1 ACPI ACPI support is enabled. 2 AGP AGP (Accelerated Graphics Port) is enabled. 3 ALSA ALSA sound support is enabled. 4 APIC APIC support is enabled. 5 APM Advanced Power Management support is enabled. 6 APPARMOR AppArmor support is enabled. 7 ARM ARM architecture is enabled. 8 ARM64 ARM64 architecture is enabled. 9 AX25 Appropriate AX.25 support is enabled. 10 CLK Common clock infrastructure is enabled. 11 CMA Contiguous Memory Area support is enabled. 12 DRM Direct Rendering Management support is enabled. 13 DYNAMIC_DEBUG Build in debug messages and enable them at runtime 14 EARLY Parameter processed too early to be embedded in initrd. 15 EDD BIOS Enhanced Disk Drive Services (EDD) is enabled 16 EFI EFI Partitioning (GPT) is enabled 17 EVM Extended Verification Module 18 FB The frame buffer device is enabled. 19 FTRACE Function tracing enabled. 20 GCOV GCOV profiling is enabled. 21 HIBERNATION HIBERNATION is enabled. 22 HW Appropriate hardware is enabled. 23 HYPER_V HYPERV support is enabled. 24 IMA Integrity measurement architecture is enabled. 25 IP_PNP IP DHCP, BOOTP, or RARP is enabled. 26 IPV6 IPv6 support is enabled. 27 ISAPNP ISA PnP code is enabled. 28 ISDN Appropriate ISDN support is enabled. 29 ISOL CPU Isolation is enabled. 30 JOY Appropriate joystick support is enabled. 31 KGDB Kernel debugger support is enabled. 32 KVM Kernel Virtual Machine support is enabled. 33 LIBATA Libata driver is enabled 34 LOONGARCH LoongArch architecture is enabled. 35 LOOP Loopback device support is enabled. 36 LP Printer support is enabled. 37 M68k M68k architecture is enabled. 38 These options have more detailed description inside of 39 Documentation/arch/m68k/kernel-options.rst. 40 MDA MDA console support is enabled. 41 MIPS MIPS architecture is enabled. 42 MOUSE Appropriate mouse support is enabled. 43 MSI Message Signaled Interrupts (PCI). 44 MTD MTD (Memory Technology Device) support is enabled. 45 NET Appropriate network support is enabled. 46 NFS Appropriate NFS support is enabled. 47 NUMA NUMA support is enabled. 48 OF Devicetree is enabled. 49 PARISC The PA-RISC architecture is enabled. 50 PCI PCI bus support is enabled. 51 PCIE PCI Express support is enabled. 52 PCMCIA The PCMCIA subsystem is enabled. 53 PNP Plug & Play support is enabled. 54 PPC PowerPC architecture is enabled. 55 PPT Parallel port support is enabled. 56 PS2 Appropriate PS/2 support is enabled. 57 PV_OPS A paravirtualized kernel is enabled. 58 RAM RAM disk support is enabled. 59 RDT Intel Resource Director Technology. 60 RISCV RISCV architecture is enabled. 61 S390 S390 architecture is enabled. 62 SCSI Appropriate SCSI support is enabled. 63 A lot of drivers have their options described inside 64 the Documentation/scsi/ sub-directory. 65 SDW SoundWire support is enabled. 66 SECURITY Different security models are enabled. 67 SELINUX SELinux support is enabled. 68 SERIAL Serial support is enabled. 69 SH SuperH architecture is enabled. 70 SMP The kernel is an SMP kernel. 71 SPARC Sparc architecture is enabled. 72 SUSPEND System suspend states are enabled. 73 SWSUSP Software suspend (hibernation) is enabled. 74 TPM TPM drivers are enabled. 75 UMS USB Mass Storage support is enabled. 76 USB USB support is enabled. 77 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_enforce_resources= [ACPI] 193 { strict | lax | no } 194 Check for resource conflicts between native drivers 195 and ACPI OperationRegions (SystemIO and SystemMemory 196 only). IO ports and memory declared in ACPI might be 197 used by the ACPI subsystem in arbitrary AML code and 198 can interfere with legacy drivers. 199 strict (default): access to resources claimed by ACPI 200 is denied; legacy drivers trying to access reserved 201 resources will fail to bind to device using them. 202 lax: access to resources claimed by ACPI is allowed; 203 legacy drivers trying to access reserved resources 204 will bind successfully but a warning message is logged. 205 no: ACPI OperationRegions are not marked as reserved, 206 no further checks are performed. 207 208 acpi_force_table_verification [HW,ACPI,EARLY] 209 Enable table checksum verification during early stage. 210 By default, this is disabled due to x86 early mapping 211 size limitation. 212 213 acpi_irq_balance [HW,ACPI] 214 ACPI will balance active IRQs 215 default in APIC mode 216 217 acpi_irq_nobalance [HW,ACPI] 218 ACPI will not move active IRQs (default) 219 default in PIC mode 220 221 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 222 Format: <irq>,<irq>... 223 224 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 225 use by PCI 226 Format: <irq>,<irq>... 227 228 acpi_mask_gpe= [HW,ACPI] 229 Due to the existence of _Lxx/_Exx, some GPEs triggered 230 by unsupported hardware/firmware features can result in 231 GPE floodings that cannot be automatically disabled by 232 the GPE dispatcher. 233 This facility can be used to prevent such uncontrolled 234 GPE floodings. 235 Format: <byte> or <bitmap-list> 236 237 acpi_no_auto_serialize [HW,ACPI] 238 Disable auto-serialization of AML methods 239 AML control methods that contain the opcodes to create 240 named objects will be marked as "Serialized" by the 241 auto-serialization feature. 242 This feature is enabled by default. 243 This option allows to turn off the feature. 244 245 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 246 kernels. 247 248 acpi_no_static_ssdt [HW,ACPI,EARLY] 249 Disable installation of static SSDTs at early boot time 250 By default, SSDTs contained in the RSDT/XSDT will be 251 installed automatically and they will appear under 252 /sys/firmware/acpi/tables. 253 This option turns off this feature. 254 Note that specifying this option does not affect 255 dynamic table installation which will install SSDT 256 tables to /sys/firmware/acpi/tables/dynamic. 257 258 acpi_no_watchdog [HW,ACPI,WDT] 259 Ignore the ACPI-based watchdog interface (WDAT) and let 260 a native driver control the watchdog device instead. 261 262 acpi_rsdp= [ACPI,EFI,KEXEC,EARLY] 263 Pass the RSDP address to the kernel, mostly used 264 on machines running EFI runtime service to boot the 265 second kernel for kdump. 266 267 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 268 Format: To spoof as Windows 98: ="Microsoft Windows" 269 270 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 271 of 2 which is mandated by ACPI 6) as the supported ACPI 272 specification revision (when using this switch, it may 273 be necessary to carry out a cold reboot _twice_ in a 274 row to make it take effect on the platform firmware). 275 276 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 277 acpi_osi="string1" # add string1 278 acpi_osi="!string2" # remove string2 279 acpi_osi=!* # remove all strings 280 acpi_osi=! # disable all built-in OS vendor 281 strings 282 acpi_osi=!! # enable all built-in OS vendor 283 strings 284 acpi_osi= # disable all strings 285 286 'acpi_osi=!' can be used in combination with single or 287 multiple 'acpi_osi="string1"' to support specific OS 288 vendor string(s). Note that such command can only 289 affect the default state of the OS vendor strings, thus 290 it cannot affect the default state of the feature group 291 strings and the current state of the OS vendor strings, 292 specifying it multiple times through kernel command line 293 is meaningless. This command is useful when one do not 294 care about the state of the feature group strings which 295 should be controlled by the OSPM. 296 Examples: 297 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 298 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 299 can make '_OSI("Windows 2000")' TRUE. 300 301 'acpi_osi=' cannot be used in combination with other 302 'acpi_osi=' command lines, the _OSI method will not 303 exist in the ACPI namespace. NOTE that such command can 304 only affect the _OSI support state, thus specifying it 305 multiple times through kernel command line is also 306 meaningless. 307 Examples: 308 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 309 FALSE. 310 311 'acpi_osi=!*' can be used in combination with single or 312 multiple 'acpi_osi="string1"' to support specific 313 string(s). Note that such command can affect the 314 current state of both the OS vendor strings and the 315 feature group strings, thus specifying it multiple times 316 through kernel command line is meaningful. But it may 317 still not able to affect the final state of a string if 318 there are quirks related to this string. This command 319 is useful when one want to control the state of the 320 feature group strings to debug BIOS issues related to 321 the OSPM features. 322 Examples: 323 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 324 '_OSI("Module Device")' FALSE. 325 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 326 '_OSI("Module Device")' TRUE. 327 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 328 equivalent to 329 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 330 and 331 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 332 they all will make '_OSI("Windows 2000")' TRUE. 333 334 acpi_pm_good [X86] 335 Override the pmtimer bug detection: force the kernel 336 to assume that this machine's pmtimer latches its value 337 and always returns good values. 338 339 acpi_sci= [HW,ACPI,EARLY] ACPI System Control Interrupt trigger mode 340 Format: { level | edge | high | low } 341 342 acpi_skip_timer_override [HW,ACPI,EARLY] 343 Recognize and ignore IRQ0/pin2 Interrupt Override. 344 For broken nForce2 BIOS resulting in XT-PIC timer. 345 346 acpi_sleep= [HW,ACPI] Sleep options 347 Format: { s3_bios, s3_mode, s3_beep, s4_hwsig, 348 s4_nohwsig, old_ordering, nonvs, 349 sci_force_enable, nobl } 350 See Documentation/power/video.rst for information on 351 s3_bios and s3_mode. 352 s3_beep is for debugging; it makes the PC's speaker beep 353 as soon as the kernel's real-mode entry point is called. 354 s4_hwsig causes the kernel to check the ACPI hardware 355 signature during resume from hibernation, and gracefully 356 refuse to resume if it has changed. This complies with 357 the ACPI specification but not with reality, since 358 Windows does not do this and many laptops do change it 359 on docking. So the default behaviour is to allow resume 360 and simply warn when the signature changes, unless the 361 s4_hwsig option is enabled. 362 s4_nohwsig prevents ACPI hardware signature from being 363 used (or even warned about) during resume. 364 old_ordering causes the ACPI 1.0 ordering of the _PTS 365 control method, with respect to putting devices into 366 low power states, to be enforced (the ACPI 2.0 ordering 367 of _PTS is used by default). 368 nonvs prevents the kernel from saving/restoring the 369 ACPI NVS memory during suspend/hibernation and resume. 370 sci_force_enable causes the kernel to set SCI_EN directly 371 on resume from S1/S3 (which is against the ACPI spec, 372 but some broken systems don't work without it). 373 nobl causes the internal blacklist of systems known to 374 behave incorrectly in some ways with respect to system 375 suspend and resume to be ignored (use wisely). 376 377 acpi_use_timer_override [HW,ACPI,EARLY] 378 Use timer override. For some broken Nvidia NF5 boards 379 that require a timer override, but don't have HPET 380 381 add_efi_memmap [EFI,X86,EARLY] Include EFI memory map in 382 kernel's map of available physical RAM. 383 384 agp= [AGP] 385 { off | try_unsupported } 386 off: disable AGP support 387 try_unsupported: try to drive unsupported chipsets 388 (may crash computer or cause data corruption) 389 390 ALSA [HW,ALSA] 391 See Documentation/sound/alsa-configuration.rst 392 393 alignment= [KNL,ARM] 394 Allow the default userspace alignment fault handler 395 behaviour to be specified. Bit 0 enables warnings, 396 bit 1 enables fixups, and bit 2 sends a segfault. 397 398 align_va_addr= [X86-64] 399 Align virtual addresses by clearing slice [14:12] when 400 allocating a VMA at process creation time. This option 401 gives you up to 3% performance improvement on AMD F15h 402 machines (where it is enabled by default) for a 403 CPU-intensive style benchmark, and it can vary highly in 404 a microbenchmark depending on workload and compiler. 405 406 32: only for 32-bit processes 407 64: only for 64-bit processes 408 on: enable for both 32- and 64-bit processes 409 off: disable for both 32- and 64-bit processes 410 411 alloc_snapshot [FTRACE] 412 Allocate the ftrace snapshot buffer on boot up when the 413 main buffer is allocated. This is handy if debugging 414 and you need to use tracing_snapshot() on boot up, and 415 do not want to use tracing_snapshot_alloc() as it needs 416 to be done where GFP_KERNEL allocations are allowed. 417 418 allow_mismatched_32bit_el0 [ARM64,EARLY] 419 Allow execve() of 32-bit applications and setting of the 420 PER_LINUX32 personality on systems where only a strict 421 subset of the CPUs support 32-bit EL0. When this 422 parameter is present, the set of CPUs supporting 32-bit 423 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0 424 and hot-unplug operations may be restricted. 425 426 See Documentation/arch/arm64/asymmetric-32bit.rst for more 427 information. 428 429 amd_iommu= [HW,X86-64] 430 Pass parameters to the AMD IOMMU driver in the system. 431 Possible values are: 432 fullflush - Deprecated, equivalent to iommu.strict=1 433 off - do not initialize any AMD IOMMU found in 434 the system 435 force_isolation - Force device isolation for all 436 devices. The IOMMU driver is not 437 allowed anymore to lift isolation 438 requirements as needed. This option 439 does not override iommu=pt 440 force_enable - Force enable the IOMMU on platforms known 441 to be buggy with IOMMU enabled. Use this 442 option with care. 443 pgtbl_v1 - Use v1 page table for DMA-API (Default). 444 pgtbl_v2 - Use v2 page table for DMA-API. 445 irtcachedis - Disable Interrupt Remapping Table (IRT) caching. 446 nohugepages - Limit page-sizes used for v1 page-tables 447 to 4 KiB. 448 v2_pgsizes_only - Limit page-sizes used for v1 page-tables 449 to 4KiB/2Mib/1GiB. 450 451 452 amd_iommu_dump= [HW,X86-64] 453 Enable AMD IOMMU driver option to dump the ACPI table 454 for AMD IOMMU. With this option enabled, AMD IOMMU 455 driver will print ACPI tables for AMD IOMMU during 456 IOMMU initialization. 457 458 amd_iommu_intr= [HW,X86-64] 459 Specifies one of the following AMD IOMMU interrupt 460 remapping modes: 461 legacy - Use legacy interrupt remapping mode. 462 vapic - Use virtual APIC mode, which allows IOMMU 463 to inject interrupts directly into guest. 464 This mode requires kvm-amd.avic=1. 465 (Default when IOMMU HW support is present.) 466 467 amd_pstate= [X86,EARLY] 468 disable 469 Do not enable amd_pstate as the default 470 scaling driver for the supported processors 471 passive 472 Use amd_pstate with passive mode as a scaling driver. 473 In this mode autonomous selection is disabled. 474 Driver requests a desired performance level and platform 475 tries to match the same performance level if it is 476 satisfied by guaranteed performance level. 477 active 478 Use amd_pstate_epp driver instance as the scaling driver, 479 driver provides a hint to the hardware if software wants 480 to bias toward performance (0x0) or energy efficiency (0xff) 481 to the CPPC firmware. then CPPC power algorithm will 482 calculate the runtime workload and adjust the realtime cores 483 frequency. 484 guided 485 Activate guided autonomous mode. Driver requests minimum and 486 maximum performance level and the platform autonomously 487 selects a performance level in this range and appropriate 488 to the current workload. 489 490 amd_prefcore= 491 [X86] 492 disable 493 Disable amd-pstate preferred core. 494 495 amijoy.map= [HW,JOY] Amiga joystick support 496 Map of devices attached to JOY0DAT and JOY1DAT 497 Format: <a>,<b> 498 See also Documentation/input/joydev/joystick.rst 499 500 analog.map= [HW,JOY] Analog joystick and gamepad support 501 Specifies type or capabilities of an analog joystick 502 connected to one of 16 gameports 503 Format: <type1>,<type2>,..<type16> 504 505 apc= [HW,SPARC] 506 Power management functions (SPARCstation-4/5 + deriv.) 507 Format: noidle 508 Disable APC CPU standby support. SPARCstation-Fox does 509 not play well with APC CPU idle - disable it if you have 510 APC and your system crashes randomly. 511 512 apic [APIC,X86-64] Use IO-APIC. Default. 513 514 apic= [APIC,X86,EARLY] Advanced Programmable Interrupt Controller 515 Change the output verbosity while booting 516 Format: { quiet (default) | verbose | debug } 517 Change the amount of debugging information output 518 when initialising the APIC and IO-APIC components. 519 520 apic_extnmi= [APIC,X86,EARLY] External NMI delivery setting 521 Format: { bsp (default) | all | none } 522 bsp: External NMI is delivered only to CPU 0 523 all: External NMIs are broadcast to all CPUs as a 524 backup of CPU 0 525 none: External NMI is masked for all CPUs. This is 526 useful so that a dump capture kernel won't be 527 shot down by NMI 528 529 apicpmtimer Do APIC timer calibration using the pmtimer. Implies 530 apicmaintimer. Useful when your PIT timer is totally 531 broken. 532 533 autoconf= [IPV6] 534 See Documentation/networking/ipv6.rst. 535 536 apm= [APM] Advanced Power Management 537 See header of arch/x86/kernel/apm_32.c. 538 539 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 540 Format: { "0" | "1" } 541 See security/apparmor/Kconfig help text 542 0 -- disable. 543 1 -- enable. 544 Default value is set via kernel config option. 545 546 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards 547 Format: <io>,<irq>,<nodeID> 548 549 arm64.no32bit_el0 [ARM64] Unconditionally disable the execution of 550 32 bit applications. 551 552 arm64.nobti [ARM64] Unconditionally disable Branch Target 553 Identification support 554 555 arm64.nogcs [ARM64] Unconditionally disable Guarded Control Stack 556 support 557 558 arm64.nomops [ARM64] Unconditionally disable Memory Copy and Memory 559 Set instructions support 560 561 arm64.nompam [ARM64] Unconditionally disable Memory Partitioning And 562 Monitoring support 563 564 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension 565 support 566 567 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 568 support 569 570 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix 571 Extension support 572 573 arm64.nosve [ARM64] Unconditionally disable Scalable Vector 574 Extension support 575 576 ataflop= [HW,M68k] 577 578 atarimouse= [HW,MOUSE] Atari Mouse 579 580 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 581 EzKey and similar keyboards 582 583 atkbd.reset= [HW] Reset keyboard during initialization 584 585 atkbd.set= [HW] Select keyboard code set 586 Format: <int> (2 = AT (default), 3 = PS/2) 587 588 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 589 keyboards 590 591 atkbd.softraw= [HW] Choose between synthetic and real raw mode 592 Format: <bool> (0 = real, 1 = synthetic (default)) 593 594 atkbd.softrepeat= [HW] 595 Use software keyboard repeat 596 597 audit= [KNL] Enable the audit sub-system 598 Format: { "0" | "1" | "off" | "on" } 599 0 | off - kernel audit is disabled and can not be 600 enabled until the next reboot 601 unset - kernel audit is initialized but disabled and 602 will be fully enabled by the userspace auditd. 603 1 | on - kernel audit is initialized and partially 604 enabled, storing at most audit_backlog_limit 605 messages in RAM until it is fully enabled by the 606 userspace auditd. 607 Default: unset 608 609 audit_backlog_limit= [KNL] Set the audit queue size limit. 610 Format: <int> (must be >=0) 611 Default: 64 612 613 bau= [X86_UV] Enable the BAU on SGI UV. The default 614 behavior is to disable the BAU (i.e. bau=0). 615 Format: { "0" | "1" } 616 0 - Disable the BAU. 617 1 - Enable the BAU. 618 unset - Disable the BAU. 619 620 baycom_epp= [HW,AX25] 621 Format: <io>,<mode> 622 623 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem 624 Format: <io>,<mode> 625 See header of drivers/net/hamradio/baycom_par.c. 626 627 baycom_ser_fdx= [HW,AX25] 628 BayCom Serial Port AX.25 Modem (Full Duplex Mode) 629 Format: <io>,<irq>,<mode>[,<baud>] 630 See header of drivers/net/hamradio/baycom_ser_fdx.c. 631 632 baycom_ser_hdx= [HW,AX25] 633 BayCom Serial Port AX.25 Modem (Half Duplex Mode) 634 Format: <io>,<irq>,<mode> 635 See header of drivers/net/hamradio/baycom_ser_hdx.c. 636 637 bdev_allow_write_mounted= 638 Format: <bool> 639 Control the ability to open a mounted block device 640 for writing, i.e., allow / disallow writes that bypass 641 the FS. This was implemented as a means to prevent 642 fuzzers from crashing the kernel by overwriting the 643 metadata underneath a mounted FS without its awareness. 644 This also prevents destructive formatting of mounted 645 filesystems by naive storage tooling that don't use 646 O_EXCL. Default is Y and can be changed through the 647 Kconfig option CONFIG_BLK_DEV_WRITE_MOUNTED. 648 649 bert_disable [ACPI] 650 Disable BERT OS support on buggy BIOSes. 651 652 bgrt_disable [ACPI,X86,EARLY] 653 Disable BGRT to avoid flickering OEM logo. 654 655 blkdevparts= Manual partition parsing of block device(s) for 656 embedded devices based on command line input. 657 See Documentation/block/cmdline-partition.rst 658 659 boot_delay= [KNL,EARLY] 660 Milliseconds to delay each printk during boot. 661 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled, 662 and you may also have to specify "lpj=". Boot_delay 663 values larger than 10 seconds (10000) are assumed 664 erroneous and ignored. 665 Format: integer 666 667 bootconfig [KNL,EARLY] 668 Extended command line options can be added to an initrd 669 and this will cause the kernel to look for it. 670 671 See Documentation/admin-guide/bootconfig.rst 672 673 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 674 bttv.radio= Most important insmod options are available as 675 kernel args too. 676 bttv.pll= See Documentation/admin-guide/media/bttv.rst 677 bttv.tuner= 678 679 bulk_remove=off [PPC] This parameter disables the use of the pSeries 680 firmware feature for flushing multiple hpte entries 681 at a time. 682 683 c101= [NET] Moxa C101 synchronous serial card 684 685 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 686 Sometimes CPU hardware bugs make them report the cache 687 size incorrectly. The kernel will attempt work arounds 688 to fix known problems, but for some CPUs it is not 689 possible to determine what the correct size should be. 690 This option provides an override for these situations. 691 692 carrier_timeout= 693 [NET] Specifies amount of time (in seconds) that 694 the kernel should wait for a network carrier. By default 695 it waits 120 seconds. 696 697 ca_keys= [KEYS] This parameter identifies a specific key(s) on 698 the system trusted keyring to be used for certificate 699 trust validation. 700 format: { id:<keyid> | builtin } 701 702 cca= [MIPS,EARLY] Override the kernel pages' cache coherency 703 algorithm. Accepted values range from 0 to 7 704 inclusive. See arch/mips/include/asm/pgtable-bits.h 705 for platform specific values (SB1, Loongson3 and 706 others). 707 708 ccw_timeout_log [S390] 709 See Documentation/arch/s390/common_io.rst for details. 710 711 cfi= [X86-64] Set Control Flow Integrity checking features 712 when CONFIG_FINEIBT is enabled. 713 Format: feature[,feature...] 714 Default: auto 715 716 auto: Use FineIBT if IBT available, otherwise kCFI. 717 Under FineIBT, enable "paranoid" mode when 718 FRED is not available. 719 off: Turn off CFI checking. 720 kcfi: Use kCFI (disable FineIBT). 721 fineibt: Use FineIBT (even if IBT not available). 722 norand: Do not re-randomize CFI hashes. 723 paranoid: Add caller hash checking under FineIBT. 724 bhi: Enable register poisoning to stop speculation 725 across FineIBT. (Disabled by default.) 726 warn: Do not enforce CFI checking: warn only. 727 debug: Report CFI initialization details. 728 729 cgroup_disable= [KNL] Disable a particular controller or optional feature 730 Format: {name of the controller(s) or feature(s) to disable} 731 The effects of cgroup_disable=foo are: 732 - foo isn't auto-mounted if you mount all cgroups in 733 a single hierarchy 734 - foo isn't visible as an individually mountable 735 subsystem 736 - if foo is an optional feature then the feature is 737 disabled and corresponding cgroup files are not 738 created 739 {Currently only "memory" controller deal with this and 740 cut the overhead, others just disable the usage. So 741 only cgroup_disable=memory is actually worthy} 742 Specifying "pressure" disables per-cgroup pressure 743 stall information accounting feature 744 745 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 746 Format: { { controller | "all" | "named" } 747 [,{ controller | "all" | "named" }...] } 748 Like cgroup_disable, but only applies to cgroup v1; 749 the blacklisted controllers remain available in cgroup2. 750 "all" blacklists all controllers and "named" disables 751 named mounts. Specifying both "all" and "named" disables 752 all v1 hierarchies. 753 754 cgroup_v1_proc= [KNL] Show also missing controllers in /proc/cgroups 755 Format: { "true" | "false" } 756 /proc/cgroups lists only v1 controllers by default. 757 This compatibility option enables listing also v2 758 controllers (whose v1 code is not compiled!), so that 759 semi-legacy software can check this file to decide 760 about usage of v2 (sic) controllers. 761 762 cgroup_favordynmods= [KNL] Enable or Disable favordynmods. 763 Format: { "true" | "false" } 764 Defaults to the value of CONFIG_CGROUP_FAVOR_DYNMODS. 765 766 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 767 Format: <string> 768 nosocket -- Disable socket memory accounting. 769 nokmem -- Disable kernel memory accounting. 770 nobpf -- Disable BPF memory accounting. 771 772 check_pages= [MM,EARLY] Enable sanity checking of pages after 773 allocations / before freeing. This adds checks to catch 774 double-frees, use-after-frees, and other sources of 775 page corruption by inspecting page internals (flags, 776 mapcount/refcount, memcg_data, etc.). 777 Format: { "0" | "1" } 778 Default: 0 (1 if CONFIG_DEBUG_VM is set) 779 780 checkreqprot= [SELINUX] Set initial checkreqprot flag value. 781 Format: { "0" | "1" } 782 See security/selinux/Kconfig help text. 783 0 -- check protection applied by kernel (includes 784 any implied execute protection). 785 1 -- check protection requested by application. 786 Default value is set via a kernel config option. 787 Value can be changed at runtime via 788 /sys/fs/selinux/checkreqprot. 789 Setting checkreqprot to 1 is deprecated. 790 791 cio_ignore= [S390] 792 See Documentation/arch/s390/common_io.rst for details. 793 794 clearcpuid=X[,X...] [X86] 795 Disable CPUID feature X for the kernel. See 796 arch/x86/include/asm/cpufeatures.h for the valid bit 797 numbers X. Note the Linux-specific bits are not necessarily 798 stable over kernel options, but the vendor-specific 799 ones should be. 800 X can also be a string as appearing in the flags: line 801 in /proc/cpuinfo which does not have the above 802 instability issue. However, not all features have names 803 in /proc/cpuinfo. 804 Note that using this option will taint your kernel. 805 Also note that user programs calling CPUID directly 806 or using the feature without checking anything 807 will still see it. This just prevents it from 808 being used by the kernel or shown in /proc/cpuinfo. 809 Also note the kernel might malfunction if you disable 810 some critical bits. 811 812 clk_ignore_unused 813 [CLK] 814 Prevents the clock framework from automatically gating 815 clocks that have not been explicitly enabled by a Linux 816 device driver but are enabled in hardware at reset or 817 by the bootloader/firmware. Note that this does not 818 force such clocks to be always-on nor does it reserve 819 those clocks in any way. This parameter is useful for 820 debug and development, but should not be needed on a 821 platform with proper driver support. For more 822 information, see Documentation/driver-api/clk.rst. 823 824 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 825 [Deprecated] 826 Forces specified clocksource (if available) to be used 827 when calculating gettimeofday(). If specified 828 clocksource is not available, it defaults to PIT. 829 Format: { pit | tsc | cyclone | pmtmr } 830 831 clocksource= Override the default clocksource 832 Format: <string> 833 Override the default clocksource and use the clocksource 834 with the name specified. 835 Some clocksource names to choose from, depending on 836 the platform: 837 [all] jiffies (this is the base, fallback clocksource) 838 [ACPI] acpi_pm 839 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 840 pxa_timer,timer3,32k_counter,timer0_1 841 [X86-32] pit,hpet,tsc; 842 scx200_hrt on Geode; cyclone on IBM x440 843 [MIPS] MIPS 844 [PARISC] cr16 845 [S390] tod 846 [SH] SuperH 847 [SPARC64] tick 848 [X86-64] hpet,tsc 849 850 clocksource.arm_arch_timer.evtstrm= 851 [ARM,ARM64,EARLY] 852 Format: <bool> 853 Enable/disable the eventstream feature of the ARM 854 architected timer so that code using WFE-based polling 855 loops can be debugged more effectively on production 856 systems. 857 858 clocksource.verify_n_cpus= [KNL] 859 Limit the number of CPUs checked for clocksources 860 marked with CLOCK_SOURCE_VERIFY_PERCPU that 861 are marked unstable due to excessive skew. 862 A negative value says to check all CPUs, while 863 zero says not to check any. Values larger than 864 nr_cpu_ids are silently truncated to nr_cpu_ids. 865 The actual CPUs are chosen randomly, with 866 no replacement if the same CPU is chosen twice. 867 868 clocksource-wdtest.holdoff= [KNL] 869 Set the time in seconds that the clocksource 870 watchdog test waits before commencing its tests. 871 Defaults to zero when built as a module and to 872 10 seconds when built into the kernel. 873 874 cma=nn[MG]@[start[MG][-end[MG]]] 875 [KNL,CMA,EARLY] 876 Sets the size of kernel global memory area for 877 contiguous memory allocations and optionally the 878 placement constraint by the physical address range of 879 memory allocations. A value of 0 disables CMA 880 altogether. For more information, see 881 kernel/dma/contiguous.c 882 883 cma_pernuma=nn[MG] 884 [KNL,CMA,EARLY] 885 Sets the size of kernel per-numa memory area for 886 contiguous memory allocations. A value of 0 disables 887 per-numa CMA altogether. And If this option is not 888 specified, the default value is 0. 889 With per-numa CMA enabled, DMA users on node nid will 890 first try to allocate buffer from the pernuma area 891 which is located in node nid, if the allocation fails, 892 they will fallback to the global default memory area. 893 894 numa_cma=<node>:nn[MG][,<node>:nn[MG]] 895 [KNL,CMA,EARLY] 896 Sets the size of kernel numa memory area for 897 contiguous memory allocations. It will reserve CMA 898 area for the specified node. 899 900 With numa CMA enabled, DMA users on node nid will 901 first try to allocate buffer from the numa area 902 which is located in node nid, if the allocation fails, 903 they will fallback to the global default memory area. 904 905 cmo_free_hint= [PPC] Format: { yes | no } 906 Specify whether pages are marked as being inactive 907 when they are freed. This is used in CMO environments 908 to determine OS memory pressure for page stealing by 909 a hypervisor. 910 Default: yes 911 912 coherent_pool=nn[KMG] [ARM,KNL,EARLY] 913 Sets the size of memory pool for coherent, atomic dma 914 allocations, by default set to 256K. 915 916 com20020= [HW,NET] ARCnet - COM20020 chipset 917 Format: 918 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 919 920 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 921 Format: <io>[,<irq>] 922 923 com90xx= [HW,NET] 924 ARCnet - COM90xx chipset (memory-mapped buffers) 925 Format: <io>[,<irq>[,<memstart>]] 926 927 condev= [HW,S390] console device 928 conmode= 929 930 con3215_drop= [S390,EARLY] 3215 console drop mode. 931 Format: y|n|Y|N|1|0 932 When set to true, drop data on the 3215 console when 933 the console buffer is full. In this case the 934 operator using a 3270 terminal emulator (for example 935 x3270) does not have to enter the clear key for the 936 console output to advance and the kernel to continue. 937 This leads to a much faster boot time when a 3270 938 terminal emulator is active. If no 3270 terminal 939 emulator is used, this parameter has no effect. 940 941 console= [KNL] Output console device and options. 942 943 tty<n> Use the virtual console device <n>. 944 945 ttyS<n>[,options] 946 ttyUSB0[,options] 947 Use the specified serial port. The options are of 948 the form "bbbbpnf", where "bbbb" is the baud rate, 949 "p" is parity ("n", "o", or "e"), "n" is number of 950 bits, and "f" is flow control ("r" for RTS or 951 omit it). Default is "9600n8". 952 953 See Documentation/admin-guide/serial-console.rst for more 954 information. See 955 Documentation/networking/netconsole.rst for an 956 alternative. 957 958 <DEVNAME>:<n>.<n>[,options] 959 Use the specified serial port on the serial core bus. 960 The addressing uses DEVNAME of the physical serial port 961 device, followed by the serial core controller instance, 962 and the serial port instance. The options are the same 963 as documented for the ttyS addressing above. 964 965 The mapping of the serial ports to the tty instances 966 can be viewed with: 967 968 $ ls -d /sys/bus/serial-base/devices/*:*.*/tty/* 969 /sys/bus/serial-base/devices/00:04:0.0/tty/ttyS0 970 971 In the above example, the console can be addressed with 972 console=00:04:0.0. Note that a console addressed this 973 way will only get added when the related device driver 974 is ready. The use of an earlycon parameter in addition to 975 the console may be desired for console output early on. 976 977 uart[8250],io,<addr>[,options] 978 uart[8250],mmio,<addr>[,options] 979 uart[8250],mmio16,<addr>[,options] 980 uart[8250],mmio32,<addr>[,options] 981 uart[8250],0x<addr>[,options] 982 Start an early, polled-mode console on the 8250/16550 983 UART at the specified I/O port or MMIO address, 984 switching to the matching ttyS device later. 985 MMIO inter-register address stride is either 8-bit 986 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 987 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 988 to be equivalent to 'mmio'. 'options' are specified in 989 the same format described for ttyS above; if unspecified, 990 the h/w is not re-initialized. 991 992 hvc<n> Use the hypervisor console device <n>. This is for 993 both Xen and PowerPC hypervisors. 994 995 { null | "" } 996 Use to disable console output, i.e., to have kernel 997 console messages discarded. 998 This must be the only console= parameter used on the 999 kernel command line. 1000 1001 If the device connected to the port is not a TTY but a braille 1002 device, prepend "brl," before the device type, for instance 1003 console=brl,ttyS0 1004 For now, only VisioBraille is supported. 1005 1006 console_msg_format= 1007 [KNL] Change console messages format 1008 default 1009 By default we print messages on consoles in 1010 "[time stamp] text\n" format (time stamp may not be 1011 printed, depending on CONFIG_PRINTK_TIME or 1012 `printk_time' param). 1013 syslog 1014 Switch to syslog format: "<%u>[time stamp] text\n" 1015 IOW, each message will have a facility and loglevel 1016 prefix. The format is similar to one used by syslog() 1017 syscall, or to executing "dmesg -S --raw" or to reading 1018 from /proc/kmsg. 1019 1020 consoleblank= [KNL] The console blank (screen saver) timeout in 1021 seconds. A value of 0 disables the blank timer. 1022 Defaults to 0. 1023 1024 coredump_filter= 1025 [KNL] Change the default value for 1026 /proc/<pid>/coredump_filter. 1027 See also Documentation/filesystems/proc.rst. 1028 1029 coresight_cpu_debug.enable 1030 [ARM,ARM64] 1031 Format: <bool> 1032 Enable/disable the CPU sampling based debugging. 1033 0: default value, disable debugging 1034 1: enable debugging at boot time 1035 1036 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 1037 Format: 1038 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 1039 1040 cpuidle.off=1 [CPU_IDLE] 1041 disable the cpuidle sub-system 1042 1043 cpuidle.governor= 1044 [CPU_IDLE] Name of the cpuidle governor to use. 1045 1046 cpufreq.off=1 [CPU_FREQ] 1047 disable the cpufreq sub-system 1048 1049 cpufreq.default_governor= 1050 [CPU_FREQ] Name of the default cpufreq governor or 1051 policy to use. This governor must be registered in the 1052 kernel before the cpufreq driver probes. 1053 1054 cpu_init_udelay=N 1055 [X86,EARLY] Delay for N microsec between assert and de-assert 1056 of APIC INIT to start processors. This delay occurs 1057 on every CPU online, such as boot, and resume from suspend. 1058 Default: 10000 1059 1060 cpuhp.parallel= 1061 [SMP] Enable/disable parallel bringup of secondary CPUs 1062 Format: <bool> 1063 Default is enabled if CONFIG_HOTPLUG_PARALLEL=y. Otherwise 1064 the parameter has no effect. 1065 1066 crash_kexec_post_notifiers 1067 Only jump to kdump kernel after running the panic 1068 notifiers and dumping kmsg. This option increases 1069 the risks of a kdump failure, since some panic 1070 notifiers can make the crashed kernel more unstable. 1071 In configurations where kdump may not be reliable, 1072 running the panic notifiers could allow collecting 1073 more data on dmesg, like stack traces from other CPUS 1074 or extra data dumped by panic_print. Note that some 1075 configurations enable this option unconditionally, 1076 like Hyper-V, PowerPC (fadump) and AMD SEV-SNP. 1077 1078 crashkernel=size[KMG][@offset[KMG]] 1079 [KNL,EARLY] Using kexec, Linux can switch to a 'crash kernel' 1080 upon panic. This parameter reserves the physical 1081 memory region [offset, offset + size] for that kernel 1082 image. If '@offset' is omitted, then a suitable offset 1083 is selected automatically. 1084 [KNL, X86-64, ARM64, RISCV, LoongArch] Select a region 1085 under 4G first, and fall back to reserve region above 1086 4G when '@offset' hasn't been specified. 1087 See Documentation/admin-guide/kdump/kdump.rst for further details. 1088 1089 crashkernel=range1:size1[,range2:size2,...][@offset] 1090 [KNL] Same as above, but depends on the memory 1091 in the running system. The syntax of range is 1092 start-[end] where start and end are both 1093 a memory unit (amount[KMG]). See also 1094 Documentation/admin-guide/kdump/kdump.rst for an example. 1095 1096 crashkernel=size[KMG],high 1097 [KNL, X86-64, ARM64, RISCV, LoongArch] range could be 1098 above 4G. 1099 Allow kernel to allocate physical memory region from top, 1100 so could be above 4G if system have more than 4G ram 1101 installed. Otherwise memory region will be allocated 1102 below 4G, if available. 1103 It will be ignored if crashkernel=X is specified. 1104 crashkernel=size[KMG],low 1105 [KNL, X86-64, ARM64, RISCV, LoongArch] range under 4G. 1106 When crashkernel=X,high is passed, kernel could allocate 1107 physical memory region above 4G, that cause second kernel 1108 crash on system that require some amount of low memory, 1109 e.g. swiotlb requires at least 64M+32K low memory, also 1110 enough extra low memory is needed to make sure DMA buffers 1111 for 32-bit devices won't run out. Kernel would try to allocate 1112 default size of memory below 4G automatically. The default 1113 size is platform dependent. 1114 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB) 1115 --> arm64: 128MiB 1116 --> riscv: 128MiB 1117 --> loongarch: 128MiB 1118 This one lets the user specify own low range under 4G 1119 for second kernel instead. 1120 0: to disable low allocation. 1121 It will be ignored when crashkernel=X,high is not used 1122 or memory reserved is below 4G. 1123 crashkernel=size[KMG],cma 1124 [KNL, X86, ppc] Reserve additional crash kernel memory from 1125 CMA. This reservation is usable by the first system's 1126 userspace memory and kernel movable allocations (memory 1127 balloon, zswap). Pages allocated from this memory range 1128 will not be included in the vmcore so this should not 1129 be used if dumping of userspace memory is intended and 1130 it has to be expected that some movable kernel pages 1131 may be missing from the dump. 1132 1133 A standard crashkernel reservation, as described above, 1134 is still needed to hold the crash kernel and initrd. 1135 1136 This option increases the risk of a kdump failure: DMA 1137 transfers configured by the first kernel may end up 1138 corrupting the second kernel's memory. 