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 For ARM64, ONLY "acpi=off", "acpi=on", "acpi=force" or 129 "acpi=nospcr" are available 130 For RISCV64, ONLY "acpi=off", "acpi=on" or "acpi=force" 131 are available 132 133 See also Documentation/power/runtime_pm.rst, pci=noacpi 134 135 acpi_apic_instance= [ACPI,IOAPIC,EARLY] 136 Format: <int> 137 2: use 2nd APIC table, if available 138 1,0: use 1st APIC table 139 default: 0 140 141 acpi_backlight= [HW,ACPI] 142 { vendor | video | native | none } 143 If set to vendor, prefer vendor-specific driver 144 (e.g. thinkpad_acpi, sony_acpi, etc.) instead 145 of the ACPI video.ko driver. 146 If set to video, use the ACPI video.ko driver. 147 If set to native, use the device's native backlight mode. 148 If set to none, disable the ACPI backlight interface. 149 150 acpi_force_32bit_fadt_addr [ACPI,EARLY] 151 force FADT to use 32 bit addresses rather than the 152 64 bit X_* addresses. Some firmware have broken 64 153 bit addresses for force ACPI ignore these and use 154 the older legacy 32 bit addresses. 155 156 acpica_no_return_repair [HW, ACPI] 157 Disable AML predefined validation mechanism 158 This mechanism can repair the evaluation result to make 159 the return objects more ACPI specification compliant. 160 This option is useful for developers to identify the 161 root cause of an AML interpreter issue when the issue 162 has something to do with the repair mechanism. 163 164 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG] 165 acpi.debug_level= [HW,ACPI,ACPI_DEBUG] 166 Format: <int> 167 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI 168 debug output. Bits in debug_layer correspond to a 169 _COMPONENT in an ACPI source file, e.g., 170 #define _COMPONENT ACPI_EVENTS 171 Bits in debug_level correspond to a level in 172 ACPI_DEBUG_PRINT statements, e.g., 173 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ... 174 The debug_level mask defaults to "info". See 175 Documentation/firmware-guide/acpi/debug.rst for more information about 176 debug layers and levels. 177 178 Enable processor driver info messages: 179 acpi.debug_layer=0x20000000 180 Enable AML "Debug" output, i.e., stores to the Debug 181 object while interpreting AML: 182 acpi.debug_layer=0xffffffff acpi.debug_level=0x2 183 Enable all messages related to ACPI hardware: 184 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff 185 186 Some values produce so much output that the system is 187 unusable. The "log_buf_len" parameter may be useful 188 if you need to capture more output. 189 190 acpi_enforce_resources= [ACPI] 191 { strict | lax | no } 192 Check for resource conflicts between native drivers 193 and ACPI OperationRegions (SystemIO and SystemMemory 194 only). IO ports and memory declared in ACPI might be 195 used by the ACPI subsystem in arbitrary AML code and 196 can interfere with legacy drivers. 197 strict (default): access to resources claimed by ACPI 198 is denied; legacy drivers trying to access reserved 199 resources will fail to bind to device using them. 200 lax: access to resources claimed by ACPI is allowed; 201 legacy drivers trying to access reserved resources 202 will bind successfully but a warning message is logged. 203 no: ACPI OperationRegions are not marked as reserved, 204 no further checks are performed. 205 206 acpi_force_table_verification [HW,ACPI,EARLY] 207 Enable table checksum verification during early stage. 208 By default, this is disabled due to x86 early mapping 209 size limitation. 210 211 acpi_irq_balance [HW,ACPI] 212 ACPI will balance active IRQs 213 default in APIC mode 214 215 acpi_irq_nobalance [HW,ACPI] 216 ACPI will not move active IRQs (default) 217 default in PIC mode 218 219 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 220 Format: <irq>,<irq>... 221 222 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 223 use by PCI 224 Format: <irq>,<irq>... 225 226 acpi_mask_gpe= [HW,ACPI] 227 Due to the existence of _Lxx/_Exx, some GPEs triggered 228 by unsupported hardware/firmware features can result in 229 GPE floodings that cannot be automatically disabled by 230 the GPE dispatcher. 231 This facility can be used to prevent such uncontrolled 232 GPE floodings. 233 Format: <byte> or <bitmap-list> 234 235 acpi_no_auto_serialize [HW,ACPI] 236 Disable auto-serialization of AML methods 237 AML control methods that contain the opcodes to create 238 named objects will be marked as "Serialized" by the 239 auto-serialization feature. 240 This feature is enabled by default. 241 This option allows to turn off the feature. 242 243 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 244 kernels. 245 246 acpi_no_static_ssdt [HW,ACPI,EARLY] 247 Disable installation of static SSDTs at early boot time 248 By default, SSDTs contained in the RSDT/XSDT will be 249 installed automatically and they will appear under 250 /sys/firmware/acpi/tables. 251 This option turns off this feature. 252 Note that specifying this option does not affect 253 dynamic table installation which will install SSDT 254 tables to /sys/firmware/acpi/tables/dynamic. 255 256 acpi_no_watchdog [HW,ACPI,WDT] 257 Ignore the ACPI-based watchdog interface (WDAT) and let 258 a native driver control the watchdog device instead. 259 260 acpi_rsdp= [ACPI,EFI,KEXEC,EARLY] 261 Pass the RSDP address to the kernel, mostly used 262 on machines running EFI runtime service to boot the 263 second kernel for kdump. 264 265 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 266 Format: To spoof as Windows 98: ="Microsoft Windows" 267 268 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 269 of 2 which is mandated by ACPI 6) as the supported ACPI 270 specification revision (when using this switch, it may 271 be necessary to carry out a cold reboot _twice_ in a 272 row to make it take effect on the platform firmware). 273 274 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 275 acpi_osi="string1" # add string1 276 acpi_osi="!string2" # remove string2 277 acpi_osi=!* # remove all strings 278 acpi_osi=! # disable all built-in OS vendor 279 strings 280 acpi_osi=!! # enable all built-in OS vendor 281 strings 282 acpi_osi= # disable all strings 283 284 'acpi_osi=!' can be used in combination with single or 285 multiple 'acpi_osi="string1"' to support specific OS 286 vendor string(s). Note that such command can only 287 affect the default state of the OS vendor strings, thus 288 it cannot affect the default state of the feature group 289 strings and the current state of the OS vendor strings, 290 specifying it multiple times through kernel command line 291 is meaningless. This command is useful when one do not 292 care about the state of the feature group strings which 293 should be controlled by the OSPM. 294 Examples: 295 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 296 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 297 can make '_OSI("Windows 2000")' TRUE. 298 299 'acpi_osi=' cannot be used in combination with other 300 'acpi_osi=' command lines, the _OSI method will not 301 exist in the ACPI namespace. NOTE that such command can 302 only affect the _OSI support state, thus specifying it 303 multiple times through kernel command line is also 304 meaningless. 305 Examples: 306 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 307 FALSE. 308 309 'acpi_osi=!*' can be used in combination with single or 310 multiple 'acpi_osi="string1"' to support specific 311 string(s). Note that such command can affect the 312 current state of both the OS vendor strings and the 313 feature group strings, thus specifying it multiple times 314 through kernel command line is meaningful. But it may 315 still not able to affect the final state of a string if 316 there are quirks related to this string. This command 317 is useful when one want to control the state of the 318 feature group strings to debug BIOS issues related to 319 the OSPM features. 320 Examples: 321 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 322 '_OSI("Module Device")' FALSE. 323 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 324 '_OSI("Module Device")' TRUE. 325 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 326 equivalent to 327 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 328 and 329 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 330 they all will make '_OSI("Windows 2000")' TRUE. 331 332 acpi_pm_good [X86] 333 Override the pmtimer bug detection: force the kernel 334 to assume that this machine's pmtimer latches its value 335 and always returns good values. 336 337 acpi_sci= [HW,ACPI,EARLY] ACPI System Control Interrupt trigger mode 338 Format: { level | edge | high | low } 339 340 acpi_skip_timer_override [HW,ACPI,EARLY] 341 Recognize and ignore IRQ0/pin2 Interrupt Override. 342 For broken nForce2 BIOS resulting in XT-PIC timer. 343 344 acpi_sleep= [HW,ACPI] Sleep options 345 Format: { s3_bios, s3_mode, s3_beep, s4_hwsig, 346 s4_nohwsig, old_ordering, nonvs, 347 sci_force_enable, nobl } 348 See Documentation/power/video.rst for information on 349 s3_bios and s3_mode. 350 s3_beep is for debugging; it makes the PC's speaker beep 351 as soon as the kernel's real-mode entry point is called. 352 s4_hwsig causes the kernel to check the ACPI hardware 353 signature during resume from hibernation, and gracefully 354 refuse to resume if it has changed. This complies with 355 the ACPI specification but not with reality, since 356 Windows does not do this and many laptops do change it 357 on docking. So the default behaviour is to allow resume 358 and simply warn when the signature changes, unless the 359 s4_hwsig option is enabled. 360 s4_nohwsig prevents ACPI hardware signature from being 361 used (or even warned about) during resume. 362 old_ordering causes the ACPI 1.0 ordering of the _PTS 363 control method, with respect to putting devices into 364 low power states, to be enforced (the ACPI 2.0 ordering 365 of _PTS is used by default). 366 nonvs prevents the kernel from saving/restoring the 367 ACPI NVS memory during suspend/hibernation and resume. 368 sci_force_enable causes the kernel to set SCI_EN directly 369 on resume from S1/S3 (which is against the ACPI spec, 370 but some broken systems don't work without it). 371 nobl causes the internal blacklist of systems known to 372 behave incorrectly in some ways with respect to system 373 suspend and resume to be ignored (use wisely). 374 375 acpi_use_timer_override [HW,ACPI,EARLY] 376 Use timer override. For some broken Nvidia NF5 boards 377 that require a timer override, but don't have HPET 378 379 add_efi_memmap [EFI,X86,EARLY] Include EFI memory map in 380 kernel's map of available physical RAM. 381 382 agp= [AGP] 383 { off | try_unsupported } 384 off: disable AGP support 385 try_unsupported: try to drive unsupported chipsets 386 (may crash computer or cause data corruption) 387 388 ALSA [HW,ALSA] 389 See Documentation/sound/alsa-configuration.rst 390 391 alignment= [KNL,ARM] 392 Allow the default userspace alignment fault handler 393 behaviour to be specified. Bit 0 enables warnings, 394 bit 1 enables fixups, and bit 2 sends a segfault. 395 396 align_va_addr= [X86-64] 397 Align virtual addresses by clearing slice [14:12] when 398 allocating a VMA at process creation time. This option 399 gives you up to 3% performance improvement on AMD F15h 400 machines (where it is enabled by default) for a 401 CPU-intensive style benchmark, and it can vary highly in 402 a microbenchmark depending on workload and compiler. 