches 7106 retpoline,generic - Retpolines 7107 retpoline,lfence - LFENCE; indirect branch 7108 retpoline,amd - alias for retpoline,lfence 7109 eibrs - Enhanced/Auto IBRS 7110 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines 7111 eibrs,lfence - Enhanced/Auto IBRS + LFENCE 7112 ibrs - use IBRS to protect kernel 7113 7114 Not specifying this option is equivalent to 7115 spectre_v2=auto. 7116 7117 spectre_v2_user= 7118 [X86] Control mitigation of Spectre variant 2 7119 (indirect branch speculation) vulnerability between 7120 user space tasks 7121 7122 on - Unconditionally enable mitigations. Is 7123 enforced by spectre_v2=on 7124 7125 off - Unconditionally disable mitigations. Is 7126 enforced by spectre_v2=off 7127 7128 prctl - Indirect branch speculation is enabled, 7129 but mitigation can be enabled via prctl 7130 per thread. The mitigation control state 7131 is inherited on fork. 7132 7133 prctl,ibpb 7134 - Like "prctl" above, but only STIBP is 7135 controlled per thread. IBPB is issued 7136 always when switching between different user 7137 space processes. 7138 7139 seccomp 7140 - Same as "prctl" above, but all seccomp 7141 threads will enable the mitigation unless 7142 they explicitly opt out. 7143 7144 seccomp,ibpb 7145 - Like "seccomp" above, but only STIBP is 7146 controlled per thread. IBPB is issued 7147 always when switching between different 7148 user space processes. 7149 7150 auto - Kernel selects the mitigation depending on 7151 the available CPU features and vulnerability. 7152 7153 Default mitigation: "prctl" 7154 7155 Not specifying this option is equivalent to 7156 spectre_v2_user=auto. 7157 7158 spec_rstack_overflow= 7159 [X86,EARLY] Control RAS overflow mitigation on AMD Zen CPUs 7160 7161 off - Disable mitigation 7162 microcode - Enable microcode mitigation only 7163 safe-ret - Enable sw-only safe RET mitigation (default) 7164 ibpb - Enable mitigation by issuing IBPB on 7165 kernel entry 7166 ibpb-vmexit - Issue IBPB only on VMEXIT 7167 (cloud-specific mitigation) 7168 7169 spec_store_bypass_disable= 7170 [HW,EARLY] Control Speculative Store Bypass (SSB) Disable mitigation 7171 (Speculative Store Bypass vulnerability) 7172 7173 Certain CPUs are vulnerable to an exploit against a 7174 a common industry wide performance optimization known 7175 as "Speculative Store Bypass" in which recent stores 7176 to the same memory location may not be observed by 7177 later loads during speculative execution. The idea 7178 is that such stores are unlikely and that they can 7179 be detected prior to instruction retirement at the 7180 end of a particular speculation execution window. 7181 7182 In vulnerable processors, the speculatively forwarded 7183 store can be used in a cache side channel attack, for 7184 example to read memory to which the attacker does not 7185 directly have access (e.g. inside sandboxed code). 7186 7187 This parameter controls whether the Speculative Store 7188 Bypass optimization is used. 7189 7190 On x86 the options are: 7191 7192 on - Unconditionally disable Speculative Store Bypass 7193 off - Unconditionally enable Speculative Store Bypass 7194 auto - Kernel detects whether the CPU model contains an 7195 implementation of Speculative Store Bypass and 7196 picks the most appropriate mitigation. If the 7197 CPU is not vulnerable, "off" is selected. If the 7198 CPU is vulnerable the default mitigation is 7199 architecture and Kconfig dependent. See below. 7200 prctl - Control Speculative Store Bypass per thread 7201 via prctl. Speculative Store Bypass is enabled 7202 for a process by default. The state of the control 7203 is inherited on fork. 7204 seccomp - Same as "prctl" above, but all seccomp threads 7205 will disable SSB unless they explicitly opt out. 7206 7207 Default mitigations: 7208 X86: "prctl" 7209 7210 On powerpc the options are: 7211 7212 on,auto - On Power8 and Power9 insert a store-forwarding 7213 barrier on kernel entry and exit. On Power7 7214 perform a software flush on kernel entry and 7215 exit. 7216 off - No action. 7217 7218 Not specifying this option is equivalent to 7219 spec_store_bypass_disable=auto. 7220 7221 split_lock_detect= 7222 [X86] Enable split lock detection or bus lock detection 7223 7224 When enabled (and if hardware support is present), atomic 7225 instructions that access data across cache line 7226 boundaries will result in an alignment check exception 7227 for split lock detection or a debug exception for 7228 bus lock detection. 7229 7230 off - not enabled 7231 7232 warn - the kernel will emit rate-limited warnings 7233 about applications triggering the #AC 7234 exception or the #DB exception. This mode is 7235 the default on CPUs that support split lock 7236 detection or bus lock detection. Default 7237 behavior is by #AC if both features are 7238 enabled in hardware. 7239 7240 fatal - the kernel will send SIGBUS to applications 7241 that trigger the #AC exception or the #DB 7242 exception. Default behavior is by #AC if 7243 both features are enabled in hardware. 7244 7245 ratelimit:N - 7246 Set system wide rate limit to N bus locks 7247 per second for bus lock detection. 7248 0 < N <= 1000. 7249 7250 N/A for split lock detection. 7251 7252 7253 If an #AC exception is hit in the kernel or in 7254 firmware (i.e. not while executing in user mode) 7255 the kernel will oops in either "warn" or "fatal" 7256 mode. 7257 7258 #DB exception for bus lock is triggered only when 7259 CPL > 0. 7260 7261 srbds= [X86,INTEL,EARLY] 7262 Control the Special Register Buffer Data Sampling 7263 (SRBDS) mitigation. 7264 7265 Certain CPUs are vulnerable to an MDS-like 7266 exploit which can leak bits from the random 7267 number generator. 7268 7269 By default, this issue is mitigated by 7270 microcode. However, the microcode fix can cause 7271 the RDRAND and RDSEED instructions to become 7272 much slower. Among other effects, this will 7273 result in reduced throughput from /dev/urandom. 7274 7275 The microcode mitigation can be disabled with 7276 the following option: 7277 7278 off: Disable mitigation and remove 7279 performance impact to RDRAND and RDSEED 7280 7281 srcutree.big_cpu_lim [KNL] 7282 Specifies the number of CPUs constituting a 7283 large system, such that srcu_struct structures 7284 should immediately allocate an srcu_node array. 7285 This kernel-boot parameter defaults to 128, 7286 but takes effect only when the low-order four 7287 bits of srcutree.convert_to_big is equal to 3 7288 (decide at boot). 7289 7290 srcutree.convert_to_big [KNL] 7291 Specifies under what conditions an SRCU tree 7292 srcu_struct structure will be converted to big 7293 form, that is, with an rcu_node tree: 7294 7295 0: Never. 7296 1: At init_srcu_struct() time. 7297 2: When rcutorture decides to. 7298 3: Decide at boot time (default). 7299 0x1X: Above plus if high contention. 7300 7301 Either way, the srcu_node tree will be sized based 7302 on the actual runtime number of CPUs (nr_cpu_ids) 7303 instead of the compile-time CONFIG_NR_CPUS. 7304 7305 srcutree.counter_wrap_check [KNL] 7306 Specifies how frequently to check for 7307 grace-period sequence counter wrap for the 7308 srcu_data structure's ->srcu_gp_seq_needed field. 7309 The greater the number of bits set in this kernel 7310 parameter, the less frequently counter wrap will 7311 be checked for. Note that the bottom two bits 7312 are ignored. 