1139 1140 This reservation method is intended for systems that 1141 can't afford to sacrifice enough memory for standard 1142 crashkernel reservation and where less reliable and 1143 possibly incomplete kdump is preferable to no kdump at 1144 all. 1145 1146 cryptomgr.notests 1147 [KNL] Disable crypto self-tests 1148 1149 cs89x0_dma= [HW,NET] 1150 Format: <dma> 1151 1152 cs89x0_media= [HW,NET] 1153 Format: { rj45 | aui | bnc } 1154 1155 csdlock_debug= [KNL] Enable or disable debug add-ons of cross-CPU 1156 function call handling. When switched on, 1157 additional debug data is printed to the console 1158 in case a hanging CPU is detected, and that 1159 CPU is pinged again in order to try to resolve 1160 the hang situation. The default value of this 1161 option depends on the CSD_LOCK_WAIT_DEBUG_DEFAULT 1162 Kconfig option. 1163 1164 dasd= [HW,NET] 1165 See header of drivers/s390/block/dasd_devmap.c. 1166 1167 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 1168 (one device per port) 1169 Format: <port#>,<type> 1170 See also Documentation/input/devices/joystick-parport.rst 1171 1172 debug [KNL,EARLY] Enable kernel debugging (events log level). 1173 1174 debug_boot_weak_hash 1175 [KNL,EARLY] Enable printing [hashed] pointers early in the 1176 boot sequence. If enabled, we use a weak hash instead 1177 of siphash to hash pointers. Use this option if you are 1178 seeing instances of '(___ptrval___)') and need to see a 1179 value (hashed pointer) instead. Cryptographically 1180 insecure, please do not use on production kernels. 1181 1182 debug_locks_verbose= 1183 [KNL] verbose locking self-tests 1184 Format: <int> 1185 Print debugging info while doing the locking API 1186 self-tests. 1187 Bitmask for the various LOCKTYPE_ tests. Defaults to 0 1188 (no extra messages), setting it to -1 (all bits set) 1189 will print _a_lot_ more information - normally only 1190 useful to lockdep developers. 1191 1192 debug_objects [KNL,EARLY] Enable object debugging 1193 1194 debug_guardpage_minorder= 1195 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this 1196 parameter allows control of the order of pages that will 1197 be intentionally kept free (and hence protected) by the 1198 buddy allocator. Bigger value increase the probability 1199 of catching random memory corruption, but reduce the 1200 amount of memory for normal system use. The maximum 1201 possible value is MAX_PAGE_ORDER/2. Setting this 1202 parameter to 1 or 2 should be enough to identify most 1203 random memory corruption problems caused by bugs in 1204 kernel or driver code when a CPU writes to (or reads 1205 from) a random memory location. Note that there exists 1206 a class of memory corruptions problems caused by buggy 1207 H/W or F/W or by drivers badly programming DMA 1208 (basically when memory is written at bus level and the 1209 CPU MMU is bypassed) which are not detectable by 1210 CONFIG_DEBUG_PAGEALLOC, hence this option will not 1211 help tracking down these problems. 1212 1213 debug_pagealloc= 1214 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 1215 enables the feature at boot time. By default, it is 1216 disabled and the system will work mostly the same as a 1217 kernel built without CONFIG_DEBUG_PAGEALLOC. 1218 Note: to get most of debug_pagealloc error reports, it's 1219 useful to also enable the page_owner functionality. 1220 on: enable the feature 1221 1222 debugfs= [KNL,EARLY] This parameter enables what is exposed to 1223 userspace and debugfs internal clients. 1224 Format: { on, off } 1225 on: All functions are enabled. 1226 off: Filesystem is not registered and clients 1227 get a -EPERM as result when trying to register files 1228 or directories within debugfs. 1229 This is equivalent of the runtime functionality if 1230 debugfs was not enabled in the kernel at all. 1231 Default value is set in build-time with a kernel configuration. 1232 1233 debugpat [X86] Enable PAT debugging 1234 1235 default_hugepagesz= 1236 [HW] The size of the default HugeTLB page. This is 1237 the size represented by the legacy /proc/ hugepages 1238 APIs. In addition, this is the default hugetlb size 1239 used for shmget(), mmap() and mounting hugetlbfs 1240 filesystems. If not specified, defaults to the 1241 architecture's default huge page size. Huge page 1242 sizes are architecture dependent. See also 1243 Documentation/admin-guide/mm/hugetlbpage.rst. 1244 Format: size[KMG] 1245 1246 deferred_probe_timeout= 1247 [KNL] Debugging option to set a timeout in seconds for 1248 deferred probe to give up waiting on dependencies to 1249 probe. Only specific dependencies (subsystems or 1250 drivers) that have opted in will be ignored. A timeout 1251 of 0 will timeout at the end of initcalls. If the time 1252 out hasn't expired, it'll be restarted by each 1253 successful driver registration. This option will also 1254 dump out devices still on the deferred probe list after 1255 retrying. 1256 1257 delayacct [KNL] Enable per-task delay accounting 1258 1259 dell_smm_hwmon.ignore_dmi= 1260 [HW] Continue probing hardware even if DMI data 1261 indicates that the driver is running on unsupported 1262 hardware. 1263 1264 dell_smm_hwmon.force= 1265 [HW] Activate driver even if SMM BIOS signature does 1266 not match list of supported models and enable otherwise 1267 blacklisted features. 1268 1269 dell_smm_hwmon.power_status= 1270 [HW] Report power status in /proc/i8k 1271 (disabled by default). 1272 1273 dell_smm_hwmon.restricted= 1274 [HW] Allow controlling fans only if SYS_ADMIN 1275 capability is set. 1276 1277 dell_smm_hwmon.fan_mult= 1278 [HW] Factor to multiply fan speed with. 1279 1280 dell_smm_hwmon.fan_max= 1281 [HW] Maximum configurable fan speed. 1282 1283 dfltcc= [HW,S390] 1284 Format: { on | off | def_only | inf_only | always } 1285 on: s390 zlib hardware support for compression on 1286 level 1 and decompression (default) 1287 off: No s390 zlib hardware support 1288 def_only: s390 zlib hardware support for deflate 1289 only (compression on level 1) 1290 inf_only: s390 zlib hardware support for inflate 1291 only (decompression) 1292 always: Same as 'on' but ignores the selected compression 1293 level always using hardware support (used for debugging) 1294 1295 dhash_entries= [KNL] 1296 Set number of hash buckets for dentry cache. 1297 1298 disable_1tb_segments [PPC,EARLY] 1299 Disables the use of 1TB hash page table segments. This 1300 causes the kernel to fall back to 256MB segments which 1301 can be useful when debugging issues that require an SLB 1302 miss to occur. 1303 1304 disable= [IPV6] 1305 See Documentation/networking/ipv6.rst. 1306 1307 disable_radix [PPC,EARLY] 1308 Disable RADIX MMU mode on POWER9 1309 1310 disable_tlbie [PPC] 1311 Disable TLBIE instruction. Currently does not work 1312 with KVM, with HASH MMU, or with coherent accelerators. 1313 1314 disable_ddw [PPC/PSERIES,EARLY] 1315 Disable Dynamic DMA Window support. Use this 1316 to workaround buggy firmware. 1317 1318 disable_ipv6= [IPV6] 1319 See Documentation/networking/ipv6.rst. 1320 1321 disable_mtrr_cleanup [X86,EARLY] 1322 The kernel tries to adjust MTRR layout from continuous 1323 to discrete, to make X server driver able to add WB 1324 entry later. This parameter disables that. 1325 1326 disable_mtrr_trim [X86, Intel and AMD only,EARLY] 1327 By default the kernel will trim any uncacheable 1328 memory out of your available memory pool based on 1329 MTRR settings. This parameter disables that behavior, 1330 possibly causing your machine to run very slowly. 1331 1332 disable_timer_pin_1 [X86,EARLY] 1333 Disable PIN 1 of APIC timer 1334 Can be useful to work around chipset bugs. 1335 1336 dis_ucode_ldr [X86] Disable the microcode loader. 1337 1338 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 1339 this option disables the debugging code at boot. 1340 1341 dma_debug_entries=<number> 1342 This option allows to tune the number of preallocated 1343 entries for DMA-API debugging code. One entry is 1344 required per DMA-API allocation. Use this if the 1345 DMA-API debugging code disables itself because the 1346 architectural default is too low. 1347 1348 dma_debug_driver=<driver_name> 1349 With this option the DMA-API debugging driver 1350 filter feature can be enabled at boot time. Just 1351 pass the driver to filter for as the parameter. 1352 The filter can be disabled or changed to another 1353 driver later using sysfs. 1354 1355 reg_file_data_sampling= 1356 [X86] Controls mitigation for Register File Data 1357 Sampling (RFDS) vulnerability. RFDS is a CPU 1358 vulnerability which may allow userspace to infer 1359 kernel data values previously stored in floating point 1360 registers, vector registers, or integer registers. 1361 RFDS only affects Intel Atom processors. 1362 1363 on: Turns ON the mitigation. 1364 off: Turns OFF the mitigation. 1365 1366 This parameter overrides the compile time default set 1367 by CONFIG_MITIGATION_RFDS. Mitigation cannot be 1368 disabled when other VERW based mitigations (like MDS) 1369 are enabled. In order to disable RFDS mitigation all 1370 VERW based mitigations need to be disabled. 1371 1372 For details see: 1373 Documentation/admin-guide/hw-vuln/reg-file-data-sampling.rst 1374 1375 driver_async_probe= [KNL] 1376 List of driver names to be probed asynchronously. * 1377 matches with all driver names. If * is specified, the 1378 rest of the listed driver names are those that will NOT 1379 match the *. 1380 Format: <driver_name1>,<driver_name2>... 1381 1382 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1383 Broken monitors, graphic adapters, KVMs and EDIDless 1384 panels may send no or incorrect EDID data sets. 1385 This parameter allows to specify an EDID data sets 1386 in the /lib/firmware directory that are used instead. 1387 An EDID data set will only be used for a particular 1388 connector, if its name and a colon are prepended to 1389 the EDID name. Each connector may use a unique EDID 1390 data set by separating the files with a comma. An EDID 1391 data set with no connector name will be used for 1392 any connectors not explicitly specified. 1393 1394 dscc4.setup= [NET] 1395 1396 dt_cpu_ftrs= [PPC,EARLY] 1397 Format: {"off" | "known"} 1398 Control how the dt_cpu_ftrs device-tree binding is 1399 used for CPU feature discovery and setup (if it 1400 exists). 1401 off: Do not use it, fall back to legacy cpu table. 1402 known: Do not pass through unknown features to guests 1403 or userspace, only those that the kernel is aware of. 1404 1405 dump_apple_properties [X86] 1406 Dump name and content of EFI device properties on 1407 x86 Macs. Useful for driver authors to determine 1408 what data is available or for reverse-engineering. 1409 1410 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1411 <module>.dyndbg[="val"] 1412 Enable debug messages at boot time. See 1413 Documentation/admin-guide/dynamic-debug-howto.rst 1414 for details. 1415 1416 early_ioremap_debug [KNL,EARLY] 1417 Enable debug messages in early_ioremap support. This 1418 is useful for tracking down temporary early mappings 1419 which are not unmapped. 1420 1421 earlycon= [KNL,EARLY] Output early console device and options. 1422 1423 When used with no options, the early console is 1424 determined by stdout-path property in device tree's 1425 chosen node or the ACPI SPCR table if supported by 1426 the platform. 1427 1428 cdns,<addr>[,options] 1429 Start an early, polled-mode console on a Cadence 1430 (xuartps) serial port at the specified address. Only 1431 supported option is baud rate. If baud rate is not 1432 specified, the serial port must already be setup and 1433 configured. 1434 1435 uart[8250],io,<addr>[,options[,uartclk]] 1436 uart[8250],mmio,<addr>[,options[,uartclk]] 1437 uart[8250],mmio32,<addr>[,options[,uartclk]] 1438 uart[8250],mmio32be,<addr>[,options[,uartclk]] 1439 uart[8250],0x<addr>[,options] 1440 Start an early, polled-mode console on the 8250/16550 1441 UART at the specified I/O port or MMIO address. 1442 MMIO inter-register address stride is either 8-bit 1443 (mmio) or 32-bit (mmio32 or mmio32be). 1444 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1445 to be equivalent to 'mmio'. 'options' are specified 1446 in the same format described for "console=ttyS<n>"; if 1447 unspecified, the h/w is not initialized. 'uartclk' is 1448 the uart clock frequency; if unspecified, it is set 1449 to 'BASE_BAUD' * 16. 1450 1451 pl011,<addr> 1452 pl011,mmio32,<addr> 1453 Start an early, polled-mode console on a pl011 serial 1454 port at the specified address. The pl011 serial port 1455 must already be setup and configured. Options are not 1456 yet supported. If 'mmio32' is specified, then only 1457 the driver will use only 32-bit accessors to read/write 1458 the device registers. 1459 1460 liteuart,<addr> 1461 Start an early console on a litex serial port at the 1462 specified address. The serial port must already be 1463 setup and configured. Options are not yet supported. 1464 1465 meson,<addr> 1466 Start an early, polled-mode console on a meson serial 1467 port at the specified address. The serial port must 1468 already be setup and configured. Options are not yet 1469 supported. 1470 1471 msm_serial,<addr> 1472 Start an early, polled-mode console on an msm serial 1473 port at the specified address. The serial port 1474 must already be setup and configured. Options are not 1475 yet supported. 1476 1477 msm_serial_dm,<addr> 1478 Start an early, polled-mode console on an msm serial 1479 dm port at the specified address. The serial port 1480 must already be setup and configured. Options are not 1481 yet supported. 1482 1483 owl,<addr> 1484 Start an early, polled-mode console on a serial port 1485 of an Actions Semi SoC, such as S500 or S900, at the 1486 specified address. The serial port must already be 1487 setup and configured. Options are not yet supported. 1488 1489 rda,<addr> 1490 Start an early, polled-mode console on a serial port 1491 of an RDA Micro SoC, such as RDA8810PL, at the 1492 specified address. The serial port must already be 1493 setup and configured. Options are not yet supported. 1494 1495 sbi 1496 Use RISC-V SBI (Supervisor Binary Interface) for early 1497 console. 1498 1499 smh Use ARM semihosting calls for early console. 1500 1501 s3c2410,<addr> 1502 s3c2412,<addr> 1503 s3c2440,<addr> 1504 s3c6400,<addr> 1505 s5pv210,<addr> 1506 exynos4210,<addr> 1507 Use early console provided by serial driver available 1508 on Samsung SoCs, requires selecting proper type and 1509 a correct base address of the selected UART port. The 1510 serial port must already be setup and configured. 1511 Options are not yet supported. 1512 1513 lantiq,<addr> 1514 Start an early, polled-mode console on a lantiq serial 1515 (lqasc) port at the specified address. The serial port 1516 must already be setup and configured. Options are not 1517 yet supported. 1518 1519 lpuart,<addr> 1520 lpuart32,<addr> 1521 Use early console provided by Freescale LP UART driver 1522 found on Freescale Vybrid and QorIQ LS1021A processors. 1523 A valid base address must be provided, and the serial 1524 port must already be setup and configured. 1525 1526 ec_imx21,<addr> 1527 ec_imx6q,<addr> 1528 Start an early, polled-mode, output-only console on the 1529 Freescale i.MX UART at the specified address. The UART 1530 must already be setup and configured. 1531 1532 ar3700_uart,<addr> 1533 Start an early, polled-mode console on the 1534 Armada 3700 serial port at the specified 1535 address. The serial port must already be setup 1536 and configured. Options are not yet supported. 1537 1538 qcom_geni,<addr> 1539 Start an early, polled-mode console on a Qualcomm 1540 Generic Interface (GENI) based serial port at the 1541 specified address. The serial port must already be 1542 setup and configured. Options are not yet supported. 1543 1544 efifb,[options] 1545 Start an early, unaccelerated console on the EFI 1546 memory mapped framebuffer (if available). On cache 1547 coherent non-x86 systems that use system memory for 1548 the framebuffer, pass the 'ram' option so that it is 1549 mapped with the correct attributes. 1550 1551 linflex,<addr> 1552 Use early console provided by Freescale LINFlexD UART 1553 serial driver for NXP S32V234 SoCs. A valid base 1554 address must be provided, and the serial port must 1555 already be setup and configured. 1556 1557 earlyprintk= [X86,SH,ARM,M68k,S390,UM,EARLY] 1558 earlyprintk=vga 1559 earlyprintk=sclp 1560 earlyprintk=xen 1561 earlyprintk=serial[,ttySn[,baudrate]] 1562 earlyprintk=serial[,0x...[,baudrate]] 1563 earlyprintk=ttySn[,baudrate] 1564 earlyprintk=dbgp[debugController#] 1565 earlyprintk=mmio32,membase[,{nocfg|baudrate}] 1566 earlyprintk=pciserial[,force],bus:device.function[,{nocfg|baudrate}] 1567 earlyprintk=xdbc[xhciController#] 1568 earlyprintk=bios 1569 1570 earlyprintk is useful when the kernel crashes before 1571 the normal console is initialized. It is not enabled by 1572 default because it has some cosmetic problems. 1573 1574 Use "nocfg" to skip UART configuration, assume 1575 BIOS/firmware has configured UART correctly. 1576 1577 Append ",keep" to not disable it when the real console 1578 takes over. 1579 1580 Only one of vga, serial, or usb debug port can 1581 be used at a time. 1582 1583 Currently only ttyS0 and ttyS1 may be specified by 1584 name. Other I/O ports may be explicitly specified 1585 on some architectures (x86 and arm at least) by 1586 replacing ttySn with an I/O port address, like this: 1587 earlyprintk=serial,0x1008,115200 1588 You can find the port for a given device in 1589 /proc/tty/driver/serial: 1590 2: uart:ST16650V2 port:00001008 irq:18 ... 1591 1592 Interaction with the standard serial driver is not 1593 very good. 1594 1595 The VGA output is eventually overwritten by 1596 the real console. 1597 1598 The xen option can only be used in Xen domains. 1599 1600 The sclp output can only be used on s390. 1601 1602 The bios output can only be used on SuperH. 1603 1604 The optional "force" to "pciserial" enables use of a 1605 PCI device even when its classcode is not of the 1606 UART class. 1607 1608 edac_report= [HW,EDAC] Control how to report EDAC event 1609 Format: {"on" | "off" | "force"} 1610 on: enable EDAC to report H/W event. May be overridden 1611 by other higher priority error reporting module. 1612 off: disable H/W event reporting through EDAC. 1613 force: enforce the use of EDAC to report H/W event. 1614 default: on. 1615 1616 edd= [EDD] 1617 Format: {"off" | "on" | "skip[mbr]"} 1618 1619 efi= [EFI,EARLY] 1620 Format: { "debug", "disable_early_pci_dma", 1621 "nochunk", "noruntime", "nosoftreserve", 1622 "novamap", "no_disable_early_pci_dma" } 1623 debug: enable misc debug output. 1624 disable_early_pci_dma: disable the busmaster bit on all 1625 PCI bridges while in the EFI boot stub. 1626 nochunk: disable reading files in "chunks" in the EFI 1627 boot stub, as chunking can cause problems with some 1628 firmware implementations. 1629 noruntime : disable EFI runtime services support 1630 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1631 attribute may cause the kernel to reserve the 1632 memory range for a memory mapping driver to 1633 claim. Specify efi=nosoftreserve to disable this 1634 reservation and treat the memory by its base type 1635 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1636 novamap: do not call SetVirtualAddressMap(). 1637 no_disable_early_pci_dma: Leave the busmaster bit set 1638 on all PCI bridges while in the EFI boot stub 1639 1640 efi_no_storage_paranoia [EFI,X86,EARLY] 1641 Using this parameter you can use more than 50% of 1642 your efi variable storage. Use this parameter only if 1643 you are really sure that your UEFI does sane gc and 1644 fulfills the spec otherwise your board may brick. 1645 1646 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1647 that is to be dynamically loaded by Linux. If there are 1648 multiple variables with the same name but with different 1649 vendor GUIDs, all of them will be loaded. See 1650 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1651 1652 1653 eisa_irq_edge= [PARISC,HW] 1654 See header of drivers/parisc/eisa.c. 1655 1656 ekgdboc= [X86,KGDB,EARLY] Allow early kernel console debugging 1657 Format: ekgdboc=kbd 1658 1659 This is designed to be used in conjunction with 1660 the boot argument: earlyprintk=vga 1661 1662 This parameter works in place of the kgdboc parameter 1663 but can only be used if the backing tty is available 1664 very early in the boot process. For early debugging 1665 via a serial port see kgdboc_earlycon instead. 1666 1667 elanfreq= [X86-32] 1668 See comment before function elanfreq_setup() in 1669 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1670 1671 elfcorehdr=[size[KMG]@]offset[KMG] [PPC,SH,X86,S390,EARLY] 1672 Specifies physical address of start of kernel core 1673 image elf header and optionally the size. Generally 1674 kexec loader will pass this option to capture kernel. 1675 See Documentation/admin-guide/kdump/kdump.rst for details. 1676 1677 enable_mtrr_cleanup [X86,EARLY] 1678 The kernel tries to adjust MTRR layout from continuous 1679 to discrete, to make X server driver able to add WB 1680 entry later. This parameter enables that. 1681 1682 enable_timer_pin_1 [X86] 1683 Enable PIN 1 of APIC timer 1684 Can be useful to work around chipset bugs 1685 (in particular on some ATI chipsets). 1686 The kernel tries to set a reasonable default. 1687 1688 enforcing= [SELINUX] Set initial enforcing status. 1689 Format: {"0" | "1"} 1690 See security/selinux/Kconfig help text. 1691 0 -- permissive (log only, no denials). 1692 1 -- enforcing (deny and log). 1693 Default value is 0. 1694 Value can be changed at runtime via 1695 /sys/fs/selinux/enforce. 1696 1697 erst_disable [ACPI] 1698 Disable Error Record Serialization Table (ERST) 1699 support. 1700 1701 ether= [HW,NET] Ethernet cards parameters 1702 This option is obsoleted by the "netdev=" option, which 1703 has equivalent usage. See its documentation for details. 1704 1705 evm= [EVM] 1706 Format: { "fix" } 1707 Permit 'security.evm' to be updated regardless of 1708 current integrity status. 1709 1710 early_page_ext [KNL,EARLY] Enforces page_ext initialization to earlier 1711 stages so cover more early boot allocations. 1712 Please note that as side effect some optimizations 1713 might be disabled to achieve that (e.g. parallelized 1714 memory initialization is disabled) so the boot process 1715 might take longer, especially on systems with a lot of 1716 memory. Available with CONFIG_PAGE_EXTENSION=y. 1717 1718 failslab= 1719 fail_usercopy= 1720 fail_page_alloc= 1721 fail_skb_realloc= 1722 fail_make_request=[KNL] 1723 General fault injection mechanism. 1724 Format: <interval>,<probability>,<space>,<times> 1725 See also Documentation/fault-injection/. 1726 1727 fb_tunnels= [NET] 1728 Format: { initns | none } 1729 See Documentation/admin-guide/sysctl/net.rst for 1730 fb_tunnels_only_for_init_ns 1731 1732 floppy= [HW] 1733 See Documentation/admin-guide/blockdev/floppy.rst. 1734 1735 forcepae [X86-32] 1736 Forcefully enable Physical Address Extension (PAE). 1737 Many Pentium M systems disable PAE but may have a 1738 functionally usable PAE implementation. 1739 Warning: use of this parameter will taint the kernel 1740 and may cause unknown problems. 1741 1742 fred= [X86-64] 1743 Enable/disable Flexible Return and Event Delivery. 1744 Format: { on | off } 1745 on: enable FRED when it's present. 1746 off: disable FRED, the default setting. 1747 1748 ftrace=[tracer] 1749 [FTRACE] will set and start the specified tracer 1750 as early as possible in order to facilitate early 1751 boot debugging. 1752 1753 ftrace_boot_snapshot 1754 [FTRACE] On boot up, a snapshot will be taken of the 1755 ftrace ring buffer that can be read at: 1756 /sys/kernel/tracing/snapshot. 1757 This is useful if you need tracing information from kernel 1758 boot up that is likely to be overridden by user space 1759 start up functionality. 1760 1761 Optionally, the snapshot can also be defined for a tracing 1762 instance that was created by the trace_instance= command 1763 line parameter. 1764 1765 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo 1766 1767 The above will cause the "foo" tracing instance to trigger 1768 a snapshot at the end of boot up. 1769 1770 ftrace_dump_on_oops[=2(orig_cpu) | =<instance>][,<instance> | 1771 ,<instance>=2(orig_cpu)] 1772 [FTRACE] will dump the trace buffers on oops. 1773 If no parameter is passed, ftrace will dump global 1774 buffers of all CPUs, if you pass 2 or orig_cpu, it 1775 will dump only the buffer of the CPU that triggered 1776 the oops, or the specific instance will be dumped if 1777 its name is passed. Multiple instance dump is also 1778 supported, and instances are separated by commas. Each 1779 instance supports only dump on CPU that triggered the 1780 oops by passing 2 or orig_cpu to it. 1781 1782 ftrace_dump_on_oops=foo=orig_cpu 1783 1784 The above will dump only the buffer of "foo" instance 1785 on CPU that triggered the oops. 1786 1787 ftrace_dump_on_oops,foo,bar=orig_cpu 1788 1789 The above will dump global buffer on all CPUs, the 1790 buffer of "foo" instance on all CPUs and the buffer 1791 of "bar" instance on CPU that triggered the oops. 1792 1793 ftrace_filter=[function-list] 1794 [FTRACE] Limit the functions traced by the function 1795 tracer at boot up. function-list is a comma-separated 1796 list of functions. This list can be changed at run 1797 time by the set_ftrace_filter file in the debugfs 1798 tracing directory. 1799 1800 ftrace_notrace=[function-list] 1801 [FTRACE] Do not trace the functions specified in 1802 function-list. This list can be changed at run time 1803 by the set_ftrace_notrace file in the debugfs 1804 tracing directory. 1805 1806 ftrace_graph_filter=[function-list] 1807 [FTRACE] Limit the top level callers functions traced 1808 by the function graph tracer at boot up. 1809 function-list is a comma-separated list of functions 1810 that can be changed at run time by the 1811 set_graph_function file in the debugfs tracing directory. 1812 1813 ftrace_graph_notrace=[function-list] 1814 [FTRACE] Do not trace from the functions specified in 1815 function-list. This list is a comma-separated list of 1816 functions that can be changed at run time by the 1817 set_graph_notrace file in the debugfs tracing directory. 1818 1819 ftrace_graph_max_depth=<uint> 1820 [FTRACE] Used with the function graph tracer. This is 1821 the max depth it will trace into a function. This value 1822 can be changed at run time by the max_graph_depth file 1823 in the tracefs tracing directory. default: 0 (no limit) 1824 1825 fw_devlink= [KNL,EARLY] Create device links between consumer and supplier 1826 devices by scanning the firmware to infer the 1827 consumer/supplier relationships. This feature is 1828 especially useful when drivers are loaded as modules as 1829 it ensures proper ordering of tasks like device probing 1830 (suppliers first, then consumers), supplier boot state 1831 clean up (only after all consumers have probed), 1832 suspend/resume & runtime PM (consumers first, then 1833 suppliers). 1834 Format: { off | permissive | on | rpm } 1835 off -- Don't create device links from firmware info. 1836 permissive -- Create device links from firmware info 1837 but use it only for ordering boot state clean 1838 up (sync_state() calls). 1839 on -- Create device links from firmware info and use it 1840 to enforce probe and suspend/resume ordering. 1841 rpm -- Like "on", but also use to order runtime PM. 1842 1843 fw_devlink.strict=<bool> 1844 [KNL,EARLY] Treat all inferred dependencies as mandatory 1845 dependencies. This only applies for fw_devlink=on|rpm. 1846 Format: <bool> 1847 1848 fw_devlink.sync_state = 1849 [KNL,EARLY] When all devices that could probe have finished 1850 probing, this parameter controls what to do with 1851 devices that haven't yet received their sync_state() 1852 calls. 1853 Format: { strict | timeout } 1854 strict -- Default. Continue waiting on consumers to 1855 probe successfully. 1856 timeout -- Give up waiting on consumers and call 1857 sync_state() on any devices that haven't yet 1858 received their sync_state() calls after 1859 deferred_probe_timeout has expired or by 1860 late_initcall() if !CONFIG_MODULES. 1861 1862 gamecon.map[2|3]= 1863 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1864 support via parallel port (up to 5 devices per port) 1865 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1866 See also Documentation/input/devices/joystick-parport.rst 1867 1868 gamma= [HW,DRM] 1869 1870 gart_fix_e820= [X86-64,EARLY] disable the fix e820 for K8 GART 1871 Format: off | on 1872 default: on 1873 1874 gather_data_sampling= 1875 [X86,INTEL,EARLY] Control the Gather Data Sampling (GDS) 1876 mitigation. 1877 1878 Gather Data Sampling is a hardware vulnerability which 1879 allows unprivileged speculative access to data which was 1880 previously stored in vector registers. 1881 1882 This issue is mitigated by default in updated microcode. 1883 The mitigation may have a performance impact but can be 1884 disabled. On systems without the microcode mitigation 1885 disabling AVX serves as a mitigation. 1886 1887 force: Disable AVX to mitigate systems without 1888 microcode mitigation. No effect if the microcode 1889 mitigation is present. Known to cause crashes in 1890 userspace with buggy AVX enumeration. 1891 1892 off: Disable GDS mitigation. 1893 1894 gbpages [X86] Use GB pages for kernel direct mappings. 1895 1896 gcov_persist= [GCOV] When non-zero (default), profiling data for 1897 kernel modules is saved and remains accessible via 1898 debugfs, even when the module is unloaded/reloaded. 1899 When zero, profiling data is discarded and associated 1900 debugfs files are removed at module unload time. 1901 1902 goldfish [X86] Enable the goldfish android emulator platform. 1903 Don't use this when you are not running on the 1904 android emulator 1905 1906 gpio-mockup.gpio_mockup_ranges 1907 [HW] Sets the ranges of gpiochip of for this device. 1908 Format: <start1>,<end1>,<start2>,<end2>... 1909 gpio-mockup.gpio_mockup_named_lines 1910 [HW] Let the driver know GPIO lines should be named. 1911 1912 gpt [EFI] Forces disk with valid GPT signature but 1913 invalid Protective MBR to be treated as GPT. If the 1914 primary GPT is corrupted, it enables the backup/alternate 1915 GPT to be used instead. 1916 1917 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1918 the "Enable 0" bit of the configuration register. 1919 Format: 0 | 1 1920 Default: 0 1921 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1922 the "Enable 0" bit of the configuration register. 1923 Format: 0 | 1 1924 Default: 0 1925 grcan.select= [HW] Select which physical interface to use. 1926 Format: 0 | 1 1927 Default: 0 1928 grcan.txsize= [HW] Sets the size of the tx buffer. 1929 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1930 Default: 1024 1931 grcan.rxsize= [HW] Sets the size of the rx buffer. 1932 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1933 Default: 1024 1934 1935 hardened_usercopy= 1936 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 1937 hardening is enabled for this boot. Hardened 1938 usercopy checking is used to protect the kernel 1939 from reading or writing beyond known memory 1940 allocation boundaries as a proactive defense 1941 against bounds-checking flaws in the kernel's 1942 copy_to_user()/copy_from_user() interface. 1943 The default is determined by 1944 CONFIG_HARDENED_USERCOPY_DEFAULT_ON. 1945 on Perform hardened usercopy checks. 1946 off Disable hardened usercopy checks. 1947 1948 hardlockup_all_cpu_backtrace= 1949 [KNL] Should the hard-lockup detector generate 1950 backtraces on all cpus. 1951 Format: 0 | 1 1952 1953 hash_pointers= 1954 [KNL,EARLY] 1955 By default, when pointers are printed to the console 1956 or buffers via the %p format string, that pointer is 1957 "hashed", i.e. obscured by hashing the pointer value. 1958 This is a security feature that hides actual kernel 1959 addresses from unprivileged users, but it also makes 1960 debugging the kernel more difficult since unequal 1961 pointers can no longer be compared. The choices are: 1962 Format: { auto | always | never } 1963 Default: auto 1964 1965 auto - Hash pointers unless slab_debug is enabled. 1966 always - Always hash pointers (even if slab_debug is 1967 enabled). 1968 never - Never hash pointers. This option should only 1969 be specified when debugging the kernel. Do 1970 not use on production kernels. The boot 1971 param "no_hash_pointers" is an alias for 1972 this mode. 1973 1974 For controlling hashing dynamically at runtime, 1975 use the "kernel.kptr_restrict" sysctl instead. 1976 1977 hashdist= [KNL,NUMA] Large hashes allocated during boot 1978 are distributed across NUMA nodes. Defaults on 1979 for 64-bit NUMA, off otherwise. 1980 Format: 0 | 1 (for off | on) 1981 1982 hd= [EIDE] (E)IDE hard drive subsystem geometry 1983 Format: <cyl>,<head>,<sect> 1984 1985 hest_disable [ACPI] 1986 Disable Hardware Error Source Table (HEST) support; 1987 corresponding firmware-first mode error processing 1988 logic will be disabled. 1989 1990 hibernate= [HIBERNATION] 1991 noresume Don't check if there's a hibernation image 1992 present during boot. 1993 nocompress Don't compress/decompress hibernation images. 1994 no Disable hibernation and resume. 1995 protect_image Turn on image protection during restoration 1996 (that will set all pages holding image data 1997 during restoration read-only). 1998 1999 hibernate.compressor= [HIBERNATION] Compression algorithm to be 2000 used with hibernation. 2001 Format: { lzo | lz4 } 2002 Default: lzo 2003 2004 lzo: Select LZO compression algorithm to 2005 compress/decompress hibernation image. 2006 2007 lz4: Select LZ4 compression algorithm to 2008 compress/decompress hibernation image. 2009 2010 hibernate.pm_test_delay= 2011 [HIBERNATION] 2012 Sets the number of seconds to remain in a hibernation test 2013 mode before resuming the system (see 2014 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 2015 is set. Default value is 5. 2016 2017 hibernate_compression_threads= 2018 [HIBERNATION] 2019 Set the number of threads used for compressing or decompressing 2020 hibernation images. 2021 2022 Format: <integer> 2023 Default: 3 2024 Minimum: 1 2025 Example: hibernate_compression_threads=4 2026 2027 highmem=nn[KMG] [KNL,BOOT,EARLY] forces the highmem zone to have an exact 2028 size of <nn>. This works even on boxes that have no 2029 highmem otherwise. This also works to reduce highmem 2030 size on bigger boxes. 2031 2032 highres= [KNL] Enable/disable high resolution timer mode. 2033 Valid parameters: "on", "off" 2034 Default: "on" 2035 2036 hlt [BUGS=ARM,SH] 2037 2038 hostname= [KNL,EARLY] Set the hostname (aka UTS nodename). 2039 Format: <string> 2040 This allows setting the system's hostname during early 2041 startup. This sets the name returned by gethostname. 2042 Using this parameter to set the hostname makes it 2043 possible to ensure the hostname is correctly set before 2044 any userspace processes run, avoiding the possibility 2045 that a process may call gethostname before the hostname 2046 has been explicitly set, resulting in the calling 2047 process getting an incorrect result. The string must 2048 not exceed the maximum allowed hostname length (usually 2049 64 characters) and will be truncated otherwise. 2050 2051 hpet= [X86-32,HPET] option to control HPET usage 2052 Format: { enable (default) | disable | force | 2053 verbose } 2054 disable: disable HPET and use PIT instead 2055 force: allow force enabled of undocumented chips (ICH4, 2056 VIA, nVidia) 2057 verbose: show contents of HPET registers during setup 2058 2059 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 2060 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 2061 2062 hugepages= [HW,EARLY] Number of HugeTLB pages to allocate at boot. 2063 If this follows hugepagesz (below), it specifies 2064 the number of pages of hugepagesz to be allocated. 2065 If this is the first HugeTLB parameter on the command 2066 line, it specifies the number of pages to allocate for 2067 the default huge page size. If using node format, the 2068 number of pages to allocate per-node can be specified. 2069 See also Documentation/admin-guide/mm/hugetlbpage.rst. 2070 Format: <integer> or (node format) 2071 <node>:<integer>[,<node>:<integer>] 2072 2073 hugepagesz= 2074 [HW,EARLY] The size of the HugeTLB pages. This is 2075 used in conjunction with hugepages (above) to 2076 allocate huge pages of a specific size at boot. The 2077 pair hugepagesz=X hugepages=Y can be specified once 2078 for each supported huge page size. Huge page sizes 2079 are architecture dependent. See also 2080 Documentation/admin-guide/mm/hugetlbpage.rst. 2081 Format: size[KMG] 2082 2083 hugepage_alloc_threads= 2084 [HW] The number of threads that should be used to 2085 allocate hugepages during boot. This option can be 2086 used to improve system bootup time when allocating 2087 a large amount of huge pages. 2088 The default value is 25% of the available hardware threads. 2089 2090 Note that this parameter only applies to non-gigantic huge pages. 2091 2092 hugetlb_cma= [HW,CMA,EARLY] The size of a CMA area used for allocation 2093 of gigantic hugepages. Or using node format, the size 2094 of a CMA area per node can be specified. 2095 Format: nn[KMGTPE] or (node format) 2096 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]] 2097 2098 Reserve a CMA area of given size and allocate gigantic 2099 hugepages using the CMA allocator. If enabled, the 2100 boot-time allocation of gigantic hugepages is skipped. 2101 2102 hugetlb_cma_only= 2103 [HW,CMA,EARLY] When allocating new HugeTLB pages, only 2104 try to allocate from the CMA areas. 2105 2106 This option does nothing if hugetlb_cma= is not also 2107 specified. 2108 2109 hugetlb_free_vmemmap= 2110 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP 2111 enabled. 2112 Control if HugeTLB Vmemmap Optimization (HVO) is enabled. 2113 Allows heavy hugetlb users to free up some more 2114 memory (7 * PAGE_SIZE for each 2MB hugetlb page). 2115 Format: { on | off (default) } 2116 2117 on: enable HVO 2118 off: disable HVO 2119 2120 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y, 2121 the default is on. 2122 2123 Note that the vmemmap pages may be allocated from the added 2124 memory block itself when memory_hotplug.memmap_on_memory is 2125 enabled, those vmemmap pages cannot be optimized even if this 2126 feature is enabled. Other vmemmap pages not allocated from 2127 the added memory block itself do not be affected. 2128 2129 hung_task_panic= 2130 [KNL] Number of hung tasks to trigger kernel panic. 2131 Format: <int> 2132 2133 When set to a non-zero value, a kernel panic will be triggered if 2134 the number of detected hung tasks reaches this value. 2135 2136 0: don't panic 2137 1: panic immediately on first hung task 2138 N: panic after N hung tasks are detected in a single scan 2139 2140 The default value is controlled by the 2141 CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time option. The value 2142 selected by this boot parameter can be changed later by the 2143 kernel.hung_task_panic sysctl. 2144 2145 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 2146 terminal devices. Valid values: 0..8 2147 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 2148 If specified, z/VM IUCV HVC accepts connections 2149 from listed z/VM user IDs only. 2150 2151 hv_nopvspin [X86,HYPER_V,EARLY] 2152 Disables the paravirt spinlock optimizations 2153 which allow the hypervisor to 'idle' the guest 2154 on lock contention. 2155 2156 hw_protection= [HW] 2157 Format: reboot | shutdown 2158 2159 Hardware protection action taken on critical events like 2160 overtemperature or imminent voltage loss. 2161 2162 i2c_bus= [HW] Override the default board specific I2C bus speed 2163 or register an additional I2C bus that is not 2164 registered from board initialization code. 2165 Format: 2166 <bus_id>,<clkrate> 2167 2168 i2c_touchscreen_props= [HW,ACPI,X86] 2169 Set device-properties for ACPI-enumerated I2C-attached 2170 touchscreen, to e.g. fix coordinates of upside-down 2171 mounted touchscreens. If you need this option please 2172 submit a drivers/platform/x86/touchscreen_dmi.c patch 2173 adding a DMI quirk for this. 2174 2175 Format: 2176 <ACPI_HW_ID>:<prop_name>=<val>[:prop_name=val][:...] 