403 404 32: only for 32-bit processes 405 64: only for 64-bit processes 406 on: enable for both 32- and 64-bit processes 407 off: disable for both 32- and 64-bit processes 408 409 alloc_snapshot [FTRACE] 410 Allocate the ftrace snapshot buffer on boot up when the 411 main buffer is allocated. This is handy if debugging 412 and you need to use tracing_snapshot() on boot up, and 413 do not want to use tracing_snapshot_alloc() as it needs 414 to be done where GFP_KERNEL allocations are allowed. 415 416 allow_mismatched_32bit_el0 [ARM64,EARLY] 417 Allow execve() of 32-bit applications and setting of the 418 PER_LINUX32 personality on systems where only a strict 419 subset of the CPUs support 32-bit EL0. When this 420 parameter is present, the set of CPUs supporting 32-bit 421 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0 422 and hot-unplug operations may be restricted. 423 424 See Documentation/arch/arm64/asymmetric-32bit.rst for more 425 information. 426 427 amd_iommu= [HW,X86-64] 428 Pass parameters to the AMD IOMMU driver in the system. 429 Possible values are: 430 fullflush - Deprecated, equivalent to iommu.strict=1 431 off - do not initialize any AMD IOMMU found in 432 the system 433 force_isolation - Force device isolation for all 434 devices. The IOMMU driver is not 435 allowed anymore to lift isolation 436 requirements as needed. This option 437 does not override iommu=pt 438 force_enable - Force enable the IOMMU on platforms known 439 to be buggy with IOMMU enabled. Use this 440 option with care. 441 pgtbl_v1 - Use v1 page table for DMA-API (Default). 442 pgtbl_v2 - Use v2 page table for DMA-API. 443 irtcachedis - Disable Interrupt Remapping Table (IRT) caching. 444 nohugepages - Limit page-sizes used for v1 page-tables 445 to 4 KiB. 446 v2_pgsizes_only - Limit page-sizes used for v1 page-tables 447 to 4KiB/2Mib/1GiB. 448 449 450 amd_iommu_dump= [HW,X86-64] 451 Enable AMD IOMMU driver option to dump the ACPI table 452 for AMD IOMMU. With this option enabled, AMD IOMMU 453 driver will print ACPI tables for AMD IOMMU during 454 IOMMU initialization. 455 456 amd_iommu_intr= [HW,X86-64] 457 Specifies one of the following AMD IOMMU interrupt 458 remapping modes: 459 legacy - Use legacy interrupt remapping mode. 460 vapic - Use virtual APIC mode, which allows IOMMU 461 to inject interrupts directly into guest. 462 This mode requires kvm-amd.avic=1. 463 (Default when IOMMU HW support is present.) 464 465 amd_pstate= [X86,EARLY] 466 disable 467 Do not enable amd_pstate as the default 468 scaling driver for the supported processors 469 passive 470 Use amd_pstate with passive mode as a scaling driver. 471 In this mode autonomous selection is disabled. 472 Driver requests a desired performance level and platform 473 tries to match the same performance level if it is 474 satisfied by guaranteed performance level. 475 active 476 Use amd_pstate_epp driver instance as the scaling driver, 477 driver provides a hint to the hardware if software wants 478 to bias toward performance (0x0) or energy efficiency (0xff) 479 to the CPPC firmware. then CPPC power algorithm will 480 calculate the runtime workload and adjust the realtime cores 481 frequency. 482 guided 483 Activate guided autonomous mode. Driver requests minimum and 484 maximum performance level and the platform autonomously 485 selects a performance level in this range and appropriate 486 to the current workload. 487 488 amd_prefcore= 489 [X86] 490 disable 491 Disable amd-pstate preferred core. 492 493 amijoy.map= [HW,JOY] Amiga joystick support 494 Map of devices attached to JOY0DAT and JOY1DAT 495 Format: <a>,<b> 496 See also Documentation/input/joydev/joystick.rst 497 498 analog.map= [HW,JOY] Analog joystick and gamepad support 499 Specifies type or capabilities of an analog joystick 500 connected to one of 16 gameports 501 Format: <type1>,<type2>,..<type16> 502 503 apc= [HW,SPARC] 504 Power management functions (SPARCstation-4/5 + deriv.) 505 Format: noidle 506 Disable APC CPU standby support. SPARCstation-Fox does 507 not play well with APC CPU idle - disable it if you have 508 APC and your system crashes randomly. 509 510 apic [APIC,X86-64] Use IO-APIC. Default. 511 512 apic= [APIC,X86,EARLY] Advanced Programmable Interrupt Controller 513 Change the output verbosity while booting 514 Format: { quiet (default) | verbose | debug } 515 Change the amount of debugging information output 516 when initialising the APIC and IO-APIC components. 517 518 apic_extnmi= [APIC,X86,EARLY] External NMI delivery setting 519 Format: { bsp (default) | all | none } 520 bsp: External NMI is delivered only to CPU 0 521 all: External NMIs are broadcast to all CPUs as a 522 backup of CPU 0 523 none: External NMI is masked for all CPUs. This is 524 useful so that a dump capture kernel won't be 525 shot down by NMI 526 527 apicpmtimer Do APIC timer calibration using the pmtimer. Implies 528 apicmaintimer. Useful when your PIT timer is totally 529 broken. 530 531 autoconf= [IPV6] 532 See Documentation/networking/ipv6.rst. 533 534 apm= [APM] Advanced Power Management 535 See header of arch/x86/kernel/apm_32.c. 536 537 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 538 Format: { "0" | "1" } 539 See security/apparmor/Kconfig help text 540 0 -- disable. 541 1 -- enable. 542 Default value is set via kernel config option. 543 544 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards 545 Format: <io>,<irq>,<nodeID> 546 547 arm64.no32bit_el0 [ARM64] Unconditionally disable the execution of 548 32 bit applications. 549 550 arm64.nobti [ARM64] Unconditionally disable Branch Target 551 Identification support 552 553 arm64.nogcs [ARM64] Unconditionally disable Guarded Control Stack 554 support 555 556 arm64.nomops [ARM64] Unconditionally disable Memory Copy and Memory 557 Set instructions support 558 559 arm64.nompam [ARM64] Unconditionally disable Memory Partitioning And 560 Monitoring support 561 562 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension 563 support 564 565 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 566 support 567 568 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix 569 Extension support 570 571 arm64.nosve [ARM64] Unconditionally disable Scalable Vector 572 Extension support 573 574 ataflop= [HW,M68k] 575 576 atarimouse= [HW,MOUSE] Atari Mouse 577 578 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 579 EzKey and similar keyboards 580 581 atkbd.reset= [HW] Reset keyboard during initialization 582 583 atkbd.set= [HW] Select keyboard code set 584 Format: <int> (2 = AT (default), 3 = PS/2) 585 586 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 587 keyboards 588 589 atkbd.softraw= [HW] Choose between synthetic and real raw mode 590 Format: <bool> (0 = real, 1 = synthetic (default)) 591 592 atkbd.softrepeat= [HW] 593 Use software keyboard repeat 594 595 audit= [KNL] Enable the audit sub-system 596 Format: { "0" | "1" | "off" | "on" } 597 0 | off - kernel audit is disabled and can not be 598 enabled until the next reboot 599 unset - kernel audit is initialized but disabled and 600 will be fully enabled by the userspace auditd. 601 1 | on - kernel audit is initialized and partially 602 enabled, storing at most audit_backlog_limit 603 messages in RAM until it is fully enabled by the 604 userspace auditd. 605 Default: unset 606 607 audit_backlog_limit= [KNL] Set the audit queue size limit. 608 Format: <int> (must be >=0) 609 Default: 64 610 611 bau= [X86_UV] Enable the BAU on SGI UV. The default 612 behavior is to disable the BAU (i.e. bau=0). 613 Format: { "0" | "1" } 614 0 - Disable the BAU. 615 1 - Enable the BAU. 616 unset - Disable the BAU. 617 618 baycom_epp= [HW,AX25] 619 Format: <io>,<mode> 620 621 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem 622 Format: <io>,<mode> 623 See header of drivers/net/hamradio/baycom_par.c. 624 625 baycom_ser_fdx= [HW,AX25] 626 BayCom Serial Port AX.25 Modem (Full Duplex Mode) 627 Format: <io>,<irq>,<mode>[,<baud>] 628 See header of drivers/net/hamradio/baycom_ser_fdx.c. 629 630 baycom_ser_hdx= [HW,AX25] 631 BayCom Serial Port AX.25 Modem (Half Duplex Mode) 632 Format: <io>,<irq>,<mode> 633 See header of drivers/net/hamradio/baycom_ser_hdx.c. 634 635 bdev_allow_write_mounted= 636 Format: <bool> 637 Control the ability to open a mounted block device 638 for writing, i.e., allow / disallow writes that bypass 639 the FS. This was implemented as a means to prevent 640 fuzzers from crashing the kernel by overwriting the 641 metadata underneath a mounted FS without its awareness. 642 This also prevents destructive formatting of mounted 643 filesystems by naive storage tooling that don't use 644 O_EXCL. Default is Y and can be changed through the 645 Kconfig option CONFIG_BLK_DEV_WRITE_MOUNTED. 646 647 bert_disable [ACPI] 648 Disable BERT OS support on buggy BIOSes. 649 650 bgrt_disable [ACPI,X86,EARLY] 651 Disable BGRT to avoid flickering OEM logo. 652 653 blkdevparts= Manual partition parsing of block device(s) for 654 embedded devices based on command line input. 655 See Documentation/block/cmdline-partition.rst 656 657 boot_delay= [KNL,EARLY] 658 Milliseconds to delay each printk during boot. 659 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled, 660 and you may also have to specify "lpj=". Boot_delay 661 values larger than 10 seconds (10000) are assumed 662 erroneous and ignored. 663 Format: integer 664 665 bootconfig [KNL,EARLY] 666 Extended command line options can be added to an initrd 667 and this will cause the kernel to look for it. 668 669 See Documentation/admin-guide/bootconfig.rst 670 671 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 672 bttv.radio= Most important insmod options are available as 673 kernel args too. 674 bttv.pll= See Documentation/admin-guide/media/bttv.rst 675 bttv.tuner= 676 677 bulk_remove=off [PPC] This parameter disables the use of the pSeries 678 firmware feature for flushing multiple hpte entries 679 at a time. 680 681 c101= [NET] Moxa C101 synchronous serial card 682 683 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 684 Sometimes CPU hardware bugs make them report the cache 685 size incorrectly. The kernel will attempt work arounds 686 to fix known problems, but for some CPUs it is not 687 possible to determine what the correct size should be. 688 This option provides an override for these situations. 689 690 carrier_timeout= 691 [NET] Specifies amount of time (in seconds) that 692 the kernel should wait for a network carrier. By default 693 it waits 120 seconds. 694 695 ca_keys= [KEYS] This parameter identifies a specific key(s) on 696 the system trusted keyring to be used for certificate 697 trust validation. 698 format: { id:<keyid> | builtin } 699 700 cca= [MIPS,EARLY] Override the kernel pages' cache coherency 701 algorithm. Accepted values range from 0 to 7 702 inclusive. See arch/mips/include/asm/pgtable-bits.h 703 for platform specific values (SB1, Loongson3 and 704 others). 705 706 ccw_timeout_log [S390] 707 See Documentation/arch/s390/common_io.rst for details. 708 709 cfi= [X86-64] Set Control Flow Integrity checking features 710 when CONFIG_FINEIBT is enabled. 