7313 7314 srcutree.exp_holdoff [KNL] 7315 Specifies how many nanoseconds must elapse 7316 since the end of the last SRCU grace period for 7317 a given srcu_struct until the next normal SRCU 7318 grace period will be considered for automatic 7319 expediting. Set to zero to disable automatic 7320 expediting. 7321 7322 srcutree.srcu_max_nodelay [KNL] 7323 Specifies the number of no-delay instances 7324 per jiffy for which the SRCU grace period 7325 worker thread will be rescheduled with zero 7326 delay. Beyond this limit, worker thread will 7327 be rescheduled with a sleep delay of one jiffy. 7328 7329 srcutree.srcu_max_nodelay_phase [KNL] 7330 Specifies the per-grace-period phase, number of 7331 non-sleeping polls of readers. Beyond this limit, 7332 grace period worker thread will be rescheduled 7333 with a sleep delay of one jiffy, between each 7334 rescan of the readers, for a grace period phase. 7335 7336 srcutree.srcu_retry_check_delay [KNL] 7337 Specifies number of microseconds of non-sleeping 7338 delay between each non-sleeping poll of readers. 7339 7340 srcutree.small_contention_lim [KNL] 7341 Specifies the number of update-side contention 7342 events per jiffy will be tolerated before 7343 initiating a conversion of an srcu_struct 7344 structure to big form. Note that the value of 7345 srcutree.convert_to_big must have the 0x10 bit 7346 set for contention-based conversions to occur. 7347 7348 ssbd= [ARM64,HW,EARLY] 7349 Speculative Store Bypass Disable control 7350 7351 On CPUs that are vulnerable to the Speculative 7352 Store Bypass vulnerability and offer a 7353 firmware based mitigation, this parameter 7354 indicates how the mitigation should be used: 7355 7356 force-on: Unconditionally enable mitigation for 7357 for both kernel and userspace 7358 force-off: Unconditionally disable mitigation for 7359 for both kernel and userspace 7360 kernel: Always enable mitigation in the 7361 kernel, and offer a prctl interface 7362 to allow userspace to register its 7363 interest in being mitigated too. 7364 7365 stack_guard_gap= [MM] 7366 override the default stack gap protection. The value 7367 is in page units and it defines how many pages prior 7368 to (for stacks growing down) resp. after (for stacks 7369 growing up) the main stack are reserved for no other 7370 mapping. Default value is 256 pages. 7371 7372 stack_depot_disable= [KNL,EARLY] 7373 Setting this to true through kernel command line will 7374 disable the stack depot thereby saving the static memory 7375 consumed by the stack hash table. By default this is set 7376 to false. 7377 7378 stack_depot_max_pools= [KNL,EARLY] 7379 Specify the maximum number of pools to use for storing 7380 stack traces. Pools are allocated on-demand up to this 7381 limit. Default value is 8191 pools. 7382 7383 stacktrace [FTRACE] 7384 Enable the stack tracer on boot up. 7385 7386 stacktrace_filter=[function-list] 7387 [FTRACE] Limit the functions that the stack tracer 7388 will trace at boot up. function-list is a comma-separated 7389 list of functions. This list can be changed at run 7390 time by the stack_trace_filter file in the debugfs 7391 tracing directory. Note, this enables stack tracing 7392 and the stacktrace above is not needed. 7393 7394 sti= [PARISC,HW] 7395 Format: <num> 7396 Set the STI (builtin display/keyboard on the HP-PARISC 7397 machines) console (graphic card) which should be used 7398 as the initial boot-console. 7399 See also comment in drivers/video/console/sticore.c. 7400 7401 sti_font= [HW] 7402 See comment in drivers/video/console/sticore.c. 7403 7404 stifb= [HW] 7405 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 7406 7407 strict_sas_size= 7408 [X86] 7409 Format: <bool> 7410 Enable or disable strict sigaltstack size checks 7411 against the required signal frame size which 7412 depends on the supported FPU features. This can 7413 be used to filter out binaries which have 7414 not yet been made aware of AT_MINSIGSTKSZ. 7415 7416 stress_hpt [PPC,EARLY] 7417 Limits the number of kernel HPT entries in the hash 7418 page table to increase the rate of hash page table 7419 faults on kernel addresses. 7420 7421 stress_slb [PPC,EARLY] 7422 Limits the number of kernel SLB entries, and flushes 7423 them frequently to increase the rate of SLB faults 7424 on kernel addresses. 7425 7426 no_slb_preload [PPC,EARLY] 7427 Disables slb preloading for userspace. 7428 7429 sunrpc.min_resvport= 7430 sunrpc.max_resvport= 7431 [NFS,SUNRPC] 7432 SunRPC servers often require that client requests 7433 originate from a privileged port (i.e. a port in the 7434 range 0 < portnr < 1024). 7435 An administrator who wishes to reserve some of these 7436 ports for other uses may adjust the range that the 7437 kernel's sunrpc client considers to be privileged 7438 using these two parameters to set the minimum and 7439 maximum port values. 7440 7441 sunrpc.svc_rpc_per_connection_limit= 7442 [NFS,SUNRPC] 7443 Limit the number of requests that the server will 7444 process in parallel from a single connection. 7445 The default value is 0 (no limit). 7446 7447 sunrpc.pool_mode= 7448 [NFS] 7449 Control how the NFS server code allocates CPUs to 7450 service thread pools. Depending on how many NICs 7451 you have and where their interrupts are bound, this 7452 option will affect which CPUs will do NFS serving. 7453 Note: this parameter cannot be changed while the 7454 NFS server is running. 7455 7456 auto the server chooses an appropriate mode 7457 automatically using heuristics 7458 global a single global pool contains all CPUs 7459 percpu one pool for each CPU 7460 pernode one pool for each NUMA node (equivalent 7461 to global on non-NUMA machines) 7462 7463 sunrpc.tcp_slot_table_entries= 7464 sunrpc.udp_slot_table_entries= 7465 [NFS,SUNRPC] 7466 Sets the upper limit on the number of simultaneous 7467 RPC calls that can be sent from the client to a 7468 server. Increasing these values may allow you to 7469 improve throughput, but will also increase the 7470 amount of memory reserved for use by the client. 7471 7472 suspend.pm_test_delay= 7473 [SUSPEND] 7474 Sets the number of seconds to remain in a suspend test 7475 mode before resuming the system (see 7476 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 7477 is set. Default value is 5. 7478 7479 svm= [PPC] 7480 Format: { on | off | y | n | 1 | 0 } 7481 This parameter controls use of the Protected 7482 Execution Facility on pSeries. 7483 7484 swiotlb= [ARM,PPC,MIPS,X86,S390,EARLY] 7485 Format: { <int> [,<int>] | force | noforce } 7486 <int> -- Number of I/O TLB slabs 7487 <int> -- Second integer after comma. Number of swiotlb 7488 areas with their own lock. Will be rounded up 7489 to a power of 2. 7490 force -- force using of bounce buffers even if they 7491 wouldn't be automatically used by the kernel 7492 noforce -- Never use bounce buffers (for debugging) 7493 7494 switches= [HW,M68k,EARLY] 7495 7496 sysctl.*= [KNL] 7497 Set a sysctl parameter, right before loading the init 7498 process, as if the value was written to the respective 7499 /proc/sys/... file. Both '.' and '/' are recognized as 7500 separators. Unrecognized parameters and invalid values 7501 are reported in the kernel log. Sysctls registered 7502 later by a loaded module cannot be set this way. 7503 Example: sysctl.vm.swappiness=40 7504 7505 sysrq_always_enabled 7506 [KNL] 7507 Ignore sysrq setting - this boot parameter will 7508 neutralize any effect of /proc/sys/kernel/sysrq. 7509 Useful for debugging. 