2177 Where <val> is one of: 2178 Omit "=<val>" entirely Set a boolean device-property 2179 Unsigned number Set a u32 device-property 2180 Anything else Set a string device-property 2181 2182 Examples (split over multiple lines): 2183 i2c_touchscreen_props=GDIX1001:touchscreen-inverted-x: 2184 touchscreen-inverted-y 2185 2186 i2c_touchscreen_props=MSSL1680:touchscreen-size-x=1920: 2187 touchscreen-size-y=1080:touchscreen-inverted-y: 2188 firmware-name=gsl1680-vendor-model.fw:silead,home-button 2189 2190 i8042.debug [HW] Toggle i8042 debug mode 2191 i8042.unmask_kbd_data 2192 [HW] Enable printing of interrupt data from the KBD port 2193 (disabled by default, and as a pre-condition 2194 requires that i8042.debug=1 be enabled) 2195 i8042.direct [HW] Put keyboard port into non-translated mode 2196 i8042.dumbkbd [HW] Pretend that controller can only read data from 2197 keyboard and cannot control its state 2198 (Don't attempt to blink the leds) 2199 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 2200 i8042.nokbd [HW] Don't check/create keyboard port 2201 i8042.noloop [HW] Disable the AUX Loopback command while probing 2202 for the AUX port 2203 i8042.nomux [HW] Don't check presence of an active multiplexing 2204 controller 2205 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 2206 controllers 2207 i8042.notimeout [HW] Ignore timeout condition signalled by controller 2208 i8042.reset [HW] Reset the controller during init, cleanup and 2209 suspend-to-ram transitions, only during s2r 2210 transitions, or never reset 2211 Format: { 1 | Y | y | 0 | N | n } 2212 1, Y, y: always reset controller 2213 0, N, n: don't ever reset controller 2214 Default: only on s2r transitions on x86; most other 2215 architectures force reset to be always executed 2216 i8042.unlock [HW] Unlock (ignore) the keylock 2217 i8042.kbdreset [HW] Reset device connected to KBD port 2218 i8042.probe_defer 2219 [HW] Allow deferred probing upon i8042 probe errors 2220 2221 i810= [HW,DRM] 2222 2223 i915.invert_brightness= 2224 [DRM] Invert the sense of the variable that is used to 2225 set the brightness of the panel backlight. Normally a 2226 brightness value of 0 indicates backlight switched off, 2227 and the maximum of the brightness value sets the backlight 2228 to maximum brightness. If this parameter is set to 0 2229 (default) and the machine requires it, or this parameter 2230 is set to 1, a brightness value of 0 sets the backlight 2231 to maximum brightness, and the maximum of the brightness 2232 value switches the backlight off. 2233 -1 -- never invert brightness 2234 0 -- machine default 2235 1 -- force brightness inversion 2236 2237 ia32_emulation= [X86-64] 2238 Format: <bool> 2239 When true, allows loading 32-bit programs and executing 32-bit 2240 syscalls, essentially overriding IA32_EMULATION_DEFAULT_DISABLED at 2241 boot time. When false, unconditionally disables IA32 emulation. 2242 2243 icn= [HW,ISDN] 2244 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 2245 2246 2247 idle= [X86,EARLY] 2248 Format: idle=poll, idle=halt, idle=nomwait 2249 2250 idle=poll: Don't do power saving in the idle loop 2251 using HLT, but poll for rescheduling event. This will 2252 make the CPUs eat a lot more power, but may be useful 2253 to get slightly better performance in multiprocessor 2254 benchmarks. It also makes some profiling using 2255 performance counters more accurate. Please note that 2256 on systems with MONITOR/MWAIT support (like Intel 2257 EM64T CPUs) this option has no performance advantage 2258 over the normal idle loop. It may also interact badly 2259 with hyperthreading. 2260 2261 idle=halt: Halt is forced to be used for CPU idle. 2262 In such case C2/C3 won't be used again. 2263 2264 idle=nomwait: Disable mwait for CPU C-states 2265 2266 idxd.sva= [HW] 2267 Format: <bool> 2268 Allow force disabling of Shared Virtual Memory (SVA) 2269 support for the idxd driver. By default it is set to 2270 true (1). 2271 2272 idxd.tc_override= [HW] 2273 Format: <bool> 2274 Allow override of default traffic class configuration 2275 for the device. By default it is set to false (0). 2276 2277 ieee754= [MIPS] Select IEEE Std 754 conformance mode 2278 Format: { strict | legacy | 2008 | relaxed | emulated } 2279 Default: strict 2280 2281 Choose which programs will be accepted for execution 2282 based on the IEEE 754 NaN encoding(s) supported by 2283 the FPU and the NaN encoding requested with the value 2284 of an ELF file header flag individually set by each 2285 binary. Hardware implementations are permitted to 2286 support either or both of the legacy and the 2008 NaN 2287 encoding mode. 2288 2289 Available settings are as follows: 2290 strict accept binaries that request a NaN encoding 2291 supported by the FPU 2292 legacy only accept legacy-NaN binaries, if supported 2293 by the FPU 2294 2008 only accept 2008-NaN binaries, if supported 2295 by the FPU 2296 relaxed accept any binaries regardless of whether 2297 supported by the FPU 2298 emulated accept any binaries but enable FPU emulator 2299 if binary mode is unsupported by the FPU. 2300 2301 The FPU emulator is always able to support both NaN 2302 encodings, so if no FPU hardware is present or it has 2303 been disabled with 'nofpu', then the settings of 2304 'legacy' and '2008' strap the emulator accordingly, 2305 'relaxed' straps the emulator for both legacy-NaN and 2306 2008-NaN, whereas 'strict' enables legacy-NaN only on 2307 legacy processors and both NaN encodings on MIPS32 or 2308 MIPS64 CPUs. 2309 2310 The setting for ABS.fmt/NEG.fmt instruction execution 2311 mode generally follows that for the NaN encoding, 2312 except where unsupported by hardware. 2313 2314 ignore_loglevel [KNL,EARLY] 2315 Ignore loglevel setting - this will print /all/ 2316 kernel messages to the console. Useful for debugging. 2317 We also add it as printk module parameter, so users 2318 could change it dynamically, usually by 2319 /sys/module/printk/parameters/ignore_loglevel. 2320 2321 ignore_rlimit_data 2322 Ignore RLIMIT_DATA setting for data mappings, 2323 print warning at first misuse. Can be changed via 2324 /sys/module/kernel/parameters/ignore_rlimit_data. 2325 2326 ihash_entries= [KNL] 2327 Set number of hash buckets for inode cache. 2328 2329 ima_appraise= [IMA] appraise integrity measurements 2330 Format: { "off" | "enforce" | "fix" | "log" } 2331 default: "enforce" 2332 2333 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 2334 The builtin appraise policy appraises all files 2335 owned by uid=0. 2336 2337 ima_canonical_fmt [IMA] 2338 Use the canonical format for the binary runtime 2339 measurements, instead of host native format. 2340 2341 ima_hash= [IMA] 2342 Format: { md5 | sha1 | rmd160 | sha256 | sha384 2343 | sha512 | ... } 2344 default: "sha1" 2345 2346 The list of supported hash algorithms is defined 2347 in crypto/hash_info.h. 2348 2349 ima_policy= [IMA] 2350 The builtin policies to load during IMA setup. 2351 Format: "tcb | appraise_tcb | secure_boot | 2352 fail_securely | critical_data" 2353 2354 The "tcb" policy measures all programs exec'd, files 2355 mmap'd for exec, and all files opened with the read 2356 mode bit set by either the effective uid (euid=0) or 2357 uid=0. 2358 2359 The "appraise_tcb" policy appraises the integrity of 2360 all files owned by root. 2361 2362 The "secure_boot" policy appraises the integrity 2363 of files (eg. kexec kernel image, kernel modules, 2364 firmware, policy, etc) based on file signatures. 2365 2366 The "fail_securely" policy forces file signature 2367 verification failure also on privileged mounted 2368 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 2369 flag. 2370 2371 The "critical_data" policy measures kernel integrity 2372 critical data. 2373 2374 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 2375 Load a policy which meets the needs of the Trusted 2376 Computing Base. This means IMA will measure all 2377 programs exec'd, files mmap'd for exec, and all files 2378 opened for read by uid=0. 2379 2380 ima_template= [IMA] 2381 Select one of defined IMA measurements template formats. 2382 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" | 2383 "ima-sigv2" } 2384 Default: "ima-ng" 2385 2386 ima_template_fmt= 2387 [IMA] Define a custom template format. 2388 Format: { "field1|...|fieldN" } 2389 2390 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 2391 Format: <min_file_size> 2392 Set the minimal file size for using asynchronous hash. 2393 If left unspecified, ahash usage is disabled. 2394 2395 ahash performance varies for different data sizes on 2396 different crypto accelerators. This option can be used 2397 to achieve the best performance for a particular HW. 2398 2399 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size 2400 Format: <bufsize> 2401 Set hashing buffer size. Default: 4k. 2402 2403 ahash performance varies for different chunk sizes on 2404 different crypto accelerators. This option can be used 2405 to achieve best performance for particular HW. 2406 2407 ima= [IMA] Enable or disable IMA 2408 Format: { "off" | "on" } 2409 Default: "on" 2410 Note that disabling IMA is limited to kdump kernel. 2411 2412 indirect_target_selection= [X86,Intel] Mitigation control for Indirect 2413 Target Selection(ITS) bug in Intel CPUs. Updated 2414 microcode is also required for a fix in IBPB. 2415 2416 on: Enable mitigation (default). 2417 off: Disable mitigation. 2418 force: Force the ITS bug and deploy default 2419 mitigation. 2420 vmexit: Only deploy mitigation if CPU is affected by 2421 guest/host isolation part of ITS. 2422 stuff: Deploy RSB-fill mitigation when retpoline is 2423 also deployed. Otherwise, deploy the default 2424 mitigation. 2425 2426 For details see: 2427 Documentation/admin-guide/hw-vuln/indirect-target-selection.rst 2428 2429 init= [KNL] 2430 Format: <full_path> 2431 Run specified binary instead of /sbin/init as init 2432 process. 2433 2434 initcall_debug [KNL] Trace initcalls as they are executed. Useful 2435 for working out where the kernel is dying during 2436 startup. 2437 2438 initcall_blacklist= [KNL] Do not execute a comma-separated list of 2439 initcall functions. Useful for debugging built-in 2440 modules and initcalls. 2441 2442 initramfs_async= [KNL] 2443 Format: <bool> 2444 Default: 1 2445 This parameter controls whether the initramfs 2446 image is unpacked asynchronously, concurrently 2447 with devices being probed and 2448 initialized. This should normally just work, 2449 but as a debugging aid, one can get the 2450 historical behaviour of the initramfs 2451 unpacking being completed before device_ and 2452 late_ initcalls. 2453 2454 initrd= [BOOT,EARLY] Specify the location of the initial ramdisk 2455 2456 initrdmem= [KNL,EARLY] Specify a physical address and size from which to 2457 load the initrd. If an initrd is compiled in or 2458 specified in the bootparams, it takes priority over this 2459 setting. 2460 Format: ss[KMG],nn[KMG] 2461 Default is 0, 0 2462 2463 init_on_alloc= [MM,EARLY] Fill newly allocated pages and heap objects with 2464 zeroes. 2465 Format: 0 | 1 2466 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 2467 2468 init_on_free= [MM,EARLY] Fill freed pages and heap objects with zeroes. 2469 Format: 0 | 1 2470 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 2471 2472 init_pkru= [X86] Specify the default memory protection keys rights 2473 register contents for all processes. 0x55555554 by 2474 default (disallow access to all but pkey 0). Can 2475 override in debugfs after boot. 2476 2477 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 2478 Format: <irq> 2479 2480 int_pln_enable [X86] Enable power limit notification interrupt 2481 2482 integrity_audit=[IMA] 2483 Format: { "0" | "1" } 2484 0 -- basic integrity auditing messages. (Default) 2485 1 -- additional integrity auditing messages. 2486 2487 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 2488 on 2489 Enable intel iommu driver. 2490 off 2491 Disable intel iommu driver. 2492 igfx_off [Default Off] 2493 By default, gfx is mapped as normal device. If a gfx 2494 device has a dedicated DMAR unit, the DMAR unit is 2495 bypassed by not enabling DMAR with this option. In 2496 this case, gfx device will use physical address for 2497 DMA. 2498 strict [Default Off] 2499 Deprecated, equivalent to iommu.strict=1. 2500 sp_off [Default Off] 2501 By default, super page will be supported if Intel IOMMU 2502 has the capability. With this option, super page will 2503 not be supported. 2504 sm_on 2505 Enable the Intel IOMMU scalable mode if the hardware 2506 advertises that it has support for the scalable mode 2507 translation. 2508 sm_off 2509 Disallow use of the Intel IOMMU scalable mode. 2510 tboot_noforce [Default Off] 2511 Do not force the Intel IOMMU enabled under tboot. 2512 By default, tboot will force Intel IOMMU on, which 2513 could harm performance of some high-throughput 2514 devices like 40GBit network cards, even if identity 2515 mapping is enabled. 2516 Note that using this option lowers the security 2517 provided by tboot because it makes the system 2518 vulnerable to DMA attacks. 2519 2520 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 2521 0 disables intel_idle and fall back on acpi_idle. 2522 1 to 9 specify maximum depth of C-state. 2523 2524 intel_pstate= [X86,EARLY] 2525 disable 2526 Do not enable intel_pstate as the default 2527 scaling driver for the supported processors 2528 active 2529 Use intel_pstate driver to bypass the scaling 2530 governors layer of cpufreq and provides it own 2531 algorithms for p-state selection. There are two 2532 P-state selection algorithms provided by 2533 intel_pstate in the active mode: powersave and 2534 performance. The way they both operate depends 2535 on whether or not the hardware managed P-states 2536 (HWP) feature has been enabled in the processor 2537 and possibly on the processor model. 2538 passive 2539 Use intel_pstate as a scaling driver, but configure it 2540 to work with generic cpufreq governors (instead of 2541 enabling its internal governor). This mode cannot be 2542 used along with the hardware-managed P-states (HWP) 2543 feature. 2544 force 2545 Enable intel_pstate on systems that prohibit it by default 2546 in favor of acpi-cpufreq. Forcing the intel_pstate driver 2547 instead of acpi-cpufreq may disable platform features, such 2548 as thermal controls and power capping, that rely on ACPI 2549 P-States information being indicated to OSPM and therefore 2550 should be used with caution. This option does not work with 2551 processors that aren't supported by the intel_pstate driver 2552 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 2553 no_hwp 2554 Do not enable hardware P state control (HWP) 2555 if available. 2556 hwp_only 2557 Only load intel_pstate on systems which support 2558 hardware P state control (HWP) if available. 2559 support_acpi_ppc 2560 Enforce ACPI _PPC performance limits. If the Fixed ACPI 2561 Description Table, specifies preferred power management 2562 profile as "Enterprise Server" or "Performance Server", 2563 then this feature is turned on by default. 2564 per_cpu_perf_limits 2565 Allow per-logical-CPU P-State performance control limits using 2566 cpufreq sysfs interface 2567 no_cas 2568 Do not enable capacity-aware scheduling (CAS) on 2569 hybrid systems 2570 2571 intremap= [X86-64,Intel-IOMMU,EARLY] 2572 on enable Interrupt Remapping (default) 2573 off disable Interrupt Remapping 2574 nosid disable Source ID checking 2575 no_x2apic_optout 2576 BIOS x2APIC opt-out request will be ignored 2577 nopost disable Interrupt Posting 2578 posted_msi 2579 enable MSIs delivered as posted interrupts 2580 2581 iomem= Disable strict checking of access to MMIO memory 2582 strict regions from userspace. 2583 relaxed 2584 2585 iommu= [X86,EARLY] 2586 2587 off 2588 Don't initialize and use any kind of IOMMU. 2589 2590 force 2591 Force the use of the hardware IOMMU even when 2592 it is not actually needed (e.g. because < 3 GB 2593 memory). 2594 2595 noforce 2596 Don't force hardware IOMMU usage when it is not 2597 needed. (default). 2598 2599 biomerge 2600 panic 2601 nopanic 2602 merge 2603 nomerge 2604 2605 soft 2606 Use software bounce buffering (SWIOTLB) (default for 2607 Intel machines). This can be used to prevent the usage 2608 of an available hardware IOMMU. 2609 2610 [X86] 2611 pt 2612 [X86] 2613 nopt 2614 [PPC/POWERNV] 2615 nobypass 2616 Disable IOMMU bypass, using IOMMU for PCI devices. 2617 2618 [X86] 2619 AMD Gart HW IOMMU-specific options: 2620 2621 <size> 2622 Set the size of the remapping area in bytes. 2623 2624 allowed 2625 Overwrite iommu off workarounds for specific chipsets 2626 2627 fullflush 2628 Flush IOMMU on each allocation (default). 2629 2630 nofullflush 2631 Don't use IOMMU fullflush. 2632 2633 memaper[=<order>] 2634 Allocate an own aperture over RAM with size 2635 32MB<<order. (default: order=1, i.e. 64MB) 2636 2637 merge 2638 Do scatter-gather (SG) merging. Implies "force" 2639 (experimental). 2640 2641 nomerge 2642 Don't do scatter-gather (SG) merging. 2643 2644 noaperture 2645 Ask the IOMMU not to touch the aperture for AGP. 2646 2647 noagp 2648 Don't initialize the AGP driver and use full aperture. 2649 2650 panic 2651 Always panic when IOMMU overflows. 2652 2653 iommu.forcedac= [ARM64,X86,EARLY] Control IOVA allocation for PCI devices. 2654 Format: { "0" | "1" } 2655 0 - Try to allocate a 32-bit DMA address first, before 2656 falling back to the full range if needed. 2657 1 - Allocate directly from the full usable range, 2658 forcing Dual Address Cycle for PCI cards supporting 2659 greater than 32-bit addressing. 2660 2661 iommu.strict= [ARM64,X86,S390,EARLY] Configure TLB invalidation behaviour 2662 Format: { "0" | "1" } 2663 0 - Lazy mode. 2664 Request that DMA unmap operations use deferred 2665 invalidation of hardware TLBs, for increased 2666 throughput at the cost of reduced device isolation. 2667 Will fall back to strict mode if not supported by 2668 the relevant IOMMU driver. 2669 1 - Strict mode. 2670 DMA unmap operations invalidate IOMMU hardware TLBs 2671 synchronously. 2672 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}. 2673 Note: on x86, strict mode specified via one of the 2674 legacy driver-specific options takes precedence. 2675 2676 iommu.passthrough= 2677 [ARM64,X86,EARLY] Configure DMA to bypass the IOMMU by default. 2678 Format: { "0" | "1" } 2679 0 - Use IOMMU translation for DMA. 2680 1 - Bypass the IOMMU for DMA. 2681 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2682 2683 io7= [HW] IO7 for Marvel-based Alpha systems 2684 See comment before marvel_specify_io7 in 2685 arch/alpha/kernel/core_marvel.c. 2686 2687 io_delay= [X86,EARLY] I/O delay method 2688 0x80 2689 Standard port 0x80 based delay 2690 0xed 2691 Alternate port 0xed based delay (needed on some systems) 2692 udelay 2693 Simple two microseconds delay 2694 none 2695 No delay 2696 2697 ip= [IP_PNP] 2698 See Documentation/admin-guide/nfs/nfsroot.rst. 2699 2700 ipcmni_extend [KNL,EARLY] Extend the maximum number of unique System V 2701 IPC identifiers from 32,768 to 16,777,216. 2702 2703 ipe.enforce= [IPE] 2704 Format: <bool> 2705 Determine whether IPE starts in permissive (0) or 2706 enforce (1) mode. The default is enforce. 2707 2708 ipe.success_audit= 2709 [IPE] 2710 Format: <bool> 2711 Start IPE with success auditing enabled, emitting 2712 an audit event when a binary is allowed. The default 2713 is 0. 2714 2715 irqaffinity= [SMP] Set the default irq affinity mask 2716 The argument is a cpu list, as described above. 2717 2718 irqchip.gicv2_force_probe= 2719 [ARM,ARM64,EARLY] 2720 Format: <bool> 2721 Force the kernel to look for the second 4kB page 2722 of a GICv2 controller even if the memory range 2723 exposed by the device tree is too small. 2724 2725 irqchip.gicv3_nolpi= 2726 [ARM,ARM64,EARLY] 2727 Force the kernel to ignore the availability of 2728 LPIs (and by consequence ITSs). Intended for system 2729 that use the kernel as a bootloader, and thus want 2730 to let secondary kernels in charge of setting up 2731 LPIs. 2732 2733 irqchip.gicv3_pseudo_nmi= [ARM64,EARLY] 2734 Enables support for pseudo-NMIs in the kernel. This 2735 requires the kernel to be built with 2736 CONFIG_ARM64_PSEUDO_NMI. 2737 2738 irqchip.riscv_imsic_noipi 2739 [RISC-V,EARLY] 2740 Force the kernel to not use IMSIC software injected MSIs 2741 as IPIs. Intended for system where IMSIC is trap-n-emulated, 2742 and thus want to reduce MMIO traps when triggering IPIs 2743 to multiple harts. 2744 2745 irqfixup [HW] 2746 When an interrupt is not handled search all handlers 2747 for it. Intended to get systems with badly broken 2748 firmware running. 2749 2750 irqhandler.duration_warn_us= [KNL] 2751 Warn if an IRQ handler exceeds the specified duration 2752 threshold in microseconds. Useful for identifying 2753 long-running IRQs in the system. 2754 2755 irqpoll [HW] 2756 When an interrupt is not handled search all handlers 2757 for it. Also check all handlers each timer 2758 interrupt. Intended to get systems with badly broken 2759 firmware running. 2760 2761 isapnp= [ISAPNP] 2762 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2763 2764 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2765 [Deprecated - use cpusets instead] 2766 Format: [flag-list,]<cpu-list> 2767 2768 Specify one or more CPUs to isolate from disturbances 2769 specified in the flag list (default: domain): 2770 2771 nohz 2772 Disable the tick when a single task runs as well as 2773 disabling other kernel noises like having RCU callbacks 2774 offloaded. This is equivalent to the nohz_full parameter. 2775 2776 A residual 1Hz tick is offloaded to workqueues, which you 2777 need to affine to housekeeping through the global 2778 workqueue's affinity configured via the 2779 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2780 by using the 'domain' flag described below. 2781 2782 NOTE: by default the global workqueue runs on all CPUs, 2783 so to protect individual CPUs the 'cpumask' file has to 2784 be configured manually after bootup. 2785 2786 domain 2787 Isolate from the general SMP balancing and scheduling 2788 algorithms. Note that performing domain isolation this way 2789 is irreversible: it's not possible to bring back a CPU to 2790 the domains once isolated through isolcpus. It's strongly 2791 advised to use cpusets instead to disable scheduler load 2792 balancing through the "cpuset.sched_load_balance" file. 2793 It offers a much more flexible interface where CPUs can 2794 move in and out of an isolated set anytime. 2795 2796 You can move a process onto or off an "isolated" CPU via 2797 the CPU affinity syscalls or cpuset. 2798 <cpu number> begins at 0 and the maximum value is 2799 "number of CPUs in system - 1". 2800 2801 managed_irq 2802 2803 Isolate from being targeted by managed interrupts 2804 which have an interrupt mask containing isolated 2805 CPUs. The affinity of managed interrupts is 2806 handled by the kernel and cannot be changed via 2807 the /proc/irq/* interfaces. 2808 2809 This isolation is best effort and only effective 2810 if the automatically assigned interrupt mask of a 2811 device queue contains isolated and housekeeping 2812 CPUs. If housekeeping CPUs are online then such 2813 interrupts are directed to the housekeeping CPU 2814 so that IO submitted on the housekeeping CPU 2815 cannot disturb the isolated CPU. 2816 2817 If a queue's affinity mask contains only isolated 2818 CPUs then this parameter has no effect on the 2819 interrupt routing decision, though interrupts are 2820 only delivered when tasks running on those 2821 isolated CPUs submit IO. IO submitted on 2822 housekeeping CPUs has no influence on those 2823 queues. 2824 2825 The format of <cpu-list> is described above. 2826 2827 iucv= [HW,NET] 2828 2829 ivrs_ioapic [HW,X86-64] 2830 Provide an override to the IOAPIC-ID<->DEVICE-ID 2831 mapping provided in the IVRS ACPI table. 2832 By default, PCI segment is 0, and can be omitted. 2833 2834 For example, to map IOAPIC-ID decimal 10 to 2835 PCI segment 0x1 and PCI device 00:14.0, 2836 write the parameter as: 2837 ivrs_ioapic=10@0001:00:14.0 2838 2839 Deprecated formats: 2840 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0 2841 write the parameter as: 2842 ivrs_ioapic[10]=00:14.0 2843 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and 2844 PCI device 00:14.0 write the parameter as: 2845 ivrs_ioapic[10]=0001:00:14.0 2846 2847 ivrs_hpet [HW,X86-64] 2848 Provide an override to the HPET-ID<->DEVICE-ID 2849 mapping provided in the IVRS ACPI table. 2850 By default, PCI segment is 0, and can be omitted. 2851 2852 For example, to map HPET-ID decimal 10 to 2853 PCI segment 0x1 and PCI device 00:14.0, 2854 write the parameter as: 2855 ivrs_hpet=10@0001:00:14.0 2856 2857 Deprecated formats: 2858 * To map HPET-ID decimal 0 to PCI device 00:14.0 2859 write the parameter as: 2860 ivrs_hpet[0]=00:14.0 2861 * To map HPET-ID decimal 10 to PCI segment 0x1 and 2862 PCI device 00:14.0 write the parameter as: 2863 ivrs_ioapic[10]=0001:00:14.0 2864 2865 ivrs_acpihid [HW,X86-64] 2866 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2867 mapping provided in the IVRS ACPI table. 2868 By default, PCI segment is 0, and can be omitted. 2869 2870 For example, to map UART-HID:UID AMD0020:0 to 2871 PCI segment 0x1 and PCI device ID 00:14.5, 2872 write the parameter as: 2873 ivrs_acpihid=AMD0020:0@0001:00:14.5 2874 2875 Deprecated formats: 2876 * To map UART-HID:UID AMD0020:0 to PCI segment is 0, 2877 PCI device ID 00:14.5, write the parameter as: 2878 ivrs_acpihid[00:14.5]=AMD0020:0 2879 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and 2880 PCI device ID 00:14.5, write the parameter as: 2881 ivrs_acpihid[0001:00:14.5]=AMD0020:0 2882 2883 js= [HW,JOY] Analog joystick 2884 See Documentation/input/joydev/joystick.rst. 2885 2886 kasan_multi_shot 2887 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2888 report on every invalid memory access. Without this 2889 parameter KASAN will print report only for the first 2890 invalid access. 2891 2892 keep_bootcon [KNL,EARLY] 2893 Do not unregister boot console at start. This is only 2894 useful for debugging when something happens in the window 2895 between unregistering the boot console and initializing 2896 the real console. 2897 2898 keepinitrd [HW,ARM] See retain_initrd. 2899 2900 kernelcore= [KNL,X86,PPC,EARLY] 2901 Format: nn[KMGTPE] | nn% | "mirror" 2902 This parameter specifies the amount of memory usable by 2903 the kernel for non-movable allocations. The requested 2904 amount is spread evenly throughout all nodes in the 2905 system as ZONE_NORMAL. The remaining memory is used for 2906 movable memory in its own zone, ZONE_MOVABLE. In the 2907 event, a node is too small to have both ZONE_NORMAL and 2908 ZONE_MOVABLE, kernelcore memory will take priority and 2909 other nodes will have a larger ZONE_MOVABLE. 2910 2911 ZONE_MOVABLE is used for the allocation of pages that 2912 may be reclaimed or moved by the page migration 2913 subsystem. Note that allocations like PTEs-from-HighMem 2914 still use the HighMem zone if it exists, and the Normal 2915 zone if it does not. 2916 2917 It is possible to specify the exact amount of memory in 2918 the form of "nn[KMGTPE]", a percentage of total system 2919 memory in the form of "nn%", or "mirror". If "mirror" 2920 option is specified, mirrored (reliable) memory is used 2921 for non-movable allocations and remaining memory is used 2922 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2923 are exclusive, so you cannot specify multiple forms. 2924 2925 kfence.burst= [MM,KFENCE] The number of additional successive 2926 allocations to be attempted through KFENCE for each 2927 sample interval. 2928 Format: <unsigned integer> 2929 Default: 0 2930 2931 kfence.check_on_panic= 2932 [MM,KFENCE] Whether to check all KFENCE-managed objects' 2933 canaries on panic. 2934 Format: <bool> 2935 Default: false 2936 2937 kfence.deferrable= 2938 [MM,KFENCE] Whether to use a deferrable timer to trigger 2939 allocations. This avoids forcing CPU wake-ups if the 2940 system is idle, at the risk of a less predictable 2941 sample interval. 2942 Format: <bool> 2943 Default: CONFIG_KFENCE_DEFERRABLE 2944 2945 kfence.sample_interval= 2946 [MM,KFENCE] KFENCE's sample interval in milliseconds. 2947 Format: <unsigned integer> 2948 0 - Disable KFENCE. 2949 >0 - Enabled KFENCE with given sample interval. 2950 Default: CONFIG_KFENCE_SAMPLE_INTERVAL 2951 2952 kfence.skip_covered_thresh= 2953 [MM,KFENCE] If pool utilization reaches this threshold 2954 (pool usage%), KFENCE limits currently covered 2955 allocations of the same source from further filling 2956 up the pool. 2957 Format: <unsigned integer> 2958 Default: 75 2959 2960 kgdbdbgp= [KGDB,HW,EARLY] kgdb over EHCI usb debug port. 2961 Format: <Controller#>[,poll interval] 2962 The controller # is the number of the ehci usb debug 2963 port as it is probed via PCI. The poll interval is 2964 optional and is the number seconds in between 2965 each poll cycle to the debug port in case you need 2966 the functionality for interrupting the kernel with 2967 gdb or control-c on the dbgp connection. When 2968 not using this parameter you use sysrq-g to break into 2969 the kernel debugger. 2970 2971 kgdboc= [KGDB,HW] kgdb over consoles. 2972 Requires a tty driver that supports console polling, 2973 or a supported polling keyboard driver (non-usb). 2974 Serial only format: <serial_device>[,baud] 2975 keyboard only format: kbd 2976 keyboard and serial format: kbd,<serial_device>[,baud] 2977 Optional Kernel mode setting: 2978 kms, kbd format: kms,kbd 2979 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2980 2981 kgdboc_earlycon= [KGDB,HW,EARLY] 2982 If the boot console provides the ability to read 2983 characters and can work in polling mode, you can use 2984 this parameter to tell kgdb to use it as a backend 2985 until the normal console is registered. Intended to 2986 be used together with the kgdboc parameter which 2987 specifies the normal console to transition to. 2988 2989 The name of the early console should be specified 2990 as the value of this parameter. Note that the name of 2991 the early console might be different than the tty 2992 name passed to kgdboc. It's OK to leave the value 2993 blank and the first boot console that implements 2994 read() will be picked. 2995 2996 kgdbwait [KGDB,EARLY] Stop kernel execution and enter the 2997 kernel debugger at the earliest opportunity. 2998 2999 kho= [KEXEC,EARLY] 3000 Format: { "0" | "1" | "off" | "on" | "y" | "n" } 3001 Enables or disables Kexec HandOver. 3002 "0" | "off" | "n" - kexec handover is disabled 3003 "1" | "on" | "y" - kexec handover is enabled 3004 3005 kho_scratch= [KEXEC,EARLY] 3006 Format: ll[KMG],mm[KMG],nn[KMG] | nn% 3007 Defines the size of the KHO scratch region. The KHO 3008 scratch regions are physically contiguous memory 3009 ranges that can only be used for non-kernel 3010 allocations. That way, even when memory is heavily 3011 fragmented with handed over memory, the kexeced 3012 kernel will always have enough contiguous ranges to 3013 bootstrap itself. 3014 3015 It is possible to specify the exact amount of 3016 memory in the form of "ll[KMG],mm[KMG],nn[KMG]" 3017 where the first parameter defines the size of a low 3018 memory scratch area, the second parameter defines 3019 the size of a global scratch area and the third 3020 parameter defines the size of additional per-node 3021 scratch areas. The form "nn%" defines scale factor 3022 (in percents) of memory that was used during boot. 3023 3024 kmac= [MIPS] Korina ethernet MAC address. 3025 Configure the RouterBoard 532 series on-chip 3026 Ethernet adapter MAC address. 3027 3028 kmemleak= [KNL,EARLY] Boot-time kmemleak enable/disable 3029 Valid arguments: on, off 3030 Default: on 3031 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 3032 the default is off. 3033 3034 kprobe_event=[probe-list] 3035 [FTRACE] Add kprobe events and enable at boot time. 3036 The probe-list is a semicolon delimited list of probe 3037 definitions. Each definition is same as kprobe_events 3038 interface, but the parameters are comma delimited. 3039 For example, to add a kprobe event on vfs_read with 3040 arg1 and arg2, add to the command line; 3041 3042 kprobe_event=p,vfs_read,$arg1,$arg2 3043 3044 See also Documentation/trace/kprobetrace.rst "Kernel 3045 Boot Parameter" section. 3046 3047 kpti= [ARM64,EARLY] Control page table isolation of 3048 user and kernel address spaces. 3049 Default: enabled on cores which need mitigation. 3050 0: force disabled 3051 1: force enabled 3052 3053 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires 3054 CONFIG_KUNIT to be set to be fully enabled. The 3055 default value can be overridden via 3056 KUNIT_DEFAULT_ENABLED. 3057 Default is 1 (enabled) 3058 3059 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 3060 Default is 0 (don't ignore, but inject #GP) 3061 3062 kvm.eager_page_split= 3063 [KVM,X86] Controls whether or not KVM will try to 3064 proactively split all huge pages during dirty logging. 3065 Eager page splitting reduces interruptions to vCPU 3066 execution by eliminating the write-protection faults 3067 and MMU lock contention that would otherwise be 3068 required to split huge pages lazily. 3069 3070 VM workloads that rarely perform writes or that write 3071 only to a small region of VM memory may benefit from 3072 disabling eager page splitting to allow huge pages to 3073 still be used for reads. 3074 3075 The behavior of eager page splitting depends on whether 3076 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If 3077 disabled, all huge pages in a memslot will be eagerly 3078 split when dirty logging is enabled on that memslot. If 3079 enabled, eager page splitting will be performed during 3080 the KVM_CLEAR_DIRTY ioctl, and only for the pages being 3081 cleared. 3082 3083 Eager page splitting is only supported when kvm.tdp_mmu=Y. 3084 3085 Default is Y (on). 3086 3087 kvm.enable_virt_at_load=[KVM,ARM64,LOONGARCH,MIPS,RISCV,X86] 3088 If enabled, KVM will enable virtualization in hardware 3089 when KVM is loaded, and disable virtualization when KVM 3090 is unloaded (if KVM is built as a module). 3091 3092 If disabled, KVM will dynamically enable and disable 3093 virtualization on-demand when creating and destroying 3094 VMs, i.e. on the 0=>1 and 1=>0 transitions of the 3095 number of VMs. 3096 3097 Enabling virtualization at module load avoids potential 3098 latency for creation of the 0=>1 VM, as KVM serializes 3099 virtualization enabling across all online CPUs. The 3100 "cost" of enabling virtualization when KVM is loaded, 3101 is that doing so may interfere with using out-of-tree 3102 hypervisors that want to "own" virtualization hardware. 3103 3104 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 3105 Default is false (don't support). 3106 3107 kvm.nx_huge_pages= 3108 [KVM] Controls the software workaround for the 3109 X86_BUG_ITLB_MULTIHIT bug. 3110 force : Always deploy workaround. 3111 off : Never deploy workaround. 3112 auto : Deploy workaround based on the presence of 3113 X86_BUG_ITLB_MULTIHIT. 3114 3115 Default is 'auto'. 3116 3117 If the software workaround is enabled for the host, 3118 guests do need not to enable it for nested guests. 3119 3120 kvm.nx_huge_pages_recovery_ratio= 3121 [KVM] Controls how many 4KiB pages are periodically zapped 3122 back to huge pages. 0 disables the recovery, otherwise if 3123 the value is N KVM will zap 1/Nth of the 4KiB pages every 3124 period (see below). The default is 60. 3125 3126 kvm.nx_huge_pages_recovery_period_ms= 3127 [KVM] Controls the time period at which KVM zaps 4KiB pages 3128 back to huge pages. If the value is a non-zero N, KVM will 3129 zap a portion (see ratio above) of the pages every N msecs. 3130 If the value is 0 (the default), KVM will pick a period based 3131 on the ratio, such that a page is zapped after 1 hour on average. 3132 3133 kvm-amd.nested= [KVM,AMD] Control nested virtualization feature in 3134 KVM/SVM. Default is 1 (enabled). 3135 3136 kvm-amd.npt= [KVM,AMD] Control KVM's use of Nested Page Tables, 3137 a.k.a. Two-Dimensional Page Tables. Default is 1 3138 (enabled). Disable by KVM if hardware lacks support 3139 for NPT. 3140 3141 kvm-amd.ciphertext_hiding_asids= 3142 [KVM,AMD] Ciphertext hiding prevents disallowed accesses 3143 to SNP private memory from reading ciphertext. Instead, 3144 reads will see constant default values (0xff). 3145 3146 If ciphertext hiding is enabled, the joint SEV-ES and 3147 SEV-SNP ASID space is partitioned into separate SEV-ES 3148 and SEV-SNP ASID ranges, with the SEV-SNP range being 3149 [1..max_snp_asid] and the SEV-ES range being 3150 (max_snp_asid..min_sev_asid), where min_sev_asid is 3151 enumerated by CPUID.0x.8000_001F[EDX]. 3152 3153 A non-zero value enables SEV-SNP ciphertext hiding and 3154 adjusts the ASID ranges for SEV-ES and SEV-SNP guests. 3155 KVM caps the number of SEV-SNP ASIDs at the maximum 3156 possible value, e.g. specifying -1u will assign all 3157 joint SEV-ES and SEV-SNP ASIDs to SEV-SNP. Note, 3158 assigning all joint ASIDs to SEV-SNP, i.e. configuring 3159 max_snp_asid == min_sev_asid-1, will effectively make 3160 SEV-ES unusable. 3161 3162 kvm-arm.mode= 3163 [KVM,ARM,EARLY] Select one of KVM/arm64's modes of 3164 operation. 3165 3166 none: Forcefully disable KVM. 3167 3168 nvhe: Standard nVHE-based mode, without support for 3169 protected guests. 3170 3171 protected: Mode with support for guests whose state is 3172 kept private from the host, using VHE or 3173 nVHE depending on HW support. 3174 3175 nested: VHE-based mode with support for nested 3176 virtualization. Requires at least ARMv8.4 3177 hardware (with FEAT_NV2). 3178 3179 Defaults to VHE/nVHE based on hardware support. Setting 3180 mode to "protected" will disable kexec and hibernation 3181 for the host. To force nVHE on VHE hardware, add 3182 "arm64_sw.hvhe=0 id_aa64mmfr1.vh=0" to the 3183 command-line. 3184 "nested" is experimental and should be used with 3185 extreme caution. 3186 3187 kvm-arm.vgic_v3_group0_trap= 3188 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-0 3189 system registers 3190 3191 kvm-arm.vgic_v3_group1_trap= 3192 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-1 3193 system registers 3194 3195 kvm-arm.vgic_v3_common_trap= 3196 [KVM,ARM,EARLY] Trap guest accesses to GICv3 common 3197 system registers 3198 3199 kvm-arm.vgic_v4_enable= 3200 [KVM,ARM,EARLY] Allow use of GICv4 for direct 3201 injection of LPIs. 3202 3203 kvm-arm.wfe_trap_policy= 3204 [KVM,ARM] Control when to set WFE instruction trap for 3205 KVM VMs. Traps are allowed but not guaranteed by the 3206 CPU architecture. 3207 3208 trap: set WFE instruction trap 3209 3210 notrap: clear WFE instruction trap 3211 3212 kvm-arm.wfi_trap_policy= 3213 [KVM,ARM] Control when to set WFI instruction trap for 3214 KVM VMs. Traps are allowed but not guaranteed by the 3215 CPU architecture. 3216 3217 trap: set WFI instruction trap 3218 3219 notrap: clear WFI instruction trap 3220 3221 kvm_cma_resv_ratio=n [PPC,EARLY] 3222 Reserves given percentage from system memory area for 3223 contiguous memory allocation for KVM hash pagetable 3224 allocation. 3225 By default it reserves 5% of total system memory. 3226 Format: <integer> 3227 Default: 5 3228 3229 kvm-intel.ept= [KVM,Intel] Control KVM's use of Extended Page Tables, 3230 a.k.a. Two-Dimensional Page Tables. Default is 1 3231 (enabled). Disable by KVM if hardware lacks support 3232 for EPT. 3233 3234 kvm-intel.emulate_invalid_guest_state= 3235 [KVM,Intel] Control whether to emulate invalid guest 3236 state. Ignored if kvm-intel.enable_unrestricted_guest=1, 3237 as guest state is never invalid for unrestricted 3238 guests. This param doesn't apply to nested guests (L2), 3239 as KVM never emulates invalid L2 guest state. 3240 Default is 1 (enabled). 3241 3242 kvm-intel.flexpriority= 3243 [KVM,Intel] Control KVM's use of FlexPriority feature 3244 (TPR shadow). Default is 1 (enabled). Disable by KVM if 3245 hardware lacks support for it. 3246 3247 kvm-intel.nested= 3248 [KVM,Intel] Control nested virtualization feature in 3249 KVM/VMX. Default is 1 (enabled). 3250 3251 kvm-intel.unrestricted_guest= 3252 [KVM,Intel] Control KVM's use of unrestricted guest 3253 feature (virtualized real and unpaged mode). Default 3254 is 1 (enabled). Disable by KVM if EPT is disabled or 3255 hardware lacks support for it. 3256 3257 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 3258 CVE-2018-3620. 3259 3260 Valid arguments: never, cond, always 3261 3262 always: L1D cache flush on every VMENTER. 3263 cond: Flush L1D on VMENTER only when the code between 3264 VMEXIT and VMENTER can leak host memory. 3265 never: Disables the mitigation 3266 3267 Default is cond (do L1 cache flush in specific instances) 3268 3269 kvm-intel.vpid= [KVM,Intel] Control KVM's use of Virtual Processor 3270 Identification feature (tagged TLBs). Default is 1 3271 (enabled). Disable by KVM if hardware lacks support 3272 for it. 3273 3274 l1d_flush= [X86,INTEL,EARLY] 3275 Control mitigation for L1D based snooping vulnerability. 