711 Format: feature[,feature...] 712 Default: auto 713 714 auto: Use FineIBT if IBT available, otherwise kCFI. 715 Under FineIBT, enable "paranoid" mode when 716 FRED is not available. 717 off: Turn off CFI checking. 718 kcfi: Use kCFI (disable FineIBT). 719 fineibt: Use FineIBT (even if IBT not available). 720 norand: Do not re-randomize CFI hashes. 721 paranoid: Add caller hash checking under FineIBT. 722 bhi: Enable register poisoning to stop speculation 723 across FineIBT. (Disabled by default.) 724 warn: Do not enforce CFI checking: warn only. 725 debug: Report CFI initialization details. 726 727 cgroup_disable= [KNL] Disable a particular controller or optional feature 728 Format: {name of the controller(s) or feature(s) to disable} 729 The effects of cgroup_disable=foo are: 730 - foo isn't auto-mounted if you mount all cgroups in 731 a single hierarchy 732 - foo isn't visible as an individually mountable 733 subsystem 734 - if foo is an optional feature then the feature is 735 disabled and corresponding cgroup files are not 736 created 737 {Currently only "memory" controller deal with this and 738 cut the overhead, others just disable the usage. So 739 only cgroup_disable=memory is actually worthy} 740 Specifying "pressure" disables per-cgroup pressure 741 stall information accounting feature 742 743 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 744 Format: { { controller | "all" | "named" } 745 [,{ controller | "all" | "named" }...] } 746 Like cgroup_disable, but only applies to cgroup v1; 747 the blacklisted controllers remain available in cgroup2. 748 "all" blacklists all controllers and "named" disables 749 named mounts. Specifying both "all" and "named" disables 750 all v1 hierarchies. 751 752 cgroup_v1_proc= [KNL] Show also missing controllers in /proc/cgroups 753 Format: { "true" | "false" } 754 /proc/cgroups lists only v1 controllers by default. 755 This compatibility option enables listing also v2 756 controllers (whose v1 code is not compiled!), so that 757 semi-legacy software can check this file to decide 758 about usage of v2 (sic) controllers. 759 760 cgroup_favordynmods= [KNL] Enable or Disable favordynmods. 761 Format: { "true" | "false" } 762 Defaults to the value of CONFIG_CGROUP_FAVOR_DYNMODS. 763 764 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 765 Format: <string> 766 nosocket -- Disable socket memory accounting. 767 nokmem -- Disable kernel memory accounting. 768 nobpf -- Disable BPF memory accounting. 769 770 checkreqprot= [SELINUX] Set initial checkreqprot flag value. 771 Format: { "0" | "1" } 772 See security/selinux/Kconfig help text. 773 0 -- check protection applied by kernel (includes 774 any implied execute protection). 775 1 -- check protection requested by application. 776 Default value is set via a kernel config option. 777 Value can be changed at runtime via 778 /sys/fs/selinux/checkreqprot. 779 Setting checkreqprot to 1 is deprecated. 780 781 cio_ignore= [S390] 782 See Documentation/arch/s390/common_io.rst for details. 783 784 clearcpuid=X[,X...] [X86] 785 Disable CPUID feature X for the kernel. See 786 arch/x86/include/asm/cpufeatures.h for the valid bit 787 numbers X. Note the Linux-specific bits are not necessarily 788 stable over kernel options, but the vendor-specific 789 ones should be. 790 X can also be a string as appearing in the flags: line 791 in /proc/cpuinfo which does not have the above 792 instability issue. However, not all features have names 793 in /proc/cpuinfo. 794 Note that using this option will taint your kernel. 795 Also note that user programs calling CPUID directly 796 or using the feature without checking anything 797 will still see it. This just prevents it from 798 being used by the kernel or shown in /proc/cpuinfo. 799 Also note the kernel might malfunction if you disable 800 some critical bits. 801 802 clk_ignore_unused 803 [CLK] 804 Prevents the clock framework from automatically gating 805 clocks that have not been explicitly enabled by a Linux 806 device driver but are enabled in hardware at reset or 807 by the bootloader/firmware. Note that this does not 808 force such clocks to be always-on nor does it reserve 809 those clocks in any way. This parameter is useful for 810 debug and development, but should not be needed on a 811 platform with proper driver support. For more 812 information, see Documentation/driver-api/clk.rst. 813 814 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 815 [Deprecated] 816 Forces specified clocksource (if available) to be used 817 when calculating gettimeofday(). If specified 818 clocksource is not available, it defaults to PIT. 819 Format: { pit | tsc | cyclone | pmtmr } 820 821 clocksource= Override the default clocksource 822 Format: <string> 823 Override the default clocksource and use the clocksource 824 with the name specified. 825 Some clocksource names to choose from, depending on 826 the platform: 827 [all] jiffies (this is the base, fallback clocksource) 828 [ACPI] acpi_pm 829 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 830 pxa_timer,timer3,32k_counter,timer0_1 831 [X86-32] pit,hpet,tsc; 832 scx200_hrt on Geode; cyclone on IBM x440 833 [MIPS] MIPS 834 [PARISC] cr16 835 [S390] tod 836 [SH] SuperH 837 [SPARC64] tick 838 [X86-64] hpet,tsc 839 840 clocksource.arm_arch_timer.evtstrm= 841 [ARM,ARM64,EARLY] 842 Format: <bool> 843 Enable/disable the eventstream feature of the ARM 844 architected timer so that code using WFE-based polling 845 loops can be debugged more effectively on production 846 systems. 847 848 clocksource.verify_n_cpus= [KNL] 849 Limit the number of CPUs checked for clocksources 850 marked with CLOCK_SOURCE_VERIFY_PERCPU that 851 are marked unstable due to excessive skew. 852 A negative value says to check all CPUs, while 853 zero says not to check any. Values larger than 854 nr_cpu_ids are silently truncated to nr_cpu_ids. 855 The actual CPUs are chosen randomly, with 856 no replacement if the same CPU is chosen twice. 857 858 clocksource-wdtest.holdoff= [KNL] 859 Set the time in seconds that the clocksource 860 watchdog test waits before commencing its tests. 861 Defaults to zero when built as a module and to 862 10 seconds when built into the kernel. 863 864 cma=nn[MG]@[start[MG][-end[MG]]] 865 [KNL,CMA,EARLY] 866 Sets the size of kernel global memory area for 867 contiguous memory allocations and optionally the 868 placement constraint by the physical address range of 869 memory allocations. A value of 0 disables CMA 870 altogether. For more information, see 871 kernel/dma/contiguous.c 872 873 cma_pernuma=nn[MG] 874 [KNL,CMA,EARLY] 875 Sets the size of kernel per-numa memory area for 876 contiguous memory allocations. A value of 0 disables 877 per-numa CMA altogether. And If this option is not 878 specified, the default value is 0. 879 With per-numa CMA enabled, DMA users on node nid will 880 first try to allocate buffer from the pernuma area 881 which is located in node nid, if the allocation fails, 882 they will fallback to the global default memory area. 883 884 numa_cma=<node>:nn[MG][,<node>:nn[MG]] 885 [KNL,CMA,EARLY] 886 Sets the size of kernel numa memory area for 887 contiguous memory allocations. It will reserve CMA 888 area for the specified node. 889 890 With numa CMA enabled, DMA users on node nid will 891 first try to allocate buffer from the numa area 892 which is located in node nid, if the allocation fails, 893 they will fallback to the global default memory area. 894 895 cmo_free_hint= [PPC] Format: { yes | no } 896 Specify whether pages are marked as being inactive 897 when they are freed. This is used in CMO environments 898 to determine OS memory pressure for page stealing by 899 a hypervisor. 900 Default: yes 901 902 coherent_pool=nn[KMG] [ARM,KNL,EARLY] 903 Sets the size of memory pool for coherent, atomic dma 904 allocations, by default set to 256K. 905 906 com20020= [HW,NET] ARCnet - COM20020 chipset 907 Format: 908 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 909 910 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 911 Format: <io>[,<irq>] 912 913 com90xx= [HW,NET] 914 ARCnet - COM90xx chipset (memory-mapped buffers) 915 Format: <io>[,<irq>[,<memstart>]] 916 917 condev= [HW,S390] console device 918 conmode= 919 920 con3215_drop= [S390,EARLY] 3215 console drop mode. 921 Format: y|n|Y|N|1|0 922 When set to true, drop data on the 3215 console when 923 the console buffer is full. In this case the 924 operator using a 3270 terminal emulator (for example 925 x3270) does not have to enter the clear key for the 926 console output to advance and the kernel to continue. 927 This leads to a much faster boot time when a 3270 928 terminal emulator is active. If no 3270 terminal 929 emulator is used, this parameter has no effect. 930 931 console= [KNL] Output console device and options. 932 933 tty<n> Use the virtual console device <n>. 934 935 ttyS<n>[,options] 936 ttyUSB0[,options] 937 Use the specified serial port. The options are of 938 the form "bbbbpnf", where "bbbb" is the baud rate, 939 "p" is parity ("n", "o", or "e"), "n" is number of 940 bits, and "f" is flow control ("r" for RTS or 941 omit it). Default is "9600n8". 942 943 See Documentation/admin-guide/serial-console.rst for more 944 information. See 945 Documentation/networking/netconsole.rst for an 946 alternative. 947 948 <DEVNAME>:<n>.<n>[,options] 949 Use the specified serial port on the serial core bus. 950 The addressing uses DEVNAME of the physical serial port 951 device, followed by the serial core controller instance, 952 and the serial port instance. The options are the same 953 as documented for the ttyS addressing above. 954 955 The mapping of the serial ports to the tty instances 956 can be viewed with: 957 958 $ ls -d /sys/bus/serial-base/devices/*:*.*/tty/* 959 /sys/bus/serial-base/devices/00:04:0.0/tty/ttyS0 960 961 In the above example, the console can be addressed with 962 console=00:04:0.0. Note that a console addressed this 963 way will only get added when the related device driver 964 is ready. The use of an earlycon parameter in addition to 965 the console may be desired for console output early on. 966 967 uart[8250],io,<addr>[,options] 968 uart[8250],mmio,<addr>[,options] 969 uart[8250],mmio16,<addr>[,options] 970 uart[8250],mmio32,<addr>[,options] 971 uart[8250],0x<addr>[,options] 972 Start an early, polled-mode console on the 8250/16550 973 UART at the specified I/O port or MMIO address, 974 switching to the matching ttyS device later. 975 MMIO inter-register address stride is either 8-bit 976 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 977 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 978 to be equivalent to 'mmio'. 'options' are specified in 979 the same format described for ttyS above; if unspecified, 980 the h/w is not re-initialized. 981 982 hvc<n> Use the hypervisor console device <n>. This is for 983 both Xen and PowerPC hypervisors. 984 985 { null | "" } 986 Use to disable console output, i.e., to have kernel 987 console messages discarded. 988 This must be the only console= parameter used on the 989 kernel command line. 990 991 If the device connected to the port is not a TTY but a braille 992 device, prepend "brl," before the device type, for instance 993 console=brl,ttyS0 994 For now, only VisioBraille is supported. 