7510 7511 tcpmhash_entries= [KNL,NET] 7512 Set the number of tcp_metrics_hash slots. 7513 Default value is 8192 or 16384 depending on total 7514 ram pages. This is used to specify the TCP metrics 7515 cache size. See Documentation/networking/ip-sysctl.rst 7516 "tcp_no_metrics_save" section for more details. 7517 7518 tdfx= [HW,DRM] 7519 7520 test_suspend= [SUSPEND] 7521 Format: { "mem" | "standby" | "freeze" }[,N] 7522 Specify "mem" (for Suspend-to-RAM) or "standby" (for 7523 standby suspend) or "freeze" (for suspend type freeze) 7524 as the system sleep state during system startup with 7525 the optional capability to repeat N number of times. 7526 The system is woken from this state using a 7527 wakeup-capable RTC alarm. 7528 7529 thash_entries= [KNL,NET] 7530 Set number of hash buckets for TCP connection 7531 7532 thermal.act= [HW,ACPI] 7533 -1: disable all active trip points in all thermal zones 7534 <degrees C>: override all lowest active trip points 7535 7536 thermal.crt= [HW,ACPI] 7537 -1: disable all critical trip points in all thermal zones 7538 <degrees C>: override all critical trip points 7539 7540 thermal.off= [HW,ACPI] 7541 1: disable ACPI thermal control 7542 7543 thermal.psv= [HW,ACPI] 7544 -1: disable all passive trip points 7545 <degrees C>: override all passive trip points to this 7546 value 7547 7548 thermal.tzp= [HW,ACPI] 7549 Specify global default ACPI thermal zone polling rate 7550 <deci-seconds>: poll all this frequency 7551 0: no polling (default) 7552 7553 thp_anon= [KNL] 7554 Format: <size>[KMG],<size>[KMG]:<state>;<size>[KMG]-<size>[KMG]:<state> 7555 state is one of "always", "madvise", "never" or "inherit". 7556 Control the default behavior of the system with respect 7557 to anonymous transparent hugepages. 7558 Can be used multiple times for multiple anon THP sizes. 7559 See Documentation/admin-guide/mm/transhuge.rst for more 7560 details. 7561 7562 threadirqs [KNL,EARLY] 7563 Force threading of all interrupt handlers except those 7564 marked explicitly IRQF_NO_THREAD. 7565 7566 thp_shmem= [KNL] 7567 Format: <size>[KMG],<size>[KMG]:<policy>;<size>[KMG]-<size>[KMG]:<policy> 7568 Control the default policy of each hugepage size for the 7569 internal shmem mount. <policy> is one of policies available 7570 for the shmem mount ("always", "inherit", "never", "within_size", 7571 and "advise"). 7572 It can be used multiple times for multiple shmem THP sizes. 7573 See Documentation/admin-guide/mm/transhuge.rst for more 7574 details. 7575 7576 topology= [S390,EARLY] 7577 Format: {off | on} 7578 Specify if the kernel should make use of the cpu 7579 topology information if the hardware supports this. 7580 The scheduler will make use of this information and 7581 e.g. base its process migration decisions on it. 7582 Default is on. 7583 7584 torture.disable_onoff_at_boot= [KNL] 7585 Prevent the CPU-hotplug component of torturing 7586 until after init has spawned. 7587 7588 torture.ftrace_dump_at_shutdown= [KNL] 7589 Dump the ftrace buffer at torture-test shutdown, 7590 even if there were no errors. This can be a 7591 very costly operation when many torture tests 7592 are running concurrently, especially on systems 7593 with rotating-rust storage. 7594 7595 torture.verbose_sleep_frequency= [KNL] 7596 Specifies how many verbose printk()s should be 7597 emitted between each sleep. The default of zero 7598 disables verbose-printk() sleeping. 7599 7600 torture.verbose_sleep_duration= [KNL] 7601 Duration of each verbose-printk() sleep in jiffies. 7602 7603 tpm.disable_pcr_integrity= [HW,TPM] 7604 Do not protect PCR registers from unintended physical 7605 access, or interposers in the bus by the means of 7606 having an integrity protected session wrapped around 7607 TPM2_PCR_Extend command. Consider this in a situation 7608 where TPM is heavily utilized by IMA, thus protection 7609 causing a major performance hit, and the space where 7610 machines are deployed is by other means guarded. 7611 7612 tpm_crb_ffa.busy_timeout_ms= [ARM64,TPM] 7613 Maximum time in milliseconds to retry sending a message 7614 to the TPM service before giving up. This parameter controls 7615 how long the system will continue retrying when the TPM 7616 service is busy. 7617 Format: <unsigned int> 7618 Default: 2000 (2 seconds) 7619 7620 tpm_suspend_pcr=[HW,TPM] 7621 Format: integer pcr id 7622 Specify that at suspend time, the tpm driver 7623 should extend the specified pcr with zeros, 7624 as a workaround for some chips which fail to 7625 flush the last written pcr on TPM_SaveState. 7626 This will guarantee that all the other pcrs 7627 are saved. 7628 7629 tpm_tis.interrupts= [HW,TPM] 7630 Enable interrupts for the MMIO based physical layer 7631 for the FIFO interface. By default it is set to false 7632 (0). For more information about TPM hardware interfaces 7633 defined by Trusted Computing Group (TCG) see 7634 https://trustedcomputinggroup.org/resource/pc-client-platform-tpm-profile-ptp-specification/ 7635 7636 tp_printk [FTRACE] 7637 Have the tracepoints sent to printk as well as the 7638 tracing ring buffer. This is useful for early boot up 7639 where the system hangs or reboots and does not give the 7640 option for reading the tracing buffer or performing a 7641 ftrace_dump_on_oops. 7642 7643 To turn off having tracepoints sent to printk, 7644 echo 0 > /proc/sys/kernel/tracepoint_printk 7645 Note, echoing 1 into this file without the 7646 tp_printk kernel cmdline option has no effect. 7647 7648 The tp_printk_stop_on_boot (see below) can also be used 7649 to stop the printing of events to console at 7650 late_initcall_sync. 7651 7652 ** CAUTION ** 7653 7654 Having tracepoints sent to printk() and activating high 7655 frequency tracepoints such as irq or sched, can cause 7656 the system to live lock. 7657 7658 tp_printk_stop_on_boot [FTRACE] 7659 When tp_printk (above) is set, it can cause a lot of noise 7660 on the console. It may be useful to only include the 7661 printing of events during boot up, as user space may 7662 make the system inoperable. 7663 7664 This command line option will stop the printing of events 7665 to console at the late_initcall_sync() time frame. 7666 7667 trace_buf_size=nn[KMG] 7668 [FTRACE] will set tracing buffer size on each cpu. 7669 7670 trace_clock= [FTRACE] Set the clock used for tracing events 7671 at boot up. 7672 local - Use the per CPU time stamp counter 7673 (converted into nanoseconds). Fast, but 7674 depending on the architecture, may not be 7675 in sync between CPUs. 7676 global - Event time stamps are synchronized across 7677 CPUs. May be slower than the local clock, 7678 but better for some race conditions. 7679 counter - Simple counting of events (1, 2, ..) 7680 note, some counts may be skipped due to the 7681 infrastructure grabbing the clock more than 7682 once per event. 7683 uptime - Use jiffies as the time stamp. 7684 perf - Use the same clock that perf uses. 7685 mono - Use ktime_get_mono_fast_ns() for time stamps. 7686 mono_raw - Use ktime_get_raw_fast_ns() for time 7687 stamps. 7688 boot - Use ktime_get_boot_fast_ns() for time stamps. 7689 Architectures may add more clocks. See 7690 Documentation/trace/ftrace.rst for more details. 7691 7692 trace_event=[event-list] 7693 [FTRACE] Set and start specified trace events in order 7694 to facilitate early boot debugging. The event-list is a 7695 comma-separated list of trace events to enable. See 7696 also Documentation/trace/events.rst 7697 7698 To enable modules, use :mod: keyword: 7699 7700 trace_event=:mod:<module> 7701 7702 The value before :mod: will only enable specific events 7703 that are part of the module. See the above mentioned 7704 document for more information. 