3276 3277 Certain CPUs are vulnerable to an exploit against CPU 3278 internal buffers which can forward information to a 3279 disclosure gadget under certain conditions. 3280 3281 In vulnerable processors, the speculatively 3282 forwarded data can be used in a cache side channel 3283 attack, to access data to which the attacker does 3284 not have direct access. 3285 3286 This parameter controls the mitigation. The 3287 options are: 3288 3289 on - enable the interface for the mitigation 3290 3291 l1tf= [X86,EARLY] Control mitigation of the L1TF vulnerability on 3292 affected CPUs 3293 3294 The kernel PTE inversion protection is unconditionally 3295 enabled and cannot be disabled. 3296 3297 full 3298 Provides all available mitigations for the 3299 L1TF vulnerability. Disables SMT and 3300 enables all mitigations in the 3301 hypervisors, i.e. unconditional L1D flush. 3302 3303 SMT control and L1D flush control via the 3304 sysfs interface is still possible after 3305 boot. Hypervisors will issue a warning 3306 when the first VM is started in a 3307 potentially insecure configuration, 3308 i.e. SMT enabled or L1D flush disabled. 3309 3310 full,force 3311 Same as 'full', but disables SMT and L1D 3312 flush runtime control. Implies the 3313 'nosmt=force' command line option. 3314 (i.e. sysfs control of SMT is disabled.) 3315 3316 flush 3317 Leaves SMT enabled and enables the default 3318 hypervisor mitigation, i.e. conditional 3319 L1D flush. 3320 3321 SMT control and L1D flush control via the 3322 sysfs interface is still possible after 3323 boot. Hypervisors will issue a warning 3324 when the first VM is started in a 3325 potentially insecure configuration, 3326 i.e. SMT enabled or L1D flush disabled. 3327 3328 flush,nosmt 3329 3330 Disables SMT and enables the default 3331 hypervisor mitigation. 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 flush,nowarn 3341 Same as 'flush', but hypervisors will not 3342 warn when a VM is started in a potentially 3343 insecure configuration. 3344 3345 off 3346 Disables hypervisor mitigations and doesn't 3347 emit any warnings. 3348 It also drops the swap size and available 3349 RAM limit restriction on both hypervisor and 3350 bare metal. 3351 3352 Default is 'flush'. 3353 3354 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 3355 3356 l2cr= [PPC] 3357 3358 l3cr= [PPC] 3359 3360 lapic [X86-32,APIC,EARLY] Enable the local APIC even if BIOS 3361 disabled it. 3362 3363 lapic= [X86,APIC] Do not use TSC deadline 3364 value for LAPIC timer one-shot implementation. Default 3365 back to the programmable timer unit in the LAPIC. 3366 Format: notscdeadline 3367 3368 lapic_timer_c2_ok [X86,APIC,EARLY] trust the local apic timer 3369 in C2 power state. 3370 3371 libata.dma= [LIBATA] DMA control 3372 libata.dma=0 Disable all PATA and SATA DMA 3373 libata.dma=1 PATA and SATA Disk DMA only 3374 libata.dma=2 ATAPI (CDROM) DMA only 3375 libata.dma=4 Compact Flash DMA only 3376 Combinations also work, so libata.dma=3 enables DMA 3377 for disks and CDROMs, but not CFs. 3378 3379 libata.ignore_hpa= [LIBATA] Ignore HPA limit 3380 libata.ignore_hpa=0 keep BIOS limits (default) 3381 libata.ignore_hpa=1 ignore limits, using full disk 3382 3383 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 3384 when set. 3385 Format: <int> 3386 3387 libata.force= [LIBATA] Force configurations. The format is a comma- 3388 separated list of "[ID:]VAL" where ID is PORT[.DEVICE]. 3389 PORT and DEVICE are decimal numbers matching port, link 3390 or device. Basically, it matches the ATA ID string 3391 printed on console by libata. If the whole ID part is 3392 omitted, the last PORT and DEVICE values are used. If 3393 ID hasn't been specified yet, the configuration applies 3394 to all ports, links and devices. 3395 3396 If only DEVICE is omitted, the parameter applies to 3397 the port and all links and devices behind it. DEVICE 3398 number of 0 either selects the first device or the 3399 first fan-out link behind PMP device. It does not 3400 select the host link. DEVICE number of 15 selects the 3401 host link and device attached to it. 3402 3403 The VAL specifies the configuration to force. As long 3404 as there is no ambiguity, shortcut notation is allowed. 3405 For example, both 1.5 and 1.5G would work for 1.5Gbps. 3406 The following configurations can be forced. 3407 3408 * Cable type: 40c, 80c, short40c, unk, ign or sata. 3409 Any ID with matching PORT is used. 3410 3411 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 3412 3413 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 3414 udma[/][16,25,33,44,66,100,133] notation is also 3415 allowed. 3416 3417 * nohrst, nosrst, norst: suppress hard, soft and both 3418 resets. 3419 3420 * rstonce: only attempt one reset during hot-unplug 3421 link recovery. 3422 3423 * [no]dbdelay: Enable or disable the extra 200ms delay 3424 before debouncing a link PHY and device presence 3425 detection. 3426 3427 * [no]ncq: Turn on or off NCQ. 3428 3429 * [no]ncqtrim: Enable or disable queued DSM TRIM. 3430 3431 * [no]ncqati: Enable or disable NCQ trim on ATI chipset. 3432 3433 * [no]trim: Enable or disable (unqueued) TRIM. 3434 3435 * trim_zero: Indicate that TRIM command zeroes data. 3436 3437 * max_trim_128m: Set 128M maximum trim size limit. 3438 3439 * [no]dma: Turn on or off DMA transfers. 3440 3441 * atapi_dmadir: Enable ATAPI DMADIR bridge support. 3442 3443 * atapi_mod16_dma: Enable the use of ATAPI DMA for 3444 commands that are not a multiple of 16 bytes. 3445 3446 * [no]dmalog: Enable or disable the use of the 3447 READ LOG DMA EXT command to access logs. 3448 3449 * [no]iddevlog: Enable or disable access to the 3450 identify device data log. 3451 3452 * [no]logdir: Enable or disable access to the general 3453 purpose log directory. 3454 3455 * max_sec_128: Set transfer size limit to 128 sectors. 3456 3457 * max_sec_1024: Set or clear transfer size limit to 3458 1024 sectors. 3459 3460 * max_sec_lba48: Set or clear transfer size limit to 3461 65535 sectors. 3462 3463 * external: Mark port as external (hotplug-capable). 3464 3465 * [no]lpm: Enable or disable link power management. 3466 3467 * [no]setxfer: Indicate if transfer speed mode setting 3468 should be skipped. 3469 3470 * [no]fua: Disable or enable FUA (Force Unit Access) 3471 support for devices supporting this feature. 3472 3473 * dump_id: Dump IDENTIFY data. 3474 3475 * disable: Disable this device. 3476 3477 If there are multiple matching configurations changing 3478 the same attribute, the last one is used. 3479 3480 liveupdate= [KNL,EARLY] 3481 Format: <bool> 3482 Enable Live Update Orchestrator (LUO). 3483 Default: off. 3484 3485 lockd.nlm_grace_period=P [NFS] Assign grace period. 3486 Format: <integer> 3487 3488 lockd.nlm_tcpport=N [NFS] Assign TCP port. 3489 Format: <integer> 3490 3491 lockd.nlm_timeout=T [NFS] Assign timeout value. 3492 Format: <integer> 3493 3494 lockd.nlm_udpport=M [NFS] Assign UDP port. 3495 Format: <integer> 3496 3497 lockdown= [SECURITY,EARLY] 3498 { integrity | confidentiality } 3499 Enable the kernel lockdown feature. If set to 3500 integrity, kernel features that allow userland to 3501 modify the running kernel are disabled. If set to 3502 confidentiality, kernel features that allow userland 3503 to extract confidential information from the kernel 3504 are also disabled. 3505 3506 locktorture.acq_writer_lim= [KNL] 3507 Set the time limit in jiffies for a lock 3508 acquisition. Acquisitions exceeding this limit 3509 will result in a splat once they do complete. 3510 3511 locktorture.bind_readers= [KNL] 3512 Specify the list of CPUs to which the readers are 3513 to be bound. 3514 3515 locktorture.bind_writers= [KNL] 3516 Specify the list of CPUs to which the writers are 3517 to be bound. 3518 3519 locktorture.call_rcu_chains= [KNL] 3520 Specify the number of self-propagating call_rcu() 3521 chains to set up. These are used to ensure that 3522 there is a high probability of an RCU grace period 3523 in progress at any given time. Defaults to 0, 3524 which disables these call_rcu() chains. 3525 3526 locktorture.long_hold= [KNL] 3527 Specify the duration in milliseconds for the 3528 occasional long-duration lock hold time. Defaults 3529 to 100 milliseconds. Select 0 to disable. 3530 3531 locktorture.nested_locks= [KNL] 3532 Specify the maximum lock nesting depth that 3533 locktorture is to exercise, up to a limit of 8 3534 (MAX_NESTED_LOCKS). Specify zero to disable. 3535 Note that this parameter is ineffective on types 3536 of locks that do not support nested acquisition. 3537 3538 locktorture.nreaders_stress= [KNL] 3539 Set the number of locking read-acquisition kthreads. 3540 Defaults to being automatically set based on the 3541 number of online CPUs. 3542 3543 locktorture.nwriters_stress= [KNL] 3544 Set the number of locking write-acquisition kthreads. 3545 3546 locktorture.onoff_holdoff= [KNL] 3547 Set time (s) after boot for CPU-hotplug testing. 3548 3549 locktorture.onoff_interval= [KNL] 3550 Set time (s) between CPU-hotplug operations, or 3551 zero to disable CPU-hotplug testing. 3552 3553 locktorture.rt_boost= [KNL] 3554 Do periodic testing of real-time lock priority 3555 boosting. Select 0 to disable, 1 to boost 3556 only rt_mutex, and 2 to boost unconditionally. 3557 Defaults to 2, which might seem to be an 3558 odd choice, but which should be harmless for 3559 non-real-time spinlocks, due to their disabling 3560 of preemption. Note that non-realtime mutexes 3561 disable boosting. 3562 3563 locktorture.rt_boost_factor= [KNL] 3564 Number that determines how often and for how 3565 long priority boosting is exercised. This is 3566 scaled down by the number of writers, so that the 3567 number of boosts per unit time remains roughly 3568 constant as the number of writers increases. 3569 On the other hand, the duration of each boost 3570 increases with the number of writers. 3571 3572 locktorture.shuffle_interval= [KNL] 3573 Set task-shuffle interval (jiffies). Shuffling 3574 tasks allows some CPUs to go into dyntick-idle 3575 mode during the locktorture test. 3576 3577 locktorture.shutdown_secs= [KNL] 3578 Set time (s) after boot system shutdown. This 3579 is useful for hands-off automated testing. 3580 3581 locktorture.stat_interval= [KNL] 3582 Time (s) between statistics printk()s. 3583 3584 locktorture.stutter= [KNL] 3585 Time (s) to stutter testing, for example, 3586 specifying five seconds causes the test to run for 3587 five seconds, wait for five seconds, and so on. 3588 This tests the locking primitive's ability to 3589 transition abruptly to and from idle. 3590 3591 locktorture.torture_type= [KNL] 3592 Specify the locking implementation to test. 3593 3594 locktorture.verbose= [KNL] 3595 Enable additional printk() statements. 3596 3597 locktorture.writer_fifo= [KNL] 3598 Run the write-side locktorture kthreads at 3599 sched_set_fifo() real-time priority. 3600 3601 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 3602 Format: <irq> 3603 3604 loglevel= [KNL,EARLY] 3605 All Kernel Messages with a loglevel smaller than the 3606 console loglevel will be printed to the console. It can 3607 also be changed with klogd or other programs. The 3608 loglevels are defined as follows: 3609 3610 0 (KERN_EMERG) system is unusable 3611 1 (KERN_ALERT) action must be taken immediately 3612 2 (KERN_CRIT) critical conditions 3613 3 (KERN_ERR) error conditions 3614 4 (KERN_WARNING) warning conditions 3615 5 (KERN_NOTICE) normal but significant condition 3616 6 (KERN_INFO) informational 3617 7 (KERN_DEBUG) debug-level messages 3618 3619 log_buf_len=n[KMG] [KNL,EARLY] 3620 Sets the size of the printk ring buffer, in bytes. 3621 n must be a power of two and greater than the 3622 minimal size. The minimal size is defined by 3623 LOG_BUF_SHIFT kernel config parameter. There 3624 is also CONFIG_LOG_CPU_MAX_BUF_SHIFT config 3625 parameter that allows to increase the default size 3626 depending on the number of CPUs. See init/Kconfig 3627 for more details. 3628 3629 logo.nologo [FB] Disables display of the built-in Linux logo. 3630 This may be used to provide more screen space for 3631 kernel log messages and is useful when debugging 3632 kernel boot problems. 3633 3634 lp=0 [LP] Specify parallel ports to use, e.g, 3635 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 3636 lp=reset first parallel port). 'lp=0' disables the 3637 lp=auto printer driver. 'lp=reset' (which can be 3638 specified in addition to the ports) causes 3639 attached printers to be reset. Using 3640 lp=port1,port2,... specifies the parallel ports 3641 to associate lp devices with, starting with 3642 lp0. A port specification may be 'none' to skip 3643 that lp device, or a parport name such as 3644 'parport0'. Specifying 'lp=auto' instead of a 3645 port specification list means that device IDs 3646 from each port should be examined, to see if 3647 an IEEE 1284-compliant printer is attached; if 3648 so, the driver will manage that printer. 3649 See also header of drivers/char/lp.c. 3650 3651 lpj=n [KNL] 3652 Sets loops_per_jiffy to given constant, thus avoiding 3653 time-consuming boot-time autodetection (up to 250 ms per 3654 CPU). 0 enables autodetection (default). To determine 3655 the correct value for your kernel, boot with normal 3656 autodetection and see what value is printed. Note that 3657 on SMP systems the preset will be applied to all CPUs, 3658 which is likely to cause problems if your CPUs need 3659 significantly divergent settings. An incorrect value 3660 will cause delays in the kernel to be wrong, leading to 3661 unpredictable I/O errors and other breakage. Although 3662 unlikely, in the extreme case this might damage your 3663 hardware. 3664 3665 lsm.debug [SECURITY] Enable LSM initialization debugging output. 3666 3667 lsm=lsm1,...,lsmN 3668 [SECURITY] Choose order of LSM initialization. This 3669 overrides CONFIG_LSM, and the "security=" parameter. 3670 3671 machtype= [Loongson] Share the same kernel image file between 3672 different yeeloong laptops. 3673 Example: machtype=lemote-yeeloong-2f-7inch 3674 3675 maxcpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 3676 will bring up during bootup. maxcpus=n : n >= 0 limits 3677 the kernel to bring up 'n' processors. Surely after 3678 bootup you can bring up the other plugged cpu by executing 3679 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 3680 only takes effect during system bootup. 3681 While n=0 is a special case, it is equivalent to "nosmp", 3682 which also disables the IO APIC. 3683 3684 max_loop= [LOOP] The number of loop block devices that get 3685 (loop.max_loop) unconditionally pre-created at init time. The default 3686 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 3687 of statically allocating a predefined number, loop 3688 devices can be requested on-demand with the 3689 /dev/loop-control interface. 3690 3691 mce= [X86-{32,64}] 3692 3693 Please see Documentation/arch/x86/x86_64/machinecheck.rst for sysfs runtime tunables. 3694 3695 off 3696 disable machine check 3697 3698 no_cmci 3699 disable CMCI(Corrected Machine Check Interrupt) that 3700 Intel processor supports. Usually this disablement is 3701 not recommended, but it might be handy if your 3702 hardware is misbehaving. 3703 3704 Note that you'll get more problems without CMCI than 3705 with due to the shared banks, i.e. you might get 3706 duplicated error logs. 3707 3708 dont_log_ce 3709 don't make logs for corrected errors. All events 3710 reported as corrected are silently cleared by OS. This 3711 option will be useful if you have no interest in any 3712 of corrected errors. 3713 3714 ignore_ce 3715 disable features for corrected errors, e.g. 3716 polling timer and CMCI. All events reported as 3717 corrected are not cleared by OS and remained in its 3718 error banks. 3719 3720 Usually this disablement is not recommended, however 3721 if there is an agent checking/clearing corrected 3722 errors (e.g. BIOS or hardware monitoring 3723 applications), conflicting with OS's error handling, 3724 and you cannot deactivate the agent, then this option 3725 will be a help. 3726 3727 no_lmce 3728 do not opt-in to Local MCE delivery. Use legacy method 3729 to broadcast MCEs. 3730 3731 bootlog 3732 enable logging of machine checks left over from 3733 booting. Disabled by default on AMD Fam10h and older 3734 because some BIOS leave bogus ones. 3735 3736 If your BIOS doesn't do that it's a good idea to 3737 enable though to make sure you log even machine check 3738 events that result in a reboot. On Intel systems it is 3739 enabled by default. 3740 3741 nobootlog 3742 disable boot machine check logging. 3743 3744 monarchtimeout (number) 3745 sets the time in us to wait for other CPUs on machine 3746 checks. 0 to disable. 3747 3748 bios_cmci_threshold 3749 don't overwrite the bios-set CMCI threshold. This boot 3750 option prevents Linux from overwriting the CMCI 3751 threshold set by the bios. Without this option, Linux 3752 always sets the CMCI threshold to 1. Enabling this may 3753 make memory predictive failure analysis less effective 3754 if the bios sets thresholds for memory errors since we 3755 will not see details for all errors. 3756 3757 recovery 3758 force-enable recoverable machine check code paths 3759 3760 Everything else is in sysfs now. 3761 3762 3763 md= [HW] RAID subsystems devices and level 3764 See Documentation/admin-guide/md.rst. 3765 3766 mdacon= [MDA] 3767 Format: <first>,<last> 3768 Specifies range of consoles to be captured by the MDA. 3769 3770 mds= [X86,INTEL,EARLY] 3771 Control mitigation for the Micro-architectural Data 3772 Sampling (MDS) vulnerability. 3773 3774 Certain CPUs are vulnerable to an exploit against CPU 3775 internal buffers which can forward information to a 3776 disclosure gadget under certain conditions. 3777 3778 In vulnerable processors, the speculatively 3779 forwarded data can be used in a cache side channel 3780 attack, to access data to which the attacker does 3781 not have direct access. 3782 3783 This parameter controls the MDS mitigation. The 3784 options are: 3785 3786 full - Enable MDS mitigation on vulnerable CPUs 3787 full,nosmt - Enable MDS mitigation and disable 3788 SMT on vulnerable CPUs 3789 off - Unconditionally disable MDS mitigation 3790 3791 On TAA-affected machines, mds=off can be prevented by 3792 an active TAA mitigation as both vulnerabilities are 3793 mitigated with the same mechanism so in order to disable 3794 this mitigation, you need to specify tsx_async_abort=off 3795 too. 3796 3797 Not specifying this option is equivalent to 3798 mds=full. 3799 3800 For details see: Documentation/admin-guide/hw-vuln/mds.rst 3801 3802 mem=nn[KMG] [HEXAGON,EARLY] Set the memory size. 3803 Must be specified, otherwise memory size will be 0. 3804 3805 mem=nn[KMG] [KNL,BOOT,EARLY] Force usage of a specific amount 3806 of memory Amount of memory to be used in cases 3807 as follows: 3808 3809 1 for test; 3810 2 when the kernel is not able to see the whole system memory; 3811 3 memory that lies after 'mem=' boundary is excluded from 3812 the hypervisor, then assigned to KVM guests. 3813 4 to limit the memory available for kdump kernel. 3814 3815 [ARC,MICROBLAZE] - the limit applies only to low memory, 3816 high memory is not affected. 3817 3818 [ARM64] - only limits memory covered by the linear 3819 mapping. The NOMAP regions are not affected. 3820 3821 [X86] Work as limiting max address. Use together 3822 with memmap= to avoid physical address space collisions. 3823 Without memmap= PCI devices could be placed at addresses 3824 belonging to unused RAM. 3825 3826 Note that this only takes effects during boot time since 3827 in above case 3, memory may need be hot added after boot 3828 if system memory of hypervisor is not sufficient. 3829 3830 mem=nn[KMG]@ss[KMG] 3831 [ARM,MIPS,EARLY] - override the memory layout 3832 reported by firmware. 3833 Define a memory region of size nn[KMG] starting at 3834 ss[KMG]. 3835 Multiple different regions can be specified with 3836 multiple mem= parameters on the command line. 3837 3838 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 3839 memory. 3840 3841 memblock=debug [KNL,EARLY] Enable memblock debug messages. 3842 3843 memchunk=nn[KMG] 3844 [KNL,SH] Allow user to override the default size for 3845 per-device physically contiguous DMA buffers. 3846 3847 memhp_default_state=online/offline/online_kernel/online_movable 3848 [KNL] Set the initial state for the memory hotplug 3849 onlining policy. If not specified, the default value is 3850 set according to the 3851 CONFIG_MHP_DEFAULT_ONLINE_TYPE kernel config 3852 options. 3853 See Documentation/admin-guide/mm/memory-hotplug.rst. 3854 3855 memmap=exactmap [KNL,X86,EARLY] Enable setting of an exact 3856 E820 memory map, as specified by the user. 3857 Such memmap=exactmap lines can be constructed based on 3858 BIOS output or other requirements. See the memmap=nn@ss 3859 option description. 3860 3861 memmap=nn[KMG]@ss[KMG] 3862 [KNL, X86,MIPS,XTENSA,EARLY] Force usage of a specific region of memory. 3863 Region of memory to be used is from ss to ss+nn. 3864 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 3865 which limits max address to nn[KMG]. 3866 Multiple different regions can be specified, 3867 comma delimited. 3868 Example: 3869 memmap=100M@2G,100M#3G,1G!1024G 3870 3871 memmap=nn[KMG]#ss[KMG] 3872 [KNL,ACPI,EARLY] Mark specific memory as ACPI data. 3873 Region of memory to be marked is from ss to ss+nn. 3874 3875 memmap=nn[KMG]$ss[KMG] 3876 [KNL,ACPI,EARLY] Mark specific memory as reserved. 3877 Region of memory to be reserved is from ss to ss+nn. 3878 Example: Exclude memory from 0x18690000-0x1869ffff 3879 memmap=64K$0x18690000 3880 or 3881 memmap=0x10000$0x18690000 3882 Some bootloaders may need an escape character before '$', 3883 like Grub2, otherwise '$' and the following number 3884 will be eaten. 3885 3886 memmap=nn[KMG]!ss[KMG,EARLY] 3887 [KNL,X86] Mark specific memory as protected. 3888 Region of memory to be used, from ss to ss+nn. 3889 The memory region may be marked as e820 type 12 (0xc) 3890 and is NVDIMM or ADR memory. 3891 3892 memmap=<size>%<offset>-<oldtype>+<newtype> 3893 [KNL,ACPI,EARLY] Convert memory within the specified region 3894 from <oldtype> to <newtype>. If "-<oldtype>" is left 3895 out, the whole region will be marked as <newtype>, 3896 even if previously unavailable. If "+<newtype>" is left 3897 out, matching memory will be removed. Types are 3898 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 3899 3 = ACPI, 12 = PRAM. 3900 3901 memory_corruption_check=0/1 [X86,EARLY] 3902 Some BIOSes seem to corrupt the first 64k of 3903 memory when doing things like suspend/resume. 3904 Setting this option will scan the memory 3905 looking for corruption. Enabling this will 3906 both detect corruption and prevent the kernel 3907 from using the memory being corrupted. 3908 However, it's intended as a diagnostic tool; if 3909 repeatable BIOS-originated corruption always 3910 affects the same memory, you can use memmap= 3911 to prevent the kernel from using that memory. 3912 3913 memory_corruption_check_size=size [X86,EARLY] 3914 By default it checks for corruption in the low 3915 64k, making this memory unavailable for normal 3916 use. Use this parameter to scan for 3917 corruption in more or less memory. 3918 3919 memory_corruption_check_period=seconds [X86,EARLY] 3920 By default it checks for corruption every 60 3921 seconds. Use this parameter to check at some 3922 other rate. 0 disables periodic checking. 3923 3924 memory_hotplug.memmap_on_memory 3925 [KNL,X86,ARM] Boolean flag to enable this feature. 3926 Format: {on | off (default)} 3927 When enabled, runtime hotplugged memory will 3928 allocate its internal metadata (struct pages, 3929 those vmemmap pages cannot be optimized even 3930 if hugetlb_free_vmemmap is enabled) from the 3931 hotadded memory which will allow to hotadd a 3932 lot of memory without requiring additional 3933 memory to do so. 3934 This feature is disabled by default because it 3935 has some implication on large (e.g. GB) 3936 allocations in some configurations (e.g. small 3937 memory blocks). 3938 The state of the flag can be read in 3939 /sys/module/memory_hotplug/parameters/memmap_on_memory. 3940 Note that even when enabled, there are a few cases where 3941 the feature is not effective. 3942 3943 memtest= [KNL,X86,ARM,M68K,PPC,RISCV,EARLY] Enable memtest 3944 Format: <integer> 3945 default : 0 <disable> 3946 Specifies the number of memtest passes to be 3947 performed. Each pass selects another test 3948 pattern from a given set of patterns. Memtest 3949 fills the memory with this pattern, validates 3950 memory contents and reserves bad memory 3951 regions that are detected. 3952 3953 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 3954 Valid arguments: on, off 3955 Default: off 3956 mem_encrypt=on: Activate SME 3957 mem_encrypt=off: Do not activate SME 3958 3959 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst 3960 for details on when memory encryption can be activated. 3961 3962 mem_sleep_default= [SUSPEND] Default system suspend mode: 3963 s2idle - Suspend-To-Idle 3964 shallow - Power-On Suspend or equivalent (if supported) 3965 deep - Suspend-To-RAM or equivalent (if supported) 3966 See Documentation/admin-guide/pm/sleep-states.rst. 3967 3968 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 3969 the BIOS has incorrectly applied a workaround. TinyBIOS 3970 version 0.98 is known to be affected, 0.99 fixes the 3971 problem by letting the user disable the workaround. 3972 3973 mga= [HW,DRM] 3974 3975 microcode= [X86] Control the behavior of the microcode loader. 3976 Available options, comma separated: 3977 3978 base_rev=X - with <X> with format: <u32> 3979 Set the base microcode revision of each thread when in 3980 debug mode. 3981 3982 dis_ucode_ldr: disable the microcode loader 3983 3984 force_minrev: 3985 Enable or disable the microcode minimal revision 3986 enforcement for the runtime microcode loader. 3987 3988 mini2440= [ARM,HW,KNL] 3989 Format:[0..2][b][c][t] 3990 Default: "0tb" 3991 MINI2440 configuration specification: 3992 0 - The attached screen is the 3.5" TFT 3993 1 - The attached screen is the 7" TFT 3994 2 - The VGA Shield is attached (1024x768) 3995 Leaving out the screen size parameter will not load 3996 the TFT driver, and the framebuffer will be left 3997 unconfigured. 3998 b - Enable backlight. The TFT backlight pin will be 3999 linked to the kernel VESA blanking code and a GPIO 4000 LED. This parameter is not necessary when using the 4001 VGA shield. 4002 c - Enable the s3c camera interface. 4003 t - Reserved for enabling touchscreen support. The 4004 touchscreen support is not enabled in the mainstream 4005 kernel as of 2.6.30, a preliminary port can be found 4006 in the "bleeding edge" mini2440 support kernel at 4007 https://repo.or.cz/w/linux-2.6/mini2440.git 4008 4009 mitigations= 4010 [X86,PPC,S390,ARM64,EARLY] Control optional mitigations for 4011 CPU vulnerabilities. This is a set of curated, 4012 arch-independent options, each of which is an 4013 aggregation of existing arch-specific options. 4014 4015 Note, "mitigations" is supported if and only if the 4016 kernel was built with CPU_MITIGATIONS=y. 4017 4018 off 4019 Disable all optional CPU mitigations. This 4020 improves system performance, but it may also 4021 expose users to several CPU vulnerabilities. 4022 Equivalent to: if nokaslr then kpti=0 [ARM64] 4023 gather_data_sampling=off [X86] 4024 indirect_target_selection=off [X86] 4025 kvm.nx_huge_pages=off [X86] 4026 l1tf=off [X86] 4027 mds=off [X86] 4028 mmio_stale_data=off [X86] 4029 no_entry_flush [PPC] 4030 no_uaccess_flush [PPC] 4031 nobp=0 [S390] 4032 nopti [X86,PPC] 4033 nospectre_bhb [ARM64] 4034 nospectre_v1 [X86,PPC] 4035 nospectre_v2 [X86,PPC,S390,ARM64] 4036 reg_file_data_sampling=off [X86] 4037 retbleed=off [X86] 4038 spec_rstack_overflow=off [X86] 4039 spec_store_bypass_disable=off [X86,PPC] 4040 spectre_bhi=off [X86] 4041 spectre_v2_user=off [X86] 4042 srbds=off [X86,INTEL] 4043 ssbd=force-off [ARM64] 4044 tsa=off [X86,AMD] 4045 tsx_async_abort=off [X86] 4046 vmscape=off [X86] 4047 4048 Exceptions: 4049 This does not have any effect on 4050 kvm.nx_huge_pages when 4051 kvm.nx_huge_pages=force. 4052 4053 auto (default) 4054 Mitigate all CPU vulnerabilities, but leave SMT 4055 enabled, even if it's vulnerable. This is for 4056 users who don't want to be surprised by SMT 4057 getting disabled across kernel upgrades, or who 4058 have other ways of avoiding SMT-based attacks. 4059 Equivalent to: (default behavior) 4060 4061 auto,nosmt 4062 Mitigate all CPU vulnerabilities, disabling SMT 4063 if needed. This is for users who always want to 4064 be fully mitigated, even if it means losing SMT. 4065 Equivalent to: l1tf=flush,nosmt [X86] 4066 mds=full,nosmt [X86] 4067 tsx_async_abort=full,nosmt [X86] 4068 mmio_stale_data=full,nosmt [X86] 4069 retbleed=auto,nosmt [X86] 4070 4071 [X86] After one of the above options, additionally 4072 supports attack-vector based controls as documented in 4073 Documentation/admin-guide/hw-vuln/attack_vector_controls.rst 4074 4075 mminit_loglevel= 4076 [KNL,EARLY] When CONFIG_DEBUG_MEMORY_INIT is set, this 4077 parameter allows control of the logging verbosity for 4078 the additional memory initialisation checks. A value 4079 of 0 disables mminit logging and a level of 4 will 4080 log everything. Information is printed at KERN_DEBUG 4081 so loglevel=8 may also need to be specified. 4082 4083 mmio_stale_data= 4084 [X86,INTEL,EARLY] Control mitigation for the Processor 4085 MMIO Stale Data vulnerabilities. 4086 4087 Processor MMIO Stale Data is a class of 4088 vulnerabilities that may expose data after an MMIO 4089 operation. Exposed data could originate or end in 4090 the same CPU buffers as affected by MDS and TAA. 4091 Therefore, similar to MDS and TAA, the mitigation 4092 is to clear the affected CPU buffers. 4093 4094 This parameter controls the mitigation. The 4095 options are: 4096 4097 full - Enable mitigation on vulnerable CPUs 4098 4099 full,nosmt - Enable mitigation and disable SMT on 4100 vulnerable CPUs. 4101 4102 off - Unconditionally disable mitigation 4103 4104 On MDS or TAA affected machines, 4105 mmio_stale_data=off can be prevented by an active 4106 MDS or TAA mitigation as these vulnerabilities are 4107 mitigated with the same mechanism so in order to 4108 disable this mitigation, you need to specify 4109 mds=off and tsx_async_abort=off too. 4110 4111 Not specifying this option is equivalent to 4112 mmio_stale_data=full. 4113 4114 For details see: 4115 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst 4116 4117 <module>.async_probe[=<bool>] [KNL] 4118 If no <bool> value is specified or if the value 4119 specified is not a valid <bool>, enable asynchronous 4120 probe on this module. Otherwise, enable/disable 4121 asynchronous probe on this module as indicated by the 4122 <bool> value. See also: module.async_probe 4123 4124 module.async_probe=<bool> 4125 [KNL] When set to true, modules will use async probing 4126 by default. To enable/disable async probing for a 4127 specific module, use the module specific control that 4128 is documented under <module>.async_probe. When both 4129 module.async_probe and <module>.async_probe are 4130 specified, <module>.async_probe takes precedence for 4131 the specific module. 4132 4133 module.enable_dups_trace 4134 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set, 4135 this means that duplicate request_module() calls will 4136 trigger a WARN_ON() instead of a pr_warn(). Note that 4137 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s 4138 will always be issued and this option does nothing. 4139 module.sig_enforce 4140 [KNL] When CONFIG_MODULE_SIG is set, this means that 4141 modules without (valid) signatures will fail to load. 4142 Note that if CONFIG_MODULE_SIG_FORCE is set, that 4143 is always true, so this option does nothing. 4144 4145 module_blacklist= [KNL] Do not load a comma-separated list of 4146 modules. Useful for debugging problem modules. 4147 4148 mousedev.tap_time= 4149 [MOUSE] Maximum time between finger touching and 4150 leaving touchpad surface for touch to be considered 4151 a tap and be reported as a left button click (for 4152 touchpads working in absolute mode only). 4153 Format: <msecs> 4154 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 4155 reporting absolute coordinates, such as tablets 4156 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 4157 reporting absolute coordinates, such as tablets 4158 4159 movablecore= [KNL,X86,PPC,EARLY] 4160 Format: nn[KMGTPE] | nn% 4161 This parameter is the complement to kernelcore=, it 4162 specifies the amount of memory used for migratable 4163 allocations. If both kernelcore and movablecore is 4164 specified, then kernelcore will be at *least* the 4165 specified value but may be more. If movablecore on its 4166 own is specified, the administrator must be careful 4167 that the amount of memory usable for all allocations 4168 is not too small. 4169 4170 movable_node [KNL,EARLY] Boot-time switch to make hotplugable memory 4171 NUMA nodes to be movable. This means that the memory 4172 of such nodes will be usable only for movable 4173 allocations which rules out almost all kernel 4174 allocations. Use with caution! 4175 4176 MTD_Partition= [MTD] 4177 Format: <name>,<region-number>,<size>,<offset> 4178 4179 MTD_Region= [MTD] Format: 4180 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 4181 4182 mtdparts= [MTD] 4183 See drivers/mtd/parsers/cmdlinepart.c 4184 4185 mtouchusb.raw_coordinates= 4186 [HW] Make the MicroTouch USB driver use raw coordinates 4187 ('y', default) or cooked coordinates ('n') 4188 4189 mtrr=debug [X86,EARLY] 4190 Enable printing debug information related to MTRR 4191 registers at boot time. 4192 4193 mtrr_chunk_size=nn[KMG,X86,EARLY] 4194 used for mtrr cleanup. It is largest continuous chunk 4195 that could hold holes aka. UC entries. 4196 4197 mtrr_gran_size=nn[KMG,X86,EARLY] 4198 Used for mtrr cleanup. It is granularity of mtrr block. 4199 Default is 1. 4200 Large value could prevent small alignment from 4201 using up MTRRs. 4202 4203 mtrr_spare_reg_nr=n [X86,EARLY] 4204 Format: <integer> 4205 Range: 0,7 : spare reg number 4206 Default : 1 4207 Used for mtrr cleanup. It is spare mtrr entries number. 4208 Set to 2 or more if your graphical card needs more. 4209 4210 multitce=off [PPC] This parameter disables the use of the pSeries 4211 firmware feature for updating multiple TCE entries 4212 at a time. 4213 4214 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 4215 4216 netdev= [NET] Network devices parameters 4217 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 4218 Note that mem_start is often overloaded to mean 4219 something different and driver-specific. 4220 This usage is only documented in each driver source 4221 file if at all. 4222 4223 netpoll.carrier_timeout= 4224 [NET] Specifies amount of time (in seconds) that 4225 netpoll should wait for a carrier. By default netpoll 4226 waits 4 seconds. 4227 4228 nf_conntrack.acct= 4229 [NETFILTER] Enable connection tracking flow accounting 4230 0 to disable accounting 4231 1 to enable accounting 4232 Default value is 0. 4233 4234 nfs.cache_getent= 4235 [NFS] sets the pathname to the program which is used 4236 to update the NFS client cache entries. 4237 4238 nfs.cache_getent_timeout= 4239 [NFS] sets the timeout after which an attempt to 4240 update a cache entry is deemed to have failed. 4241 4242 nfs.callback_nr_threads= 4243 [NFSv4] set the total number of threads that the 4244 NFS client will assign to service NFSv4 callback 4245 requests. 4246 4247 nfs.callback_tcpport= 4248 [NFS] set the TCP port on which the NFSv4 callback 4249 channel should listen. 4250 4251 nfs.delay_retrans= 4252 [NFS] specifies the number of times the NFSv4 client 4253 retries the request before returning an EAGAIN error, 4254 after a reply of NFS4ERR_DELAY from the server. 4255 Only applies if the softerr mount option is enabled, 4256 and the specified value is >= 0. 4257 4258 nfs.enable_ino64= 4259 [NFS] enable 64-bit inode numbers. 4260 If zero, the NFS client will fake up a 32-bit inode 4261 number for the readdir() and stat() syscalls instead 4262 of returning the full 64-bit number. 4263 The default is to return 64-bit inode numbers. 4264 4265 nfs.idmap_cache_timeout= 4266 [NFS] set the maximum lifetime for idmapper cache 4267 entries. 4268 4269 nfs.max_session_cb_slots= 4270 [NFSv4.1] Sets the maximum number of session 4271 slots the client will assign to the callback 4272 channel. This determines the maximum number of 4273 callbacks the client will process in parallel for 4274 a particular server. 4275 4276 nfs.max_session_slots= 4277 [NFSv4.1] Sets the maximum number of session slots 4278 the client will attempt to negotiate with the server. 4279 This limits the number of simultaneous RPC requests 4280 that the client can send to the NFSv4.1 server. 4281 Note that there is little point in setting this 4282 value higher than the max_tcp_slot_table_limit. 4283 4284 nfs.nfs4_disable_idmapping= 4285 [NFSv4] When set to the default of '1', this option 4286 ensures that both the RPC level authentication 4287 scheme and the NFS level operations agree to use 4288 numeric uids/gids if the mount is using the 4289 'sec=sys' security flavour. In effect it is 4290 disabling idmapping, which can make migration from 4291 legacy NFSv2/v3 systems to NFSv4 easier. 4292 Servers that do not support this mode of operation 4293 will be autodetected by the client, and it will fall 4294 back to using the idmapper. 4295 To turn off this behaviour, set the value to '0'. 4296 4297 nfs.nfs4_unique_id= 4298 [NFS4] Specify an additional fixed unique ident- 4299 ification string that NFSv4 clients can insert into 4300 their nfs_client_id4 string. This is typically a 4301 UUID that is generated at system install time. 4302 4303 nfs.recover_lost_locks= 4304 [NFSv4] Attempt to recover locks that were lost due 4305 to a lease timeout on the server. Please note that 4306 doing this risks data corruption, since there are 4307 no guarantees that the file will remain unchanged 4308 after the locks are lost. 4309 If you want to enable the kernel legacy behaviour of 4310 attempting to recover these locks, then set this 4311 parameter to '1'. 4312 The default parameter value of '0' causes the kernel 4313 not to attempt recovery of lost locks. 4314 4315 nfs.send_implementation_id= 4316 [NFSv4.1] Send client implementation identification 4317 information in exchange_id requests. 4318 If zero, no implementation identification information 4319 will be sent. 4320 The default is to send the implementation identification 4321 information. 4322 4323 nfs4.layoutstats_timer= 4324 [NFSv4.2] Change the rate at which the kernel sends 4325 layoutstats to the pNFS metadata server. 4326 4327 Setting this to value to 0 causes the kernel to use 4328 whatever value is the default set by the layout 4329 driver. A non-zero value sets the minimum interval 4330 in seconds between layoutstats transmissions. 4331 4332 nfsd.inter_copy_offload_enable= 4333 [NFSv4.2] When set to 1, the server will support 4334 server-to-server copies for which this server is 4335 the destination of the copy. 4336 4337 nfsd.nfs4_disable_idmapping= 4338 [NFSv4] When set to the default of '1', the NFSv4 4339 server will return only numeric uids and gids to 4340 clients using auth_sys, and will accept numeric uids 4341 and gids from such clients. This is intended to ease 4342 migration from NFSv2/v3. 4343 4344 nfsd.nfsd4_ssc_umount_timeout= 4345 [NFSv4.2] When used as the destination of a 4346 server-to-server copy, knfsd temporarily mounts 4347 the source server. It caches the mount in case 4348 it will be needed again, and discards it if not 4349 used for the number of milliseconds specified by 4350 this parameter. 4351 4352 nfsaddrs= [NFS] Deprecated. Use ip= instead. 4353 See Documentation/admin-guide/nfs/nfsroot.rst. 4354 4355 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 4356 See Documentation/admin-guide/nfs/nfsroot.rst. 4357 4358 nfsrootdebug [NFS] enable nfsroot debugging messages. 4359 See Documentation/admin-guide/nfs/nfsroot.rst. 4360 4361 nmi_backtrace.backtrace_idle [KNL] 4362 Dump stacks even of idle CPUs in response to an 4363 NMI stack-backtrace request. 4364 4365 nmi_debug= [KNL,SH] Specify one or more actions to take 4366 when a NMI is triggered. 