995 996 console_msg_format= 997 [KNL] Change console messages format 998 default 999 By default we print messages on consoles in 1000 "[time stamp] text\n" format (time stamp may not be 1001 printed, depending on CONFIG_PRINTK_TIME or 1002 `printk_time' param). 1003 syslog 1004 Switch to syslog format: "<%u>[time stamp] text\n" 1005 IOW, each message will have a facility and loglevel 1006 prefix. The format is similar to one used by syslog() 1007 syscall, or to executing "dmesg -S --raw" or to reading 1008 from /proc/kmsg. 1009 1010 consoleblank= [KNL] The console blank (screen saver) timeout in 1011 seconds. A value of 0 disables the blank timer. 1012 Defaults to 0. 1013 1014 coredump_filter= 1015 [KNL] Change the default value for 1016 /proc/<pid>/coredump_filter. 1017 See also Documentation/filesystems/proc.rst. 1018 1019 coresight_cpu_debug.enable 1020 [ARM,ARM64] 1021 Format: <bool> 1022 Enable/disable the CPU sampling based debugging. 1023 0: default value, disable debugging 1024 1: enable debugging at boot time 1025 1026 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 1027 Format: 1028 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 1029 1030 cpuidle.off=1 [CPU_IDLE] 1031 disable the cpuidle sub-system 1032 1033 cpuidle.governor= 1034 [CPU_IDLE] Name of the cpuidle governor to use. 1035 1036 cpufreq.off=1 [CPU_FREQ] 1037 disable the cpufreq sub-system 1038 1039 cpufreq.default_governor= 1040 [CPU_FREQ] Name of the default cpufreq governor or 1041 policy to use. This governor must be registered in the 1042 kernel before the cpufreq driver probes. 1043 1044 cpu_init_udelay=N 1045 [X86,EARLY] Delay for N microsec between assert and de-assert 1046 of APIC INIT to start processors. This delay occurs 1047 on every CPU online, such as boot, and resume from suspend. 1048 Default: 10000 1049 1050 cpuhp.parallel= 1051 [SMP] Enable/disable parallel bringup of secondary CPUs 1052 Format: <bool> 1053 Default is enabled if CONFIG_HOTPLUG_PARALLEL=y. Otherwise 1054 the parameter has no effect. 1055 1056 crash_kexec_post_notifiers 1057 Only jump to kdump kernel after running the panic 1058 notifiers and dumping kmsg. This option increases 1059 the risks of a kdump failure, since some panic 1060 notifiers can make the crashed kernel more unstable. 1061 In configurations where kdump may not be reliable, 1062 running the panic notifiers could allow collecting 1063 more data on dmesg, like stack traces from other CPUS 1064 or extra data dumped by panic_print. Note that some 1065 configurations enable this option unconditionally, 1066 like Hyper-V, PowerPC (fadump) and AMD SEV-SNP. 1067 1068 crashkernel=size[KMG][@offset[KMG]] 1069 [KNL,EARLY] Using kexec, Linux can switch to a 'crash kernel' 1070 upon panic. This parameter reserves the physical 1071 memory region [offset, offset + size] for that kernel 1072 image. If '@offset' is omitted, then a suitable offset 1073 is selected automatically. 1074 [KNL, X86-64, ARM64, RISCV, LoongArch] Select a region 1075 under 4G first, and fall back to reserve region above 1076 4G when '@offset' hasn't been specified. 1077 See Documentation/admin-guide/kdump/kdump.rst for further details. 1078 1079 crashkernel=range1:size1[,range2:size2,...][@offset] 1080 [KNL] Same as above, but depends on the memory 1081 in the running system. The syntax of range is 1082 start-[end] where start and end are both 1083 a memory unit (amount[KMG]). See also 1084 Documentation/admin-guide/kdump/kdump.rst for an example. 1085 1086 crashkernel=size[KMG],high 1087 [KNL, X86-64, ARM64, RISCV, LoongArch] range could be 1088 above 4G. 1089 Allow kernel to allocate physical memory region from top, 1090 so could be above 4G if system have more than 4G ram 1091 installed. Otherwise memory region will be allocated 1092 below 4G, if available. 1093 It will be ignored if crashkernel=X is specified. 1094 crashkernel=size[KMG],low 1095 [KNL, X86-64, ARM64, RISCV, LoongArch] range under 4G. 1096 When crashkernel=X,high is passed, kernel could allocate 1097 physical memory region above 4G, that cause second kernel 1098 crash on system that require some amount of low memory, 1099 e.g. swiotlb requires at least 64M+32K low memory, also 1100 enough extra low memory is needed to make sure DMA buffers 1101 for 32-bit devices won't run out. Kernel would try to allocate 1102 default size of memory below 4G automatically. The default 1103 size is platform dependent. 1104 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB) 1105 --> arm64: 128MiB 1106 --> riscv: 128MiB 1107 --> loongarch: 128MiB 1108 This one lets the user specify own low range under 4G 1109 for second kernel instead. 1110 0: to disable low allocation. 1111 It will be ignored when crashkernel=X,high is not used 1112 or memory reserved is below 4G. 1113 crashkernel=size[KMG],cma 1114 [KNL, X86] Reserve additional crash kernel memory from 1115 CMA. This reservation is usable by the first system's 1116 userspace memory and kernel movable allocations (memory 1117 balloon, zswap). Pages allocated from this memory range 1118 will not be included in the vmcore so this should not 1119 be used if dumping of userspace memory is intended and 1120 it has to be expected that some movable kernel pages 1121 may be missing from the dump. 1122 1123 A standard crashkernel reservation, as described above, 1124 is still needed to hold the crash kernel and initrd. 1125 1126 This option increases the risk of a kdump failure: DMA 1127 transfers configured by the first kernel may end up 1128 corrupting the second kernel's memory. 1129 1130 This reservation method is intended for systems that 1131 can't afford to sacrifice enough memory for standard 1132 crashkernel reservation and where less reliable and 1133 possibly incomplete kdump is preferable to no kdump at 1134 all. 1135 1136 cryptomgr.notests 1137 [KNL] Disable crypto self-tests 1138 1139 cs89x0_dma= [HW,NET] 1140 Format: <dma> 1141 1142 cs89x0_media= [HW,NET] 1143 Format: { rj45 | aui | bnc } 1144 1145 csdlock_debug= [KNL] Enable or disable debug add-ons of cross-CPU 1146 function call handling. When switched on, 1147 additional debug data is printed to the console 1148 in case a hanging CPU is detected, and that 1149 CPU is pinged again in order to try to resolve 1150 the hang situation. The default value of this 1151 option depends on the CSD_LOCK_WAIT_DEBUG_DEFAULT 1152 Kconfig option. 1153 1154 dasd= [HW,NET] 1155 See header of drivers/s390/block/dasd_devmap.c. 1156 1157 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 1158 (one device per port) 1159 Format: <port#>,<type> 1160 See also Documentation/input/devices/joystick-parport.rst 1161 1162 debug [KNL,EARLY] Enable kernel debugging (events log level). 1163 1164 debug_boot_weak_hash 1165 [KNL,EARLY] Enable printing [hashed] pointers early in the 1166 boot sequence. If enabled, we use a weak hash instead 1167 of siphash to hash pointers. Use this option if you are 1168 seeing instances of '(___ptrval___)') and need to see a 1169 value (hashed pointer) instead. Cryptographically 1170 insecure, please do not use on production kernels. 1171 1172 debug_locks_verbose= 1173 [KNL] verbose locking self-tests 1174 Format: <int> 1175 Print debugging info while doing the locking API 1176 self-tests. 1177 Bitmask for the various LOCKTYPE_ tests. Defaults to 0 1178 (no extra messages), setting it to -1 (all bits set) 1179 will print _a_lot_ more information - normally only 1180 useful to lockdep developers. 1181 1182 debug_objects [KNL,EARLY] Enable object debugging 1183 1184 debug_guardpage_minorder= 1185 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this 1186 parameter allows control of the order of pages that will 1187 be intentionally kept free (and hence protected) by the 1188 buddy allocator. Bigger value increase the probability 1189 of catching random memory corruption, but reduce the 1190 amount of memory for normal system use. The maximum 1191 possible value is MAX_PAGE_ORDER/2. Setting this 1192 parameter to 1 or 2 should be enough to identify most 1193 random memory corruption problems caused by bugs in 1194 kernel or driver code when a CPU writes to (or reads 1195 from) a random memory location. Note that there exists 1196 a class of memory corruptions problems caused by buggy 1197 H/W or F/W or by drivers badly programming DMA 1198 (basically when memory is written at bus level and the 1199 CPU MMU is bypassed) which are not detectable by 1200 CONFIG_DEBUG_PAGEALLOC, hence this option will not 1201 help tracking down these problems. 1202 1203 debug_pagealloc= 1204 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 1205 enables the feature at boot time. By default, it is 1206 disabled and the system will work mostly the same as a 1207 kernel built without CONFIG_DEBUG_PAGEALLOC. 1208 Note: to get most of debug_pagealloc error reports, it's 1209 useful to also enable the page_owner functionality. 1210 on: enable the feature 1211 1212 debugfs= [KNL,EARLY] This parameter enables what is exposed to 1213 userspace and debugfs internal clients. 1214 Format: { on, no-mount, off } 1215 on: All functions are enabled. 1216 no-mount: 1217 Filesystem is not registered but kernel clients can 1218 access APIs and a crashkernel can be used to read 1219 its content. There is nothing to mount. 1220 off: Filesystem is not registered and clients 1221 get a -EPERM as result when trying to register files 1222 or directories within debugfs. 1223 This is equivalent of the runtime functionality if 1224 debugfs was not enabled in the kernel at all. 1225 Default value is set in build-time with a kernel configuration. 1226 1227 debugpat [X86] Enable PAT debugging 1228 1229 default_hugepagesz= 1230 [HW] The size of the default HugeTLB page. This is 1231 the size represented by the legacy /proc/ hugepages 1232 APIs. In addition, this is the default hugetlb size 1233 used for shmget(), mmap() and mounting hugetlbfs 1234 filesystems. If not specified, defaults to the 1235 architecture's default huge page size. Huge page 1236 sizes are architecture dependent. See also 1237 Documentation/admin-guide/mm/hugetlbpage.rst. 1238 Format: size[KMG] 1239 1240 deferred_probe_timeout= 1241 [KNL] Debugging option to set a timeout in seconds for 1242 deferred probe to give up waiting on dependencies to 1243 probe. Only specific dependencies (subsystems or 1244 drivers) that have opted in will be ignored. A timeout 1245 of 0 will timeout at the end of initcalls. If the time 1246 out hasn't expired, it'll be restarted by each 1247 successful driver registration. This option will also 1248 dump out devices still on the deferred probe list after 1249 retrying. 1250 1251 delayacct [KNL] Enable per-task delay accounting 1252 1253 dell_smm_hwmon.ignore_dmi= 1254 [HW] Continue probing hardware even if DMI data 1255 indicates that the driver is running on unsupported 1256 hardware. 1257 1258 dell_smm_hwmon.force= 1259 [HW] Activate driver even if SMM BIOS signature does 1260 not match list of supported models and enable otherwise 1261 blacklisted features. 1262 1263 dell_smm_hwmon.power_status= 1264 [HW] Report power status in /proc/i8k 1265 (disabled by default). 