7705 7706 trace_instance=[instance-info] 7707 [FTRACE] Create a ring buffer instance early in boot up. 7708 This will be listed in: 7709 7710 /sys/kernel/tracing/instances 7711 7712 Events can be enabled at the time the instance is created 7713 via: 7714 7715 trace_instance=<name>,<system1>:<event1>,<system2>:<event2> 7716 7717 Note, the "<system*>:" portion is optional if the event is 7718 unique. 7719 7720 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall 7721 7722 will enable the "sched_switch" event (note, the "sched:" is optional, and 7723 the same thing would happen if it was left off). The irq_handler_entry 7724 event, and all events under the "initcall" system. 7725 7726 Flags can be added to the instance to modify its behavior when it is 7727 created. The flags are separated by '^'. 7728 7729 The available flags are: 7730 7731 traceoff - Have the tracing instance tracing disabled after it is created. 7732 traceprintk - Have trace_printk() write into this trace instance 7733 (note, "printk" and "trace_printk" can also be used) 7734 7735 trace_instance=foo^traceoff^traceprintk,sched,irq 7736 7737 The flags must come before the defined events. 7738 7739 If memory has been reserved (see memmap for x86), the instance 7740 can use that memory: 7741 7742 memmap=12M$0x284500000 trace_instance=boot_map@0x284500000:12M 7743 7744 The above will create a "boot_map" instance that uses the physical 7745 memory at 0x284500000 that is 12Megs. The per CPU buffers of that 7746 instance will be split up accordingly. 7747 7748 Alternatively, the memory can be reserved by the reserve_mem option: 7749 7750 reserve_mem=12M:4096:trace trace_instance=boot_map@trace 7751 7752 This will reserve 12 megabytes at boot up with a 4096 byte alignment 7753 and place the ring buffer in this memory. Note that due to KASLR, the 7754 memory may not be the same location each time, which will not preserve 7755 the buffer content. 7756 7757 Also note that the layout of the ring buffer data may change between 7758 kernel versions where the validator will fail and reset the ring buffer 7759 if the layout is not the same as the previous kernel. 7760 7761 If the ring buffer is used for persistent bootups and has events enabled, 7762 it is recommend to disable tracing so that events from a previous boot do not 7763 mix with events of the current boot (unless you are debugging a random crash 7764 at boot up). 7765 7766 reserve_mem=12M:4096:trace trace_instance=boot_map^traceoff^traceprintk@trace,sched,irq 7767 7768 Note, saving the trace buffer across reboots does require that the system 7769 is set up to not wipe memory. For instance, CONFIG_RESET_ATTACK_MITIGATION 7770 can force a memory reset on boot which will clear any trace that was stored. 7771 This is just one of many ways that can clear memory. Make sure your system 7772 keeps the content of memory across reboots before relying on this option. 7773 7774 NB: Both the mapped address and size must be page aligned for the architecture. 7775 7776 See also Documentation/trace/debugging.rst 7777 7778 7779 trace_options=[option-list] 7780 [FTRACE] Enable or disable tracer options at boot. 7781 The option-list is a comma delimited list of options 7782 that can be enabled or disabled just as if you were 7783 to echo the option name into 7784 7785 /sys/kernel/tracing/trace_options 7786 7787 For example, to enable stacktrace option (to dump the 7788 stack trace of each event), add to the command line: 7789 7790 trace_options=stacktrace 7791 7792 See also Documentation/trace/ftrace.rst "trace options" 7793 section. 7794 7795 trace_trigger=[trigger-list] 7796 [FTRACE] Add an event trigger on specific events. 7797 Set a trigger on top of a specific event, with an optional 7798 filter. 7799 7800 The format is "trace_trigger=<event>.<trigger>[ if <filter>],..." 7801 Where more than one trigger may be specified that are comma delimited. 7802 7803 For example: 7804 7805 trace_trigger="sched_switch.stacktrace if prev_state == 2" 7806 7807 The above will enable the "stacktrace" trigger on the "sched_switch" 7808 event but only trigger it if the "prev_state" of the "sched_switch" 7809 event is "2" (TASK_UNINTERRUPTIBLE). 7810 7811 See also "Event triggers" in Documentation/trace/events.rst 7812 7813 7814 traceoff_after_boot 7815 [FTRACE] Sometimes tracing is used to debug issues 7816 during the boot process. Since the trace buffer has a 7817 limited amount of storage, it may be prudent to 7818 disable tracing after the boot is finished, otherwise 7819 the critical information may be overwritten. With this 7820 option, the main tracing buffer will be turned off at 7821 the end of the boot process. 7822 7823 traceoff_on_warning 7824 [FTRACE] enable this option to disable tracing when a 7825 warning is hit. This turns off "tracing_on". Tracing can 7826 be enabled again by echoing '1' into the "tracing_on" 7827 file located in /sys/kernel/tracing/ 7828 7829 This option is useful, as it disables the trace before 7830 the WARNING dump is called, which prevents the trace to 7831 be filled with content caused by the warning output. 7832 7833 This option can also be set at run time via the sysctl 7834 option: kernel/traceoff_on_warning 7835 7836 transparent_hugepage= 7837 [KNL] 7838 Format: [always|madvise|never] 7839 Can be used to control the default behavior of the system 7840 with respect to transparent hugepages. 7841 See Documentation/admin-guide/mm/transhuge.rst 7842 for more details. 7843 7844 transparent_hugepage_shmem= [KNL] 7845 Format: [always|within_size|advise|never|deny|force] 7846 Can be used to control the hugepage allocation policy for 7847 the internal shmem mount. 7848 See Documentation/admin-guide/mm/transhuge.rst 7849 for more details. 7850 7851 transparent_hugepage_tmpfs= [KNL] 7852 Format: [always|within_size|advise|never] 7853 Can be used to control the default hugepage allocation policy 7854 for the tmpfs mount. 7855 See Documentation/admin-guide/mm/transhuge.rst 7856 for more details. 7857 7858 trusted.source= [KEYS] 7859 Format: <string> 7860 This parameter identifies the trust source as a backend 7861 for trusted keys implementation. Supported trust 7862 sources: 7863 - "tpm" 7864 - "tee" 7865 - "caam" 7866 - "dcp" 7867 - "pkwm" 7868 If not specified then it defaults to iterating through 7869 the trust source list starting with TPM and assigns the 7870 first trust source as a backend which is initialized 7871 successfully during iteration. 7872 7873 trusted.rng= [KEYS] 7874 Format: <string> 7875 The RNG used to generate key material for trusted keys. 7876 Can be one of: 7877 - "kernel" 7878 - the same value as trusted.source: "tpm" or "tee" 7879 - "default" 7880 If not specified, "default" is used. In this case, 7881 the RNG's choice is left to each individual trust source. 7882 7883 trusted.dcp_use_otp_key 7884 This is intended to be used in combination with 7885 trusted.source=dcp and will select the DCP OTP key 7886 instead of the DCP UNIQUE key blob encryption. 7887 7888 trusted.dcp_skip_zk_test 7889 This is intended to be used in combination with 7890 trusted.source=dcp and will disable the check if the 7891 blob key is all zeros. This is helpful for situations where 7892 having this key zero'ed is acceptable. E.g. in testing 7893 scenarios. 