4367 Format: [state][,regs][,debounce][,die] 4368 4369 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 4370 Format: [panic,][nopanic,][rNNN,][num] 4371 Valid num: 0 or 1 4372 0 - turn hardlockup detector in nmi_watchdog off 4373 1 - turn hardlockup detector in nmi_watchdog on 4374 rNNN - configure the watchdog with raw perf event 0xNNN 4375 4376 When panic is specified, panic when an NMI watchdog 4377 timeout occurs (or 'nopanic' to not panic on an NMI 4378 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 4379 To disable both hard and soft lockup detectors, 4380 please see 'nowatchdog'. 4381 This is useful when you use a panic=... timeout and 4382 need the box quickly up again. 4383 4384 These settings can be accessed at runtime via 4385 the nmi_watchdog and hardlockup_panic sysctls. 4386 4387 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 4388 emulation library even if a 387 maths coprocessor 4389 is present. 4390 4391 no4lvl [RISCV,EARLY] Disable 4-level and 5-level paging modes. 4392 Forces kernel to use 3-level paging instead. 4393 4394 no5lvl [X86-64,RISCV,EARLY] Disable 5-level paging mode. Forces 4395 kernel to use 4-level paging instead. 4396 4397 noalign [KNL,ARM] 4398 4399 noapic [SMP,APIC,EARLY] Tells the kernel to not make use of any 4400 IOAPICs that may be present in the system. 4401 4402 noapictimer [APIC,X86] Don't set up the APIC timer 4403 4404 noautogroup Disable scheduler automatic task group creation. 4405 4406 nocache [ARM,EARLY] 4407 4408 no_console_suspend 4409 [HW] Never suspend the console 4410 Disable suspending of consoles during suspend and 4411 hibernate operations. Once disabled, debugging 4412 messages can reach various consoles while the rest 4413 of the system is being put to sleep (ie, while 4414 debugging driver suspend/resume hooks). This may 4415 not work reliably with all consoles, but is known 4416 to work with serial and VGA consoles. 4417 To facilitate more flexible debugging, we also add 4418 console_suspend, a printk module parameter to control 4419 it. Users could use console_suspend (usually 4420 /sys/module/printk/parameters/console_suspend) to 4421 turn on/off it dynamically. 4422 4423 no_debug_objects 4424 [KNL,EARLY] Disable object debugging 4425 4426 nodsp [SH] Disable hardware DSP at boot time. 4427 4428 noefi [EFI,EARLY] Disable EFI runtime services support. 4429 4430 no_entry_flush [PPC,EARLY] Don't flush the L1-D cache when entering the kernel. 4431 4432 noexec32 [X86-64] 4433 This affects only 32-bit executables. 4434 noexec32=on: enable non-executable mappings (default) 4435 read doesn't imply executable mappings 4436 noexec32=off: disable non-executable mappings 4437 read implies executable mappings 4438 4439 no_file_caps Tells the kernel not to honor file capabilities. The 4440 only way then for a file to be executed with privilege 4441 is to be setuid root or executed by root. 4442 4443 nofpu [MIPS,SH] Disable hardware FPU at boot time. 4444 4445 nofsgsbase [X86] Disables FSGSBASE instructions. 4446 4447 nofxsr [BUGS=X86-32] Disables x86 floating point extended 4448 register save and restore. The kernel will only save 4449 legacy floating-point registers on task switch. 4450 4451 nogbpages [X86] Do not use GB pages for kernel direct mappings. 4452 4453 no_hash_pointers 4454 [KNL,EARLY] 4455 Alias for "hash_pointers=never". 4456 4457 nohibernate [HIBERNATION] Disable hibernation and resume. 4458 4459 nohlt [ARM,ARM64,MICROBLAZE,MIPS,PPC,RISCV,SH] Forces the kernel to 4460 busy wait in do_idle() and not use the arch_cpu_idle() 4461 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP 4462 to be effective. This is useful on platforms where the 4463 sleep(SH) or wfi(ARM,ARM64) instructions do not work 4464 correctly or when doing power measurements to evaluate 4465 the impact of the sleep instructions. This is also 4466 useful when using JTAG debugger. 4467 4468 nohpet [X86] Don't use the HPET timer. 4469 4470 nohugeiomap [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge I/O mappings. 4471 4472 nohugevmalloc [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge vmalloc mappings. 4473 4474 nohz= [KNL] Boottime enable/disable dynamic ticks 4475 Valid arguments: on, off 4476 Default: on 4477 4478 nohz_full= [KNL,BOOT,SMP,ISOL] 4479 The argument is a cpu list, as described above. 4480 In kernels built with CONFIG_NO_HZ_FULL=y, set 4481 the specified list of CPUs whose tick will be stopped 4482 whenever possible. The boot CPU will be forced outside 4483 the range to maintain the timekeeping. Any CPUs 4484 in this list will have their RCU callbacks offloaded, 4485 just as if they had also been called out in the 4486 rcu_nocbs= boot parameter. 4487 4488 Note that this argument takes precedence over 4489 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 4490 4491 noinitrd [Deprecated,RAM] Tells the kernel not to load any configured 4492 initial RAM disk. Currently this parameter applies to 4493 initrd only, not to initramfs. But it applies to both 4494 in EFI mode. 4495 4496 nointremap [X86-64,Intel-IOMMU,EARLY] Do not enable interrupt 4497 remapping. 4498 [Deprecated - use intremap=off] 4499 4500 noinvpcid [X86,EARLY] Disable the INVPCID cpu feature. 4501 4502 noiotrap [SH] Disables trapped I/O port accesses. 4503 4504 noirqdebug [X86-32] Disables the code which attempts to detect and 4505 disable unhandled interrupt sources. 4506 4507 noisapnp [ISAPNP] Disables ISA PnP code. 4508 4509 nokaslr [KNL,EARLY] 4510 When CONFIG_RANDOMIZE_BASE is set, this disables 4511 kernel and module base offset ASLR (Address Space 4512 Layout Randomization). 4513 4514 no-kvmapf [X86,KVM,EARLY] Disable paravirtualized asynchronous page 4515 fault handling. 4516 4517 no-kvmclock [X86,KVM,EARLY] Disable paravirtualized KVM clock driver 4518 4519 nolapic [X86-32,APIC,EARLY] Do not enable or use the local APIC. 4520 4521 nolapic_timer [X86-32,APIC,EARLY] Do not use the local APIC timer. 4522 4523 nomce [X86-32] Disable Machine Check Exception 4524 4525 nomfgpt [X86-32] Disable Multi-Function General Purpose 4526 Timer usage (for AMD Geode machines). 4527 4528 nomodeset Disable kernel modesetting. Most systems' firmware 4529 sets up a display mode and provides framebuffer memory 4530 for output. With nomodeset, DRM and fbdev drivers will 4531 not load if they could possibly displace the pre- 4532 initialized output. Only the system framebuffer will 4533 be available for use. The respective drivers will not 4534 perform display-mode changes or accelerated rendering. 4535 4536 Useful as error fallback, or for testing and debugging. 4537 4538 nomodule Disable module load 4539 4540 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 4541 shutdown the other cpus. Instead use the REBOOT_VECTOR 4542 irq. 4543 4544 nopat [X86,EARLY] Disable PAT (page attribute table extension of 4545 pagetables) support. 4546 4547 nopcid [X86-64,EARLY] Disable the PCID cpu feature. 4548 4549 nopku [X86] Disable Memory Protection Keys CPU feature found 4550 in some Intel CPUs. 4551 4552 nopti [X86-64,EARLY] 4553 Equivalent to pti=off 4554 4555 nopv= [X86,XEN,KVM,HYPER_V,VMWARE,EARLY] 4556 Disables the PV optimizations forcing the guest to run 4557 as generic guest with no PV drivers. Currently support 4558 XEN HVM, KVM, HYPER_V and VMWARE guest. 4559 4560 nopvspin [X86,XEN,KVM,EARLY] 4561 Disables the qspinlock slow path using PV optimizations 4562 which allow the hypervisor to 'idle' the guest on lock 4563 contention. 4564 4565 norandmaps Don't use address space randomization. Equivalent to 4566 echo 0 > /proc/sys/kernel/randomize_va_space 4567 4568 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 4569 with UP alternatives 4570 4571 noresume [SWSUSP] Disables resume and restores original swap 4572 space. 4573 4574 no-scroll [VGA] Disables scrollback. 4575 This is required for the Braillex ib80-piezo Braille 4576 reader made by F.H. Papenmeier (Germany). 4577 4578 nosgx [X86-64,SGX,EARLY] Disables Intel SGX kernel support. 4579 4580 nosmap [PPC,EARLY] 4581 Disable SMAP (Supervisor Mode Access Prevention) 4582 even if it is supported by processor. 4583 4584 nosmep [PPC64s,EARLY] 4585 Disable SMEP (Supervisor Mode Execution Prevention) 4586 even if it is supported by processor. 4587 4588 nosmp [SMP,EARLY] Tells an SMP kernel to act as a UP kernel, 4589 and disable the IO APIC. legacy for "maxcpus=0". 4590 4591 nosmt [KNL,MIPS,PPC,EARLY] Disable symmetric multithreading (SMT). 4592 Equivalent to smt=1. 4593 4594 [KNL,X86,PPC,S390] Disable symmetric multithreading (SMT). 4595 nosmt=force: Force disable SMT, cannot be undone 4596 via the sysfs control file. 4597 4598 nosoftlockup [KNL] Disable the soft-lockup detector. 4599 4600 nospec_store_bypass_disable 4601 [HW,EARLY] Disable all mitigations for the Speculative 4602 Store Bypass vulnerability 4603 4604 nospectre_bhb [ARM64,EARLY] Disable all mitigations for Spectre-BHB (branch 4605 history injection) vulnerability. System may allow data leaks 4606 with this option. 4607 4608 nospectre_v1 [X86,PPC,EARLY] Disable mitigations for Spectre Variant 1 4609 (bounds check bypass). With this option data leaks are 4610 possible in the system. 4611 4612 nospectre_v2 [X86,PPC_E500,ARM64,EARLY] Disable all mitigations 4613 for the Spectre variant 2 (indirect branch 4614 prediction) vulnerability. System may allow data 4615 leaks with this option. 4616 4617 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES,RISCV,LOONGARCH,EARLY] 4618 Disable paravirtualized steal time accounting. steal time 4619 is computed, but won't influence scheduler behaviour 4620 4621 nosync [HW,M68K] Disables sync negotiation for all devices. 4622 4623 no_timer_check [X86,APIC] Disables the code which tests for broken 4624 timer IRQ sources, i.e., the IO-APIC timer. This can 4625 work around problems with incorrect timer 4626 initialization on some boards. 4627 4628 no_uaccess_flush 4629 [PPC,EARLY] Don't flush the L1-D cache after accessing user data. 4630 4631 novmcoredd [KNL,KDUMP] 4632 Disable device dump. Device dump allows drivers to 4633 append dump data to vmcore so you can collect driver 4634 specified debug info. Drivers can append the data 4635 without any limit and this data is stored in memory, 4636 so this may cause significant memory stress. Disabling 4637 device dump can help save memory but the driver debug 4638 data will be no longer available. This parameter 4639 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 4640 is set. 4641 4642 no-vmw-sched-clock 4643 [X86,PV_OPS,EARLY] Disable paravirtualized VMware 4644 scheduler clock and use the default one. 4645 4646 nowatchdog [KNL] Disable both lockup detectors, i.e. 4647 soft-lockup and NMI watchdog (hard-lockup). 4648 4649 nowb [ARM,EARLY] 4650 4651 nox2apic [X86-64,APIC,EARLY] Do not enable x2APIC mode. 4652 4653 NOTE: this parameter will be ignored on systems with the 4654 LEGACY_XAPIC_DISABLED bit set in the 4655 IA32_XAPIC_DISABLE_STATUS MSR. 4656 4657 noxsave [BUGS=X86] Disables x86 extended register state save 4658 and restore using xsave. The kernel will fallback to 4659 enabling legacy floating-point and sse state. 4660 4661 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 4662 register states. The kernel will fall back to use 4663 xsave to save the states. By using this parameter, 4664 performance of saving the states is degraded because 4665 xsave doesn't support modified optimization while 4666 xsaveopt supports it on xsaveopt enabled systems. 4667 4668 noxsaves [X86] Disables xsaves and xrstors used in saving and 4669 restoring x86 extended register state in compacted 4670 form of xsave area. The kernel will fall back to use 4671 xsaveopt and xrstor to save and restore the states 4672 in standard form of xsave area. By using this 4673 parameter, xsave area per process might occupy more 4674 memory on xsaves enabled systems. 4675 4676 nr_cpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 4677 could support. nr_cpus=n : n >= 1 limits the kernel to 4678 support 'n' processors. It could be larger than the 4679 number of already plugged CPU during bootup, later in 4680 runtime you can physically add extra cpu until it reaches 4681 n. So during boot up some boot time memory for per-cpu 4682 variables need be pre-allocated for later physical cpu 4683 hot plugging. 4684 4685 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 4686 4687 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86, EARLY] 4688 Disable NUMA, Only set up a single NUMA node 4689 spanning all memory. 4690 4691 numa=fake=<size>[MG] 4692 [KNL, ARM64, RISCV, X86, EARLY] 4693 If given as a memory unit, fills all system RAM with 4694 nodes of size interleaved over physical nodes. 4695 4696 numa=fake=<N> 4697 [KNL, ARM64, RISCV, X86, EARLY] 4698 If given as an integer, fills all system RAM with N 4699 fake nodes interleaved over physical nodes. 4700 4701 numa=fake=<N>U 4702 [KNL, ARM64, RISCV, X86, EARLY] 4703 If given as an integer followed by 'U', it will 4704 divide each physical node into N emulated nodes. 4705 4706 numa=noacpi [X86] Don't parse the SRAT table for NUMA setup 4707 4708 numa=nohmat [X86] Don't parse the HMAT table for NUMA setup, or 4709 soft-reserved memory partitioning. 4710 4711 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic 4712 NUMA balancing. 4713 Allowed values are enable and disable 4714 4715 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 4716 'node', 'default' can be specified 4717 This can be set from sysctl after boot. 4718 See Documentation/admin-guide/sysctl/vm.rst for details. 4719 4720 ohci1394_dma=early [HW,EARLY] enable debugging via the ohci1394 driver. 4721 See Documentation/core-api/debugging-via-ohci1394.rst for more 4722 info. 4723 4724 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 4725 Rather than timing out after 20 ms if an EC 4726 command is not properly ACKed, override the length 4727 of the timeout. We have interrupts disabled while 4728 waiting for the ACK, so if this is set too high 4729 interrupts *may* be lost! 4730 4731 omap_mux= [OMAP] Override bootloader pin multiplexing. 4732 Format: <mux_mode0.mode_name=value>... 4733 For example, to override I2C bus2: 4734 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 4735 4736 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 4737 4738 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 4739 4740 boundary - index of last SLC block on Flex-OneNAND. 4741 The remaining blocks are configured as MLC blocks. 4742 lock - Configure if Flex-OneNAND boundary should be locked. 4743 Once locked, the boundary cannot be changed. 4744 1 indicates lock status, 0 indicates unlock status. 4745 4746 oops=panic [KNL,EARLY] 4747 Always panic on oopses. Default is to just kill the 4748 process, but there is a small probability of 4749 deadlocking the machine. 4750 This will also cause panics on machine check exceptions. 4751 Useful together with panic=30 to trigger a reboot. 4752 4753 page_alloc.shuffle= 4754 [KNL] Boolean flag to control whether the page allocator 4755 should randomize its free lists. This parameter can be 4756 used to enable/disable page randomization. The state of 4757 the flag can be read from sysfs at: 4758 /sys/module/page_alloc/parameters/shuffle. 4759 This parameter is only available if CONFIG_SHUFFLE_PAGE_ALLOCATOR=y. 4760 4761 page_owner= [KNL,EARLY] Boot-time page_owner enabling option. 4762 Storage of the information about who allocated 4763 each page is disabled in default. With this switch, 4764 we can turn it on. 4765 on: enable the feature 4766 4767 page_poison= [KNL,EARLY] Boot-time parameter changing the state of 4768 poisoning on the buddy allocator, available with 4769 CONFIG_PAGE_POISONING=y. 4770 off: turn off poisoning (default) 4771 on: turn on poisoning 4772 4773 page_reporting.page_reporting_order= 4774 [KNL] Minimal page reporting order 4775 Format: <integer> 4776 Adjust the minimal page reporting order. The page 4777 reporting is disabled when it exceeds MAX_PAGE_ORDER. 4778 4779 panic= [KNL] Kernel behaviour on panic: delay <timeout> 4780 timeout > 0: seconds before rebooting 4781 timeout = 0: wait forever 4782 timeout < 0: reboot immediately 4783 Format: <timeout> 4784 4785 panic_on_taint= [KNL,EARLY] 4786 Bitmask for conditionally calling panic() in add_taint() 4787 Format: <hex>[,nousertaint] 4788 Hexadecimal bitmask representing the set of TAINT flags 4789 that will cause the kernel to panic when add_taint() is 4790 called with any of the flags in this set. 4791 The optional switch "nousertaint" can be utilized to 4792 prevent userspace forced crashes by writing to sysctl 4793 /proc/sys/kernel/tainted any flagset matching with the 4794 bitmask set on panic_on_taint. 4795 See Documentation/admin-guide/tainted-kernels.rst for 4796 extra details on the taint flags that users can pick 4797 to compose the bitmask to assign to panic_on_taint. 4798 4799 panic_on_warn=1 panic() instead of WARN(). Useful to cause kdump 4800 on a WARN(). 4801 4802 panic_print= Bitmask for printing system info when panic happens. 4803 User can chose combination of the following bits: 4804 bit 0: print all tasks info 4805 bit 1: print system memory info 4806 bit 2: print timer info 4807 bit 3: print locks info if CONFIG_LOCKDEP is on 4808 bit 4: print ftrace buffer 4809 bit 5: replay all kernel messages on consoles at the end of panic 4810 bit 6: print all CPUs backtrace (if available in the arch) 4811 bit 7: print only tasks in uninterruptible (blocked) state 4812 *Be aware* that this option may print a _lot_ of lines, 4813 so there are risks of losing older messages in the log. 4814 Use this option carefully, maybe worth to setup a 4815 bigger log buffer with "log_buf_len" along with this. 4816 4817 panic_sys_info= A comma separated list of extra information to be dumped 4818 on panic. 4819 Format: val[,val...] 4820 Where @val can be any of the following: 4821 4822 tasks: print all tasks info 4823 mem: print system memory info 4824 timers: print timers info 4825 locks: print locks info if CONFIG_LOCKDEP is on 4826 ftrace: print ftrace buffer 4827 all_bt: print all CPUs backtrace (if available in the arch) 4828 blocked_tasks: print only tasks in uninterruptible (blocked) state 4829 4830 This is a human readable alternative to the 'panic_print' option. 4831 4832 panic_console_replay 4833 When panic happens, replay all kernel messages on 4834 consoles at the end of panic. 4835 4836 parkbd.port= [HW] Parallel port number the keyboard adapter is 4837 connected to, default is 0. 4838 Format: <parport#> 4839 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 4840 0 for XT, 1 for AT (default is AT). 4841 Format: <mode> 4842 4843 parport= [HW,PPT] Specify parallel ports. 0 disables. 4844 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 4845 Use 'auto' to force the driver to use any 4846 IRQ/DMA settings detected (the default is to 4847 ignore detected IRQ/DMA settings because of 4848 possible conflicts). You can specify the base 4849 address, IRQ, and DMA settings; IRQ and DMA 4850 should be numbers, or 'auto' (for using detected 4851 settings on that particular port), or 'nofifo' 4852 (to avoid using a FIFO even if it is detected). 4853 Parallel ports are assigned in the order they 4854 are specified on the command line, starting 4855 with parport0. 4856 4857 parport_init_mode= [HW,PPT] 4858 Configure VIA parallel port to operate in 4859 a specific mode. This is necessary on Pegasos 4860 computer where firmware has no options for setting 4861 up parallel port mode and sets it to spp. 4862 Currently this function knows 686a and 8231 chips. 4863 Format: [spp|ps2|epp|ecp|ecpepp] 4864 4865 pata_legacy.all= [HW,LIBATA] 4866 Format: <int> 4867 Set to non-zero to probe primary and secondary ISA 4868 port ranges on PCI systems where no PCI PATA device 4869 has been found at either range. Disabled by default. 4870 4871 pata_legacy.autospeed= [HW,LIBATA] 4872 Format: <int> 4873 Set to non-zero if a chip is present that snoops speed 4874 changes. Disabled by default. 4875 4876 pata_legacy.ht6560a= [HW,LIBATA] 4877 Format: <int> 4878 Set to 1, 2, or 3 for HT 6560A on the primary channel, 4879 the secondary channel, or both channels respectively. 4880 Disabled by default. 4881 4882 pata_legacy.ht6560b= [HW,LIBATA] 4883 Format: <int> 4884 Set to 1, 2, or 3 for HT 6560B on the primary channel, 4885 the secondary channel, or both channels respectively. 4886 Disabled by default. 4887 4888 pata_legacy.iordy_mask= [HW,LIBATA] 4889 Format: <int> 4890 IORDY enable mask. Set individual bits to allow IORDY 4891 for the respective channel. Bit 0 is for the first 4892 legacy channel handled by this driver, bit 1 is for 4893 the second channel, and so on. The sequence will often 4894 correspond to the primary legacy channel, the secondary 4895 legacy channel, and so on, but the handling of a PCI 4896 bus and the use of other driver options may interfere 4897 with the sequence. By default IORDY is allowed across 4898 all channels. 4899 4900 pata_legacy.opti82c46x= [HW,LIBATA] 4901 Format: <int> 4902 Set to 1, 2, or 3 for Opti 82c611A on the primary 4903 channel, the secondary channel, or both channels 4904 respectively. Disabled by default. 4905 4906 pata_legacy.opti82c611a= [HW,LIBATA] 4907 Format: <int> 4908 Set to 1, 2, or 3 for Opti 82c465MV on the primary 4909 channel, the secondary channel, or both channels 4910 respectively. Disabled by default. 4911 4912 pata_legacy.pio_mask= [HW,LIBATA] 4913 Format: <int> 4914 PIO mode mask for autospeed devices. Set individual 4915 bits to allow the use of the respective PIO modes. 4916 Bit 0 is for mode 0, bit 1 is for mode 1, and so on. 4917 All modes allowed by default. 4918 4919 pata_legacy.probe_all= [HW,LIBATA] 4920 Format: <int> 4921 Set to non-zero to probe tertiary and further ISA 4922 port ranges on PCI systems. Disabled by default. 4923 4924 pata_legacy.probe_mask= [HW,LIBATA] 4925 Format: <int> 4926 Probe mask for legacy ISA PATA ports. Depending on 4927 platform configuration and the use of other driver 4928 options up to 6 legacy ports are supported: 0x1f0, 4929 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing 4930 of individual ports can be disabled by setting the 4931 corresponding bits in the mask to 1. Bit 0 is for 4932 the first port in the list above (0x1f0), and so on. 4933 By default all supported ports are probed. 4934 4935 pata_legacy.qdi= [HW,LIBATA] 4936 Format: <int> 4937 Set to non-zero to probe QDI controllers. By default 4938 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise. 4939 4940 pata_legacy.winbond= [HW,LIBATA] 4941 Format: <int> 4942 Set to non-zero to probe Winbond controllers. Use 4943 the standard I/O port (0x130) if 1, otherwise the 4944 value given is the I/O port to use (typically 0x1b0). 4945 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE, 4946 0 otherwise. 4947 4948 pata_platform.pio_mask= [HW,LIBATA] 4949 Format: <int> 4950 Supported PIO mode mask. Set individual bits to allow 4951 the use of the respective PIO modes. Bit 0 is for 4952 mode 0, bit 1 is for mode 1, and so on. Mode 0 only 4953 allowed by default. 4954 4955 pause_on_oops=<int> 4956 Halt all CPUs after the first oops has been printed for 4957 the specified number of seconds. This is to be used if 4958 your oopses keep scrolling off the screen. 4959 4960 pcbit= [HW,ISDN] 4961 4962 pci=option[,option...] [PCI,EARLY] various PCI subsystem options. 4963 4964 Some options herein operate on a specific device 4965 or a set of devices (<pci_dev>). These are 4966 specified in one of the following formats: 4967 4968 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 4969 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 4970 4971 Note: the first format specifies a PCI 4972 bus/device/function address which may change 4973 if new hardware is inserted, if motherboard 4974 firmware changes, or due to changes caused 4975 by other kernel parameters. If the 4976 domain is left unspecified, it is 4977 taken to be zero. Optionally, a path 4978 to a device through multiple device/function 4979 addresses can be specified after the base 4980 address (this is more robust against 4981 renumbering issues). The second format 4982 selects devices using IDs from the 4983 configuration space which may match multiple 4984 devices in the system. 4985 4986 earlydump dump PCI config space before the kernel 4987 changes anything 4988 off [X86] don't probe for the PCI bus 4989 bios [X86-32] force use of PCI BIOS, don't access 4990 the hardware directly. Use this if your machine 4991 has a non-standard PCI host bridge. 4992 nobios [X86-32] disallow use of PCI BIOS, only direct 4993 hardware access methods are allowed. Use this 4994 if you experience crashes upon bootup and you 4995 suspect they are caused by the BIOS. 4996 conf1 [X86] Force use of PCI Configuration Access 4997 Mechanism 1 (config address in IO port 0xCF8, 4998 data in IO port 0xCFC, both 32-bit). 4999 conf2 [X86] Force use of PCI Configuration Access 5000 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 5001 the function, IO port 0xCFA, also 8-bit, sets 5002 bus number. The config space is then accessed 5003 through ports 0xC000-0xCFFF). 5004 See http://wiki.osdev.org/PCI for more info 5005 on the configuration access mechanisms. 5006 noaer [PCIE] If the PCIEAER kernel config parameter is 5007 enabled, this kernel boot option can be used to 5008 disable the use of PCIE advanced error reporting. 5009 nodomains [PCI] Disable support for multiple PCI 5010 root domains (aka PCI segments, in ACPI-speak). 5011 nommconf [X86] Disable use of MMCONFIG for PCI 5012 Configuration 5013 check_enable_amd_mmconf [X86] check for and enable 5014 properly configured MMIO access to PCI 5015 config space on AMD family 10h CPU 5016 nomsi [MSI] If the PCI_MSI kernel config parameter is 5017 enabled, this kernel boot option can be used to 5018 disable the use of MSI interrupts system-wide. 5019 noioapicquirk [APIC] Disable all boot interrupt quirks. 5020 Safety option to keep boot IRQs enabled. This 5021 should never be necessary. 5022 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 5023 primary IO-APIC for bridges that cannot disable 5024 boot IRQs. This fixes a source of spurious IRQs 5025 when the system masks IRQs. 5026 noioapicreroute [APIC] Disable workaround that uses the 5027 boot IRQ equivalent of an IRQ that connects to 5028 a chipset where boot IRQs cannot be disabled. 5029 The opposite of ioapicreroute. 5030 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 5031 routing table. These calls are known to be buggy 5032 on several machines and they hang the machine 5033 when used, but on other computers it's the only 5034 way to get the interrupt routing table. Try 5035 this option if the kernel is unable to allocate 5036 IRQs or discover secondary PCI buses on your 5037 motherboard. 5038 rom [X86] Assign address space to expansion ROMs. 5039 Use with caution as certain devices share 5040 address decoders between ROMs and other 5041 resources. 5042 norom [X86] Do not assign address space to 5043 expansion ROMs that do not already have 5044 BIOS assigned address ranges. 5045 nobar [X86] Do not assign address space to the 5046 BARs that weren't assigned by the BIOS. 5047 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 5048 assigned automatically to PCI devices. You can 5049 make the kernel exclude IRQs of your ISA cards 5050 this way. 5051 pirqaddr=0xAAAAA [X86] Specify the physical address 5052 of the PIRQ table (normally generated 5053 by the BIOS) if it is outside the 5054 F0000h-100000h range. 5055 lastbus=N [X86] Scan all buses thru bus #N. Can be 5056 useful if the kernel is unable to find your 5057 secondary buses and you want to tell it 5058 explicitly which ones they are. 5059 assign-busses [X86] Always assign all PCI bus 5060 numbers ourselves, overriding 5061 whatever the firmware may have done. 5062 usepirqmask [X86] Honor the possible IRQ mask stored 5063 in the BIOS $PIR table. This is needed on 5064 some systems with broken BIOSes, notably 5065 some HP Pavilion N5400 and Omnibook XE3 5066 notebooks. This will have no effect if ACPI 5067 IRQ routing is enabled. 5068 noacpi [X86] Do not use ACPI for IRQ routing 5069 or for PCI scanning. 5070 use_crs [X86] Use PCI host bridge window information 5071 from ACPI. On BIOSes from 2008 or later, this 5072 is enabled by default. If you need to use this, 5073 please report a bug. 5074 nocrs [X86] Ignore PCI host bridge windows from ACPI. 5075 If you need to use this, please report a bug. 5076 use_e820 [X86] Use E820 reservations to exclude parts of 5077 PCI host bridge windows. This is a workaround 5078 for BIOS defects in host bridge _CRS methods. 5079 If you need to use this, please report a bug to 5080 <linux-pci@vger.kernel.org>. 5081 no_e820 [X86] Ignore E820 reservations for PCI host 5082 bridge windows. This is the default on modern 5083 hardware. If you need to use this, please report 5084 a bug to <linux-pci@vger.kernel.org>. 5085 routeirq Do IRQ routing for all PCI devices. 5086 This is normally done in pci_enable_device(), 5087 so this option is a temporary workaround 5088 for broken drivers that don't call it. 5089 skip_isa_align [X86] do not align io start addr, so can 5090 handle more pci cards 5091 noearly [X86] Don't do any early type 1 scanning. 5092 This might help on some broken boards which 5093 machine check when some devices' config space 5094 is read. But various workarounds are disabled 5095 and some IOMMU drivers will not work. 5096 bfsort Sort PCI devices into breadth-first order. 5097 This sorting is done to get a device 5098 order compatible with older (<= 2.4) kernels. 5099 nobfsort Don't sort PCI devices into breadth-first order. 5100 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 5101 tuning and use the BIOS-configured MPS defaults. 5102 pcie_bus_safe Set every device's MPS to the largest value 5103 supported by all devices below the root complex. 5104 pcie_bus_perf Set device MPS to the largest allowable MPS 5105 based on its parent bus. Also set MRRS (Max 5106 Read Request Size) to the largest supported 5107 value (no larger than the MPS that the device 5108 or bus can support) for best performance. 5109 pcie_bus_peer2peer Set every device's MPS to 128B, which 5110 every device is guaranteed to support. This 5111 configuration allows peer-to-peer DMA between 5112 any pair of devices, possibly at the cost of 5113 reduced performance. This also guarantees 5114 that hot-added devices will work. 5115 cbiosize=nn[KMG] The fixed amount of bus space which is 5116 reserved for the CardBus bridge's IO window. 5117 The default value is 256 bytes. 5118 cbmemsize=nn[KMG] The fixed amount of bus space which is 5119 reserved for the CardBus bridge's memory 5120 window. The default value is 64 megabytes. 5121 resource_alignment= 5122 Format: 5123 [<order of align>@]<pci_dev>[; ...] 5124 Specifies alignment and device to reassign 5125 aligned memory resources. How to 5126 specify the device is described above. 5127 If <order of align> is not specified, 5128 PAGE_SIZE is used as alignment. 5129 A PCI-PCI bridge can be specified if resource 5130 windows need to be expanded. 5131 To specify the alignment for several 5132 instances of a device, the PCI vendor, 5133 device, subvendor, and subdevice may be 5134 specified, e.g., 12@pci:8086:9c22:103c:198f 5135 for 4096-byte alignment. 5136 ecrc= Enable/disable PCIe ECRC (transaction layer 5137 end-to-end CRC checking). Only effective if 5138 OS has native AER control (either granted by 5139 ACPI _OSC or forced via "pcie_ports=native") 5140 bios: Use BIOS/firmware settings. This is the 5141 the default. 5142 off: Turn ECRC off 5143 on: Turn ECRC on. 5144 hpiosize=nn[KMG] The fixed amount of bus space which is 5145 reserved for hotplug bridge's IO window. 5146 Default size is 256 bytes. 5147 hpmmiosize=nn[KMG] The fixed amount of bus space which is 5148 reserved for hotplug bridge's MMIO window. 5149 Default size is 2 megabytes. 5150 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 5151 reserved for hotplug bridge's MMIO_PREF window. 5152 Default size is 2 megabytes. 5153 hpmemsize=nn[KMG] The fixed amount of bus space which is 5154 reserved for hotplug bridge's MMIO and 5155 MMIO_PREF window. 5156 Default size is 2 megabytes. 5157 hpbussize=nn The minimum amount of additional bus numbers 5158 reserved for buses below a hotplug bridge. 5159 Default is 1. 5160 realloc= Enable/disable reallocating PCI bridge resources 5161 if allocations done by BIOS are too small to 5162 accommodate resources required by all child 5163 devices. 5164 off: Turn realloc off 5165 on: Turn realloc on 5166 realloc same as realloc=on 5167 noari do not use PCIe ARI. 5168 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 5169 do not use PCIe ATS (and IOMMU device IOTLB). 5170 pcie_scan_all Scan all possible PCIe devices. Otherwise we 5171 only look for one device below a PCIe downstream 5172 port. 5173 big_root_window Try to add a big 64bit memory window to the PCIe 5174 root complex on AMD CPUs. Some GFX hardware 5175 can resize a BAR to allow access to all VRAM. 5176 Adding the window is slightly risky (it may 5177 conflict with unreported devices), so this 5178 taints the kernel. 5179 disable_acs_redir=<pci_dev>[; ...] 5180 Specify one or more PCI devices (in the format 5181 specified above) separated by semicolons. 5182 Each device specified will have the PCI ACS 5183 redirect capabilities forced off which will 5184 allow P2P traffic between devices through 5185 bridges without forcing it upstream. Note: 5186 this removes isolation between devices and 5187 may put more devices in an IOMMU group. 5188 config_acs= 5189 Format: 5190 <ACS flags>@<pci_dev>[; ...] 5191 Specify one or more PCI devices (in the format 5192 specified above) optionally prepended with flags 5193 and separated by semicolons. The respective 5194 capabilities will be enabled, disabled or 5195 unchanged based on what is specified in 5196 flags. 5197 5198 ACS Flags is defined as follows: 5199 bit-0 : ACS Source Validation 5200 bit-1 : ACS Translation Blocking 5201 bit-2 : ACS P2P Request Redirect 5202 bit-3 : ACS P2P Completion Redirect 5203 bit-4 : ACS Upstream Forwarding 5204 bit-5 : ACS P2P Egress Control 5205 bit-6 : ACS Direct Translated P2P 5206 Each bit can be marked as: 5207 '0' – force disabled 5208 '1' – force enabled 5209 'x' – unchanged 5210 For example, 5211 pci=config_acs=10x@pci:0:0 5212 would configure all devices that support 5213 ACS to enable P2P Request Redirect, disable 5214 Translation Blocking, and leave Source 5215 Validation unchanged from whatever power-up 5216 or firmware set it to. 5217 5218 Note: this may remove isolation between devices 5219 and may put more devices in an IOMMU group. 5220 force_floating [S390] Force usage of floating interrupts. 5221 nomio [S390] Do not use MIO instructions. 5222 norid [S390] ignore the RID field and force use of 5223 one PCI domain per PCI function 5224 notph [PCIE] If the PCIE_TPH kernel config parameter 5225 is enabled, this kernel boot option can be used 5226 to disable PCIe TLP Processing Hints support 5227 system-wide. 5228 5229 pcie_aspm= [PCIE] Forcibly enable or ignore PCIe Active State Power 5230 Management. 5231 off Don't touch ASPM configuration at all. Leave any 5232 configuration done by firmware unchanged. 5233 force Enable ASPM even on devices that claim not to support it. 5234 WARNING: Forcing ASPM on may cause system lockups. 5235 5236 pcie_ports= [PCIE] PCIe port services handling: 5237 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 5238 even if the platform doesn't give the OS permission to 5239 use them. This may cause conflicts if the platform 5240 also tries to use these services. 5241 dpc-native Use native PCIe service for DPC only. May 5242 cause conflicts if firmware uses AER or DPC. 5243 compat Disable native PCIe services (PME, AER, DPC, PCIe 5244 hotplug). 5245 5246 pcie_port_pm= [PCIE] PCIe port power management handling: 5247 off Disable power management of all PCIe ports 5248 force Forcibly enable power management of all PCIe ports 5249 5250 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 5251 nomsi Do not use MSI for native PCIe PME signaling (this makes 5252 all PCIe root ports use INTx for all services). 5253 5254 pcmv= [HW,PCMCIA] BadgePAD 4 5255 5256 pd_ignore_unused 5257 [PM] 5258 Keep all power-domains already enabled by bootloader on, 5259 even if no driver has claimed them. This is useful 5260 for debug and development, but should not be 5261 needed on a platform with proper driver support. 5262 5263 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 5264 boot time. 5265 Format: { 0 | 1 } 5266 See arch/parisc/kernel/pdc_chassis.c 5267 5268 percpu_alloc= [MM,EARLY] 5269 Select which percpu first chunk allocator to use. 5270 Currently supported values are "embed" and "page". 5271 Archs may support subset or none of the selections. 5272 See comments in mm/percpu.c for details on each 5273 allocator. This parameter is primarily for debugging 5274 and performance comparison. 5275 5276 pirq= [SMP,APIC] Manual mp-table setup 5277 See Documentation/arch/x86/i386/IO-APIC.rst. 5278 5279 plip= [PPT,NET] Parallel port network link 5280 Format: { parport<nr> | timid | 0 } 5281 See also Documentation/admin-guide/parport.rst. 5282 5283 pmtmr= [X86] Manual setup of pmtmr I/O Port. 5284 Override pmtimer IOPort with a hex value. 5285 e.g. pmtmr=0x508 5286 5287 pmu_override= [PPC] Override the PMU. 5288 This option takes over the PMU facility, so it is no 5289 longer usable by perf. Setting this option starts the 5290 PMU counters by setting MMCR0 to 0 (the FC bit is 5291 cleared). If a number is given, then MMCR1 is set to 5292 that number, otherwise (e.g., 'pmu_override=on'), MMCR1 5293 remains 0. 5294 5295 pm_async= [PM] 5296 Format: off 5297 This parameter sets the initial value of the 5298 /sys/power/pm_async sysfs knob at boot time. 5299 If set to "off", disables asynchronous suspend and 5300 resume of devices during system-wide power transitions. 5301 This can be useful on platforms where device 5302 dependencies are not well-defined, or for debugging 5303 power management issues. Asynchronous operations are 5304 enabled by default. 5305 5306 5307 pm_debug_messages [SUSPEND,KNL] 5308 Enable suspend/resume debug messages during boot up. 5309 5310 pnp.debug=1 [PNP] 5311 Enable PNP debug messages (depends on the 5312 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 5313 via /sys/module/pnp/parameters/debug. We always show 5314 current resource usage; turning this on also shows 5315 possible settings and some assignment information. 5316 5317 pnpacpi= [ACPI] 5318 { off } 5319 5320 pnpbios= [ISAPNP] 5321 { on | off | curr | res | no-curr | no-res } 5322 5323 pnp_reserve_irq= 5324 [ISAPNP] Exclude IRQs for the autoconfiguration 5325 5326 pnp_reserve_dma= 5327 [ISAPNP] Exclude DMAs for the autoconfiguration 5328 5329 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 5330 Ranges are in pairs (I/O port base and size). 5331 5332 pnp_reserve_mem= 5333 [ISAPNP] Exclude memory regions for the 5334 autoconfiguration. 5335 Ranges are in pairs (memory base and size). 5336 5337 ports= [IP_VS_FTP] IPVS ftp helper module 5338 Default is 21. 5339 Up to 8 (IP_VS_APP_MAX_PORTS) ports 5340 may be specified. 5341 Format: <port>,<port>.... 5342 5343 possible_cpus= [SMP,S390,X86] 5344 Format: <unsigned int> 5345 Set the number of possible CPUs, overriding the 5346 regular discovery mechanisms (such as ACPI/FW, etc). 5347 5348 powersave=off [PPC] This option disables power saving features. 5349 It specifically disables cpuidle and sets the 5350 platform machine description specific power_save 5351 function to NULL. On Idle the CPU just reduces 5352 execution priority. 5353 5354 ppc_strict_facility_enable 5355 [PPC,ENABLE] This option catches any kernel floating point, 5356 Altivec, VSX and SPE outside of regions specifically 5357 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 5358 There is some performance impact when enabling this. 5359 5360 ppc_tm= [PPC,EARLY] 5361 Format: {"off"} 5362 Disable Hardware Transactional Memory 5363 5364 preempt= [KNL] 5365 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC 5366 none - Limited to cond_resched() calls 5367 voluntary - Limited to cond_resched() and might_sleep() calls 5368 full - Any section that isn't explicitly preempt disabled 5369 can be preempted anytime. Tasks will also yield 5370 contended spinlocks (if the critical section isn't 5371 explicitly preempt disabled beyond the lock itself). 5372 lazy - Scheduler controlled. Similar to full but instead 5373 of preempting the task immediately, the task gets 5374 one HZ tick time to yield itself before the 5375 preemption will be forced. One preemption is when the 5376 task returns to user space. 