1266 1267 dell_smm_hwmon.restricted= 1268 [HW] Allow controlling fans only if SYS_ADMIN 1269 capability is set. 1270 1271 dell_smm_hwmon.fan_mult= 1272 [HW] Factor to multiply fan speed with. 1273 1274 dell_smm_hwmon.fan_max= 1275 [HW] Maximum configurable fan speed. 1276 1277 dfltcc= [HW,S390] 1278 Format: { on | off | def_only | inf_only | always } 1279 on: s390 zlib hardware support for compression on 1280 level 1 and decompression (default) 1281 off: No s390 zlib hardware support 1282 def_only: s390 zlib hardware support for deflate 1283 only (compression on level 1) 1284 inf_only: s390 zlib hardware support for inflate 1285 only (decompression) 1286 always: Same as 'on' but ignores the selected compression 1287 level always using hardware support (used for debugging) 1288 1289 dhash_entries= [KNL] 1290 Set number of hash buckets for dentry cache. 1291 1292 disable_1tb_segments [PPC,EARLY] 1293 Disables the use of 1TB hash page table segments. This 1294 causes the kernel to fall back to 256MB segments which 1295 can be useful when debugging issues that require an SLB 1296 miss to occur. 1297 1298 disable= [IPV6] 1299 See Documentation/networking/ipv6.rst. 1300 1301 disable_radix [PPC,EARLY] 1302 Disable RADIX MMU mode on POWER9 1303 1304 disable_tlbie [PPC] 1305 Disable TLBIE instruction. Currently does not work 1306 with KVM, with HASH MMU, or with coherent accelerators. 1307 1308 disable_ddw [PPC/PSERIES,EARLY] 1309 Disable Dynamic DMA Window support. Use this 1310 to workaround buggy firmware. 1311 1312 disable_ipv6= [IPV6] 1313 See Documentation/networking/ipv6.rst. 1314 1315 disable_mtrr_cleanup [X86,EARLY] 1316 The kernel tries to adjust MTRR layout from continuous 1317 to discrete, to make X server driver able to add WB 1318 entry later. This parameter disables that. 1319 1320 disable_mtrr_trim [X86, Intel and AMD only,EARLY] 1321 By default the kernel will trim any uncacheable 1322 memory out of your available memory pool based on 1323 MTRR settings. This parameter disables that behavior, 1324 possibly causing your machine to run very slowly. 1325 1326 disable_timer_pin_1 [X86,EARLY] 1327 Disable PIN 1 of APIC timer 1328 Can be useful to work around chipset bugs. 1329 1330 dis_ucode_ldr [X86] Disable the microcode loader. 1331 1332 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 1333 this option disables the debugging code at boot. 1334 1335 dma_debug_entries=<number> 1336 This option allows to tune the number of preallocated 1337 entries for DMA-API debugging code. One entry is 1338 required per DMA-API allocation. Use this if the 1339 DMA-API debugging code disables itself because the 1340 architectural default is too low. 1341 1342 dma_debug_driver=<driver_name> 1343 With this option the DMA-API debugging driver 1344 filter feature can be enabled at boot time. Just 1345 pass the driver to filter for as the parameter. 1346 The filter can be disabled or changed to another 1347 driver later using sysfs. 1348 1349 reg_file_data_sampling= 1350 [X86] Controls mitigation for Register File Data 1351 Sampling (RFDS) vulnerability. RFDS is a CPU 1352 vulnerability which may allow userspace to infer 1353 kernel data values previously stored in floating point 1354 registers, vector registers, or integer registers. 1355 RFDS only affects Intel Atom processors. 1356 1357 on: Turns ON the mitigation. 1358 off: Turns OFF the mitigation. 1359 1360 This parameter overrides the compile time default set 1361 by CONFIG_MITIGATION_RFDS. Mitigation cannot be 1362 disabled when other VERW based mitigations (like MDS) 1363 are enabled. In order to disable RFDS mitigation all 1364 VERW based mitigations need to be disabled. 1365 1366 For details see: 1367 Documentation/admin-guide/hw-vuln/reg-file-data-sampling.rst 1368 1369 driver_async_probe= [KNL] 1370 List of driver names to be probed asynchronously. * 1371 matches with all driver names. If * is specified, the 1372 rest of the listed driver names are those that will NOT 1373 match the *. 1374 Format: <driver_name1>,<driver_name2>... 1375 1376 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1377 Broken monitors, graphic adapters, KVMs and EDIDless 1378 panels may send no or incorrect EDID data sets. 1379 This parameter allows to specify an EDID data sets 1380 in the /lib/firmware directory that are used instead. 1381 An EDID data set will only be used for a particular 1382 connector, if its name and a colon are prepended to 1383 the EDID name. Each connector may use a unique EDID 1384 data set by separating the files with a comma. An EDID 1385 data set with no connector name will be used for 1386 any connectors not explicitly specified. 1387 1388 dscc4.setup= [NET] 1389 1390 dt_cpu_ftrs= [PPC,EARLY] 1391 Format: {"off" | "known"} 1392 Control how the dt_cpu_ftrs device-tree binding is 1393 used for CPU feature discovery and setup (if it 1394 exists). 1395 off: Do not use it, fall back to legacy cpu table. 1396 known: Do not pass through unknown features to guests 1397 or userspace, only those that the kernel is aware of. 1398 1399 dump_apple_properties [X86] 1400 Dump name and content of EFI device properties on 1401 x86 Macs. Useful for driver authors to determine 1402 what data is available or for reverse-engineering. 1403 1404 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1405 <module>.dyndbg[="val"] 1406 Enable debug messages at boot time. See 1407 Documentation/admin-guide/dynamic-debug-howto.rst 1408 for details. 1409 1410 early_ioremap_debug [KNL,EARLY] 1411 Enable debug messages in early_ioremap support. This 1412 is useful for tracking down temporary early mappings 1413 which are not unmapped. 1414 1415 earlycon= [KNL,EARLY] Output early console device and options. 1416 1417 When used with no options, the early console is 1418 determined by stdout-path property in device tree's 1419 chosen node or the ACPI SPCR table if supported by 1420 the platform. 1421 1422 cdns,<addr>[,options] 1423 Start an early, polled-mode console on a Cadence 1424 (xuartps) serial port at the specified address. Only 1425 supported option is baud rate. If baud rate is not 1426 specified, the serial port must already be setup and 1427 configured. 1428 1429 uart[8250],io,<addr>[,options[,uartclk]] 1430 uart[8250],mmio,<addr>[,options[,uartclk]] 1431 uart[8250],mmio32,<addr>[,options[,uartclk]] 1432 uart[8250],mmio32be,<addr>[,options[,uartclk]] 1433 uart[8250],0x<addr>[,options] 1434 Start an early, polled-mode console on the 8250/16550 1435 UART at the specified I/O port or MMIO address. 1436 MMIO inter-register address stride is either 8-bit 1437 (mmio) or 32-bit (mmio32 or mmio32be). 1438 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1439 to be equivalent to 'mmio'. 'options' are specified 1440 in the same format described for "console=ttyS<n>"; if 1441 unspecified, the h/w is not initialized. 'uartclk' is 1442 the uart clock frequency; if unspecified, it is set 1443 to 'BASE_BAUD' * 16. 1444 1445 pl011,<addr> 1446 pl011,mmio32,<addr> 1447 Start an early, polled-mode console on a pl011 serial 1448 port at the specified address. The pl011 serial port 1449 must already be setup and configured. Options are not 1450 yet supported. If 'mmio32' is specified, then only 1451 the driver will use only 32-bit accessors to read/write 1452 the device registers. 1453 1454 liteuart,<addr> 1455 Start an early console on a litex serial port at the 1456 specified address. The serial port must already be 1457 setup and configured. Options are not yet supported. 1458 1459 meson,<addr> 1460 Start an early, polled-mode console on a meson serial 1461 port at the specified address. The serial port must 1462 already be setup and configured. Options are not yet 1463 supported. 1464 1465 msm_serial,<addr> 1466 Start an early, polled-mode console on an msm serial 1467 port at the specified address. The serial port 1468 must already be setup and configured. Options are not 1469 yet supported. 1470 1471 msm_serial_dm,<addr> 1472 Start an early, polled-mode console on an msm serial 1473 dm port at the specified address. The serial port 1474 must already be setup and configured. Options are not 1475 yet supported. 1476 1477 owl,<addr> 1478 Start an early, polled-mode console on a serial port 1479 of an Actions Semi SoC, such as S500 or S900, at the 1480 specified address. The serial port must already be 1481 setup and configured. Options are not yet supported. 1482 1483 rda,<addr> 1484 Start an early, polled-mode console on a serial port 1485 of an RDA Micro SoC, such as RDA8810PL, at the 1486 specified address. The serial port must already be 1487 setup and configured. Options are not yet supported. 1488 1489 sbi 1490 Use RISC-V SBI (Supervisor Binary Interface) for early 1491 console. 1492 1493 smh Use ARM semihosting calls for early console. 1494 1495 s3c2410,<addr> 1496 s3c2412,<addr> 1497 s3c2440,<addr> 1498 s3c6400,<addr> 1499 s5pv210,<addr> 1500 exynos4210,<addr> 1501 Use early console provided by serial driver available 1502 on Samsung SoCs, requires selecting proper type and 1503 a correct base address of the selected UART port. The 1504 serial port must already be setup and configured. 1505 Options are not yet supported. 1506 1507 lantiq,<addr> 1508 Start an early, polled-mode console on a lantiq serial 1509 (lqasc) port at the specified address. The serial port 1510 must already be setup and configured. Options are not 1511 yet supported. 1512 1513 lpuart,<addr> 1514 lpuart32,<addr> 1515 Use early console provided by Freescale LP UART driver 1516 found on Freescale Vybrid and QorIQ LS1021A processors. 1517 A valid base address must be provided, and the serial 1518 port must already be setup and configured. 1519 1520 ec_imx21,<addr> 1521 ec_imx6q,<addr> 1522 Start an early, polled-mode, output-only console on the 1523 Freescale i.MX UART at the specified address. The UART 1524 must already be setup and configured. 1525 1526 ar3700_uart,<addr> 1527 Start an early, polled-mode console on the 1528 Armada 3700 serial port at the specified 1529 address. The serial port must already be setup 1530 and configured. Options are not yet supported. 1531 1532 qcom_geni,<addr> 1533 Start an early, polled-mode console on a Qualcomm 1534 Generic Interface (GENI) based serial port at the 1535 specified address. The serial port must already be 1536 setup and configured. Options are not yet supported. 1537 1538 efifb,[options] 1539 Start an early, unaccelerated console on the EFI 1540 memory mapped framebuffer (if available). On cache 1541 coherent non-x86 systems that use system memory for 1542 the framebuffer, pass the 'ram' option so that it is 1543 mapped with the correct attributes. 1544 1545 linflex,<addr> 1546 Use early console provided by Freescale LINFlexD UART 1547 serial driver for NXP S32V234 SoCs. A valid base 1548 address must be provided, and the serial port must 1549 already be setup and configured. 1550 1551 earlyprintk= [X86,SH,ARM,M68k,S390,UM,EARLY] 1552 earlyprintk=vga 1553 earlyprintk=sclp 1554 earlyprintk=xen 1555 earlyprintk=serial[,ttySn[,baudrate]] 1556 earlyprintk=serial[,0x...[,baudrate]] 1557 earlyprintk=ttySn[,baudrate] 1558 earlyprintk=dbgp[debugController#] 1559 earlyprintk=mmio32,membase[,{nocfg|baudrate}] 1560 earlyprintk=pciserial[,force],bus:device.function[,{nocfg|baudrate}] 1561 earlyprintk=xdbc[xhciController#] 1562 earlyprintk=bios 1563 1564 earlyprintk is useful when the kernel crashes before 1565 the normal console is initialized. It is not enabled by 1566 default because it has some cosmetic problems. 