7894 7895 tsa= [X86] Control mitigation for Transient Scheduler 7896 Attacks on AMD CPUs. Search the following in your 7897 favourite search engine for more details: 7898 7899 "Technical guidance for mitigating transient scheduler 7900 attacks". 7901 7902 off - disable the mitigation 7903 on - enable the mitigation (default) 7904 user - mitigate only user/kernel transitions 7905 vm - mitigate only guest/host transitions 7906 7907 7908 tsc= Disable clocksource stability checks for TSC. 7909 Format: <string> 7910 [x86] reliable: mark tsc clocksource as reliable, this 7911 disables clocksource verification at runtime, as well 7912 as the stability checks done at bootup. Used to enable 7913 high-resolution timer mode on older hardware, and in 7914 virtualized environment. 7915 [x86] noirqtime: Do not use TSC to do irq accounting. 7916 Used to run time disable IRQ_TIME_ACCOUNTING on any 7917 platforms where RDTSC is slow and this accounting 7918 can add overhead. 7919 [x86] unstable: mark the TSC clocksource as unstable, this 7920 marks the TSC unconditionally unstable at bootup and 7921 avoids any further wobbles once the TSC watchdog notices. 7922 [x86] nowatchdog: disable clocksource watchdog. Used 7923 in situations with strict latency requirements (where 7924 interruptions from clocksource watchdog are not 7925 acceptable). 7926 [x86] recalibrate: force recalibration against a HW timer 7927 (HPET or PM timer) on systems whose TSC frequency was 7928 obtained from HW or FW using either an MSR or CPUID(0x15). 7929 Warn if the difference is more than 500 ppm. 7930 [x86] watchdog: Enforce the clocksource watchdog on TSC 7931 7932 tsc_early_khz= [X86,EARLY] Skip early TSC calibration and use the given 7933 value instead. Useful when the early TSC frequency discovery 7934 procedure is not reliable, such as on overclocked systems 7935 with CPUID.16h support and partial CPUID.15h support. 7936 Format: <unsigned int> 7937 7938 tsx= [X86] Control Transactional Synchronization 7939 Extensions (TSX) feature in Intel processors that 7940 support TSX control. 7941 7942 This parameter controls the TSX feature. The options are: 7943 7944 on - Enable TSX on the system. Although there are 7945 mitigations for all known security vulnerabilities, 7946 TSX has been known to be an accelerator for 7947 several previous speculation-related CVEs, and 7948 so there may be unknown security risks associated 7949 with leaving it enabled. 7950 7951 off - Disable TSX on the system. (Note that this 7952 option takes effect only on newer CPUs which are 7953 not vulnerable to MDS, i.e., have 7954 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 7955 the new IA32_TSX_CTRL MSR through a microcode 7956 update. This new MSR allows for the reliable 7957 deactivation of the TSX functionality.) 7958 7959 auto - Disable TSX if X86_BUG_TAA is present, 7960 otherwise enable TSX on the system. 7961 7962 Not specifying this option is equivalent to tsx=off. 7963 7964 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 7965 for more details. 7966 7967 tsx_async_abort= [X86,INTEL,EARLY] Control mitigation for the TSX Async 7968 Abort (TAA) vulnerability. 7969 7970 Similar to Micro-architectural Data Sampling (MDS) 7971 certain CPUs that support Transactional 7972 Synchronization Extensions (TSX) are vulnerable to an 7973 exploit against CPU internal buffers which can forward 7974 information to a disclosure gadget under certain 7975 conditions. 7976 7977 In vulnerable processors, the speculatively forwarded 7978 data can be used in a cache side channel attack, to 7979 access data to which the attacker does not have direct 7980 access. 7981 7982 This parameter controls the TAA mitigation. The 7983 options are: 7984 7985 full - Enable TAA mitigation on vulnerable CPUs 7986 if TSX is enabled. 7987 7988 full,nosmt - Enable TAA mitigation and disable SMT on 7989 vulnerable CPUs. If TSX is disabled, SMT 7990 is not disabled because CPU is not 7991 vulnerable to cross-thread TAA attacks. 7992 off - Unconditionally disable TAA mitigation 7993 7994 On MDS-affected machines, tsx_async_abort=off can be 7995 prevented by an active MDS mitigation as both vulnerabilities 7996 are mitigated with the same mechanism so in order to disable 7997 this mitigation, you need to specify mds=off too. 7998 7999 Not specifying this option is equivalent to 8000 tsx_async_abort=full. On CPUs which are MDS affected 8001 and deploy MDS mitigation, TAA mitigation is not 8002 required and doesn't provide any additional 8003 mitigation. 8004 8005 For details see: 8006 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 8007 8008 turbografx.map[2|3]= [HW,JOY] 8009 TurboGraFX parallel port interface 8010 Format: 8011 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 8012 See also Documentation/input/devices/joystick-parport.rst 8013 8014 udbg-immortal [PPC] When debugging early kernel crashes that 8015 happen after console_init() and before a proper 8016 console driver takes over, this boot options might 8017 help "seeing" what's going on. 8018 8019 uhash_entries= [KNL,NET] 8020 Set number of hash buckets for UDP/UDP-Lite connections 8021 8022 uhci-hcd.ignore_oc= 8023 [USB] Ignore overcurrent events (default N). 8024 Some badly-designed motherboards generate lots of 8025 bogus events, for ports that aren't wired to 8026 anything. Set this parameter to avoid log spamming. 8027 Note that genuine overcurrent events won't be 8028 reported either. 8029 8030 unaligned_scalar_speed= 8031 [RISCV] 8032 Format: {slow | fast | unsupported} 8033 Allow skipping scalar unaligned access speed tests. This 8034 is useful for testing alternative code paths and to skip 8035 the tests in environments where they run too slowly. All 8036 CPUs must have the same scalar unaligned access speed. 8037 8038 unaligned_vector_speed= 8039 [RISCV] 8040 Format: {slow | fast | unsupported} 8041 Allow skipping vector unaligned access speed tests. This 8042 is useful for testing alternative code paths and to skip 8043 the tests in environments where they run too slowly. All 8044 CPUs must have the same vector unaligned access speed. 8045 8046 unknown_nmi_panic 8047 [X86] Cause panic on unknown NMI. 8048 8049 unwind_debug [X86-64,EARLY] 8050 Enable unwinder debug output. This can be 8051 useful for debugging certain unwinder error 8052 conditions, including corrupt stacks and 8053 bad/missing unwinder metadata. 8054 8055 usbcore.authorized_default= 8056 [USB] Default USB device authorization: 8057 (default -1 = authorized (same as 1), 8058 0 = not authorized, 1 = authorized, 2 = authorized 8059 if device connected to internal port) 8060 8061 usbcore.autosuspend= 8062 [USB] The autosuspend time delay (in seconds) used 8063 for newly-detected USB devices (default 2). This 8064 is the time required before an idle device will be 8065 autosuspended. Devices for which the delay is set 8066 to a negative value won't be autosuspended at all. 8067 8068 usbcore.usbfs_snoop= 8069 [USB] Set to log all usbfs traffic (default 0 = off). 8070 8071 usbcore.usbfs_snoop_max= 8072 [USB] Maximum number of bytes to snoop in each URB 8073 (default = 65536). 8074 8075 usbcore.blinkenlights= 8076 [USB] Set to cycle leds on hubs (default 0 = off). 8077 8078 usbcore.old_scheme_first= 8079 [USB] Start with the old device initialization 8080 scheme (default 0 = off). 