5377 5378 print-fatal-signals= 5379 [KNL] debug: print fatal signals 5380 5381 If enabled, warn about various signal handling 5382 related application anomalies: too many signals, 5383 too many POSIX.1 timers, fatal signals causing a 5384 coredump - etc. 5385 5386 If you hit the warning due to signal overflow, 5387 you might want to try "ulimit -i unlimited". 5388 5389 default: off. 5390 5391 printk.always_kmsg_dump= 5392 Trigger kmsg_dump for cases other than kernel oops or 5393 panics 5394 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5395 default: disabled 5396 5397 printk.console_no_auto_verbose= 5398 Disable console loglevel raise on oops, panic 5399 or lockdep-detected issues (only if lock debug is on). 5400 With an exception to setups with low baudrate on 5401 serial console, keeping this 0 is a good choice 5402 in order to provide more debug information. 5403 Format: <bool> 5404 default: 0 (auto_verbose is enabled) 5405 5406 printk.debug_non_panic_cpus= 5407 Allows storing messages from non-panic CPUs into 5408 the printk log buffer during panic(). They are 5409 flushed to consoles by the panic-CPU on 5410 a best-effort basis. 5411 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5412 Default: disabled 5413 5414 printk.devkmsg={on,off,ratelimit} 5415 Control writing to /dev/kmsg. 5416 on - unlimited logging to /dev/kmsg from userspace 5417 off - logging to /dev/kmsg disabled 5418 ratelimit - ratelimit the logging 5419 Default: ratelimit 5420 5421 printk.time= Show timing data prefixed to each printk message line 5422 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5423 5424 proc_mem.force_override= [KNL] 5425 Format: {always | ptrace | never} 5426 Traditionally /proc/pid/mem allows memory permissions to be 5427 overridden without restrictions. This option may be set to 5428 restrict that. Can be one of: 5429 - 'always': traditional behavior always allows mem overrides. 5430 - 'ptrace': only allow mem overrides for active ptracers. 5431 - 'never': never allow mem overrides. 5432 If not specified, default is the CONFIG_PROC_MEM_* choice. 5433 5434 processor.max_cstate= [HW,ACPI] 5435 Limit processor to maximum C-state 5436 max_cstate=9 overrides any DMI blacklist limit. 5437 5438 processor.nocst [HW,ACPI] 5439 Ignore the _CST method to determine C-states, 5440 instead using the legacy FADT method 5441 5442 profile= [KNL] Enable kernel profiling via /proc/profile 5443 Format: [<profiletype>,]<number> 5444 Param: <profiletype>: "schedule" or "kvm" 5445 [defaults to kernel profiling] 5446 Param: "schedule" - profile schedule points. 5447 Param: "kvm" - profile VM exits. 5448 Param: <number> - step/bucket size as a power of 2 for 5449 statistical time based profiling. 5450 5451 prot_virt= [S390] enable hosting protected virtual machines 5452 isolated from the hypervisor (if hardware supports 5453 that). If enabled, the default kernel base address 5454 might be overridden even when Kernel Address Space 5455 Layout Randomization is disabled. 5456 Format: <bool> 5457 5458 psi= [KNL] Enable or disable pressure stall information 5459 tracking. 5460 Format: <bool> 5461 5462 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 5463 probe for; one of (bare|imps|exps|lifebook|any). 5464 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 5465 per second. 5466 psmouse.resetafter= [HW,MOUSE] 5467 Try to reset the device after so many bad packets 5468 (0 = never). 5469 psmouse.resolution= 5470 [HW,MOUSE] Set desired mouse resolution, in dpi. 5471 psmouse.smartscroll= 5472 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 5473 0 = disabled, 1 = enabled (default). 5474 5475 pstore.backend= Specify the name of the pstore backend to use 5476 5477 pti= [X86-64] Control Page Table Isolation of user and 5478 kernel address spaces. Disabling this feature 5479 removes hardening, but improves performance of 5480 system calls and interrupts. 5481 5482 on - unconditionally enable 5483 off - unconditionally disable 5484 auto - kernel detects whether your CPU model is 5485 vulnerable to issues that PTI mitigates 5486 5487 Not specifying this option is equivalent to pti=auto. 5488 5489 pty.legacy_count= 5490 [KNL] Number of legacy pty's. Overwrites compiled-in 5491 default number. 5492 5493 quiet [KNL,EARLY] Disable most log messages 5494 5495 r128= [HW,DRM] 5496 5497 radix_hcall_invalidate=on [PPC/PSERIES] 5498 Disable RADIX GTSE feature and use hcall for TLB 5499 invalidate. 5500 5501 raid= [HW,RAID] 5502 See Documentation/admin-guide/md.rst. 5503 5504 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 5505 See Documentation/admin-guide/blockdev/ramdisk.rst. 5506 5507 ramdisk_start= [Deprecated,RAM] RAM disk image start address 5508 5509 random.trust_cpu=off 5510 [KNL,EARLY] Disable trusting the use of the CPU's 5511 random number generator (if available) to 5512 initialize the kernel's RNG. 5513 5514 random.trust_bootloader=off 5515 [KNL,EARLY] Disable trusting the use of the a seed 5516 passed by the bootloader (if available) to 5517 initialize the kernel's RNG. 5518 5519 randomize_kstack_offset= 5520 [KNL,EARLY] Enable or disable kernel stack offset 5521 randomization, which provides roughly 5 bits of 5522 entropy, frustrating memory corruption attacks 5523 that depend on stack address determinism or 5524 cross-syscall address exposures. This is only 5525 available on architectures that have defined 5526 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET. 5527 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5528 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT. 5529 5530 ras=option[,option,...] [KNL] RAS-specific options 5531 5532 cec_disable [X86] 5533 Disable the Correctable Errors Collector, 5534 see CONFIG_RAS_CEC help text. 5535 5536 rcu_nocbs[=cpu-list] 5537 [KNL] The optional argument is a cpu list, 5538 as described above. 5539 5540 In kernels built with CONFIG_RCU_NOCB_CPU=y, 5541 enable the no-callback CPU mode, which prevents 5542 such CPUs' callbacks from being invoked in 5543 softirq context. Invocation of such CPUs' RCU 5544 callbacks will instead be offloaded to "rcuox/N" 5545 kthreads created for that purpose, where "x" is 5546 "p" for RCU-preempt, "s" for RCU-sched, and "g" 5547 for the kthreads that mediate grace periods; and 5548 "N" is the CPU number. This reduces OS jitter on 5549 the offloaded CPUs, which can be useful for HPC 5550 and real-time workloads. It can also improve 5551 energy efficiency for asymmetric multiprocessors. 5552 5553 If a cpulist is passed as an argument, the specified 5554 list of CPUs is set to no-callback mode from boot. 5555 5556 Otherwise, if the '=' sign and the cpulist 5557 arguments are omitted, no CPU will be set to 5558 no-callback mode from boot but the mode may be 5559 toggled at runtime via cpusets. 5560 5561 Note that this argument takes precedence over 5562 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 5563 5564 rcu_nocb_poll [KNL] 5565 Rather than requiring that offloaded CPUs 5566 (specified by rcu_nocbs= above) explicitly 5567 awaken the corresponding "rcuoN" kthreads, 5568 make these kthreads poll for callbacks. 5569 This improves the real-time response for the 5570 offloaded CPUs by relieving them of the need to 5571 wake up the corresponding kthread, but degrades 5572 energy efficiency by requiring that the kthreads 5573 periodically wake up to do the polling. 5574 5575 rcutree.blimit= [KNL] 5576 Set maximum number of finished RCU callbacks to 5577 process in one batch. 5578 5579 rcutree.csd_lock_suppress_rcu_stall= [KNL] 5580 Do only a one-line RCU CPU stall warning when 5581 there is an ongoing too-long CSD-lock wait. 5582 5583 rcutree.do_rcu_barrier= [KNL] 5584 Request a call to rcu_barrier(). This is 5585 throttled so that userspace tests can safely 5586 hammer on the sysfs variable if they so choose. 5587 If triggered before the RCU grace-period machinery 5588 is fully active, this will error out with EAGAIN. 5589 5590 rcutree.dump_tree= [KNL] 5591 Dump the structure of the rcu_node combining tree 5592 out at early boot. This is used for diagnostic 5593 purposes, to verify correct tree setup. 5594 5595 rcutree.gp_cleanup_delay= [KNL] 5596 Set the number of jiffies to delay each step of 5597 RCU grace-period cleanup. 5598 5599 rcutree.gp_init_delay= [KNL] 5600 Set the number of jiffies to delay each step of 5601 RCU grace-period initialization. 5602 5603 rcutree.gp_preinit_delay= [KNL] 5604 Set the number of jiffies to delay each step of 5605 RCU grace-period pre-initialization, that is, 5606 the propagation of recent CPU-hotplug changes up 5607 the rcu_node combining tree. 5608 5609 rcutree.jiffies_till_first_fqs= [KNL] 5610 Set delay from grace-period initialization to 5611 first attempt to force quiescent states. 5612 Units are jiffies, minimum value is zero, 5613 and maximum value is HZ. 5614 5615 rcutree.jiffies_till_next_fqs= [KNL] 5616 Set delay between subsequent attempts to force 5617 quiescent states. Units are jiffies, minimum 5618 value is one, and maximum value is HZ. 5619 5620 rcutree.jiffies_till_sched_qs= [KNL] 5621 Set required age in jiffies for a 5622 given grace period before RCU starts 5623 soliciting quiescent-state help from 5624 rcu_note_context_switch() and cond_resched(). 5625 If not specified, the kernel will calculate 5626 a value based on the most recent settings 5627 of rcutree.jiffies_till_first_fqs 5628 and rcutree.jiffies_till_next_fqs. 5629 This calculated value may be viewed in 5630 rcutree.jiffies_to_sched_qs. Any attempt to set 5631 rcutree.jiffies_to_sched_qs will be cheerfully 5632 overwritten. 5633 5634 rcutree.kthread_prio= [KNL,BOOT] 5635 Set the SCHED_FIFO priority of the RCU per-CPU 5636 kthreads (rcuc/N). This value is also used for 5637 the priority of the RCU boost threads (rcub/N) 5638 and for the RCU grace-period kthreads (rcu_bh, 5639 rcu_preempt, and rcu_sched). If RCU_BOOST is 5640 set, valid values are 1-99 and the default is 1 5641 (the least-favored priority). Otherwise, when 5642 RCU_BOOST is not set, valid values are 0-99 and 5643 the default is zero (non-realtime operation). 5644 When RCU_NOCB_CPU is set, also adjust the 5645 priority of NOCB callback kthreads. 5646 5647 rcutree.nocb_nobypass_lim_per_jiffy= [KNL] 5648 On callback-offloaded (rcu_nocbs) CPUs, 5649 RCU reduces the lock contention that would 5650 otherwise be caused by callback floods through 5651 use of the ->nocb_bypass list. However, in the 5652 common non-flooded case, RCU queues directly to 5653 the main ->cblist in order to avoid the extra 5654 overhead of the ->nocb_bypass list and its lock. 5655 But if there are too many callbacks queued during 5656 a single jiffy, RCU pre-queues the callbacks into 5657 the ->nocb_bypass queue. The definition of "too 5658 many" is supplied by this kernel boot parameter. 5659 5660 rcutree.nohz_full_patience_delay= [KNL] 5661 On callback-offloaded (rcu_nocbs) CPUs, avoid 5662 disturbing RCU unless the grace period has 5663 reached the specified age in milliseconds. 5664 Defaults to zero. Large values will be capped 5665 at five seconds. All values will be rounded down 5666 to the nearest value representable by jiffies. 5667 5668 rcutree.qhimark= [KNL] 5669 Set threshold of queued RCU callbacks beyond which 5670 batch limiting is disabled. 5671 5672 rcutree.qlowmark= [KNL] 5673 Set threshold of queued RCU callbacks below which 5674 batch limiting is re-enabled. 5675 5676 rcutree.qovld= [KNL] 5677 Set threshold of queued RCU callbacks beyond which 5678 RCU's force-quiescent-state scan will aggressively 5679 enlist help from cond_resched() and sched IPIs to 5680 help CPUs more quickly reach quiescent states. 5681 Set to less than zero to make this be set based 5682 on rcutree.qhimark at boot time and to zero to 5683 disable more aggressive help enlistment. 5684 5685 rcutree.rcu_delay_page_cache_fill_msec= [KNL] 5686 Set the page-cache refill delay (in milliseconds) 5687 in response to low-memory conditions. The range 5688 of permitted values is in the range 0:100000. 5689 5690 rcutree.rcu_divisor= [KNL] 5691 Set the shift-right count to use to compute 5692 the callback-invocation batch limit bl from 5693 the number of callbacks queued on this CPU. 5694 The result will be bounded below by the value of 5695 the rcutree.blimit kernel parameter. Every bl 5696 callbacks, the softirq handler will exit in 5697 order to allow the CPU to do other work. 5698 5699 Please note that this callback-invocation batch 5700 limit applies only to non-offloaded callback 5701 invocation. Offloaded callbacks are instead 5702 invoked in the context of an rcuoc kthread, which 5703 scheduler will preempt as it does any other task. 5704 5705 rcutree.rcu_fanout_exact= [KNL] 5706 Disable autobalancing of the rcu_node combining 5707 tree. This is used by rcutorture, and might 5708 possibly be useful for architectures having high 5709 cache-to-cache transfer latencies. 5710 5711 rcutree.rcu_fanout_leaf= [KNL] 5712 Change the number of CPUs assigned to each 5713 leaf rcu_node structure. Useful for very 5714 large systems, which will choose the value 64, 5715 and for NUMA systems with large remote-access 5716 latencies, which will choose a value aligned 5717 with the appropriate hardware boundaries. 5718 5719 rcutree.rcu_min_cached_objs= [KNL] 5720 Minimum number of objects which are cached and 5721 maintained per one CPU. Object size is equal 5722 to PAGE_SIZE. The cache allows to reduce the 5723 pressure to page allocator, also it makes the 5724 whole algorithm to behave better in low memory 5725 condition. 5726 5727 rcutree.rcu_nocb_gp_stride= [KNL] 5728 Set the number of NOCB callback kthreads in 5729 each group, which defaults to the square root 5730 of the number of CPUs. Larger numbers reduce 5731 the wakeup overhead on the global grace-period 5732 kthread, but increases that same overhead on 5733 each group's NOCB grace-period kthread. 5734 5735 rcutree.rcu_kick_kthreads= [KNL] 5736 Cause the grace-period kthread to get an extra 5737 wake_up() if it sleeps three times longer than 5738 it should at force-quiescent-state time. 5739 This wake_up() will be accompanied by a 5740 WARN_ONCE() splat and an ftrace_dump(). 5741 5742 rcutree.rcu_resched_ns= [KNL] 5743 Limit the time spend invoking a batch of RCU 5744 callbacks to the specified number of nanoseconds. 5745 By default, this limit is checked only once 5746 every 32 callbacks in order to limit the pain 5747 inflicted by local_clock() overhead. 5748 5749 rcutree.rcu_unlock_delay= [KNL] 5750 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 5751 this specifies an rcu_read_unlock()-time delay 5752 in microseconds. This defaults to zero. 5753 Larger delays increase the probability of 5754 catching RCU pointer leaks, that is, buggy use 5755 of RCU-protected pointers after the relevant 5756 rcu_read_unlock() has completed. 5757 5758 rcutree.sysrq_rcu= [KNL] 5759 Commandeer a sysrq key to dump out Tree RCU's 5760 rcu_node tree with an eye towards determining 5761 why a new grace period has not yet started. 5762 5763 rcutree.use_softirq= [KNL] 5764 If set to zero, move all RCU_SOFTIRQ processing to 5765 per-CPU rcuc kthreads. Defaults to a non-zero 5766 value, meaning that RCU_SOFTIRQ is used by default. 5767 Specify rcutree.use_softirq=0 to use rcuc kthreads. 5768 5769 But note that CONFIG_PREEMPT_RT=y kernels disable 5770 this kernel boot parameter, forcibly setting it 5771 to zero. 5772 5773 rcutree.enable_rcu_lazy= [KNL] 5774 To save power, batch RCU callbacks and flush after 5775 delay, memory pressure or callback list growing too 5776 big. 5777 5778 rcutree.rcu_normal_wake_from_gp= [KNL] 5779 Reduces a latency of synchronize_rcu() call. This approach 5780 maintains its own track of synchronize_rcu() callers, so it 5781 does not interact with regular callbacks because it does not 5782 use a call_rcu[_hurry]() path. Please note, this is for a 5783 normal grace period. 5784 5785 How to enable it: 5786 5787 echo 1 > /sys/module/rcutree/parameters/rcu_normal_wake_from_gp 5788 or pass a boot parameter "rcutree.rcu_normal_wake_from_gp=1" 5789 5790 Default is 1 if num_possible_cpus() <= 16 and it is not explicitly 5791 disabled by the boot parameter passing 0. 5792 5793 rcuscale.gp_async= [KNL] 5794 Measure performance of asynchronous 5795 grace-period primitives such as call_rcu(). 5796 5797 rcuscale.gp_async_max= [KNL] 5798 Specify the maximum number of outstanding 5799 callbacks per writer thread. When a writer 5800 thread exceeds this limit, it invokes the 5801 corresponding flavor of rcu_barrier() to allow 5802 previously posted callbacks to drain. 5803 5804 rcuscale.gp_exp= [KNL] 5805 Measure performance of expedited synchronous 5806 grace-period primitives. 5807 5808 rcuscale.holdoff= [KNL] 5809 Set test-start holdoff period. The purpose of 5810 this parameter is to delay the start of the 5811 test until boot completes in order to avoid 5812 interference. 5813 5814 rcuscale.kfree_by_call_rcu= [KNL] 5815 In kernels built with CONFIG_RCU_LAZY=y, test 5816 call_rcu() instead of kfree_rcu(). 5817 5818 rcuscale.kfree_mult= [KNL] 5819 Instead of allocating an object of size kfree_obj, 5820 allocate one of kfree_mult * sizeof(kfree_obj). 5821 Defaults to 1. 5822 5823 rcuscale.kfree_rcu_test= [KNL] 5824 Set to measure performance of kfree_rcu() flooding. 5825 5826 rcuscale.kfree_rcu_test_double= [KNL] 5827 Test the double-argument variant of kfree_rcu(). 5828 If this parameter has the same value as 5829 rcuscale.kfree_rcu_test_single, both the single- 5830 and double-argument variants are tested. 5831 5832 rcuscale.kfree_rcu_test_single= [KNL] 5833 Test the single-argument variant of kfree_rcu(). 5834 If this parameter has the same value as 5835 rcuscale.kfree_rcu_test_double, both the single- 5836 and double-argument variants are tested. 5837 5838 rcuscale.kfree_nthreads= [KNL] 5839 The number of threads running loops of kfree_rcu(). 5840 5841 rcuscale.kfree_alloc_num= [KNL] 5842 Number of allocations and frees done in an iteration. 5843 5844 rcuscale.kfree_loops= [KNL] 5845 Number of loops doing rcuscale.kfree_alloc_num number 5846 of allocations and frees. 5847 5848 rcuscale.minruntime= [KNL] 5849 Set the minimum test run time in seconds. This 5850 does not affect the data-collection interval, 5851 but instead allows better measurement of things 5852 like CPU consumption. 5853 5854 rcuscale.nreaders= [KNL] 5855 Set number of RCU readers. The value -1 selects 5856 N, where N is the number of CPUs. A value 5857 "n" less than -1 selects N-n+1, where N is again 5858 the number of CPUs. For example, -2 selects N 5859 (the number of CPUs), -3 selects N+1, and so on. 5860 A value of "n" less than or equal to -N selects 5861 a single reader. 5862 5863 rcuscale.nwriters= [KNL] 5864 Set number of RCU writers. The values operate 5865 the same as for rcuscale.nreaders. 5866 N, where N is the number of CPUs 5867 5868 rcuscale.scale_type= [KNL] 5869 Specify the RCU implementation to test. 5870 5871 rcuscale.shutdown= [KNL] 5872 Shut the system down after performance tests 5873 complete. This is useful for hands-off automated 5874 testing. 5875 5876 rcuscale.verbose= [KNL] 5877 Enable additional printk() statements. 5878 5879 rcuscale.writer_holdoff= [KNL] 5880 Write-side holdoff between grace periods, 5881 in microseconds. The default of zero says 5882 no holdoff. 5883 5884 rcuscale.writer_holdoff_jiffies= [KNL] 5885 Additional write-side holdoff between grace 5886 periods, but in jiffies. The default of zero 5887 says no holdoff. 5888 5889 rcutorture.fqs_duration= [KNL] 5890 Set duration of force_quiescent_state bursts 5891 in microseconds. 5892 5893 rcutorture.fqs_holdoff= [KNL] 5894 Set holdoff time within force_quiescent_state bursts 5895 in microseconds. 5896 5897 rcutorture.fqs_stutter= [KNL] 5898 Set wait time between force_quiescent_state bursts 5899 in seconds. 5900 5901 rcutorture.fwd_progress= [KNL] 5902 Specifies the number of kthreads to be used 5903 for RCU grace-period forward-progress testing 5904 for the types of RCU supporting this notion. 5905 Defaults to 1 kthread, values less than zero or 5906 greater than the number of CPUs cause the number 5907 of CPUs to be used. 5908 5909 rcutorture.fwd_progress_div= [KNL] 5910 Specify the fraction of a CPU-stall-warning 5911 period to do tight-loop forward-progress testing. 5912 5913 rcutorture.fwd_progress_holdoff= [KNL] 5914 Number of seconds to wait between successive 5915 forward-progress tests. 5916 5917 rcutorture.fwd_progress_need_resched= [KNL] 5918 Enclose cond_resched() calls within checks for 5919 need_resched() during tight-loop forward-progress 5920 testing. 5921 5922 rcutorture.gp_cond= [KNL] 5923 Use conditional/asynchronous update-side 5924 normal-grace-period primitives, if available. 5925 5926 rcutorture.gp_cond_exp= [KNL] 5927 Use conditional/asynchronous update-side 5928 expedited-grace-period primitives, if available. 5929 5930 rcutorture.gp_cond_full= [KNL] 5931 Use conditional/asynchronous update-side 5932 normal-grace-period primitives that also take 5933 concurrent expedited grace periods into account, 5934 if available. 5935 5936 rcutorture.gp_cond_exp_full= [KNL] 5937 Use conditional/asynchronous update-side 5938 expedited-grace-period primitives that also take 5939 concurrent normal grace periods into account, 5940 if available. 5941 5942 rcutorture.gp_cond_wi= [KNL] 5943 Nominal wait interval for normal conditional 5944 grace periods (specified by rcutorture's 5945 gp_cond and gp_cond_full module parameters), 5946 in microseconds. The actual wait interval will 5947 be randomly selected to nanosecond granularity up 5948 to this wait interval. Defaults to 16 jiffies, 5949 for example, 16,000 microseconds on a system 5950 with HZ=1000. 5951 5952 rcutorture.gp_cond_wi_exp= [KNL] 5953 Nominal wait interval for expedited conditional 5954 grace periods (specified by rcutorture's 5955 gp_cond_exp and gp_cond_exp_full module 5956 parameters), in microseconds. The actual wait 5957 interval will be randomly selected to nanosecond 5958 granularity up to this wait interval. Defaults to 5959 128 microseconds. 5960 5961 rcutorture.gp_exp= [KNL] 5962 Use expedited update-side primitives, if available. 5963 5964 rcutorture.gp_normal= [KNL] 5965 Use normal (non-expedited) asynchronous 5966 update-side primitives, if available. 5967 5968 rcutorture.gp_poll= [KNL] 5969 Use polled update-side normal-grace-period 5970 primitives, if available. 5971 5972 rcutorture.gp_poll_exp= [KNL] 5973 Use polled update-side expedited-grace-period 5974 primitives, if available. 5975 5976 rcutorture.gp_poll_full= [KNL] 5977 Use polled update-side normal-grace-period 5978 primitives that also take concurrent expedited 5979 grace periods into account, if available. 5980 5981 rcutorture.gp_poll_exp_full= [KNL] 5982 Use polled update-side expedited-grace-period 5983 primitives that also take concurrent normal 5984 grace periods into account, if available. 5985 5986 rcutorture.gp_poll_wi= [KNL] 5987 Nominal wait interval for normal conditional 5988 grace periods (specified by rcutorture's 5989 gp_poll and gp_poll_full module parameters), 5990 in microseconds. The actual wait interval will 5991 be randomly selected to nanosecond granularity up 5992 to this wait interval. Defaults to 16 jiffies, 5993 for example, 16,000 microseconds on a system 5994 with HZ=1000. 5995 5996 rcutorture.gp_poll_wi_exp= [KNL] 5997 Nominal wait interval for expedited conditional 5998 grace periods (specified by rcutorture's 5999 gp_poll_exp and gp_poll_exp_full module 6000 parameters), in microseconds. The actual wait 6001 interval will be randomly selected to nanosecond 6002 granularity up to this wait interval. Defaults to 6003 128 microseconds. 6004 6005 rcutorture.gp_sync= [KNL] 6006 Use normal (non-expedited) synchronous 6007 update-side primitives, if available. If all 6008 of rcutorture.gp_cond=, rcutorture.gp_exp=, 6009 rcutorture.gp_normal=, and rcutorture.gp_sync= 6010 are zero, rcutorture acts as if is interpreted 6011 they are all non-zero. 6012 6013 rcutorture.gpwrap_lag= [KNL] 6014 Enable grace-period wrap lag testing. Setting 6015 to false prevents the gpwrap lag test from 6016 running. Default is true. 6017 6018 rcutorture.gpwrap_lag_gps= [KNL] 6019 Set the value for grace-period wrap lag during 6020 active lag testing periods. This controls how many 6021 grace periods differences we tolerate between 6022 rdp and rnp's gp_seq before setting overflow flag. 6023 The default is always set to 8. 6024 6025 rcutorture.gpwrap_lag_cycle_mins= [KNL] 6026 Set the total cycle duration for gpwrap lag 6027 testing in minutes. This is the total time for 6028 one complete cycle of active and inactive 6029 testing periods. Default is 30 minutes. 6030 6031 rcutorture.gpwrap_lag_active_mins= [KNL] 6032 Set the duration for which gpwrap lag is active 6033 within each cycle, in minutes. During this time, 6034 the grace-period wrap lag will be set to the 6035 value specified by gpwrap_lag_gps. Default is 6036 5 minutes. 6037 6038 rcutorture.irqreader= [KNL] 6039 Run RCU readers from irq handlers, or, more 6040 accurately, from a timer handler. Not all RCU 6041 flavors take kindly to this sort of thing. 6042 6043 rcutorture.leakpointer= [KNL] 6044 Leak an RCU-protected pointer out of the reader. 6045 This can of course result in splats, and is 6046 intended to test the ability of things like 6047 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 6048 such leaks. 6049 6050 rcutorture.n_barrier_cbs= [KNL] 6051 Set callbacks/threads for rcu_barrier() testing. 6052 6053 rcutorture.nfakewriters= [KNL] 6054 Set number of concurrent RCU writers. These just 6055 stress RCU, they don't participate in the actual 6056 test, hence the "fake". 6057 6058 rcutorture.nocbs_nthreads= [KNL] 6059 Set number of RCU callback-offload togglers. 6060 Zero (the default) disables toggling. 6061 6062 rcutorture.nocbs_toggle= [KNL] 6063 Set the delay in milliseconds between successive 6064 callback-offload toggling attempts. 6065 6066 rcutorture.nreaders= [KNL] 6067 Set number of RCU readers. The value -1 selects 6068 N-1, where N is the number of CPUs. A value 6069 "n" less than -1 selects N-n-2, where N is again 6070 the number of CPUs. For example, -2 selects N 6071 (the number of CPUs), -3 selects N+1, and so on. 6072 6073 rcutorture.object_debug= [KNL] 6074 Enable debug-object double-call_rcu() testing. 6075 6076 rcutorture.onoff_holdoff= [KNL] 6077 Set time (s) after boot for CPU-hotplug testing. 6078 6079 rcutorture.onoff_interval= [KNL] 6080 Set time (jiffies) between CPU-hotplug operations, 6081 or zero to disable CPU-hotplug testing. 6082 6083 rcutorture.preempt_duration= [KNL] 6084 Set duration (in milliseconds) of preemptions 6085 by a high-priority FIFO real-time task. Set to 6086 zero (the default) to disable. The CPUs to 6087 preempt are selected randomly from the set that 6088 are online at a given point in time. Races with 6089 CPUs going offline are ignored, with that attempt 6090 at preemption skipped. 6091 6092 rcutorture.preempt_interval= [KNL] 6093 Set interval (in milliseconds, defaulting to one 6094 second) between preemptions by a high-priority 6095 FIFO real-time task. This delay is mediated 6096 by an hrtimer and is further fuzzed to avoid 6097 inadvertent synchronizations. 6098 6099 rcutorture.read_exit_burst= [KNL] 6100 The number of times in a given read-then-exit 6101 episode that a set of read-then-exit kthreads 6102 is spawned. 6103 6104 rcutorture.read_exit_delay= [KNL] 6105 The delay, in seconds, between successive 6106 read-then-exit testing episodes. 6107 6108 rcutorture.reader_flavor= [KNL] 6109 A bit mask indicating which readers to use. 6110 If there is more than one bit set, the readers 6111 are entered from low-order bit up, and are 6112 exited in the opposite order. For SRCU, the 6113 0x1 bit is normal readers, 0x2 NMI-safe readers, 6114 and 0x4 light-weight readers. 6115 6116 rcutorture.shuffle_interval= [KNL] 6117 Set task-shuffle interval (s). Shuffling tasks 6118 allows some CPUs to go into dyntick-idle mode 6119 during the rcutorture test. 6120 6121 rcutorture.shutdown_secs= [KNL] 6122 Set time (s) after boot system shutdown. This 6123 is useful for hands-off automated testing. 6124 6125 rcutorture.stall_cpu= [KNL] 6126 Duration of CPU stall (s) to test RCU CPU stall 6127 warnings, zero to disable. 6128 6129 rcutorture.stall_cpu_block= [KNL] 6130 Sleep while stalling if set. This will result 6131 in warnings from preemptible RCU in addition to 6132 any other stall-related activity. Note that 6133 in kernels built with CONFIG_PREEMPTION=n and 6134 CONFIG_PREEMPT_COUNT=y, this parameter will 6135 cause the CPU to pass through a quiescent state. 6136 Given CONFIG_PREEMPTION=n, this will suppress 6137 RCU CPU stall warnings, but will instead result 6138 in scheduling-while-atomic splats. 6139 6140 Use of this module parameter results in splats. 6141 6142 6143 rcutorture.stall_cpu_holdoff= [KNL] 6144 Time to wait (s) after boot before inducing stall. 6145 6146 rcutorture.stall_cpu_irqsoff= [KNL] 6147 Disable interrupts while stalling if set, but only 6148 on the first stall in the set. 6149 6150 rcutorture.stall_cpu_repeat= [KNL] 6151 Number of times to repeat the stall sequence, 6152 so that rcutorture.stall_cpu_repeat=3 will result 6153 in four stall sequences. 6154 6155 rcutorture.stall_gp_kthread= [KNL] 6156 Duration (s) of forced sleep within RCU 6157 grace-period kthread to test RCU CPU stall 6158 warnings, zero to disable. If both stall_cpu 6159 and stall_gp_kthread are specified, the 6160 kthread is starved first, then the CPU. 6161 6162 rcutorture.stat_interval= [KNL] 6163 Time (s) between statistics printk()s. 6164 6165 rcutorture.stutter= [KNL] 6166 Time (s) to stutter testing, for example, specifying 6167 five seconds causes the test to run for five seconds, 6168 wait for five seconds, and so on. This tests RCU's 6169 ability to transition abruptly to and from idle. 6170 6171 rcutorture.test_boost= [KNL] 6172 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 6173 "Maybe" means test if the RCU implementation 6174 under test support RCU priority boosting. 6175 6176 rcutorture.test_boost_duration= [KNL] 6177 Duration (s) of each individual boost test. 6178 6179 rcutorture.test_boost_holdoff= [KNL] 6180 Holdoff time (s) from start of test to the start 6181 of RCU priority-boost testing. Defaults to zero, 6182 that is, no holdoff. 6183 6184 rcutorture.test_boost_interval= [KNL] 6185 Interval (s) between each boost test. 6186 6187 rcutorture.test_no_idle_hz= [KNL] 6188 Test RCU's dyntick-idle handling. See also the 6189 rcutorture.shuffle_interval parameter. 6190 6191 rcutorture.torture_type= [KNL] 6192 Specify the RCU implementation to test. 6193 6194 rcutorture.verbose= [KNL] 6195 Enable additional printk() statements. 6196 6197 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 6198 Dump ftrace buffer after reporting RCU CPU 6199 stall warning. 6200 6201 rcupdate.rcu_cpu_stall_notifiers= [KNL] 6202 Provide RCU CPU stall notifiers, but see the 6203 warnings in the RCU_CPU_STALL_NOTIFIER Kconfig 6204 option's help text. TL;DR: You almost certainly 6205 do not want rcupdate.rcu_cpu_stall_notifiers. 6206 6207 rcupdate.rcu_cpu_stall_suppress= [KNL] 6208 Suppress RCU CPU stall warning messages. 6209 6210 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 6211 Suppress RCU CPU stall warning messages and 6212 rcutorture writer stall warnings that occur 6213 during early boot, that is, during the time 6214 before the init task is spawned. 6215 6216 rcupdate.rcu_cpu_stall_timeout= [KNL] 6217 Set timeout for RCU CPU stall warning messages. 6218 The value is in seconds and the maximum allowed 6219 value is 300 seconds. 6220 6221 rcupdate.rcu_exp_cpu_stall_timeout= [KNL] 6222 Set timeout for expedited RCU CPU stall warning 6223 messages. The value is in milliseconds 6224 and the maximum allowed value is 21000 6225 milliseconds. Please note that this value is 6226 adjusted to an arch timer tick resolution. 6227 Setting this to zero causes the value from 6228 rcupdate.rcu_cpu_stall_timeout to be used (after 6229 conversion from seconds to milliseconds). 6230 6231 rcupdate.rcu_cpu_stall_cputime= [KNL] 6232 Provide statistics on the cputime and count of 6233 interrupts and tasks during the sampling period. For 6234 multiple continuous RCU stalls, all sampling periods 6235 begin at half of the first RCU stall timeout. 6236 6237 rcupdate.rcu_exp_stall_task_details= [KNL] 6238 Print stack dumps of any tasks blocking the 6239 current expedited RCU grace period during an 6240 expedited RCU CPU stall warning. 6241 6242 rcupdate.rcu_expedited= [KNL] 6243 Use expedited grace-period primitives, for 6244 example, synchronize_rcu_expedited() instead 6245 of synchronize_rcu(). This reduces latency, 6246 but can increase CPU utilization, degrade 6247 real-time latency, and degrade energy efficiency. 6248 No effect on CONFIG_TINY_RCU kernels. 6249 6250 rcupdate.rcu_normal= [KNL] 6251 Use only normal grace-period primitives, 6252 for example, synchronize_rcu() instead of 6253 synchronize_rcu_expedited(). This improves 6254 real-time latency, CPU utilization, and 6255 energy efficiency, but can expose users to 6256 increased grace-period latency. This parameter 6257 overrides rcupdate.rcu_expedited. No effect on 6258 CONFIG_TINY_RCU kernels. 6259 6260 rcupdate.rcu_normal_after_boot= [KNL] 6261 Once boot has completed (that is, after 6262 rcu_end_inkernel_boot() has been invoked), use 6263 only normal grace-period primitives. No effect 6264 on CONFIG_TINY_RCU kernels. 6265 6266 But note that CONFIG_PREEMPT_RT=y kernels enables 6267 this kernel boot parameter, forcibly setting 6268 it to the value one, that is, converting any 6269 post-boot attempt at an expedited RCU grace 6270 period to instead use normal non-expedited 6271 grace-period processing. 6272 6273 rcupdate.rcu_task_collapse_lim= [KNL] 6274 Set the maximum number of callbacks present 6275 at the beginning of a grace period that allows 6276 the RCU Tasks flavors to collapse back to using 6277 a single callback queue. This switching only 6278 occurs when rcupdate.rcu_task_enqueue_lim is 6279 set to the default value of -1. 6280 6281 rcupdate.rcu_task_contend_lim= [KNL] 6282 Set the minimum number of callback-queuing-time 6283 lock-contention events per jiffy required to 6284 cause the RCU Tasks flavors to switch to per-CPU 6285 callback queuing. This switching only occurs 6286 when rcupdate.rcu_task_enqueue_lim is set to 6287 the default value of -1. 6288 6289 rcupdate.rcu_task_enqueue_lim= [KNL] 6290 Set the number of callback queues to use for the 6291 RCU Tasks family of RCU flavors. The default 6292 of -1 allows this to be automatically (and 6293 dynamically) adjusted. This parameter is intended 6294 for use in testing. 6295 6296 rcupdate.rcu_task_lazy_lim= [KNL] 6297 Number of callbacks on a given CPU that will 6298 cancel laziness on that CPU. Use -1 to disable 6299 cancellation of laziness, but be advised that 6300 doing so increases the danger of OOM due to 6301 callback flooding. 6302 6303 rcupdate.rcu_task_stall_info= [KNL] 6304 Set initial timeout in jiffies for RCU task stall 6305 informational messages, which give some indication 6306 of the problem for those not patient enough to 6307 wait for ten minutes. Informational messages are 6308 only printed prior to the stall-warning message 6309 for a given grace period. Disable with a value 6310 less than or equal to zero. Defaults to ten 6311 seconds. A change in value does not take effect 6312 until the beginning of the next grace period. 6313 6314 rcupdate.rcu_task_stall_info_mult= [KNL] 6315 Multiplier for time interval between successive 6316 RCU task stall informational messages for a given 6317 RCU tasks grace period. This value is clamped 6318 to one through ten, inclusive. It defaults to 6319 the value three, so that the first informational 6320 message is printed 10 seconds into the grace 6321 period, the second at 40 seconds, the third at 6322 160 seconds, and then the stall warning at 600 6323 seconds would prevent a fourth at 640 seconds. 6324 6325 rcupdate.rcu_task_stall_timeout= [KNL] 6326 Set timeout in jiffies for RCU task stall 6327 warning messages. Disable with a value less 6328 than or equal to zero. Defaults to ten minutes. 6329 A change in value does not take effect until 6330 the beginning of the next grace period. 6331 6332 rcupdate.rcu_tasks_lazy_ms= [KNL] 6333 Set timeout in milliseconds RCU Tasks asynchronous 6334 callback batching for call_rcu_tasks(). 6335 A negative value will take the default. A value 6336 of zero will disable batching. Batching is 6337 always disabled for synchronize_rcu_tasks(). 6338 6339 rcupdate.rcu_self_test= [KNL] 6340 Run the RCU early boot self tests 6341 6342 rdinit= [KNL] 6343 Format: <full_path> 6344 Run specified binary instead of /init from the ramdisk, 6345 used for early userspace startup. See initrd. 6346 6347 rdrand= [X86,EARLY] 6348 force - Override the decision by the kernel to hide the 6349 advertisement of RDRAND support (this affects 6350 certain AMD processors because of buggy BIOS 6351 support, specifically around the suspend/resume 6352 path). 6353 6354 rdt= [HW,X86,RDT] 6355 Turn on/off individual RDT features. List is: 6356 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 6357 mba, smba, bmec, abmc, sdciae, energy[:guid], 6358 perf[:guid]. 6359 E.g. to turn on cmt and turn off mba use: 6360 rdt=cmt,!mba 6361 To turn off all energy telemetry monitoring and ensure that 6362 perf telemetry monitoring associated with guid 0x12345 6363 is enabled use: 6364 rdt=!energy,perf:0x12345 6365 6366 reboot= [KNL] 6367 Format (x86 or x86_64): 6368 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \ 6369 [[,]s[mp]#### \ 6370 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 6371 [[,]f[orce] 6372 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 6373 (prefix with 'panic_' to set mode for panic 6374 reboot only), 6375 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 6376 reboot_force is either force or not specified, 6377 reboot_cpu is s[mp]#### with #### being the processor 6378 to be used for rebooting. 6379 6380 acpi 6381 Use the ACPI RESET_REG in the FADT. If ACPI is not 6382 configured or the ACPI reset does not work, the reboot 6383 path attempts the reset using the keyboard controller. 6384 6385 bios 6386 Use the CPU reboot vector for warm reset 6387 6388 cold 6389 Set the cold reboot flag 6390 6391 default 6392 There are some built-in platform specific "quirks" 6393 - you may see: "reboot: <name> series board detected. 6394 Selecting <type> for reboots." In the case where you 6395 think the quirk is in error (e.g. you have newer BIOS, 6396 or newer board) using this option will ignore the 6397 built-in quirk table, and use the generic default 6398 reboot actions. 6399 6400 efi 6401 Use efi reset_system runtime service. If EFI is not 6402 configured or the EFI reset does not work, the reboot 6403 path attempts the reset using the keyboard controller. 6404 6405 force 6406 Don't stop other CPUs on reboot. This can make reboot 6407 more reliable in some cases. 6408 6409 kbd 6410 Use the keyboard controller. cold reset (default) 6411 6412 pci 6413 Use a write to the PCI config space register 0xcf9 to 6414 trigger reboot. 6415 6416 triple 6417 Force a triple fault (init) 6418 6419 warm 6420 Don't set the cold reboot flag 6421 6422 Using warm reset will be much faster especially on big 6423 memory systems because the BIOS will not go through 6424 the memory check. Disadvantage is that not all 6425 hardware will be completely reinitialized on reboot so 6426 there may be boot problems on some systems. 6427 6428 6429 refscale.holdoff= [KNL] 6430 Set test-start holdoff period. The purpose of 6431 this parameter is to delay the start of the 6432 test until boot completes in order to avoid 6433 interference. 6434 6435 refscale.lookup_instances= [KNL] 6436 Number of data elements to use for the forms of 6437 SLAB_TYPESAFE_BY_RCU testing. A negative number 6438 is negated and multiplied by nr_cpu_ids, while 6439 zero specifies nr_cpu_ids. 6440 6441 refscale.loops= [KNL] 6442 Set the number of loops over the synchronization 6443 primitive under test. Increasing this number 6444 reduces noise due to loop start/end overhead, 6445 but the default has already reduced the per-pass 6446 noise to a handful of picoseconds on ca. 2020 6447 x86 laptops. 6448 6449 refscale.nreaders= [KNL] 6450 Set number of readers. The default value of -1 6451 selects N, where N is roughly 75% of the number 6452 of CPUs. A value of zero is an interesting choice. 