1567 1568 Use "nocfg" to skip UART configuration, assume 1569 BIOS/firmware has configured UART correctly. 1570 1571 Append ",keep" to not disable it when the real console 1572 takes over. 1573 1574 Only one of vga, serial, or usb debug port can 1575 be used at a time. 1576 1577 Currently only ttyS0 and ttyS1 may be specified by 1578 name. Other I/O ports may be explicitly specified 1579 on some architectures (x86 and arm at least) by 1580 replacing ttySn with an I/O port address, like this: 1581 earlyprintk=serial,0x1008,115200 1582 You can find the port for a given device in 1583 /proc/tty/driver/serial: 1584 2: uart:ST16650V2 port:00001008 irq:18 ... 1585 1586 Interaction with the standard serial driver is not 1587 very good. 1588 1589 The VGA output is eventually overwritten by 1590 the real console. 1591 1592 The xen option can only be used in Xen domains. 1593 1594 The sclp output can only be used on s390. 1595 1596 The bios output can only be used on SuperH. 1597 1598 The optional "force" to "pciserial" enables use of a 1599 PCI device even when its classcode is not of the 1600 UART class. 1601 1602 edac_report= [HW,EDAC] Control how to report EDAC event 1603 Format: {"on" | "off" | "force"} 1604 on: enable EDAC to report H/W event. May be overridden 1605 by other higher priority error reporting module. 1606 off: disable H/W event reporting through EDAC. 1607 force: enforce the use of EDAC to report H/W event. 1608 default: on. 1609 1610 edd= [EDD] 1611 Format: {"off" | "on" | "skip[mbr]"} 1612 1613 efi= [EFI,EARLY] 1614 Format: { "debug", "disable_early_pci_dma", 1615 "nochunk", "noruntime", "nosoftreserve", 1616 "novamap", "no_disable_early_pci_dma" } 1617 debug: enable misc debug output. 1618 disable_early_pci_dma: disable the busmaster bit on all 1619 PCI bridges while in the EFI boot stub. 1620 nochunk: disable reading files in "chunks" in the EFI 1621 boot stub, as chunking can cause problems with some 1622 firmware implementations. 1623 noruntime : disable EFI runtime services support 1624 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1625 attribute may cause the kernel to reserve the 1626 memory range for a memory mapping driver to 1627 claim. Specify efi=nosoftreserve to disable this 1628 reservation and treat the memory by its base type 1629 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1630 novamap: do not call SetVirtualAddressMap(). 1631 no_disable_early_pci_dma: Leave the busmaster bit set 1632 on all PCI bridges while in the EFI boot stub 1633 1634 efi_no_storage_paranoia [EFI,X86,EARLY] 1635 Using this parameter you can use more than 50% of 1636 your efi variable storage. Use this parameter only if 1637 you are really sure that your UEFI does sane gc and 1638 fulfills the spec otherwise your board may brick. 1639 1640 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1641 that is to be dynamically loaded by Linux. If there are 1642 multiple variables with the same name but with different 1643 vendor GUIDs, all of them will be loaded. See 1644 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1645 1646 1647 eisa_irq_edge= [PARISC,HW] 1648 See header of drivers/parisc/eisa.c. 1649 1650 ekgdboc= [X86,KGDB,EARLY] Allow early kernel console debugging 1651 Format: ekgdboc=kbd 1652 1653 This is designed to be used in conjunction with 1654 the boot argument: earlyprintk=vga 1655 1656 This parameter works in place of the kgdboc parameter 1657 but can only be used if the backing tty is available 1658 very early in the boot process. For early debugging 1659 via a serial port see kgdboc_earlycon instead. 1660 1661 elanfreq= [X86-32] 1662 See comment before function elanfreq_setup() in 1663 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1664 1665 elfcorehdr=[size[KMG]@]offset[KMG] [PPC,SH,X86,S390,EARLY] 1666 Specifies physical address of start of kernel core 1667 image elf header and optionally the size. Generally 1668 kexec loader will pass this option to capture kernel. 1669 See Documentation/admin-guide/kdump/kdump.rst for details. 1670 1671 enable_mtrr_cleanup [X86,EARLY] 1672 The kernel tries to adjust MTRR layout from continuous 1673 to discrete, to make X server driver able to add WB 1674 entry later. This parameter enables that. 1675 1676 enable_timer_pin_1 [X86] 1677 Enable PIN 1 of APIC timer 1678 Can be useful to work around chipset bugs 1679 (in particular on some ATI chipsets). 1680 The kernel tries to set a reasonable default. 1681 1682 enforcing= [SELINUX] Set initial enforcing status. 1683 Format: {"0" | "1"} 1684 See security/selinux/Kconfig help text. 1685 0 -- permissive (log only, no denials). 1686 1 -- enforcing (deny and log). 1687 Default value is 0. 1688 Value can be changed at runtime via 1689 /sys/fs/selinux/enforce. 1690 1691 erst_disable [ACPI] 1692 Disable Error Record Serialization Table (ERST) 1693 support. 1694 1695 ether= [HW,NET] Ethernet cards parameters 1696 This option is obsoleted by the "netdev=" option, which 1697 has equivalent usage. See its documentation for details. 1698 1699 evm= [EVM] 1700 Format: { "fix" } 1701 Permit 'security.evm' to be updated regardless of 1702 current integrity status. 1703 1704 early_page_ext [KNL,EARLY] Enforces page_ext initialization to earlier 1705 stages so cover more early boot allocations. 1706 Please note that as side effect some optimizations 1707 might be disabled to achieve that (e.g. parallelized 1708 memory initialization is disabled) so the boot process 1709 might take longer, especially on systems with a lot of 1710 memory. Available with CONFIG_PAGE_EXTENSION=y. 1711 1712 failslab= 1713 fail_usercopy= 1714 fail_page_alloc= 1715 fail_skb_realloc= 1716 fail_make_request=[KNL] 1717 General fault injection mechanism. 1718 Format: <interval>,<probability>,<space>,<times> 1719 See also Documentation/fault-injection/. 1720 1721 fb_tunnels= [NET] 1722 Format: { initns | none } 1723 See Documentation/admin-guide/sysctl/net.rst for 1724 fb_tunnels_only_for_init_ns 1725 1726 floppy= [HW] 1727 See Documentation/admin-guide/blockdev/floppy.rst. 1728 1729 forcepae [X86-32] 1730 Forcefully enable Physical Address Extension (PAE). 1731 Many Pentium M systems disable PAE but may have a 1732 functionally usable PAE implementation. 1733 Warning: use of this parameter will taint the kernel 1734 and may cause unknown problems. 1735 1736 fred= [X86-64] 1737 Enable/disable Flexible Return and Event Delivery. 1738 Format: { on | off } 1739 on: enable FRED when it's present. 1740 off: disable FRED, the default setting. 1741 1742 ftrace=[tracer] 1743 [FTRACE] will set and start the specified tracer 1744 as early as possible in order to facilitate early 1745 boot debugging. 1746 1747 ftrace_boot_snapshot 1748 [FTRACE] On boot up, a snapshot will be taken of the 1749 ftrace ring buffer that can be read at: 1750 /sys/kernel/tracing/snapshot. 1751 This is useful if you need tracing information from kernel 1752 boot up that is likely to be overridden by user space 1753 start up functionality. 1754 1755 Optionally, the snapshot can also be defined for a tracing 1756 instance that was created by the trace_instance= command 1757 line parameter. 1758 1759 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo 1760 1761 The above will cause the "foo" tracing instance to trigger 1762 a snapshot at the end of boot up. 1763 1764 ftrace_dump_on_oops[=2(orig_cpu) | =<instance>][,<instance> | 1765 ,<instance>=2(orig_cpu)] 1766 [FTRACE] will dump the trace buffers on oops. 1767 If no parameter is passed, ftrace will dump global 1768 buffers of all CPUs, if you pass 2 or orig_cpu, it 1769 will dump only the buffer of the CPU that triggered 1770 the oops, or the specific instance will be dumped if 1771 its name is passed. Multiple instance dump is also 1772 supported, and instances are separated by commas. Each 1773 instance supports only dump on CPU that triggered the 1774 oops by passing 2 or orig_cpu to it. 1775 1776 ftrace_dump_on_oops=foo=orig_cpu 1777 1778 The above will dump only the buffer of "foo" instance 1779 on CPU that triggered the oops. 1780 1781 ftrace_dump_on_oops,foo,bar=orig_cpu 1782 1783 The above will dump global buffer on all CPUs, the 1784 buffer of "foo" instance on all CPUs and the buffer 1785 of "bar" instance on CPU that triggered the oops. 1786 1787 ftrace_filter=[function-list] 1788 [FTRACE] Limit the functions traced by the function 1789 tracer at boot up. function-list is a comma-separated 1790 list of functions. This list can be changed at run 1791 time by the set_ftrace_filter file in the debugfs 1792 tracing directory. 1793 1794 ftrace_notrace=[function-list] 1795 [FTRACE] Do not trace the functions specified in 1796 function-list. This list can be changed at run time 1797 by the set_ftrace_notrace file in the debugfs 1798 tracing directory. 1799 1800 ftrace_graph_filter=[function-list] 1801 [FTRACE] Limit the top level callers functions traced 1802 by the function graph tracer at boot up. 1803 function-list is a comma-separated list of functions 1804 that can be changed at run time by the 1805 set_graph_function file in the debugfs tracing directory. 1806 1807 ftrace_graph_notrace=[function-list] 1808 [FTRACE] Do not trace from the functions specified in 1809 function-list. This list is a comma-separated list of 1810 functions that can be changed at run time by the 1811 set_graph_notrace file in the debugfs tracing directory. 1812 1813 ftrace_graph_max_depth=<uint> 1814 [FTRACE] Used with the function graph tracer. This is 1815 the max depth it will trace into a function. This value 1816 can be changed at run time by the max_graph_depth file 1817 in the tracefs tracing directory. default: 0 (no limit) 1818 1819 fw_devlink= [KNL,EARLY] Create device links between consumer and supplier 1820 devices by scanning the firmware to infer the 1821 consumer/supplier relationships. This feature is 1822 especially useful when drivers are loaded as modules as 1823 it ensures proper ordering of tasks like device probing 1824 (suppliers first, then consumers), supplier boot state 1825 clean up (only after all consumers have probed), 1826 suspend/resume & runtime PM (consumers first, then 1827 suppliers). 1828 Format: { off | permissive | on | rpm } 1829 off -- Don't create device links from firmware info. 1830 permissive -- Create device links from firmware info 1831 but use it only for ordering boot state clean 1832 up (sync_state() calls). 1833 on -- Create device links from firmware info and use it 1834 to enforce probe and suspend/resume ordering. 1835 rpm -- Like "on", but also use to order runtime PM. 1836 1837 fw_devlink.strict=<bool> 1838 [KNL,EARLY] Treat all inferred dependencies as mandatory 1839 dependencies. This only applies for fw_devlink=on|rpm. 1840 Format: <bool> 1841 1842 fw_devlink.sync_state = 1843 [KNL,EARLY] When all devices that could probe have finished 1844 probing, this parameter controls what to do with 1845 devices that haven't yet received their sync_state() 1846 calls. 1847 Format: { strict | timeout } 1848 strict -- Default. Continue waiting on consumers to 1849 probe successfully. 