8081 8082 usbcore.usbfs_memory_mb= 8083 [USB] Memory limit (in MB) for buffers allocated by 8084 usbfs (default = 16, 0 = max = 2047). 8085 8086 usbcore.use_both_schemes= 8087 [USB] Try the other device initialization scheme 8088 if the first one fails (default 1 = enabled). 8089 8090 usbcore.initial_descriptor_timeout= 8091 [USB] Specifies timeout for the initial 64-byte 8092 USB_REQ_GET_DESCRIPTOR request in milliseconds 8093 (default 5000 = 5.0 seconds). 8094 8095 usbcore.nousb [USB] Disable the USB subsystem 8096 8097 usbcore.quirks= 8098 [USB] A list of quirk entries to augment the built-in 8099 usb core quirk list. List entries are separated by 8100 commas. Each entry has the form 8101 VendorID:ProductID:Flags. The IDs are 4-digit hex 8102 numbers and Flags is a set of letters. Each letter 8103 will change the built-in quirk; setting it if it is 8104 clear and clearing it if it is set. The letters have 8105 the following meanings: 8106 a = USB_QUIRK_STRING_FETCH_255 (string 8107 descriptors must not be fetched using 8108 a 255-byte read); 8109 b = USB_QUIRK_RESET_RESUME (device can't resume 8110 correctly so reset it instead); 8111 c = USB_QUIRK_NO_SET_INTF (device can't handle 8112 Set-Interface requests); 8113 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 8114 handle its Configuration or Interface 8115 strings); 8116 e = USB_QUIRK_RESET (device can't be reset 8117 (e.g morph devices), don't use reset); 8118 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 8119 more interface descriptions than the 8120 bNumInterfaces count, and can't handle 8121 talking to these interfaces); 8122 g = USB_QUIRK_DELAY_INIT (device needs a pause 8123 during initialization, after we read 8124 the device descriptor); 8125 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 8126 high speed and super speed interrupt 8127 endpoints, the USB 2.0 and USB 3.0 spec 8128 require the interval in microframes (1 8129 microframe = 125 microseconds) to be 8130 calculated as interval = 2 ^ 8131 (bInterval-1). 8132 Devices with this quirk report their 8133 bInterval as the result of this 8134 calculation instead of the exponent 8135 variable used in the calculation); 8136 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 8137 handle device_qualifier descriptor 8138 requests); 8139 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 8140 generates spurious wakeup, ignore 8141 remote wakeup capability); 8142 k = USB_QUIRK_NO_LPM (device can't handle Link 8143 Power Management); 8144 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 8145 (Device reports its bInterval as linear 8146 frames instead of the USB 2.0 8147 calculation); 8148 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 8149 to be disconnected before suspend to 8150 prevent spurious wakeup); 8151 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 8152 pause after every control message); 8153 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 8154 delay after resetting its port); 8155 p = USB_QUIRK_SHORT_SET_ADDRESS_REQ_TIMEOUT 8156 (Reduce timeout of the SET_ADDRESS 8157 request from 5000 ms to 500 ms); 8158 q = USB_QUIRK_FORCE_ONE_CONFIG (Device 8159 claims zero configurations, 8160 forcing to 1); 8161 Example: quirks=0781:5580:bk,0a5c:5834:gij 8162 8163 usbhid.mousepoll= 8164 [USBHID] The interval which mice are to be polled at. 8165 8166 usbhid.jspoll= 8167 [USBHID] The interval which joysticks are to be polled at. 8168 8169 usbhid.kbpoll= 8170 [USBHID] The interval which keyboards are to be polled at. 8171 8172 usb-storage.delay_use= 8173 [UMS] The delay in seconds before a new device is 8174 scanned for Logical Units (default 1). 8175 Optionally the delay in milliseconds if the value has 8176 suffix with "ms". 8177 Example: delay_use=2567ms 8178 8179 usb-storage.quirks= 8180 [UMS] A list of quirks entries to supplement or 8181 override the built-in unusual_devs list. List 8182 entries are separated by commas. Each entry has 8183 the form VID:PID:Flags where VID and PID are Vendor 8184 and Product ID values (4-digit hex numbers) and 8185 Flags is a set of characters, each corresponding 8186 to a common usb-storage quirk flag as follows: 8187 a = SANE_SENSE (collect more than 18 bytes 8188 of sense data, not on uas); 8189 b = BAD_SENSE (don't collect more than 18 8190 bytes of sense data, not on uas); 8191 c = FIX_CAPACITY (decrease the reported 8192 device capacity by one sector); 8193 d = NO_READ_DISC_INFO (don't use 8194 READ_DISC_INFO command, not on uas); 8195 e = NO_READ_CAPACITY_16 (don't use 8196 READ_CAPACITY_16 command); 8197 f = NO_REPORT_OPCODES (don't use report opcodes 8198 command, uas only); 8199 g = MAX_SECTORS_240 (don't transfer more than 8200 240 sectors at a time, uas only); 8201 h = CAPACITY_HEURISTICS (decrease the 8202 reported device capacity by one 8203 sector if the number is odd); 8204 i = IGNORE_DEVICE (don't bind to this 8205 device); 8206 j = NO_REPORT_LUNS (don't use report luns 8207 command, uas only); 8208 k = NO_SAME (do not use WRITE_SAME, uas only) 8209 l = NOT_LOCKABLE (don't try to lock and 8210 unlock ejectable media, not on uas); 8211 m = MAX_SECTORS_64 (don't transfer more 8212 than 64 sectors = 32 KB at a time, 8213 not on uas); 8214 n = INITIAL_READ10 (force a retry of the 8215 initial READ(10) command, not on uas); 8216 o = CAPACITY_OK (accept the capacity 8217 reported by the device, not on uas); 8218 p = WRITE_CACHE (the device cache is ON 8219 by default, not on uas); 8220 r = IGNORE_RESIDUE (the device reports 8221 bogus residue values, not on uas); 8222 s = SINGLE_LUN (the device has only one 8223 Logical Unit); 8224 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 8225 commands, uas only); 8226 u = IGNORE_UAS (don't bind to the uas driver); 8227 w = NO_WP_DETECT (don't test whether the 8228 medium is write-protected). 8229 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 8230 even if the device claims no cache, 8231 not on uas) 8232 Example: quirks=0419:aaf5:rl,0421:0433:rc 8233 8234 user_debug= [KNL,ARM] 8235 Format: <int> 8236 See arch/arm/Kconfig.debug help text. 8237 1 - undefined instruction events 8238 2 - system calls 8239 4 - invalid data aborts 8240 8 - SIGSEGV faults 8241 16 - SIGBUS faults 8242 Example: user_debug=31 8243 8244 vdso= [X86,SH,SPARC] 8245 On X86_32, this is an alias for vdso32=. Otherwise: 8246 8247 vdso=1: enable VDSO (the default) 8248 vdso=0: disable VDSO mapping 8249 8250 vdso32= [X86] Control the 32-bit vDSO 8251 vdso32=1: enable 32-bit VDSO 8252 vdso32=0 or vdso32=2: disable 32-bit VDSO 8253 8254 See the help text for CONFIG_COMPAT_VDSO for more 8255 details. If CONFIG_COMPAT_VDSO is set, the default is 8256 vdso32=0; otherwise, the default is vdso32=1. 8257 8258 For compatibility with older kernels, vdso32=2 is an 8259 alias for vdso32=0. 8260 8261 Try vdso32=0 if you encounter an error that says: 8262 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 8263 8264 video= [FB,EARLY] Frame buffer configuration 8265 See Documentation/fb/modedb.rst. 8266 8267 video.brightness_switch_enabled= [ACPI] 8268 Format: [0|1] 8269 If set to 1, on receiving an ACPI notify event 8270 generated by hotkey, video driver will adjust brightness 8271 level and then send out the event to user space through 8272 the allocated input device. If set to 0, video driver 8273 will only send out the event without touching backlight 8274 brightness level. 