6453 6454 refscale.nruns= [KNL] 6455 Set number of runs, each of which is dumped onto 6456 the console log. 6457 6458 refscale.readdelay= [KNL] 6459 Set the read-side critical-section duration, 6460 measured in microseconds. 6461 6462 refscale.scale_type= [KNL] 6463 Specify the read-protection implementation to test. 6464 6465 refscale.shutdown= [KNL] 6466 Shut down the system at the end of the performance 6467 test. This defaults to 1 (shut it down) when 6468 refscale is built into the kernel and to 0 (leave 6469 it running) when refscale is built as a module. 6470 6471 refscale.verbose= [KNL] 6472 Enable additional printk() statements. 6473 6474 refscale.verbose_batched= [KNL] 6475 Batch the additional printk() statements. If zero 6476 (the default) or negative, print everything. Otherwise, 6477 print every Nth verbose statement, where N is the value 6478 specified. 6479 6480 regulator_ignore_unused 6481 [REGULATOR] 6482 Prevents regulator framework from disabling regulators 6483 that are unused, due no driver claiming them. This may 6484 be useful for debug and development, but should not be 6485 needed on a platform with proper driver support. 6486 6487 relax_domain_level= 6488 [KNL, SMP] Set scheduler's default relax_domain_level. 6489 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 6490 6491 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 6492 Format: <base1>,<size1>[,<base2>,<size2>,...] 6493 Reserve I/O ports or memory so the kernel won't use 6494 them. If <base> is less than 0x10000, the region 6495 is assumed to be I/O ports; otherwise it is memory. 6496 6497 reserve_mem= [RAM] 6498 Format: nn[KMG]:<align>:<label> 6499 Reserve physical memory and label it with a name that 6500 other subsystems can use to access it. This is typically 6501 used for systems that do not wipe the RAM, and this command 6502 line will try to reserve the same physical memory on 6503 soft reboots. Note, it is not guaranteed to be the same 6504 location. For example, if anything about the system changes 6505 or if booting a different kernel. It can also fail if KASLR 6506 places the kernel at the location of where the RAM reservation 6507 was from a previous boot, the new reservation will be at a 6508 different location. 6509 Any subsystem using this feature must add a way to verify 6510 that the contents of the physical memory is from a previous 6511 boot, as there may be cases where the memory will not be 6512 located at the same location. 6513 6514 The format is size:align:label for example, to request 6515 12 megabytes of 4096 alignment for ramoops: 6516 6517 reserve_mem=12M:4096:oops ramoops.mem_name=oops 6518 6519 reservetop= [X86-32,EARLY] 6520 Format: nn[KMG] 6521 Reserves a hole at the top of the kernel virtual 6522 address space. 6523 6524 reset_devices [KNL] Force drivers to reset the underlying device 6525 during initialization. 6526 6527 resume= [SWSUSP] 6528 Specify the partition device for software suspend 6529 Format: 6530 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 6531 6532 resume_offset= [SWSUSP] 6533 Specify the offset from the beginning of the partition 6534 given by "resume=" at which the swap header is located, 6535 in <PAGE_SIZE> units (needed only for swap files). 6536 See Documentation/power/swsusp-and-swap-files.rst 6537 6538 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 6539 read the resume files 6540 6541 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 6542 Useful for devices that are detected asynchronously 6543 (e.g. USB and MMC devices). 6544 6545 retain_initrd [RAM] Keep initrd memory after extraction. After boot, it will 6546 be accessible via /sys/firmware/initrd. 6547 6548 retbleed= [X86] Control mitigation of RETBleed (Arbitrary 6549 Speculative Code Execution with Return Instructions) 6550 vulnerability. 6551 6552 AMD-based UNRET and IBPB mitigations alone do not stop 6553 sibling threads from influencing the predictions of other 6554 sibling threads. For that reason, STIBP is used on pro- 6555 cessors that support it, and mitigate SMT on processors 6556 that don't. 6557 6558 off - no mitigation 6559 auto - automatically select a mitigation 6560 auto,nosmt - automatically select a mitigation, 6561 disabling SMT if necessary for 6562 the full mitigation (only on Zen1 6563 and older without STIBP). 6564 ibpb - On AMD, mitigate short speculation 6565 windows on basic block boundaries too. 6566 Safe, highest perf impact. It also 6567 enables STIBP if present. Not suitable 6568 on Intel. 6569 ibpb,nosmt - Like "ibpb" above but will disable SMT 6570 when STIBP is not available. This is 6571 the alternative for systems which do not 6572 have STIBP. 6573 unret - Force enable untrained return thunks, 6574 only effective on AMD f15h-f17h based 6575 systems. 6576 unret,nosmt - Like unret, but will disable SMT when STIBP 6577 is not available. This is the alternative for 6578 systems which do not have STIBP. 6579 6580 Selecting 'auto' will choose a mitigation method at run 6581 time according to the CPU. 6582 6583 Not specifying this option is equivalent to retbleed=auto. 6584 6585 rfkill.default_state= 6586 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 6587 etc. communication is blocked by default. 6588 1 Unblocked. 6589 6590 rfkill.master_switch_mode= 6591 0 The "airplane mode" button does nothing. 6592 1 The "airplane mode" button toggles between everything 6593 blocked and the previous configuration. 6594 2 The "airplane mode" button toggles between everything 6595 blocked and everything unblocked. 6596 6597 ring3mwait=disable 6598 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 6599 CPUs. 6600 6601 riscv_isa_fallback [RISCV,EARLY] 6602 When CONFIG_RISCV_ISA_FALLBACK is not enabled, permit 6603 falling back to detecting extension support by parsing 6604 "riscv,isa" property on devicetree systems when the 6605 replacement properties are not found. See the Kconfig 6606 entry for RISCV_ISA_FALLBACK. 6607 6608 ro [KNL] Mount root device read-only on boot 6609 6610 rodata= [KNL,EARLY] 6611 on Mark read-only kernel memory as read-only (default). 6612 off Leave read-only kernel memory writable for debugging. 6613 noalias Mark read-only kernel memory as read-only but retain 6614 writable aliases in the direct map for regions outside 6615 of the kernel image. [arm64] 6616 6617 rockchip.usb_uart 6618 [EARLY] 6619 Enable the uart passthrough on the designated usb port 6620 on Rockchip SoCs. When active, the signals of the 6621 debug-uart get routed to the D+ and D- pins of the usb 6622 port and the regular usb controller gets disabled. 6623 6624 root= [KNL] Root filesystem 6625 Usually this is a block device specifier of some kind, 6626 see the early_lookup_bdev comment in 6627 block/early-lookup.c for details. 6628 Alternatively this can be "ram" for the legacy initial 6629 ramdisk, "nfs" and "cifs" for root on a network file 6630 system, or "mtd" and "ubi" for mounting from raw flash. 6631 6632 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 6633 mount the root filesystem 6634 6635 rootflags= [KNL] Set root filesystem mount option string 6636 6637 rseq_slice_ext= [KNL] RSEQ based time slice extension 6638 Format: boolean 6639 Control enablement of RSEQ based time slice extension. 6640 Default is 'on'. 6641 6642 initramfs_options= [KNL] 6643 Specify mount options for for the initramfs mount. 6644 6645 rootfstype= [KNL] Set root filesystem type 6646 6647 rootwait [KNL] Wait (indefinitely) for root device to show up. 6648 Useful for devices that are detected asynchronously 6649 (e.g. USB and MMC devices). 6650 6651 rootwait= [KNL] Maximum time (in seconds) to wait for root device 6652 to show up before attempting to mount the root 6653 filesystem. 6654 6655 rproc_mem=nn[KMG][@address] 6656 [KNL,ARM,CMA] Remoteproc physical memory block. 6657 Memory area to be used by remote processor image, 6658 managed by CMA. 6659 6660 rseq_debug= [KNL] Enable or disable restartable sequence 6661 debug mode. Defaults to CONFIG_RSEQ_DEBUG_DEFAULT_ENABLE. 6662 Format: <bool> 6663 6664 rt_group_sched= [KNL] Enable or disable SCHED_RR/FIFO group scheduling 6665 when CONFIG_RT_GROUP_SCHED=y. Defaults to 6666 !CONFIG_RT_GROUP_SCHED_DEFAULT_DISABLED. 6667 Format: <bool> 6668 6669 rw [KNL] Mount root device read-write on boot 6670 6671 S [KNL] Run init in single mode 6672 6673 s390_iommu= [HW,S390] 6674 Set s390 IOTLB flushing mode 6675 strict 6676 With strict flushing every unmap operation will result 6677 in an IOTLB flush. Default is lazy flushing before 6678 reuse, which is faster. Deprecated, equivalent to 6679 iommu.strict=1. 6680 6681 s390_iommu_aperture= [KNL,S390] 6682 Specifies the size of the per device DMA address space 6683 accessible through the DMA and IOMMU APIs as a decimal 6684 factor of the size of main memory. 6685 The default is 1 meaning that one can concurrently use 6686 as many DMA addresses as physical memory is installed, 6687 if supported by hardware, and thus map all of memory 6688 once. With a value of 2 one can map all of memory twice 6689 and so on. As a special case a factor of 0 imposes no 6690 restrictions other than those given by hardware at the 6691 cost of significant additional memory use for tables. 6692 6693 sa1100ir [NET] 6694 See drivers/net/irda/sa1100_ir.c. 6695 6696 sched_proxy_exec= [KNL] 6697 Enables or disables "proxy execution" style 6698 solution to mutex-based priority inversion. 6699 Format: <bool> 6700 6701 sched_verbose [KNL,EARLY] Enables verbose scheduler debug messages. 6702 6703 schedstats= [KNL,X86] Enable or disable scheduled statistics. 6704 Allowed values are enable and disable. This feature 6705 incurs a small amount of overhead in the scheduler 6706 but is useful for debugging and performance tuning. 6707 6708 sched_thermal_decay_shift= 6709 [Deprecated] 6710 [KNL, SMP] Set a decay shift for scheduler thermal 6711 pressure signal. Thermal pressure signal follows the 6712 default decay period of other scheduler pelt 6713 signals(usually 32 ms but configurable). Setting 6714 sched_thermal_decay_shift will left shift the decay 6715 period for the thermal pressure signal by the shift 6716 value. 6717 i.e. with the default pelt decay period of 32 ms 6718 sched_thermal_decay_shift thermal pressure decay pr 6719 1 64 ms 6720 2 128 ms 6721 and so on. 6722 Format: integer between 0 and 10 6723 Default is 0. 6724 6725 scftorture.holdoff= [KNL] 6726 Number of seconds to hold off before starting 6727 test. Defaults to zero for module insertion and 6728 to 10 seconds for built-in smp_call_function() 6729 tests. 6730 6731 scftorture.longwait= [KNL] 6732 Request ridiculously long waits randomly selected 6733 up to the chosen limit in seconds. Zero (the 6734 default) disables this feature. Please note 6735 that requesting even small non-zero numbers of 6736 seconds can result in RCU CPU stall warnings, 6737 softlockup complaints, and so on. 6738 6739 scftorture.nthreads= [KNL] 6740 Number of kthreads to spawn to invoke the 6741 smp_call_function() family of functions. 6742 The default of -1 specifies a number of kthreads 6743 equal to the number of CPUs. 6744 6745 scftorture.onoff_holdoff= [KNL] 6746 Number seconds to wait after the start of the 6747 test before initiating CPU-hotplug operations. 6748 6749 scftorture.onoff_interval= [KNL] 6750 Number seconds to wait between successive 6751 CPU-hotplug operations. Specifying zero (which 6752 is the default) disables CPU-hotplug operations. 6753 6754 scftorture.shutdown_secs= [KNL] 6755 The number of seconds following the start of the 6756 test after which to shut down the system. The 6757 default of zero avoids shutting down the system. 6758 Non-zero values are useful for automated tests. 6759 6760 scftorture.stat_interval= [KNL] 6761 The number of seconds between outputting the 6762 current test statistics to the console. A value 6763 of zero disables statistics output. 6764 6765 scftorture.stutter_cpus= [KNL] 6766 The number of jiffies to wait between each change 6767 to the set of CPUs under test. 6768 6769 scftorture.use_cpus_read_lock= [KNL] 6770 Use use_cpus_read_lock() instead of the default 6771 preempt_disable() to disable CPU hotplug 6772 while invoking one of the smp_call_function*() 6773 functions. 6774 6775 scftorture.verbose= [KNL] 6776 Enable additional printk() statements. 6777 6778 scftorture.weight_single= [KNL] 6779 The probability weighting to use for the 6780 smp_call_function_single() function with a zero 6781 "wait" parameter. A value of -1 selects the 6782 default if all other weights are -1. However, 6783 if at least one weight has some other value, a 6784 value of -1 will instead select a weight of zero. 6785 6786 scftorture.weight_single_wait= [KNL] 6787 The probability weighting to use for the 6788 smp_call_function_single() function with a 6789 non-zero "wait" parameter. See weight_single. 6790 6791 scftorture.weight_many= [KNL] 6792 The probability weighting to use for the 6793 smp_call_function_many() function with a zero 6794 "wait" parameter. See weight_single. 6795 Note well that setting a high probability for 6796 this weighting can place serious IPI load 6797 on the system. 6798 6799 scftorture.weight_many_wait= [KNL] 6800 The probability weighting to use for the 6801 smp_call_function_many() function with a 6802 non-zero "wait" parameter. See weight_single 6803 and weight_many. 6804 6805 scftorture.weight_all= [KNL] 6806 The probability weighting to use for the 6807 smp_call_function_all() function with a zero 6808 "wait" parameter. See weight_single and 6809 weight_many. 6810 6811 scftorture.weight_all_wait= [KNL] 6812 The probability weighting to use for the 6813 smp_call_function_all() function with a 6814 non-zero "wait" parameter. See weight_single 6815 and weight_many. 6816 6817 sdw_mclk_divider=[SDW] 6818 Specify the MCLK divider for Intel SoundWire buses in 6819 case the BIOS does not provide the clock rate properly. 6820 6821 skew_tick= [KNL,EARLY] Offset the periodic timer tick per cpu to mitigate 6822 xtime_lock contention on larger systems, and/or RCU lock 6823 contention on all systems with CONFIG_MAXSMP set. 6824 Format: { "0" | "1" } 6825 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 6826 1 -- enable. 6827 Note: increases power consumption, thus should only be 6828 enabled if running jitter sensitive (HPC/RT) workloads. 6829 6830 security= [SECURITY] Choose a legacy "major" security module to 6831 enable at boot. This has been deprecated by the 6832 "lsm=" parameter. 6833 6834 selinux= [SELINUX] Disable or enable SELinux at boot time. 6835 Format: { "0" | "1" } 6836 See security/selinux/Kconfig help text. 6837 0 -- disable. 6838 1 -- enable. 6839 Default value is 1. 6840 6841 serialnumber [BUGS=X86-32] 6842 6843 sev=option[,option...] [X86-64] 6844 6845 debug 6846 Enable debug messages. 6847 6848 nosnp 6849 Do not enable SEV-SNP (applies to host/hypervisor 6850 only). Setting 'nosnp' avoids the RMP check overhead 6851 in memory accesses when users do not want to run 6852 SEV-SNP guests. 6853 6854 shapers= [NET] 6855 Maximal number of shapers. 6856 6857 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 6858 Limit apic dumping. The parameter defines the maximal 6859 number of local apics being dumped. Also it is possible 6860 to set it to "all" by meaning -- no limit here. 6861 Format: { 1 (default) | 2 | ... | all }. 6862 The parameter valid if only apic=debug or 6863 apic=verbose is specified. 6864 Example: apic=debug show_lapic=all 6865 6866 slab_debug[=options[,slabs][;[options[,slabs]]...] [MM] 6867 Enabling slab_debug allows one to determine the 6868 culprit if slab objects become corrupted. Enabling 6869 slab_debug can create guard zones around objects and 6870 may poison objects when not in use. Also tracks the 6871 last alloc / free. For more information see 6872 Documentation/admin-guide/mm/slab.rst. 6873 (slub_debug legacy name also accepted for now) 6874 6875 Using this option implies the "no_hash_pointers" 6876 option which can be undone by adding the 6877 "hash_pointers=always" option. 6878 6879 slab_max_order= [MM] 6880 Determines the maximum allowed order for slabs. 6881 A high setting may cause OOMs due to memory 6882 fragmentation. For more information see 6883 Documentation/admin-guide/mm/slab.rst. 6884 (slub_max_order legacy name also accepted for now) 6885 6886 slab_merge [MM] 6887 Enable merging of slabs with similar size when the 6888 kernel is built without CONFIG_SLAB_MERGE_DEFAULT. 6889 (slub_merge legacy name also accepted for now) 6890 6891 slab_min_objects= [MM] 6892 The minimum number of objects per slab. SLUB will 6893 increase the slab order up to slab_max_order to 6894 generate a sufficiently large slab able to contain 6895 the number of objects indicated. The higher the number 6896 of objects the smaller the overhead of tracking slabs 6897 and the less frequently locks need to be acquired. 6898 For more information see 6899 Documentation/admin-guide/mm/slab.rst. 6900 (slub_min_objects legacy name also accepted for now) 6901 6902 slab_min_order= [MM] 6903 Determines the minimum page order for slabs. Must be 6904 lower or equal to slab_max_order. For more information see 6905 Documentation/admin-guide/mm/slab.rst. 6906 (slub_min_order legacy name also accepted for now) 6907 6908 slab_nomerge [MM] 6909 Disable merging of slabs with similar size. May be 6910 necessary if there is some reason to distinguish 6911 allocs to different slabs, especially in hardened 6912 environments where the risk of heap overflows and 6913 layout control by attackers can usually be 6914 frustrated by disabling merging. This will reduce 6915 most of the exposure of a heap attack to a single 6916 cache (risks via metadata attacks are mostly 6917 unchanged). Debug options disable merging on their 6918 own. 6919 For more information see 6920 Documentation/admin-guide/mm/slab.rst. 6921 (slub_nomerge legacy name also accepted for now) 6922 6923 slab_strict_numa [MM] 6924 Support memory policies on a per object level 6925 in the slab allocator. The default is for memory 6926 policies to be applied at the folio level when 6927 a new folio is needed or a partial folio is 6928 retrieved from the lists. Increases overhead 6929 in the slab fastpaths but gains more accurate 6930 NUMA kernel object placement which helps with slow 6931 interconnects in NUMA systems. 6932 6933 slram= [HW,MTD] 6934 6935 smart2= [HW] 6936 Format: <io1>[,<io2>[,...,<io8>]] 6937 6938 smp.csd_lock_timeout= [KNL] 6939 Specify the period of time in milliseconds 6940 that smp_call_function() and friends will wait 6941 for a CPU to release the CSD lock. This is 6942 useful when diagnosing bugs involving CPUs 6943 disabling interrupts for extended periods 6944 of time. Defaults to 5,000 milliseconds, and 6945 setting a value of zero disables this feature. 6946 This feature may be more efficiently disabled 6947 using the csdlock_debug- kernel parameter. 6948 6949 smp.panic_on_ipistall= [KNL] 6950 If a csd_lock_timeout extends for more than 6951 the specified number of milliseconds, panic the 6952 system. By default, let CSD-lock acquisition 6953 take as long as they take. Specifying 300,000 6954 for this value provides a 5-minute timeout. 6955 6956 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 6957 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 6958 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 6959 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 6960 smsc-ircc2.ircc_irq= [HW] IRQ line 6961 smsc-ircc2.ircc_dma= [HW] DMA channel 6962 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 6963 0: Toshiba Satellite 1800 (GP data pin select) 6964 1: Fast pin select (default) 6965 2: ATC IRMode 6966 6967 smt= [KNL,MIPS,S390,EARLY] Set the maximum number of threads 6968 (logical CPUs) to use per physical CPU on systems 6969 capable of symmetric multithreading (SMT). Will 6970 be capped to the actual hardware limit. 6971 Format: <integer> 6972 Default: -1 (no limit) 6973 6974 softlockup_panic= 6975 [KNL] Should the soft-lockup detector generate panics. 6976 Format: 0 | 1 6977 6978 A value of 1 instructs the soft-lockup detector 6979 to panic the machine when a soft-lockup occurs. It is 6980 also controlled by the kernel.softlockup_panic sysctl 6981 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 6982 respective build-time switch to that functionality. 6983 6984 softlockup_all_cpu_backtrace= 6985 [KNL] Should the soft-lockup detector generate 6986 backtraces on all cpus. 6987 Format: 0 | 1 6988 6989 sonypi.*= [HW] Sony Programmable I/O Control Device driver 6990 See Documentation/admin-guide/laptops/sonypi.rst 6991 6992 spectre_bhi= [X86] Control mitigation of Branch History Injection 6993 (BHI) vulnerability. This setting affects the 6994 deployment of the HW BHI control and the SW BHB 6995 clearing sequence. 6996 6997 on - (default) Enable the HW or SW mitigation as 6998 needed. This protects the kernel from 6999 both syscalls and VMs. 7000 vmexit - On systems which don't have the HW mitigation 7001 available, enable the SW mitigation on vmexit 7002 ONLY. On such systems, the host kernel is 7003 protected from VM-originated BHI attacks, but 7004 may still be vulnerable to syscall attacks. 7005 off - Disable the mitigation. 7006 7007 spectre_v2= [X86,EARLY] Control mitigation of Spectre variant 2 7008 (indirect branch speculation) vulnerability. 7009 The default operation protects the kernel from 7010 user space attacks. 7011 7012 on - unconditionally enable, implies 7013 spectre_v2_user=on 7014 off - unconditionally disable, implies 7015 spectre_v2_user=off 7016 auto - kernel detects whether your CPU model is 7017 vulnerable 7018 7019 Selecting 'on' will, and 'auto' may, choose a 7020 mitigation method at run time according to the 7021 CPU, the available microcode, the setting of the 7022 CONFIG_MITIGATION_RETPOLINE configuration option, 7023 and the compiler with which the kernel was built. 7024 7025 Selecting 'on' will also enable the mitigation 7026 against user space to user space task attacks. 7027 Selecting specific mitigation does not force enable 7028 user mitigations. 7029 7030 Selecting 'off' will disable both the kernel and 7031 the user space protections. 7032 7033 Specific mitigations can also be selected manually: 7034 7035 retpoline - replace indirect branches 7036 retpoline,generic - Retpolines 7037 retpoline,lfence - LFENCE; indirect branch 7038 retpoline,amd - alias for retpoline,lfence 7039 eibrs - Enhanced/Auto IBRS 7040 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines 7041 eibrs,lfence - Enhanced/Auto IBRS + LFENCE 7042 ibrs - use IBRS to protect kernel 7043 7044 Not specifying this option is equivalent to 7045 spectre_v2=auto. 7046 7047 spectre_v2_user= 7048 [X86] Control mitigation of Spectre variant 2 7049 (indirect branch speculation) vulnerability between 7050 user space tasks 7051 7052 on - Unconditionally enable mitigations. Is 7053 enforced by spectre_v2=on 7054 7055 off - Unconditionally disable mitigations. Is 7056 enforced by spectre_v2=off 7057 7058 prctl - Indirect branch speculation is enabled, 7059 but mitigation can be enabled via prctl 7060 per thread. The mitigation control state 7061 is inherited on fork. 7062 7063 prctl,ibpb 7064 - Like "prctl" above, but only STIBP is 7065 controlled per thread. IBPB is issued 7066 always when switching between different user 7067 space processes. 7068 7069 seccomp 7070 - Same as "prctl" above, but all seccomp 7071 threads will enable the mitigation unless 7072 they explicitly opt out. 7073 7074 seccomp,ibpb 7075 - Like "seccomp" above, but only STIBP is 7076 controlled per thread. IBPB is issued 7077 always when switching between different 7078 user space processes. 7079 7080 auto - Kernel selects the mitigation depending on 7081 the available CPU features and vulnerability. 7082 7083 Default mitigation: "prctl" 7084 7085 Not specifying this option is equivalent to 7086 spectre_v2_user=auto. 7087 7088 spec_rstack_overflow= 7089 [X86,EARLY] Control RAS overflow mitigation on AMD Zen CPUs 7090 7091 off - Disable mitigation 7092 microcode - Enable microcode mitigation only 7093 safe-ret - Enable sw-only safe RET mitigation (default) 7094 ibpb - Enable mitigation by issuing IBPB on 7095 kernel entry 7096 ibpb-vmexit - Issue IBPB only on VMEXIT 7097 (cloud-specific mitigation) 7098 7099 spec_store_bypass_disable= 7100 [HW,EARLY] Control Speculative Store Bypass (SSB) Disable mitigation 7101 (Speculative Store Bypass vulnerability) 7102 7103 Certain CPUs are vulnerable to an exploit against a 7104 a common industry wide performance optimization known 7105 as "Speculative Store Bypass" in which recent stores 7106 to the same memory location may not be observed by 7107 later loads during speculative execution. The idea 7108 is that such stores are unlikely and that they can 7109 be detected prior to instruction retirement at the 7110 end of a particular speculation execution window. 7111 7112 In vulnerable processors, the speculatively forwarded 7113 store can be used in a cache side channel attack, for 7114 example to read memory to which the attacker does not 7115 directly have access (e.g. inside sandboxed code). 7116 7117 This parameter controls whether the Speculative Store 7118 Bypass optimization is used. 7119 7120 On x86 the options are: 7121 7122 on - Unconditionally disable Speculative Store Bypass 7123 off - Unconditionally enable Speculative Store Bypass 7124 auto - Kernel detects whether the CPU model contains an 7125 implementation of Speculative Store Bypass and 7126 picks the most appropriate mitigation. If the 7127 CPU is not vulnerable, "off" is selected. If the 7128 CPU is vulnerable the default mitigation is 7129 architecture and Kconfig dependent. See below. 7130 prctl - Control Speculative Store Bypass per thread 7131 via prctl. Speculative Store Bypass is enabled 7132 for a process by default. The state of the control 7133 is inherited on fork. 7134 seccomp - Same as "prctl" above, but all seccomp threads 7135 will disable SSB unless they explicitly opt out. 7136 7137 Default mitigations: 7138 X86: "prctl" 7139 7140 On powerpc the options are: 7141 7142 on,auto - On Power8 and Power9 insert a store-forwarding 7143 barrier on kernel entry and exit. On Power7 7144 perform a software flush on kernel entry and 7145 exit. 7146 off - No action. 7147 7148 Not specifying this option is equivalent to 7149 spec_store_bypass_disable=auto. 7150 7151 split_lock_detect= 7152 [X86] Enable split lock detection or bus lock detection 7153 7154 When enabled (and if hardware support is present), atomic 7155 instructions that access data across cache line 7156 boundaries will result in an alignment check exception 7157 for split lock detection or a debug exception for 7158 bus lock detection. 7159 7160 off - not enabled 7161 7162 warn - the kernel will emit rate-limited warnings 7163 about applications triggering the #AC 7164 exception or the #DB exception. This mode is 7165 the default on CPUs that support split lock 7166 detection or bus lock detection. Default 7167 behavior is by #AC if both features are 7168 enabled in hardware. 7169 7170 fatal - the kernel will send SIGBUS to applications 7171 that trigger the #AC exception or the #DB 7172 exception. Default behavior is by #AC if 7173 both features are enabled in hardware. 7174 7175 ratelimit:N - 7176 Set system wide rate limit to N bus locks 7177 per second for bus lock detection. 7178 0 < N <= 1000. 7179 7180 N/A for split lock detection. 7181 7182 7183 If an #AC exception is hit in the kernel or in 7184 firmware (i.e. not while executing in user mode) 7185 the kernel will oops in either "warn" or "fatal" 7186 mode. 7187 7188 #DB exception for bus lock is triggered only when 7189 CPL > 0. 7190 7191 srbds= [X86,INTEL,EARLY] 7192 Control the Special Register Buffer Data Sampling 7193 (SRBDS) mitigation. 7194 7195 Certain CPUs are vulnerable to an MDS-like 7196 exploit which can leak bits from the random 7197 number generator. 7198 7199 By default, this issue is mitigated by 7200 microcode. However, the microcode fix can cause 7201 the RDRAND and RDSEED instructions to become 7202 much slower. Among other effects, this will 7203 result in reduced throughput from /dev/urandom. 7204 7205 The microcode mitigation can be disabled with 7206 the following option: 7207 7208 off: Disable mitigation and remove 7209 performance impact to RDRAND and RDSEED 7210 7211 srcutree.big_cpu_lim [KNL] 7212 Specifies the number of CPUs constituting a 7213 large system, such that srcu_struct structures 7214 should immediately allocate an srcu_node array. 7215 This kernel-boot parameter defaults to 128, 7216 but takes effect only when the low-order four 7217 bits of srcutree.convert_to_big is equal to 3 7218 (decide at boot). 7219 7220 srcutree.convert_to_big [KNL] 7221 Specifies under what conditions an SRCU tree 7222 srcu_struct structure will be converted to big 7223 form, that is, with an rcu_node tree: 7224 7225 0: Never. 7226 1: At init_srcu_struct() time. 7227 2: When rcutorture decides to. 7228 3: Decide at boot time (default). 7229 0x1X: Above plus if high contention. 7230 7231 Either way, the srcu_node tree will be sized based 7232 on the actual runtime number of CPUs (nr_cpu_ids) 7233 instead of the compile-time CONFIG_NR_CPUS. 7234 7235 srcutree.counter_wrap_check [KNL] 7236 Specifies how frequently to check for 7237 grace-period sequence counter wrap for the 7238 srcu_data structure's ->srcu_gp_seq_needed field. 7239 The greater the number of bits set in this kernel 7240 parameter, the less frequently counter wrap will 7241 be checked for. Note that the bottom two bits 7242 are ignored. 7243 7244 srcutree.exp_holdoff [KNL] 7245 Specifies how many nanoseconds must elapse 7246 since the end of the last SRCU grace period for 7247 a given srcu_struct until the next normal SRCU 7248 grace period will be considered for automatic 7249 expediting. Set to zero to disable automatic 7250 expediting. 7251 7252 srcutree.srcu_max_nodelay [KNL] 7253 Specifies the number of no-delay instances 7254 per jiffy for which the SRCU grace period 7255 worker thread will be rescheduled with zero 7256 delay. Beyond this limit, worker thread will 7257 be rescheduled with a sleep delay of one jiffy. 7258 7259 srcutree.srcu_max_nodelay_phase [KNL] 7260 Specifies the per-grace-period phase, number of 7261 non-sleeping polls of readers. Beyond this limit, 7262 grace period worker thread will be rescheduled 7263 with a sleep delay of one jiffy, between each 7264 rescan of the readers, for a grace period phase. 7265 7266 srcutree.srcu_retry_check_delay [KNL] 7267 Specifies number of microseconds of non-sleeping 7268 delay between each non-sleeping poll of readers. 7269 7270 srcutree.small_contention_lim [KNL] 7271 Specifies the number of update-side contention 7272 events per jiffy will be tolerated before 7273 initiating a conversion of an srcu_struct 7274 structure to big form. Note that the value of 7275 srcutree.convert_to_big must have the 0x10 bit 7276 set for contention-based conversions to occur. 7277 7278 ssbd= [ARM64,HW,EARLY] 7279 Speculative Store Bypass Disable control 7280 7281 On CPUs that are vulnerable to the Speculative 7282 Store Bypass vulnerability and offer a 7283 firmware based mitigation, this parameter 7284 indicates how the mitigation should be used: 7285 7286 force-on: Unconditionally enable mitigation for 7287 for both kernel and userspace 7288 force-off: Unconditionally disable mitigation for 7289 for both kernel and userspace 7290 kernel: Always enable mitigation in the 7291 kernel, and offer a prctl interface 7292 to allow userspace to register its 7293 interest in being mitigated too. 7294 7295 stack_guard_gap= [MM] 7296 override the default stack gap protection. The value 7297 is in page units and it defines how many pages prior 7298 to (for stacks growing down) resp. after (for stacks 7299 growing up) the main stack are reserved for no other 7300 mapping. Default value is 256 pages. 7301 7302 stack_depot_disable= [KNL,EARLY] 7303 Setting this to true through kernel command line will 7304 disable the stack depot thereby saving the static memory 7305 consumed by the stack hash table. By default this is set 7306 to false. 7307 7308 stack_depot_max_pools= [KNL,EARLY] 7309 Specify the maximum number of pools to use for storing 7310 stack traces. Pools are allocated on-demand up to this 7311 limit. Default value is 8191 pools. 7312 7313 stacktrace [FTRACE] 7314 Enable the stack tracer on boot up. 7315 7316 stacktrace_filter=[function-list] 7317 [FTRACE] Limit the functions that the stack tracer 7318 will trace at boot up. function-list is a comma-separated 7319 list of functions. This list can be changed at run 7320 time by the stack_trace_filter file in the debugfs 7321 tracing directory. Note, this enables stack tracing 7322 and the stacktrace above is not needed. 7323 7324 sti= [PARISC,HW] 7325 Format: <num> 7326 Set the STI (builtin display/keyboard on the HP-PARISC 7327 machines) console (graphic card) which should be used 7328 as the initial boot-console. 7329 See also comment in drivers/video/console/sticore.c. 7330 7331 sti_font= [HW] 7332 See comment in drivers/video/console/sticore.c. 7333 7334 stifb= [HW] 7335 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 7336 7337 strict_sas_size= 7338 [X86] 7339 Format: <bool> 7340 Enable or disable strict sigaltstack size checks 7341 against the required signal frame size which 7342 depends on the supported FPU features. This can 7343 be used to filter out binaries which have 7344 not yet been made aware of AT_MINSIGSTKSZ. 7345 7346 stress_hpt [PPC,EARLY] 7347 Limits the number of kernel HPT entries in the hash 7348 page table to increase the rate of hash page table 7349 faults on kernel addresses. 7350 7351 stress_slb [PPC,EARLY] 7352 Limits the number of kernel SLB entries, and flushes 7353 them frequently to increase the rate of SLB faults 7354 on kernel addresses. 7355 7356 no_slb_preload [PPC,EARLY] 7357 Disables slb preloading for userspace. 7358 7359 sunrpc.min_resvport= 7360 sunrpc.max_resvport= 7361 [NFS,SUNRPC] 7362 SunRPC servers often require that client requests 7363 originate from a privileged port (i.e. a port in the 7364 range 0 < portnr < 1024). 7365 An administrator who wishes to reserve some of these 7366 ports for other uses may adjust the range that the 7367 kernel's sunrpc client considers to be privileged 7368 using these two parameters to set the minimum and 7369 maximum port values. 7370 7371 sunrpc.svc_rpc_per_connection_limit= 7372 [NFS,SUNRPC] 7373 Limit the number of requests that the server will 7374 process in parallel from a single connection. 7375 The default value is 0 (no limit). 7376 7377 sunrpc.pool_mode= 7378 [NFS] 7379 Control how the NFS server code allocates CPUs to 7380 service thread pools. Depending on how many NICs 7381 you have and where their interrupts are bound, this 7382 option will affect which CPUs will do NFS serving. 7383 Note: this parameter cannot be changed while the 7384 NFS server is running. 7385 7386 auto the server chooses an appropriate mode 7387 automatically using heuristics 7388 global a single global pool contains all CPUs 7389 percpu one pool for each CPU 7390 pernode one pool for each NUMA node (equivalent 7391 to global on non-NUMA machines) 7392 7393 sunrpc.tcp_slot_table_entries= 7394 sunrpc.udp_slot_table_entries= 7395 [NFS,SUNRPC] 7396 Sets the upper limit on the number of simultaneous 7397 RPC calls that can be sent from the client to a 7398 server. Increasing these values may allow you to 7399 improve throughput, but will also increase the 7400 amount of memory reserved for use by the client. 7401 7402 suspend.pm_test_delay= 7403 [SUSPEND] 7404 Sets the number of seconds to remain in a suspend test 7405 mode before resuming the system (see 7406 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 7407 is set. Default value is 5. 7408 7409 svm= [PPC] 7410 Format: { on | off | y | n | 1 | 0 } 7411 This parameter controls use of the Protected 7412 Execution Facility on pSeries. 7413 7414 swiotlb= [ARM,PPC,MIPS,X86,S390,EARLY] 7415 Format: { <int> [,<int>] | force | noforce } 7416 <int> -- Number of I/O TLB slabs 7417 <int> -- Second integer after comma. Number of swiotlb 7418 areas with their own lock. Will be rounded up 7419 to a power of 2. 7420 force -- force using of bounce buffers even if they 7421 wouldn't be automatically used by the kernel 7422 noforce -- Never use bounce buffers (for debugging) 7423 7424 switches= [HW,M68k,EARLY] 7425 7426 sysctl.*= [KNL] 7427 Set a sysctl parameter, right before loading the init 7428 process, as if the value was written to the respective 7429 /proc/sys/... file. Both '.' and '/' are recognized as 7430 separators. Unrecognized parameters and invalid values 7431 are reported in the kernel log. Sysctls registered 7432 later by a loaded module cannot be set this way. 7433 Example: sysctl.vm.swappiness=40 7434 7435 sysrq_always_enabled 7436 [KNL] 7437 Ignore sysrq setting - this boot parameter will 7438 neutralize any effect of /proc/sys/kernel/sysrq. 7439 Useful for debugging. 7440 7441 tcpmhash_entries= [KNL,NET] 7442 Set the number of tcp_metrics_hash slots. 7443 Default value is 8192 or 16384 depending on total 7444 ram pages. This is used to specify the TCP metrics 7445 cache size. See Documentation/networking/ip-sysctl.rst 7446 "tcp_no_metrics_save" section for more details. 7447 7448 tdfx= [HW,DRM] 7449 7450 test_suspend= [SUSPEND] 7451 Format: { "mem" | "standby" | "freeze" }[,N] 7452 Specify "mem" (for Suspend-to-RAM) or "standby" (for 7453 standby suspend) or "freeze" (for suspend type freeze) 7454 as the system sleep state during system startup with 7455 the optional capability to repeat N number of times. 7456 The system is woken from this state using a 7457 wakeup-capable RTC alarm. 7458 7459 thash_entries= [KNL,NET] 7460 Set number of hash buckets for TCP connection 7461 7462 thermal.act= [HW,ACPI] 7463 -1: disable all active trip points in all thermal zones 7464 <degrees C>: override all lowest active trip points 7465 7466 thermal.crt= [HW,ACPI] 7467 -1: disable all critical trip points in all thermal zones 7468 <degrees C>: override all critical trip points 7469 7470 thermal.off= [HW,ACPI] 7471 1: disable ACPI thermal control 7472 7473 thermal.psv= [HW,ACPI] 7474 -1: disable all passive trip points 7475 <degrees C>: override all passive trip points to this 7476 value 7477 7478 thermal.tzp= [HW,ACPI] 7479 Specify global default ACPI thermal zone polling rate 7480 <deci-seconds>: poll all this frequency 7481 0: no polling (default) 7482 7483 thp_anon= [KNL] 7484 Format: <size>[KMG],<size>[KMG]:<state>;<size>[KMG]-<size>[KMG]:<state> 7485 state is one of "always", "madvise", "never" or "inherit". 7486 Control the default behavior of the system with respect 7487 to anonymous transparent hugepages. 7488 Can be used multiple times for multiple anon THP sizes. 7489 See Documentation/admin-guide/mm/transhuge.rst for more 7490 details. 7491 7492 threadirqs [KNL,EARLY] 7493 Force threading of all interrupt handlers except those 7494 marked explicitly IRQF_NO_THREAD. 7495 7496 thp_shmem= [KNL] 7497 Format: <size>[KMG],<size>[KMG]:<policy>;<size>[KMG]-<size>[KMG]:<policy> 7498 Control the default policy of each hugepage size for the 7499 internal shmem mount. <policy> is one of policies available 7500 for the shmem mount ("always", "inherit", "never", "within_size", 7501 and "advise"). 7502 It can be used multiple times for multiple shmem THP sizes. 7503 See Documentation/admin-guide/mm/transhuge.rst for more 7504 details. 7505 7506 topology= [S390,EARLY] 7507 Format: {off | on} 7508 Specify if the kernel should make use of the cpu 7509 topology information if the hardware supports this. 7510 The scheduler will make use of this information and 7511 e.g. base its process migration decisions on it. 7512 Default is on. 7513 7514 torture.disable_onoff_at_boot= [KNL] 7515 Prevent the CPU-hotplug component of torturing 7516 until after init has spawned. 7517 7518 torture.ftrace_dump_at_shutdown= [KNL] 7519 Dump the ftrace buffer at torture-test shutdown, 7520 even if there were no errors. This can be a 7521 very costly operation when many torture tests 7522 are running concurrently, especially on systems 7523 with rotating-rust storage. 7524 7525 torture.verbose_sleep_frequency= [KNL] 7526 Specifies how many verbose printk()s should be 7527 emitted between each sleep. The default of zero 7528 disables verbose-printk() sleeping. 7529 7530 torture.verbose_sleep_duration= [KNL] 7531 Duration of each verbose-printk() sleep in jiffies. 