1850 timeout -- Give up waiting on consumers and call 1851 sync_state() on any devices that haven't yet 1852 received their sync_state() calls after 1853 deferred_probe_timeout has expired or by 1854 late_initcall() if !CONFIG_MODULES. 1855 1856 gamecon.map[2|3]= 1857 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1858 support via parallel port (up to 5 devices per port) 1859 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1860 See also Documentation/input/devices/joystick-parport.rst 1861 1862 gamma= [HW,DRM] 1863 1864 gart_fix_e820= [X86-64,EARLY] disable the fix e820 for K8 GART 1865 Format: off | on 1866 default: on 1867 1868 gather_data_sampling= 1869 [X86,INTEL,EARLY] Control the Gather Data Sampling (GDS) 1870 mitigation. 1871 1872 Gather Data Sampling is a hardware vulnerability which 1873 allows unprivileged speculative access to data which was 1874 previously stored in vector registers. 1875 1876 This issue is mitigated by default in updated microcode. 1877 The mitigation may have a performance impact but can be 1878 disabled. On systems without the microcode mitigation 1879 disabling AVX serves as a mitigation. 1880 1881 force: Disable AVX to mitigate systems without 1882 microcode mitigation. No effect if the microcode 1883 mitigation is present. Known to cause crashes in 1884 userspace with buggy AVX enumeration. 1885 1886 off: Disable GDS mitigation. 1887 1888 gbpages [X86] Use GB pages for kernel direct mappings. 1889 1890 gcov_persist= [GCOV] When non-zero (default), profiling data for 1891 kernel modules is saved and remains accessible via 1892 debugfs, even when the module is unloaded/reloaded. 1893 When zero, profiling data is discarded and associated 1894 debugfs files are removed at module unload time. 1895 1896 goldfish [X86] Enable the goldfish android emulator platform. 1897 Don't use this when you are not running on the 1898 android emulator 1899 1900 gpio-mockup.gpio_mockup_ranges 1901 [HW] Sets the ranges of gpiochip of for this device. 1902 Format: <start1>,<end1>,<start2>,<end2>... 1903 gpio-mockup.gpio_mockup_named_lines 1904 [HW] Let the driver know GPIO lines should be named. 1905 1906 gpt [EFI] Forces disk with valid GPT signature but 1907 invalid Protective MBR to be treated as GPT. If the 1908 primary GPT is corrupted, it enables the backup/alternate 1909 GPT to be used instead. 1910 1911 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1912 the "Enable 0" bit of the configuration register. 1913 Format: 0 | 1 1914 Default: 0 1915 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1916 the "Enable 0" bit of the configuration register. 1917 Format: 0 | 1 1918 Default: 0 1919 grcan.select= [HW] Select which physical interface to use. 1920 Format: 0 | 1 1921 Default: 0 1922 grcan.txsize= [HW] Sets the size of the tx buffer. 1923 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1924 Default: 1024 1925 grcan.rxsize= [HW] Sets the size of the rx buffer. 1926 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1927 Default: 1024 1928 1929 hardened_usercopy= 1930 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 1931 hardening is enabled for this boot. Hardened 1932 usercopy checking is used to protect the kernel 1933 from reading or writing beyond known memory 1934 allocation boundaries as a proactive defense 1935 against bounds-checking flaws in the kernel's 1936 copy_to_user()/copy_from_user() interface. 1937 The default is determined by 1938 CONFIG_HARDENED_USERCOPY_DEFAULT_ON. 1939 on Perform hardened usercopy checks. 1940 off Disable hardened usercopy checks. 1941 1942 hardlockup_all_cpu_backtrace= 1943 [KNL] Should the hard-lockup detector generate 1944 backtraces on all cpus. 1945 Format: 0 | 1 1946 1947 hash_pointers= 1948 [KNL,EARLY] 1949 By default, when pointers are printed to the console 1950 or buffers via the %p format string, that pointer is 1951 "hashed", i.e. obscured by hashing the pointer value. 1952 This is a security feature that hides actual kernel 1953 addresses from unprivileged users, but it also makes 1954 debugging the kernel more difficult since unequal 1955 pointers can no longer be compared. The choices are: 1956 Format: { auto | always | never } 1957 Default: auto 1958 1959 auto - Hash pointers unless slab_debug is enabled. 1960 always - Always hash pointers (even if slab_debug is 1961 enabled). 1962 never - Never hash pointers. This option should only 1963 be specified when debugging the kernel. Do 1964 not use on production kernels. The boot 1965 param "no_hash_pointers" is an alias for 1966 this mode. 1967 1968 hashdist= [KNL,NUMA] Large hashes allocated during boot 1969 are distributed across NUMA nodes. Defaults on 1970 for 64-bit NUMA, off otherwise. 1971 Format: 0 | 1 (for off | on) 1972 1973 hd= [EIDE] (E)IDE hard drive subsystem geometry 1974 Format: <cyl>,<head>,<sect> 1975 1976 hest_disable [ACPI] 1977 Disable Hardware Error Source Table (HEST) support; 1978 corresponding firmware-first mode error processing 1979 logic will be disabled. 1980 1981 hibernate= [HIBERNATION] 1982 noresume Don't check if there's a hibernation image 1983 present during boot. 1984 nocompress Don't compress/decompress hibernation images. 1985 no Disable hibernation and resume. 1986 protect_image Turn on image protection during restoration 1987 (that will set all pages holding image data 1988 during restoration read-only). 1989 1990 hibernate.compressor= [HIBERNATION] Compression algorithm to be 1991 used with hibernation. 1992 Format: { lzo | lz4 } 1993 Default: lzo 1994 1995 lzo: Select LZO compression algorithm to 1996 compress/decompress hibernation image. 1997 1998 lz4: Select LZ4 compression algorithm to 1999 compress/decompress hibernation image. 2000 2001 hibernate.pm_test_delay= 2002 [HIBERNATION] 2003 Sets the number of seconds to remain in a hibernation test 2004 mode before resuming the system (see 2005 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 2006 is set. Default value is 5. 2007 2008 highmem=nn[KMG] [KNL,BOOT,EARLY] forces the highmem zone to have an exact 2009 size of <nn>. This works even on boxes that have no 2010 highmem otherwise. This also works to reduce highmem 2011 size on bigger boxes. 2012 2013 highres= [KNL] Enable/disable high resolution timer mode. 2014 Valid parameters: "on", "off" 2015 Default: "on" 2016 2017 hlt [BUGS=ARM,SH] 2018 2019 hostname= [KNL,EARLY] Set the hostname (aka UTS nodename). 2020 Format: <string> 2021 This allows setting the system's hostname during early 2022 startup. This sets the name returned by gethostname. 2023 Using this parameter to set the hostname makes it 2024 possible to ensure the hostname is correctly set before 2025 any userspace processes run, avoiding the possibility 2026 that a process may call gethostname before the hostname 2027 has been explicitly set, resulting in the calling 2028 process getting an incorrect result. The string must 2029 not exceed the maximum allowed hostname length (usually 2030 64 characters) and will be truncated otherwise. 2031 2032 hpet= [X86-32,HPET] option to control HPET usage 2033 Format: { enable (default) | disable | force | 2034 verbose } 2035 disable: disable HPET and use PIT instead 2036 force: allow force enabled of undocumented chips (ICH4, 2037 VIA, nVidia) 2038 verbose: show contents of HPET registers during setup 2039 2040 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 2041 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 2042 2043 hugepages= [HW,EARLY] Number of HugeTLB pages to allocate at boot. 2044 If this follows hugepagesz (below), it specifies 2045 the number of pages of hugepagesz to be allocated. 2046 If this is the first HugeTLB parameter on the command 2047 line, it specifies the number of pages to allocate for 2048 the default huge page size. If using node format, the 2049 number of pages to allocate per-node can be specified. 2050 See also Documentation/admin-guide/mm/hugetlbpage.rst. 2051 Format: <integer> or (node format) 2052 <node>:<integer>[,<node>:<integer>] 2053 2054 hugepagesz= 2055 [HW,EARLY] The size of the HugeTLB pages. This is 2056 used in conjunction with hugepages (above) to 2057 allocate huge pages of a specific size at boot. The 2058 pair hugepagesz=X hugepages=Y can be specified once 2059 for each supported huge page size. Huge page sizes 2060 are architecture dependent. See also 2061 Documentation/admin-guide/mm/hugetlbpage.rst. 2062 Format: size[KMG] 2063 2064 hugepage_alloc_threads= 2065 [HW] The number of threads that should be used to 2066 allocate hugepages during boot. This option can be 2067 used to improve system bootup time when allocating 2068 a large amount of huge pages. 2069 The default value is 25% of the available hardware threads. 2070 2071 Note that this parameter only applies to non-gigantic huge pages. 2072 2073 hugetlb_cma= [HW,CMA,EARLY] The size of a CMA area used for allocation 2074 of gigantic hugepages. Or using node format, the size 2075 of a CMA area per node can be specified. 2076 Format: nn[KMGTPE] or (node format) 2077 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]] 2078 2079 Reserve a CMA area of given size and allocate gigantic 2080 hugepages using the CMA allocator. If enabled, the 2081 boot-time allocation of gigantic hugepages is skipped. 2082 2083 hugetlb_cma_only= 2084 [HW,CMA,EARLY] When allocating new HugeTLB pages, only 2085 try to allocate from the CMA areas. 2086 2087 This option does nothing if hugetlb_cma= is not also 2088 specified. 2089 2090 hugetlb_free_vmemmap= 2091 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP 2092 enabled. 2093 Control if HugeTLB Vmemmap Optimization (HVO) is enabled. 2094 Allows heavy hugetlb users to free up some more 2095 memory (7 * PAGE_SIZE for each 2MB hugetlb page). 2096 Format: { on | off (default) } 2097 2098 on: enable HVO 2099 off: disable HVO 2100 2101 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y, 2102 the default is on. 2103 2104 Note that the vmemmap pages may be allocated from the added 2105 memory block itself when memory_hotplug.memmap_on_memory is 2106 enabled, those vmemmap pages cannot be optimized even if this 2107 feature is enabled. Other vmemmap pages not allocated from 2108 the added memory block itself do not be affected. 2109 2110 hung_task_panic= 2111 [KNL] Should the hung task detector generate panics. 2112 Format: 0 | 1 2113 2114 A value of 1 instructs the kernel to panic when a 2115 hung task is detected. The default value is controlled 2116 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time 2117 option. The value selected by this boot parameter can 2118 be changed later by the kernel.hung_task_panic sysctl. 2119 2120 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 2121 terminal devices. Valid values: 0..8 2122 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 2123 If specified, z/VM IUCV HVC accepts connections 2124 from listed z/VM user IDs only. 2125 2126 hv_nopvspin [X86,HYPER_V,EARLY] 2127 Disables the paravirt spinlock optimizations 2128 which allow the hypervisor to 'idle' the guest 2129 on lock contention. 