8275 default: 1 8276 8277 virtio_mmio.device= 8278 [VMMIO] Memory mapped virtio (platform) device. 8279 8280 <size>@<baseaddr>:<irq>[:<id>] 8281 where: 8282 <size> := size (can use standard suffixes 8283 like K, M and G) 8284 <baseaddr> := physical base address 8285 <irq> := interrupt number (as passed to 8286 request_irq()) 8287 <id> := (optional) platform device id 8288 example: 8289 virtio_mmio.device=1K@0x100b0000:48:7 8290 8291 Can be used multiple times for multiple devices. 8292 8293 vga= [BOOT,X86-32] Select a particular video mode 8294 See Documentation/arch/x86/boot.rst and 8295 Documentation/admin-guide/svga.rst. 8296 Use vga=ask for menu. 8297 This is actually a boot loader parameter; the value is 8298 passed to the kernel using a special protocol. 8299 8300 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 8301 May slow down system boot speed, especially when 8302 enabled on systems with a large amount of memory. 8303 All options are enabled by default, and this 8304 interface is meant to allow for selectively 8305 enabling or disabling specific virtual memory 8306 debugging features. 8307 8308 Available options are: 8309 P Enable page structure init time poisoning 8310 - Disable all of the above options 8311 8312 vmalloc=nn[KMG] [KNL,BOOT,EARLY] Forces the vmalloc area to have an 8313 exact size of <nn>. This can be used to increase 8314 the minimum size (128MB on x86, arm32 platforms). 8315 It can also be used to decrease the size and leave more room 8316 for directly mapped kernel RAM. Note that this parameter does 8317 not exist on many other platforms (including arm64, alpha, 8318 loongarch, arc, csky, hexagon, microblaze, mips, nios2, openrisc, 8319 parisc, m64k, powerpc, riscv, sh, um, xtensa, s390, sparc). 8320 8321 vmcp_cma=nn[MG] [KNL,S390,EARLY] 8322 Sets the memory size reserved for contiguous memory 8323 allocations for the vmcp device driver. 8324 8325 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 8326 Format: <command> 8327 8328 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 8329 Format: <command> 8330 8331 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 8332 Format: <command> 8333 8334 vmscape= [X86] Controls mitigation for VMscape attacks. 8335 VMscape attacks can leak information from a userspace 8336 hypervisor to a guest via speculative side-channels. 8337 8338 off - disable the mitigation 8339 ibpb - use Indirect Branch Prediction Barrier 8340 (IBPB) mitigation (default) 8341 force - force vulnerability detection even on 8342 unaffected processors 8343 8344 vsyscall= [X86-64,EARLY] 8345 Controls the behavior of vsyscalls (i.e. calls to 8346 fixed addresses of 0xffffffffff600x00 from legacy 8347 code). Most statically-linked binaries and older 8348 versions of glibc use these calls. Because these 8349 functions are at fixed addresses, they make nice 8350 targets for exploits that can control RIP. 8351 8352 emulate Vsyscalls turn into traps and are emulated 8353 reasonably safely. The vsyscall page is 8354 readable. This disables the Linear 8355 Address Space Separation (LASS) security 8356 feature and makes the system less secure. 8357 8358 xonly [default] Vsyscalls turn into traps and are 8359 emulated reasonably safely. The vsyscall 8360 page is not readable. 8361 8362 none Vsyscalls don't work at all. This makes 8363 them quite hard to use for exploits but 8364 might break your system. 8365 8366 vt.color= [VT] Default text color. 8367 Format: 0xYX, X = foreground, Y = background. 8368 Default: 0x07 = light gray on black. 8369 8370 vt.cur_default= [VT] Default cursor shape. 8371 Format: 0xCCBBAA, where AA, BB, and CC are the same as 8372 the parameters of the <Esc>[?A;B;Cc escape sequence; 8373 see vga-softcursor.rst. Default: 2 = underline. 8374 8375 vt.default_blu= [VT] 8376 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 8377 Change the default blue palette of the console. 8378 This is a 16-member array composed of values 8379 ranging from 0-255. 8380 8381 vt.default_grn= [VT] 8382 Format: <green0>,<green1>,<green2>,...,<green15> 8383 Change the default green palette of the console. 8384 This is a 16-member array composed of values 8385 ranging from 0-255. 8386 8387 vt.default_red= [VT] 8388 Format: <red0>,<red1>,<red2>,...,<red15> 8389 Change the default red palette of the console. 8390 This is a 16-member array composed of values 8391 ranging from 0-255. 8392 8393 vt.default_utf8= 8394 [VT] 8395 Format=<0|1> 8396 Set system-wide default UTF-8 mode for all tty's. 8397 Default is 1, i.e. UTF-8 mode is enabled for all 8398 newly opened terminals. 8399 8400 vt.global_cursor_default= 8401 [VT] 8402 Format=<-1|0|1> 8403 Set system-wide default for whether a cursor 8404 is shown on new VTs. Default is -1, 8405 i.e. cursors will be created by default unless 8406 overridden by individual drivers. 0 will hide 8407 cursors, 1 will display them. 8408 8409 vt.italic= [VT] Default color for italic text; 0-15. 8410 Default: 2 = green. 8411 8412 vt.underline= [VT] Default color for underlined text; 0-15. 8413 Default: 3 = cyan. 8414 8415 watchdog timers [HW,WDT] For information on watchdog timers, 8416 see Documentation/watchdog/watchdog-parameters.rst 8417 or other driver-specific files in the 8418 Documentation/watchdog/ directory. 8419 8420 watchdog_thresh= 8421 [KNL] 8422 Set the hard lockup detector stall duration 8423 threshold in seconds. The soft lockup detector 8424 threshold is set to twice the value. A value of 0 8425 disables both lockup detectors. Default is 10 8426 seconds. 8427 8428 workqueue.unbound_cpus= 8429 [KNL,SMP] Specify to constrain one or some CPUs 8430 to use in unbound workqueues. 8431 Format: <cpu-list> 8432 By default, all online CPUs are available for 8433 unbound workqueues. 8434 8435 workqueue.watchdog_thresh= 8436 If CONFIG_WQ_WATCHDOG is configured, workqueue can 8437 warn stall conditions and dump internal state to 8438 help debugging. 0 disables workqueue stall 8439 detection; otherwise, it's the stall threshold 8440 duration in seconds. The default value is 30 and 8441 it can be updated at runtime by writing to the 8442 corresponding sysfs file. 8443 8444 workqueue.panic_on_stall=<uint> 8445 Panic when workqueue stall is detected by 8446 CONFIG_WQ_WATCHDOG. It sets the number times of the 8447 stall to trigger panic. 8448 8449 The default is set by CONFIG_BOOTPARAM_WQ_STALL_PANIC, 8450 which is 0 (disabled) if not configured. 8451 8452 workqueue.panic_on_stall_time=<uint> 8453 Panic when a workqueue stall has been continuous for 8454 the specified number of seconds. Unlike panic_on_stall 8455 which counts accumulated stall events, this triggers 8456 based on the duration of a single continuous stall. 8457 8458 The default is 0, which disables the time-based panic. 8459 8460 workqueue.cpu_intensive_thresh_us= 8461 Per-cpu work items which run for longer than this 8462 threshold are automatically considered CPU intensive 8463 and excluded from concurrency management to prevent 8464 them from noticeably delaying other per-cpu work 8465 items. Default is 10000 (10ms). 8466 8467 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8468 will report the work functions which violate this 8469 threshold repeatedly. They are likely good 8470 candidates for using WQ_UNBOUND workqueues instead. 8471 8472 workqueue.cpu_intensive_warning_thresh=<uint> 8473 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8474 will report the work functions which violate the 8475 intensive_threshold_us repeatedly. In order to prevent 8476 spurious warnings, start printing only after a work 8477 function has violated this threshold number of times. 