7532 7533 tpm.disable_pcr_integrity= [HW,TPM] 7534 Do not protect PCR registers from unintended physical 7535 access, or interposers in the bus by the means of 7536 having an integrity protected session wrapped around 7537 TPM2_PCR_Extend command. Consider this in a situation 7538 where TPM is heavily utilized by IMA, thus protection 7539 causing a major performance hit, and the space where 7540 machines are deployed is by other means guarded. 7541 7542 tpm_crb_ffa.busy_timeout_ms= [ARM64,TPM] 7543 Maximum time in milliseconds to retry sending a message 7544 to the TPM service before giving up. This parameter controls 7545 how long the system will continue retrying when the TPM 7546 service is busy. 7547 Format: <unsigned int> 7548 Default: 2000 (2 seconds) 7549 7550 tpm_suspend_pcr=[HW,TPM] 7551 Format: integer pcr id 7552 Specify that at suspend time, the tpm driver 7553 should extend the specified pcr with zeros, 7554 as a workaround for some chips which fail to 7555 flush the last written pcr on TPM_SaveState. 7556 This will guarantee that all the other pcrs 7557 are saved. 7558 7559 tpm_tis.interrupts= [HW,TPM] 7560 Enable interrupts for the MMIO based physical layer 7561 for the FIFO interface. By default it is set to false 7562 (0). For more information about TPM hardware interfaces 7563 defined by Trusted Computing Group (TCG) see 7564 https://trustedcomputinggroup.org/resource/pc-client-platform-tpm-profile-ptp-specification/ 7565 7566 tp_printk [FTRACE] 7567 Have the tracepoints sent to printk as well as the 7568 tracing ring buffer. This is useful for early boot up 7569 where the system hangs or reboots and does not give the 7570 option for reading the tracing buffer or performing a 7571 ftrace_dump_on_oops. 7572 7573 To turn off having tracepoints sent to printk, 7574 echo 0 > /proc/sys/kernel/tracepoint_printk 7575 Note, echoing 1 into this file without the 7576 tp_printk kernel cmdline option has no effect. 7577 7578 The tp_printk_stop_on_boot (see below) can also be used 7579 to stop the printing of events to console at 7580 late_initcall_sync. 7581 7582 ** CAUTION ** 7583 7584 Having tracepoints sent to printk() and activating high 7585 frequency tracepoints such as irq or sched, can cause 7586 the system to live lock. 7587 7588 tp_printk_stop_on_boot [FTRACE] 7589 When tp_printk (above) is set, it can cause a lot of noise 7590 on the console. It may be useful to only include the 7591 printing of events during boot up, as user space may 7592 make the system inoperable. 7593 7594 This command line option will stop the printing of events 7595 to console at the late_initcall_sync() time frame. 7596 7597 trace_buf_size=nn[KMG] 7598 [FTRACE] will set tracing buffer size on each cpu. 7599 7600 trace_clock= [FTRACE] Set the clock used for tracing events 7601 at boot up. 7602 local - Use the per CPU time stamp counter 7603 (converted into nanoseconds). Fast, but 7604 depending on the architecture, may not be 7605 in sync between CPUs. 7606 global - Event time stamps are synchronized across 7607 CPUs. May be slower than the local clock, 7608 but better for some race conditions. 7609 counter - Simple counting of events (1, 2, ..) 7610 note, some counts may be skipped due to the 7611 infrastructure grabbing the clock more than 7612 once per event. 7613 uptime - Use jiffies as the time stamp. 7614 perf - Use the same clock that perf uses. 7615 mono - Use ktime_get_mono_fast_ns() for time stamps. 7616 mono_raw - Use ktime_get_raw_fast_ns() for time 7617 stamps. 7618 boot - Use ktime_get_boot_fast_ns() for time stamps. 7619 Architectures may add more clocks. See 7620 Documentation/trace/ftrace.rst for more details. 7621 7622 trace_event=[event-list] 7623 [FTRACE] Set and start specified trace events in order 7624 to facilitate early boot debugging. The event-list is a 7625 comma-separated list of trace events to enable. See 7626 also Documentation/trace/events.rst 7627 7628 To enable modules, use :mod: keyword: 7629 7630 trace_event=:mod:<module> 7631 7632 The value before :mod: will only enable specific events 7633 that are part of the module. See the above mentioned 7634 document for more information. 7635 7636 trace_instance=[instance-info] 7637 [FTRACE] Create a ring buffer instance early in boot up. 7638 This will be listed in: 7639 7640 /sys/kernel/tracing/instances 7641 7642 Events can be enabled at the time the instance is created 7643 via: 7644 7645 trace_instance=<name>,<system1>:<event1>,<system2>:<event2> 7646 7647 Note, the "<system*>:" portion is optional if the event is 7648 unique. 7649 7650 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall 7651 7652 will enable the "sched_switch" event (note, the "sched:" is optional, and 7653 the same thing would happen if it was left off). The irq_handler_entry 7654 event, and all events under the "initcall" system. 7655 7656 Flags can be added to the instance to modify its behavior when it is 7657 created. The flags are separated by '^'. 7658 7659 The available flags are: 7660 7661 traceoff - Have the tracing instance tracing disabled after it is created. 7662 traceprintk - Have trace_printk() write into this trace instance 7663 (note, "printk" and "trace_printk" can also be used) 7664 7665 trace_instance=foo^traceoff^traceprintk,sched,irq 7666 7667 The flags must come before the defined events. 7668 7669 If memory has been reserved (see memmap for x86), the instance 7670 can use that memory: 7671 7672 memmap=12M$0x284500000 trace_instance=boot_map@0x284500000:12M 7673 7674 The above will create a "boot_map" instance that uses the physical 7675 memory at 0x284500000 that is 12Megs. The per CPU buffers of that 7676 instance will be split up accordingly. 7677 7678 Alternatively, the memory can be reserved by the reserve_mem option: 7679 7680 reserve_mem=12M:4096:trace trace_instance=boot_map@trace 7681 7682 This will reserve 12 megabytes at boot up with a 4096 byte alignment 7683 and place the ring buffer in this memory. Note that due to KASLR, the 7684 memory may not be the same location each time, which will not preserve 7685 the buffer content. 7686 7687 Also note that the layout of the ring buffer data may change between 7688 kernel versions where the validator will fail and reset the ring buffer 7689 if the layout is not the same as the previous kernel. 7690 7691 If the ring buffer is used for persistent bootups and has events enabled, 7692 it is recommend to disable tracing so that events from a previous boot do not 7693 mix with events of the current boot (unless you are debugging a random crash 7694 at boot up). 7695 7696 reserve_mem=12M:4096:trace trace_instance=boot_map^traceoff^traceprintk@trace,sched,irq 7697 7698 Note, saving the trace buffer across reboots does require that the system 7699 is set up to not wipe memory. For instance, CONFIG_RESET_ATTACK_MITIGATION 7700 can force a memory reset on boot which will clear any trace that was stored. 7701 This is just one of many ways that can clear memory. Make sure your system 7702 keeps the content of memory across reboots before relying on this option. 7703 7704 NB: Both the mapped address and size must be page aligned for the architecture. 7705 7706 See also Documentation/trace/debugging.rst 7707 7708 7709 trace_options=[option-list] 7710 [FTRACE] Enable or disable tracer options at boot. 7711 The option-list is a comma delimited list of options 7712 that can be enabled or disabled just as if you were 7713 to echo the option name into 7714 7715 /sys/kernel/tracing/trace_options 7716 7717 For example, to enable stacktrace option (to dump the 7718 stack trace of each event), add to the command line: 7719 7720 trace_options=stacktrace 7721 7722 See also Documentation/trace/ftrace.rst "trace options" 7723 section. 7724 7725 trace_trigger=[trigger-list] 7726 [FTRACE] Add an event trigger on specific events. 7727 Set a trigger on top of a specific event, with an optional 7728 filter. 7729 7730 The format is "trace_trigger=<event>.<trigger>[ if <filter>],..." 7731 Where more than one trigger may be specified that are comma delimited. 7732 7733 For example: 7734 7735 trace_trigger="sched_switch.stacktrace if prev_state == 2" 7736 7737 The above will enable the "stacktrace" trigger on the "sched_switch" 7738 event but only trigger it if the "prev_state" of the "sched_switch" 7739 event is "2" (TASK_UNINTERRUPTIBLE). 7740 7741 See also "Event triggers" in Documentation/trace/events.rst 7742 7743 7744 traceoff_after_boot 7745 [FTRACE] Sometimes tracing is used to debug issues 7746 during the boot process. Since the trace buffer has a 7747 limited amount of storage, it may be prudent to 7748 disable tracing after the boot is finished, otherwise 7749 the critical information may be overwritten. With this 7750 option, the main tracing buffer will be turned off at 7751 the end of the boot process. 7752 7753 traceoff_on_warning 7754 [FTRACE] enable this option to disable tracing when a 7755 warning is hit. This turns off "tracing_on". Tracing can 7756 be enabled again by echoing '1' into the "tracing_on" 7757 file located in /sys/kernel/tracing/ 7758 7759 This option is useful, as it disables the trace before 7760 the WARNING dump is called, which prevents the trace to 7761 be filled with content caused by the warning output. 7762 7763 This option can also be set at run time via the sysctl 7764 option: kernel/traceoff_on_warning 7765 7766 transparent_hugepage= 7767 [KNL] 7768 Format: [always|madvise|never] 7769 Can be used to control the default behavior of the system 7770 with respect to transparent hugepages. 7771 See Documentation/admin-guide/mm/transhuge.rst 7772 for more details. 7773 7774 transparent_hugepage_shmem= [KNL] 7775 Format: [always|within_size|advise|never|deny|force] 7776 Can be used to control the hugepage allocation policy for 7777 the internal shmem mount. 7778 See Documentation/admin-guide/mm/transhuge.rst 7779 for more details. 7780 7781 transparent_hugepage_tmpfs= [KNL] 7782 Format: [always|within_size|advise|never] 7783 Can be used to control the default hugepage allocation policy 7784 for the tmpfs mount. 7785 See Documentation/admin-guide/mm/transhuge.rst 7786 for more details. 7787 7788 trusted.source= [KEYS] 7789 Format: <string> 7790 This parameter identifies the trust source as a backend 7791 for trusted keys implementation. Supported trust 7792 sources: 7793 - "tpm" 7794 - "tee" 7795 - "caam" 7796 - "dcp" 7797 If not specified then it defaults to iterating through 7798 the trust source list starting with TPM and assigns the 7799 first trust source as a backend which is initialized 7800 successfully during iteration. 7801 7802 trusted.rng= [KEYS] 7803 Format: <string> 7804 The RNG used to generate key material for trusted keys. 7805 Can be one of: 7806 - "kernel" 7807 - the same value as trusted.source: "tpm" or "tee" 7808 - "default" 7809 If not specified, "default" is used. In this case, 7810 the RNG's choice is left to each individual trust source. 7811 7812 trusted.dcp_use_otp_key 7813 This is intended to be used in combination with 7814 trusted.source=dcp and will select the DCP OTP key 7815 instead of the DCP UNIQUE key blob encryption. 7816 7817 trusted.dcp_skip_zk_test 7818 This is intended to be used in combination with 7819 trusted.source=dcp and will disable the check if the 7820 blob key is all zeros. This is helpful for situations where 7821 having this key zero'ed is acceptable. E.g. in testing 7822 scenarios. 7823 7824 tsa= [X86] Control mitigation for Transient Scheduler 7825 Attacks on AMD CPUs. Search the following in your 7826 favourite search engine for more details: 7827 7828 "Technical guidance for mitigating transient scheduler 7829 attacks". 7830 7831 off - disable the mitigation 7832 on - enable the mitigation (default) 7833 user - mitigate only user/kernel transitions 7834 vm - mitigate only guest/host transitions 7835 7836 7837 tsc= Disable clocksource stability checks for TSC. 7838 Format: <string> 7839 [x86] reliable: mark tsc clocksource as reliable, this 7840 disables clocksource verification at runtime, as well 7841 as the stability checks done at bootup. Used to enable 7842 high-resolution timer mode on older hardware, and in 7843 virtualized environment. 7844 [x86] noirqtime: Do not use TSC to do irq accounting. 7845 Used to run time disable IRQ_TIME_ACCOUNTING on any 7846 platforms where RDTSC is slow and this accounting 7847 can add overhead. 7848 [x86] unstable: mark the TSC clocksource as unstable, this 7849 marks the TSC unconditionally unstable at bootup and 7850 avoids any further wobbles once the TSC watchdog notices. 7851 [x86] nowatchdog: disable clocksource watchdog. Used 7852 in situations with strict latency requirements (where 7853 interruptions from clocksource watchdog are not 7854 acceptable). 7855 [x86] recalibrate: force recalibration against a HW timer 7856 (HPET or PM timer) on systems whose TSC frequency was 7857 obtained from HW or FW using either an MSR or CPUID(0x15). 7858 Warn if the difference is more than 500 ppm. 7859 [x86] watchdog: Use TSC as the watchdog clocksource with 7860 which to check other HW timers (HPET or PM timer), but 7861 only on systems where TSC has been deemed trustworthy. 7862 This will be suppressed by an earlier tsc=nowatchdog and 7863 can be overridden by a later tsc=nowatchdog. A console 7864 message will flag any such suppression or overriding. 7865 7866 tsc_early_khz= [X86,EARLY] Skip early TSC calibration and use the given 7867 value instead. Useful when the early TSC frequency discovery 7868 procedure is not reliable, such as on overclocked systems 7869 with CPUID.16h support and partial CPUID.15h support. 7870 Format: <unsigned int> 7871 7872 tsx= [X86] Control Transactional Synchronization 7873 Extensions (TSX) feature in Intel processors that 7874 support TSX control. 7875 7876 This parameter controls the TSX feature. The options are: 7877 7878 on - Enable TSX on the system. Although there are 7879 mitigations for all known security vulnerabilities, 7880 TSX has been known to be an accelerator for 7881 several previous speculation-related CVEs, and 7882 so there may be unknown security risks associated 7883 with leaving it enabled. 7884 7885 off - Disable TSX on the system. (Note that this 7886 option takes effect only on newer CPUs which are 7887 not vulnerable to MDS, i.e., have 7888 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 7889 the new IA32_TSX_CTRL MSR through a microcode 7890 update. This new MSR allows for the reliable 7891 deactivation of the TSX functionality.) 7892 7893 auto - Disable TSX if X86_BUG_TAA is present, 7894 otherwise enable TSX on the system. 7895 7896 Not specifying this option is equivalent to tsx=off. 7897 7898 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 7899 for more details. 7900 7901 tsx_async_abort= [X86,INTEL,EARLY] Control mitigation for the TSX Async 7902 Abort (TAA) vulnerability. 7903 7904 Similar to Micro-architectural Data Sampling (MDS) 7905 certain CPUs that support Transactional 7906 Synchronization Extensions (TSX) are vulnerable to an 7907 exploit against CPU internal buffers which can forward 7908 information to a disclosure gadget under certain 7909 conditions. 7910 7911 In vulnerable processors, the speculatively forwarded 7912 data can be used in a cache side channel attack, to 7913 access data to which the attacker does not have direct 7914 access. 7915 7916 This parameter controls the TAA mitigation. The 7917 options are: 7918 7919 full - Enable TAA mitigation on vulnerable CPUs 7920 if TSX is enabled. 7921 7922 full,nosmt - Enable TAA mitigation and disable SMT on 7923 vulnerable CPUs. If TSX is disabled, SMT 7924 is not disabled because CPU is not 7925 vulnerable to cross-thread TAA attacks. 7926 off - Unconditionally disable TAA mitigation 7927 7928 On MDS-affected machines, tsx_async_abort=off can be 7929 prevented by an active MDS mitigation as both vulnerabilities 7930 are mitigated with the same mechanism so in order to disable 7931 this mitigation, you need to specify mds=off too. 7932 7933 Not specifying this option is equivalent to 7934 tsx_async_abort=full. On CPUs which are MDS affected 7935 and deploy MDS mitigation, TAA mitigation is not 7936 required and doesn't provide any additional 7937 mitigation. 7938 7939 For details see: 7940 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 7941 7942 turbografx.map[2|3]= [HW,JOY] 7943 TurboGraFX parallel port interface 7944 Format: 7945 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 7946 See also Documentation/input/devices/joystick-parport.rst 7947 7948 udbg-immortal [PPC] When debugging early kernel crashes that 7949 happen after console_init() and before a proper 7950 console driver takes over, this boot options might 7951 help "seeing" what's going on. 7952 7953 uhash_entries= [KNL,NET] 7954 Set number of hash buckets for UDP/UDP-Lite connections 7955 7956 uhci-hcd.ignore_oc= 7957 [USB] Ignore overcurrent events (default N). 7958 Some badly-designed motherboards generate lots of 7959 bogus events, for ports that aren't wired to 7960 anything. Set this parameter to avoid log spamming. 7961 Note that genuine overcurrent events won't be 7962 reported either. 7963 7964 unaligned_scalar_speed= 7965 [RISCV] 7966 Format: {slow | fast | unsupported} 7967 Allow skipping scalar unaligned access speed tests. This 7968 is useful for testing alternative code paths and to skip 7969 the tests in environments where they run too slowly. All 7970 CPUs must have the same scalar unaligned access speed. 7971 7972 unaligned_vector_speed= 7973 [RISCV] 7974 Format: {slow | fast | unsupported} 7975 Allow skipping vector unaligned access speed tests. This 7976 is useful for testing alternative code paths and to skip 7977 the tests in environments where they run too slowly. All 7978 CPUs must have the same vector unaligned access speed. 7979 7980 unknown_nmi_panic 7981 [X86] Cause panic on unknown NMI. 7982 7983 unwind_debug [X86-64,EARLY] 7984 Enable unwinder debug output. This can be 7985 useful for debugging certain unwinder error 7986 conditions, including corrupt stacks and 7987 bad/missing unwinder metadata. 7988 7989 usbcore.authorized_default= 7990 [USB] Default USB device authorization: 7991 (default -1 = authorized (same as 1), 7992 0 = not authorized, 1 = authorized, 2 = authorized 7993 if device connected to internal port) 7994 7995 usbcore.autosuspend= 7996 [USB] The autosuspend time delay (in seconds) used 7997 for newly-detected USB devices (default 2). This 7998 is the time required before an idle device will be 7999 autosuspended. Devices for which the delay is set 8000 to a negative value won't be autosuspended at all. 8001 8002 usbcore.usbfs_snoop= 8003 [USB] Set to log all usbfs traffic (default 0 = off). 8004 8005 usbcore.usbfs_snoop_max= 8006 [USB] Maximum number of bytes to snoop in each URB 8007 (default = 65536). 8008 8009 usbcore.blinkenlights= 8010 [USB] Set to cycle leds on hubs (default 0 = off). 8011 8012 usbcore.old_scheme_first= 8013 [USB] Start with the old device initialization 8014 scheme (default 0 = off). 8015 8016 usbcore.usbfs_memory_mb= 8017 [USB] Memory limit (in MB) for buffers allocated by 8018 usbfs (default = 16, 0 = max = 2047). 8019 8020 usbcore.use_both_schemes= 8021 [USB] Try the other device initialization scheme 8022 if the first one fails (default 1 = enabled). 8023 8024 usbcore.initial_descriptor_timeout= 8025 [USB] Specifies timeout for the initial 64-byte 8026 USB_REQ_GET_DESCRIPTOR request in milliseconds 8027 (default 5000 = 5.0 seconds). 8028 8029 usbcore.nousb [USB] Disable the USB subsystem 8030 8031 usbcore.quirks= 8032 [USB] A list of quirk entries to augment the built-in 8033 usb core quirk list. List entries are separated by 8034 commas. Each entry has the form 8035 VendorID:ProductID:Flags. The IDs are 4-digit hex 8036 numbers and Flags is a set of letters. Each letter 8037 will change the built-in quirk; setting it if it is 8038 clear and clearing it if it is set. The letters have 8039 the following meanings: 8040 a = USB_QUIRK_STRING_FETCH_255 (string 8041 descriptors must not be fetched using 8042 a 255-byte read); 8043 b = USB_QUIRK_RESET_RESUME (device can't resume 8044 correctly so reset it instead); 8045 c = USB_QUIRK_NO_SET_INTF (device can't handle 8046 Set-Interface requests); 8047 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 8048 handle its Configuration or Interface 8049 strings); 8050 e = USB_QUIRK_RESET (device can't be reset 8051 (e.g morph devices), don't use reset); 8052 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 8053 more interface descriptions than the 8054 bNumInterfaces count, and can't handle 8055 talking to these interfaces); 8056 g = USB_QUIRK_DELAY_INIT (device needs a pause 8057 during initialization, after we read 8058 the device descriptor); 8059 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 8060 high speed and super speed interrupt 8061 endpoints, the USB 2.0 and USB 3.0 spec 8062 require the interval in microframes (1 8063 microframe = 125 microseconds) to be 8064 calculated as interval = 2 ^ 8065 (bInterval-1). 8066 Devices with this quirk report their 8067 bInterval as the result of this 8068 calculation instead of the exponent 8069 variable used in the calculation); 8070 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 8071 handle device_qualifier descriptor 8072 requests); 8073 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 8074 generates spurious wakeup, ignore 8075 remote wakeup capability); 8076 k = USB_QUIRK_NO_LPM (device can't handle Link 8077 Power Management); 8078 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 8079 (Device reports its bInterval as linear 8080 frames instead of the USB 2.0 8081 calculation); 8082 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 8083 to be disconnected before suspend to 8084 prevent spurious wakeup); 8085 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 8086 pause after every control message); 8087 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 8088 delay after resetting its port); 8089 p = USB_QUIRK_SHORT_SET_ADDRESS_REQ_TIMEOUT 8090 (Reduce timeout of the SET_ADDRESS 8091 request from 5000 ms to 500 ms); 8092 Example: quirks=0781:5580:bk,0a5c:5834:gij 8093 8094 usbhid.mousepoll= 8095 [USBHID] The interval which mice are to be polled at. 8096 8097 usbhid.jspoll= 8098 [USBHID] The interval which joysticks are to be polled at. 8099 8100 usbhid.kbpoll= 8101 [USBHID] The interval which keyboards are to be polled at. 8102 8103 usb-storage.delay_use= 8104 [UMS] The delay in seconds before a new device is 8105 scanned for Logical Units (default 1). 8106 Optionally the delay in milliseconds if the value has 8107 suffix with "ms". 8108 Example: delay_use=2567ms 8109 8110 usb-storage.quirks= 8111 [UMS] A list of quirks entries to supplement or 8112 override the built-in unusual_devs list. List 8113 entries are separated by commas. Each entry has 8114 the form VID:PID:Flags where VID and PID are Vendor 8115 and Product ID values (4-digit hex numbers) and 8116 Flags is a set of characters, each corresponding 8117 to a common usb-storage quirk flag as follows: 8118 a = SANE_SENSE (collect more than 18 bytes 8119 of sense data, not on uas); 8120 b = BAD_SENSE (don't collect more than 18 8121 bytes of sense data, not on uas); 8122 c = FIX_CAPACITY (decrease the reported 8123 device capacity by one sector); 8124 d = NO_READ_DISC_INFO (don't use 8125 READ_DISC_INFO command, not on uas); 8126 e = NO_READ_CAPACITY_16 (don't use 8127 READ_CAPACITY_16 command); 8128 f = NO_REPORT_OPCODES (don't use report opcodes 8129 command, uas only); 8130 g = MAX_SECTORS_240 (don't transfer more than 8131 240 sectors at a time, uas only); 8132 h = CAPACITY_HEURISTICS (decrease the 8133 reported device capacity by one 8134 sector if the number is odd); 8135 i = IGNORE_DEVICE (don't bind to this 8136 device); 8137 j = NO_REPORT_LUNS (don't use report luns 8138 command, uas only); 8139 k = NO_SAME (do not use WRITE_SAME, uas only) 8140 l = NOT_LOCKABLE (don't try to lock and 8141 unlock ejectable media, not on uas); 8142 m = MAX_SECTORS_64 (don't transfer more 8143 than 64 sectors = 32 KB at a time, 8144 not on uas); 8145 n = INITIAL_READ10 (force a retry of the 8146 initial READ(10) command, not on uas); 8147 o = CAPACITY_OK (accept the capacity 8148 reported by the device, not on uas); 8149 p = WRITE_CACHE (the device cache is ON 8150 by default, not on uas); 8151 r = IGNORE_RESIDUE (the device reports 8152 bogus residue values, not on uas); 8153 s = SINGLE_LUN (the device has only one 8154 Logical Unit); 8155 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 8156 commands, uas only); 8157 u = IGNORE_UAS (don't bind to the uas driver); 8158 w = NO_WP_DETECT (don't test whether the 8159 medium is write-protected). 8160 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 8161 even if the device claims no cache, 8162 not on uas) 8163 Example: quirks=0419:aaf5:rl,0421:0433:rc 8164 8165 user_debug= [KNL,ARM] 8166 Format: <int> 8167 See arch/arm/Kconfig.debug help text. 8168 1 - undefined instruction events 8169 2 - system calls 8170 4 - invalid data aborts 8171 8 - SIGSEGV faults 8172 16 - SIGBUS faults 8173 Example: user_debug=31 8174 8175 vdso= [X86,SH,SPARC] 8176 On X86_32, this is an alias for vdso32=. Otherwise: 8177 8178 vdso=1: enable VDSO (the default) 8179 vdso=0: disable VDSO mapping 8180 8181 vdso32= [X86] Control the 32-bit vDSO 8182 vdso32=1: enable 32-bit VDSO 8183 vdso32=0 or vdso32=2: disable 32-bit VDSO 8184 8185 See the help text for CONFIG_COMPAT_VDSO for more 8186 details. If CONFIG_COMPAT_VDSO is set, the default is 8187 vdso32=0; otherwise, the default is vdso32=1. 8188 8189 For compatibility with older kernels, vdso32=2 is an 8190 alias for vdso32=0. 8191 8192 Try vdso32=0 if you encounter an error that says: 8193 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 8194 8195 video= [FB,EARLY] Frame buffer configuration 8196 See Documentation/fb/modedb.rst. 8197 8198 video.brightness_switch_enabled= [ACPI] 8199 Format: [0|1] 8200 If set to 1, on receiving an ACPI notify event 8201 generated by hotkey, video driver will adjust brightness 8202 level and then send out the event to user space through 8203 the allocated input device. If set to 0, video driver 8204 will only send out the event without touching backlight 8205 brightness level. 8206 default: 1 8207 8208 virtio_mmio.device= 8209 [VMMIO] Memory mapped virtio (platform) device. 8210 8211 <size>@<baseaddr>:<irq>[:<id>] 8212 where: 8213 <size> := size (can use standard suffixes 8214 like K, M and G) 8215 <baseaddr> := physical base address 8216 <irq> := interrupt number (as passed to 8217 request_irq()) 8218 <id> := (optional) platform device id 8219 example: 8220 virtio_mmio.device=1K@0x100b0000:48:7 8221 8222 Can be used multiple times for multiple devices. 8223 8224 vga= [BOOT,X86-32] Select a particular video mode 8225 See Documentation/arch/x86/boot.rst and 8226 Documentation/admin-guide/svga.rst. 8227 Use vga=ask for menu. 8228 This is actually a boot loader parameter; the value is 8229 passed to the kernel using a special protocol. 8230 8231 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 8232 May slow down system boot speed, especially when 8233 enabled on systems with a large amount of memory. 8234 All options are enabled by default, and this 8235 interface is meant to allow for selectively 8236 enabling or disabling specific virtual memory 8237 debugging features. 8238 8239 Available options are: 8240 P Enable page structure init time poisoning 8241 - Disable all of the above options 8242 8243 vmalloc=nn[KMG] [KNL,BOOT,EARLY] Forces the vmalloc area to have an 8244 exact size of <nn>. This can be used to increase 8245 the minimum size (128MB on x86, arm32 platforms). 8246 It can also be used to decrease the size and leave more room 8247 for directly mapped kernel RAM. Note that this parameter does 8248 not exist on many other platforms (including arm64, alpha, 8249 loongarch, arc, csky, hexagon, microblaze, mips, nios2, openrisc, 8250 parisc, m64k, powerpc, riscv, sh, um, xtensa, s390, sparc). 8251 8252 vmcp_cma=nn[MG] [KNL,S390,EARLY] 8253 Sets the memory size reserved for contiguous memory 8254 allocations for the vmcp device driver. 8255 8256 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 8257 Format: <command> 8258 8259 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 8260 Format: <command> 8261 8262 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 8263 Format: <command> 8264 8265 vmscape= [X86] Controls mitigation for VMscape attacks. 8266 VMscape attacks can leak information from a userspace 8267 hypervisor to a guest via speculative side-channels. 8268 8269 off - disable the mitigation 8270 ibpb - use Indirect Branch Prediction Barrier 8271 (IBPB) mitigation (default) 8272 force - force vulnerability detection even on 8273 unaffected processors 8274 8275 vsyscall= [X86-64,EARLY] 8276 Controls the behavior of vsyscalls (i.e. calls to 8277 fixed addresses of 0xffffffffff600x00 from legacy 8278 code). Most statically-linked binaries and older 8279 versions of glibc use these calls. Because these 8280 functions are at fixed addresses, they make nice 8281 targets for exploits that can control RIP. 8282 8283 emulate Vsyscalls turn into traps and are emulated 8284 reasonably safely. The vsyscall page is 8285 readable. 8286 8287 xonly [default] Vsyscalls turn into traps and are 8288 emulated reasonably safely. The vsyscall 8289 page is not readable. 8290 8291 none Vsyscalls don't work at all. This makes 8292 them quite hard to use for exploits but 8293 might break your system. 8294 8295 vt.color= [VT] Default text color. 8296 Format: 0xYX, X = foreground, Y = background. 8297 Default: 0x07 = light gray on black. 8298 8299 vt.cur_default= [VT] Default cursor shape. 8300 Format: 0xCCBBAA, where AA, BB, and CC are the same as 8301 the parameters of the <Esc>[?A;B;Cc escape sequence; 8302 see vga-softcursor.rst. Default: 2 = underline. 8303 8304 vt.default_blu= [VT] 8305 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 8306 Change the default blue palette of the console. 8307 This is a 16-member array composed of values 8308 ranging from 0-255. 8309 8310 vt.default_grn= [VT] 8311 Format: <green0>,<green1>,<green2>,...,<green15> 8312 Change the default green palette of the console. 8313 This is a 16-member array composed of values 8314 ranging from 0-255. 8315 8316 vt.default_red= [VT] 8317 Format: <red0>,<red1>,<red2>,...,<red15> 8318 Change the default red palette of the console. 8319 This is a 16-member array composed of values 8320 ranging from 0-255. 8321 8322 vt.default_utf8= 8323 [VT] 8324 Format=<0|1> 8325 Set system-wide default UTF-8 mode for all tty's. 8326 Default is 1, i.e. UTF-8 mode is enabled for all 8327 newly opened terminals. 8328 8329 vt.global_cursor_default= 8330 [VT] 8331 Format=<-1|0|1> 8332 Set system-wide default for whether a cursor 8333 is shown on new VTs. Default is -1, 8334 i.e. cursors will be created by default unless 8335 overridden by individual drivers. 0 will hide 8336 cursors, 1 will display them. 8337 8338 vt.italic= [VT] Default color for italic text; 0-15. 8339 Default: 2 = green. 8340 8341 vt.underline= [VT] Default color for underlined text; 0-15. 8342 Default: 3 = cyan. 8343 8344 watchdog timers [HW,WDT] For information on watchdog timers, 8345 see Documentation/watchdog/watchdog-parameters.rst 8346 or other driver-specific files in the 8347 Documentation/watchdog/ directory. 8348 8349 watchdog_thresh= 8350 [KNL] 8351 Set the hard lockup detector stall duration 8352 threshold in seconds. The soft lockup detector 8353 threshold is set to twice the value. A value of 0 8354 disables both lockup detectors. Default is 10 8355 seconds. 8356 8357 workqueue.unbound_cpus= 8358 [KNL,SMP] Specify to constrain one or some CPUs 8359 to use in unbound workqueues. 8360 Format: <cpu-list> 8361 By default, all online CPUs are available for 8362 unbound workqueues. 8363 8364 workqueue.watchdog_thresh= 8365 If CONFIG_WQ_WATCHDOG is configured, workqueue can 8366 warn stall conditions and dump internal state to 8367 help debugging. 0 disables workqueue stall 8368 detection; otherwise, it's the stall threshold 8369 duration in seconds. The default value is 30 and 8370 it can be updated at runtime by writing to the 8371 corresponding sysfs file. 8372 8373 workqueue.panic_on_stall=<uint> 8374 Panic when workqueue stall is detected by 8375 CONFIG_WQ_WATCHDOG. It sets the number times of the 8376 stall to trigger panic. 8377 8378 The default is 0, which disables the panic on stall. 8379 8380 workqueue.cpu_intensive_thresh_us= 8381 Per-cpu work items which run for longer than this 8382 threshold are automatically considered CPU intensive 8383 and excluded from concurrency management to prevent 8384 them from noticeably delaying other per-cpu work 8385 items. Default is 10000 (10ms). 8386 8387 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8388 will report the work functions which violate this 8389 threshold repeatedly. They are likely good 8390 candidates for using WQ_UNBOUND workqueues instead. 8391 8392 workqueue.cpu_intensive_warning_thresh=<uint> 8393 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8394 will report the work functions which violate the 8395 intensive_threshold_us repeatedly. In order to prevent 8396 spurious warnings, start printing only after a work 8397 function has violated this threshold number of times. 8398 8399 The default is 4 times. 0 disables the warning. 8400 8401 workqueue.power_efficient 8402 Per-cpu workqueues are generally preferred because 8403 they show better performance thanks to cache 8404 locality; unfortunately, per-cpu workqueues tend to 8405 be more power hungry than unbound workqueues. 8406 8407 Enabling this makes the per-cpu workqueues which 8408 were observed to contribute significantly to power 8409 consumption unbound, leading to measurably lower 8410 power usage at the cost of small performance 8411 overhead. 8412 8413 The default value of this parameter is determined by 8414 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 8415 8416 workqueue.default_affinity_scope= 8417 Select the default affinity scope to use for unbound 8418 workqueues. Can be one of "cpu", "smt", "cache", 8419 "numa" and "system". Default is "cache". For more 8420 information, see the Affinity Scopes section in 8421 Documentation/core-api/workqueue.rst. 8422 8423 This can be changed after boot by writing to the 8424 matching /sys/module/workqueue/parameters file. All 8425 workqueues with the "default" affinity scope will be 8426 updated accordingly. 8427 8428 workqueue.debug_force_rr_cpu 8429 Workqueue used to implicitly guarantee that work 8430 items queued without explicit CPU specified are put 8431 on the local CPU. This guarantee is no longer true 8432 and while local CPU is still preferred work items 8433 may be put on foreign CPUs. This debug option 8434 forces round-robin CPU selection to flush out 8435 usages which depend on the now broken guarantee. 8436 When enabled, memory and cache locality will be 8437 impacted. 8438 8439 writecombine= [LOONGARCH,EARLY] Control the MAT (Memory Access 8440 Type) of ioremap_wc(). 8441 8442 on - Enable writecombine, use WUC for ioremap_wc() 8443 off - Disable writecombine, use SUC for ioremap_wc() 8444 8445 x2apic_phys [X86-64,APIC,EARLY] Use x2apic physical mode instead of 8446 default x2apic cluster mode on platforms 8447 supporting x2apic. 8448 8449 xen_512gb_limit [KNL,X86-64,XEN] 8450 Restricts the kernel running paravirtualized under Xen 8451 to use only up to 512 GB of RAM. The reason to do so is 8452 crash analysis tools and Xen tools for doing domain 8453 save/restore/migration must be enabled to handle larger 8454 domains. 8455 8456 xen_console_io [XEN,EARLY] 8457 Boolean option to enable/disable the usage of the Xen 8458 console_io hypercalls to read and write to the console. 8459 Mostly useful for debugging and development. 8460 8461 xen_emul_unplug= [HW,X86,XEN,EARLY] 8462 Unplug Xen emulated devices 8463 Format: [unplug0,][unplug1] 8464 ide-disks -- unplug primary master IDE devices 8465 aux-ide-disks -- unplug non-primary-master IDE devices 8466 nics -- unplug network devices 8467 all -- unplug all emulated devices (NICs and IDE disks) 8468 unnecessary -- unplugging emulated devices is 8469 unnecessary even if the host did not respond to 8470 the unplug protocol 8471 never -- do not unplug even if version check succeeds 8472 8473 xen_legacy_crash [X86,XEN,EARLY] 8474 Crash from Xen panic notifier, without executing late 8475 panic() code such as dumping handler. 8476 8477 xen_mc_debug [X86,XEN,EARLY] 8478 Enable multicall debugging when running as a Xen PV guest. 8479 Enabling this feature will reduce performance a little 8480 bit, so it should only be enabled for obtaining extended 8481 debug data in case of multicall errors. 8482 8483 xen_msr_safe= [X86,XEN,EARLY] 8484 Format: <bool> 8485 Select whether to always use non-faulting (safe) MSR 8486 access functions when running as Xen PV guest. The 8487 default value is controlled by CONFIG_XEN_PV_MSR_SAFE. 8488 8489 xen_nopv [X86] 8490 Disables the PV optimizations forcing the HVM guest to 8491 run as generic HVM guest with no PV drivers. 8492 This option is obsoleted by the "nopv" option, which 8493 has equivalent effect for XEN platform. 8494 8495 xen_no_vector_callback 8496 [KNL,X86,XEN,EARLY] Disable the vector callback for Xen 8497 event channel interrupts. 8498 8499 xen_scrub_pages= [XEN] 8500 Boolean option to control scrubbing pages before giving them back 8501 to Xen, for use by other domains. Can be also changed at runtime 8502 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 8503 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 8504 8505 xen_timer_slop= [X86-64,XEN,EARLY] 8506 Set the timer slop (in nanoseconds) for the virtual Xen 8507 timers (default is 100000). This adjusts the minimum 8508 delta of virtualized Xen timers, where lower values 8509 improve timer resolution at the expense of processing 8510 more timer interrupts. 8511 8512 xen.balloon_boot_timeout= [XEN] 8513 The time (in seconds) to wait before giving up to boot 8514 in case initial ballooning fails to free enough memory. 8515 Applies only when running as HVM or PVH guest and 8516 started with less memory configured than allowed at 8517 max. Default is 180. 8518 8519 xen.event_eoi_delay= [XEN] 8520 How long to delay EOI handling in case of event 8521 storms (jiffies). Default is 10. 8522 8523 xen.event_loop_timeout= [XEN] 8524 After which time (jiffies) the event handling loop 8525 should start to delay EOI handling. Default is 2. 8526 8527 xen.fifo_events= [XEN] 8528 Boolean parameter to disable using fifo event handling 8529 even if available. Normally fifo event handling is 8530 preferred over the 2-level event handling, as it is 8531 fairer and the number of possible event channels is 8532 much higher. Default is on (use fifo events). 8533 8534 xirc2ps_cs= [NET,PCMCIA] 8535 Format: 8536 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 8537 8538 xive= [PPC] 8539 By default on POWER9 and above, the kernel will 8540 natively use the XIVE interrupt controller. This option 8541 allows the fallback firmware mode to be used: 8542 8543 off Fallback to firmware control of XIVE interrupt 8544 controller on both pseries and powernv 8545 platforms. Only useful on POWER9 and above. 8546 8547 xive.store-eoi=off [PPC] 8548 By default on POWER10 and above, the kernel will use 8549 stores for EOI handling when the XIVE interrupt mode 8550 is active. This option allows the XIVE driver to use 8551 loads instead, as on POWER9. 8552 8553 xhci-hcd.quirks [USB,KNL] 8554 A hex value specifying bitmask with supplemental xhci 8555 host controller quirks. Meaning of each bit can be 8556 consulted in header drivers/usb/host/xhci.h. 8557 8558 xmon [PPC,EARLY] 8559 Format: { early | on | rw | ro | off } 8560 Controls if xmon debugger is enabled. Default is off. 8561 Passing only "xmon" is equivalent to "xmon=early". 8562 early Call xmon as early as possible on boot; xmon 8563 debugger is called from setup_arch(). 8564 on xmon debugger hooks will be installed so xmon 8565 is only called on a kernel crash. Default mode, 8566 i.e. either "ro" or "rw" mode, is controlled 8567 with CONFIG_XMON_DEFAULT_RO_MODE. 8568 rw xmon debugger hooks will be installed so xmon 8569 is called only on a kernel crash, mode is write, 8570 meaning SPR registers, memory and, other data 8571 can be written using xmon commands. 8572 ro same as "rw" option above but SPR registers, 8573 memory, and other data can't be written using 8574 xmon commands. 8575 off xmon is disabled. 8576