8478 8479 The default is 4 times. 0 disables the warning. 8480 8481 workqueue.power_efficient 8482 Per-cpu workqueues are generally preferred because 8483 they show better performance thanks to cache 8484 locality; unfortunately, per-cpu workqueues tend to 8485 be more power hungry than unbound workqueues. 8486 8487 Enabling this makes the per-cpu workqueues which 8488 were observed to contribute significantly to power 8489 consumption unbound, leading to measurably lower 8490 power usage at the cost of small performance 8491 overhead. 8492 8493 The default value of this parameter is determined by 8494 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 8495 8496 workqueue.default_affinity_scope= 8497 Select the default affinity scope to use for unbound 8498 workqueues. Can be one of "cpu", "smt", "cache", 8499 "cache_shard", "numa" and "system". Default is 8500 "cache_shard". For more 8501 information, see the Affinity Scopes section in 8502 Documentation/core-api/workqueue.rst. 8503 8504 This can be changed after boot by writing to the 8505 matching /sys/module/workqueue/parameters file. All 8506 workqueues with the "default" affinity scope will be 8507 updated accordingly. 8508 8509 workqueue.debug_force_rr_cpu 8510 Workqueue used to implicitly guarantee that work 8511 items queued without explicit CPU specified are put 8512 on the local CPU. This guarantee is no longer true 8513 and while local CPU is still preferred work items 8514 may be put on foreign CPUs. This debug option 8515 forces round-robin CPU selection to flush out 8516 usages which depend on the now broken guarantee. 8517 When enabled, memory and cache locality will be 8518 impacted. 8519 8520 writecombine= [LOONGARCH,EARLY] Control the MAT (Memory Access 8521 Type) of ioremap_wc(). 8522 8523 on - Enable writecombine, use WUC for ioremap_wc() 8524 off - Disable writecombine, use SUC for ioremap_wc() 8525 8526 x2apic_phys [X86-64,APIC,EARLY] Use x2apic physical mode instead of 8527 default x2apic cluster mode on platforms 8528 supporting x2apic. 8529 8530 xen_512gb_limit [KNL,X86-64,XEN] 8531 Restricts the kernel running paravirtualized under Xen 8532 to use only up to 512 GB of RAM. The reason to do so is 8533 crash analysis tools and Xen tools for doing domain 8534 save/restore/migration must be enabled to handle larger 8535 domains. 8536 8537 xen_console_io [XEN,EARLY] 8538 Boolean option to enable/disable the usage of the Xen 8539 console_io hypercalls to read and write to the console. 8540 Mostly useful for debugging and development. 8541 8542 xen_emul_unplug= [HW,X86,XEN,EARLY] 8543 Unplug Xen emulated devices 8544 Format: [unplug0,][unplug1] 8545 ide-disks -- unplug primary master IDE devices 8546 aux-ide-disks -- unplug non-primary-master IDE devices 8547 nics -- unplug network devices 8548 all -- unplug all emulated devices (NICs and IDE disks) 8549 unnecessary -- unplugging emulated devices is 8550 unnecessary even if the host did not respond to 8551 the unplug protocol 8552 never -- do not unplug even if version check succeeds 8553 8554 xen_legacy_crash [X86,XEN,EARLY] 8555 Crash from Xen panic notifier, without executing late 8556 panic() code such as dumping handler. 8557 8558 xen_mc_debug [X86,XEN,EARLY] 8559 Enable multicall debugging when running as a Xen PV guest. 8560 Enabling this feature will reduce performance a little 8561 bit, so it should only be enabled for obtaining extended 8562 debug data in case of multicall errors. 8563 8564 xen_msr_safe= [X86,XEN,EARLY] 8565 Format: <bool> 8566 Select whether to always use non-faulting (safe) MSR 8567 access functions when running as Xen PV guest. The 8568 default value is controlled by CONFIG_XEN_PV_MSR_SAFE. 8569 8570 xen_nopv [X86] 8571 Disables the PV optimizations forcing the HVM guest to 8572 run as generic HVM guest with no PV drivers. 8573 This option is obsoleted by the "nopv" option, which 8574 has equivalent effect for XEN platform. 8575 8576 xen_no_vector_callback 8577 [KNL,X86,XEN,EARLY] Disable the vector callback for Xen 8578 event channel interrupts. 8579 8580 xen_scrub_pages= [XEN] 8581 Boolean option to control scrubbing pages before giving them back 8582 to Xen, for use by other domains. Can be also changed at runtime 8583 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 8584 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 8585 8586 xen_timer_slop= [X86-64,XEN,EARLY] 8587 Set the timer slop (in nanoseconds) for the virtual Xen 8588 timers (default is 100000). This adjusts the minimum 8589 delta of virtualized Xen timers, where lower values 8590 improve timer resolution at the expense of processing 8591 more timer interrupts. 8592 8593 xen.balloon_boot_timeout= [XEN] 8594 The time (in seconds) to wait before giving up to boot 8595 in case initial ballooning fails to free enough memory. 8596 Applies only when running as HVM or PVH guest and 8597 started with less memory configured than allowed at 8598 max. Default is 180. 8599 8600 xen.event_eoi_delay= [XEN] 8601 How long to delay EOI handling in case of event 8602 storms (jiffies). Default is 10. 8603 8604 xen.event_loop_timeout= [XEN] 8605 After which time (jiffies) the event handling loop 8606 should start to delay EOI handling. Default is 2. 8607 8608 xen.fifo_events= [XEN] 8609 Boolean parameter to disable using fifo event handling 8610 even if available. Normally fifo event handling is 8611 preferred over the 2-level event handling, as it is 8612 fairer and the number of possible event channels is 8613 much higher. Default is on (use fifo events). 8614 8615 xirc2ps_cs= [NET,PCMCIA] 8616 Format: 8617 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 8618 8619 xive= [PPC] 8620 By default on POWER9 and above, the kernel will 8621 natively use the XIVE interrupt controller. This option 8622 allows the fallback firmware mode to be used: 8623 8624 off Fallback to firmware control of XIVE interrupt 8625 controller on both pseries and powernv 8626 platforms. Only useful on POWER9 and above. 8627 8628 xive.store-eoi=off [PPC] 8629 By default on POWER10 and above, the kernel will use 8630 stores for EOI handling when the XIVE interrupt mode 8631 is active. This option allows the XIVE driver to use 8632 loads instead, as on POWER9. 8633 8634 xhci-hcd.quirks [USB,KNL] 8635 A hex value specifying bitmask with supplemental xhci 8636 host controller quirks. Meaning of each bit can be 8637 consulted in header drivers/usb/host/xhci.h. 8638 8639 xmon [PPC,EARLY] 8640 Format: { early | on | rw | ro | off } 8641 Controls if xmon debugger is enabled. Default is off. 8642 Passing only "xmon" is equivalent to "xmon=early". 8643 early Call xmon as early as possible on boot; xmon 8644 debugger is called from setup_arch(). 8645 on xmon debugger hooks will be installed so xmon 8646 is only called on a kernel crash. Default mode, 8647 i.e. either "ro" or "rw" mode, is controlled 8648 with CONFIG_XMON_DEFAULT_RO_MODE. 8649 rw xmon debugger hooks will be installed so xmon 8650 is called only on a kernel crash, mode is write, 8651 meaning SPR registers, memory and, other data 8652 can be written using xmon commands. 8653 ro same as "rw" option above but SPR registers, 8654 memory, and other data can't be written using 8655 xmon commands. 8656 off xmon is disabled. 8657