2130 2131 hw_protection= [HW] 2132 Format: reboot | shutdown 2133 2134 Hardware protection action taken on critical events like 2135 overtemperature or imminent voltage loss. 2136 2137 i2c_bus= [HW] Override the default board specific I2C bus speed 2138 or register an additional I2C bus that is not 2139 registered from board initialization code. 2140 Format: 2141 <bus_id>,<clkrate> 2142 2143 i2c_touchscreen_props= [HW,ACPI,X86] 2144 Set device-properties for ACPI-enumerated I2C-attached 2145 touchscreen, to e.g. fix coordinates of upside-down 2146 mounted touchscreens. If you need this option please 2147 submit a drivers/platform/x86/touchscreen_dmi.c patch 2148 adding a DMI quirk for this. 2149 2150 Format: 2151 <ACPI_HW_ID>:<prop_name>=<val>[:prop_name=val][:...] 2152 Where <val> is one of: 2153 Omit "=<val>" entirely Set a boolean device-property 2154 Unsigned number Set a u32 device-property 2155 Anything else Set a string device-property 2156 2157 Examples (split over multiple lines): 2158 i2c_touchscreen_props=GDIX1001:touchscreen-inverted-x: 2159 touchscreen-inverted-y 2160 2161 i2c_touchscreen_props=MSSL1680:touchscreen-size-x=1920: 2162 touchscreen-size-y=1080:touchscreen-inverted-y: 2163 firmware-name=gsl1680-vendor-model.fw:silead,home-button 2164 2165 i8042.debug [HW] Toggle i8042 debug mode 2166 i8042.unmask_kbd_data 2167 [HW] Enable printing of interrupt data from the KBD port 2168 (disabled by default, and as a pre-condition 2169 requires that i8042.debug=1 be enabled) 2170 i8042.direct [HW] Put keyboard port into non-translated mode 2171 i8042.dumbkbd [HW] Pretend that controller can only read data from 2172 keyboard and cannot control its state 2173 (Don't attempt to blink the leds) 2174 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 2175 i8042.nokbd [HW] Don't check/create keyboard port 2176 i8042.noloop [HW] Disable the AUX Loopback command while probing 2177 for the AUX port 2178 i8042.nomux [HW] Don't check presence of an active multiplexing 2179 controller 2180 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 2181 controllers 2182 i8042.notimeout [HW] Ignore timeout condition signalled by controller 2183 i8042.reset [HW] Reset the controller during init, cleanup and 2184 suspend-to-ram transitions, only during s2r 2185 transitions, or never reset 2186 Format: { 1 | Y | y | 0 | N | n } 2187 1, Y, y: always reset controller 2188 0, N, n: don't ever reset controller 2189 Default: only on s2r transitions on x86; most other 2190 architectures force reset to be always executed 2191 i8042.unlock [HW] Unlock (ignore) the keylock 2192 i8042.kbdreset [HW] Reset device connected to KBD port 2193 i8042.probe_defer 2194 [HW] Allow deferred probing upon i8042 probe errors 2195 2196 i810= [HW,DRM] 2197 2198 i915.invert_brightness= 2199 [DRM] Invert the sense of the variable that is used to 2200 set the brightness of the panel backlight. Normally a 2201 brightness value of 0 indicates backlight switched off, 2202 and the maximum of the brightness value sets the backlight 2203 to maximum brightness. If this parameter is set to 0 2204 (default) and the machine requires it, or this parameter 2205 is set to 1, a brightness value of 0 sets the backlight 2206 to maximum brightness, and the maximum of the brightness 2207 value switches the backlight off. 2208 -1 -- never invert brightness 2209 0 -- machine default 2210 1 -- force brightness inversion 2211 2212 ia32_emulation= [X86-64] 2213 Format: <bool> 2214 When true, allows loading 32-bit programs and executing 32-bit 2215 syscalls, essentially overriding IA32_EMULATION_DEFAULT_DISABLED at 2216 boot time. When false, unconditionally disables IA32 emulation. 2217 2218 icn= [HW,ISDN] 2219 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 2220 2221 2222 idle= [X86,EARLY] 2223 Format: idle=poll, idle=halt, idle=nomwait 2224 2225 idle=poll: Don't do power saving in the idle loop 2226 using HLT, but poll for rescheduling event. This will 2227 make the CPUs eat a lot more power, but may be useful 2228 to get slightly better performance in multiprocessor 2229 benchmarks. It also makes some profiling using 2230 performance counters more accurate. Please note that 2231 on systems with MONITOR/MWAIT support (like Intel 2232 EM64T CPUs) this option has no performance advantage 2233 over the normal idle loop. It may also interact badly 2234 with hyperthreading. 2235 2236 idle=halt: Halt is forced to be used for CPU idle. 2237 In such case C2/C3 won't be used again. 2238 2239 idle=nomwait: Disable mwait for CPU C-states 2240 2241 idxd.sva= [HW] 2242 Format: <bool> 2243 Allow force disabling of Shared Virtual Memory (SVA) 2244 support for the idxd driver. By default it is set to 2245 true (1). 2246 2247 idxd.tc_override= [HW] 2248 Format: <bool> 2249 Allow override of default traffic class configuration 2250 for the device. By default it is set to false (0). 2251 2252 ieee754= [MIPS] Select IEEE Std 754 conformance mode 2253 Format: { strict | legacy | 2008 | relaxed | emulated } 2254 Default: strict 2255 2256 Choose which programs will be accepted for execution 2257 based on the IEEE 754 NaN encoding(s) supported by 2258 the FPU and the NaN encoding requested with the value 2259 of an ELF file header flag individually set by each 2260 binary. Hardware implementations are permitted to 2261 support either or both of the legacy and the 2008 NaN 2262 encoding mode. 2263 2264 Available settings are as follows: 2265 strict accept binaries that request a NaN encoding 2266 supported by the FPU 2267 legacy only accept legacy-NaN binaries, if supported 2268 by the FPU 2269 2008 only accept 2008-NaN binaries, if supported 2270 by the FPU 2271 relaxed accept any binaries regardless of whether 2272 supported by the FPU 2273 emulated accept any binaries but enable FPU emulator 2274 if binary mode is unsupported by the FPU. 2275 2276 The FPU emulator is always able to support both NaN 2277 encodings, so if no FPU hardware is present or it has 2278 been disabled with 'nofpu', then the settings of 2279 'legacy' and '2008' strap the emulator accordingly, 2280 'relaxed' straps the emulator for both legacy-NaN and 2281 2008-NaN, whereas 'strict' enables legacy-NaN only on 2282 legacy processors and both NaN encodings on MIPS32 or 2283 MIPS64 CPUs. 2284 2285 The setting for ABS.fmt/NEG.fmt instruction execution 2286 mode generally follows that for the NaN encoding, 2287 except where unsupported by hardware. 2288 2289 ignore_loglevel [KNL,EARLY] 2290 Ignore loglevel setting - this will print /all/ 2291 kernel messages to the console. Useful for debugging. 2292 We also add it as printk module parameter, so users 2293 could change it dynamically, usually by 2294 /sys/module/printk/parameters/ignore_loglevel. 2295 2296 ignore_rlimit_data 2297 Ignore RLIMIT_DATA setting for data mappings, 2298 print warning at first misuse. Can be changed via 2299 /sys/module/kernel/parameters/ignore_rlimit_data. 2300 2301 ihash_entries= [KNL] 2302 Set number of hash buckets for inode cache. 2303 2304 ima_appraise= [IMA] appraise integrity measurements 2305 Format: { "off" | "enforce" | "fix" | "log" } 2306 default: "enforce" 2307 2308 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 2309 The builtin appraise policy appraises all files 2310 owned by uid=0. 2311 2312 ima_canonical_fmt [IMA] 2313 Use the canonical format for the binary runtime 2314 measurements, instead of host native format. 2315 2316 ima_hash= [IMA] 2317 Format: { md5 | sha1 | rmd160 | sha256 | sha384 2318 | sha512 | ... } 2319 default: "sha1" 2320 2321 The list of supported hash algorithms is defined 2322 in crypto/hash_info.h. 2323 2324 ima_policy= [IMA] 2325 The builtin policies to load during IMA setup. 2326 Format: "tcb | appraise_tcb | secure_boot | 2327 fail_securely | critical_data" 2328 2329 The "tcb" policy measures all programs exec'd, files 2330 mmap'd for exec, and all files opened with the read 2331 mode bit set by either the effective uid (euid=0) or 2332 uid=0. 2333 2334 The "appraise_tcb" policy appraises the integrity of 2335 all files owned by root. 2336 2337 The "secure_boot" policy appraises the integrity 2338 of files (eg. kexec kernel image, kernel modules, 2339 firmware, policy, etc) based on file signatures. 2340 2341 The "fail_securely" policy forces file signature 2342 verification failure also on privileged mounted 2343 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 2344 flag. 2345 2346 The "critical_data" policy measures kernel integrity 2347 critical data. 2348 2349 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 2350 Load a policy which meets the needs of the Trusted 2351 Computing Base. This means IMA will measure all 2352 programs exec'd, files mmap'd for exec, and all files 2353 opened for read by uid=0. 2354 2355 ima_template= [IMA] 2356 Select one of defined IMA measurements template formats. 2357 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" | 2358 "ima-sigv2" } 2359 Default: "ima-ng" 2360 2361 ima_template_fmt= 2362 [IMA] Define a custom template format. 2363 Format: { "field1|...|fieldN" } 2364 2365 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 2366 Format: <min_file_size> 2367 Set the minimal file size for using asynchronous hash. 2368 If left unspecified, ahash usage is disabled. 2369 2370 ahash performance varies for different data sizes on 2371 different crypto accelerators. This option can be used 2372 to achieve the best performance for a particular HW. 2373 2374 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size 2375 Format: <bufsize> 2376 Set hashing buffer size. Default: 4k. 2377 2378 ahash performance varies for different chunk sizes on 2379 different crypto accelerators. This option can be used 2380 to achieve best performance for particular HW. 2381 2382 ima= [IMA] Enable or disable IMA 2383 Format: { "off" | "on" } 2384 Default: "on" 2385 Note that disabling IMA is limited to kdump kernel. 2386 2387 indirect_target_selection= [X86,Intel] Mitigation control for Indirect 2388 Target Selection(ITS) bug in Intel CPUs. Updated 2389 microcode is also required for a fix in IBPB. 2390 2391 on: Enable mitigation (default). 2392 off: Disable mitigation. 2393 force: Force the ITS bug and deploy default 2394 mitigation. 2395 vmexit: Only deploy mitigation if CPU is affected by 2396 guest/host isolation part of ITS. 2397 stuff: Deploy RSB-fill mitigation when retpoline is 2398 also deployed. Otherwise, deploy the default 2399 mitigation. 2400 2401 For details see: 2402 Documentation/admin-guide/hw-vuln/indirect-target-selection.rst 2403 2404 init= [KNL] 2405 Format: <full_path> 2406 Run specified binary instead of /sbin/init as init 2407 process. 2408 2409 initcall_debug [KNL] Trace initcalls as they are executed. Useful 2410 for working out where the kernel is dying during 2411 startup. 2412 2413 initcall_blacklist= [KNL] Do not execute a comma-separated list of 2414 initcall functions. Useful for debugging built-in 2415 modules and initcalls. 2416 2417 initramfs_async= [KNL] 2418 Format: <bool> 2419 Default: 1 2420 This parameter controls whether the initramfs 2421 image is unpacked asynchronously, concurrently 2422 with devices being probed and 2423 initialized. This should normally just work, 2424 but as a debugging aid, one can get the 2425 historical behaviour of the initramfs 2426 unpacking being completed before device_ and 2427 late_ initcalls. 2428 2429 initrd= [BOOT,EARLY] Specify the location of the initial ramdisk 2430 2431 initrdmem= [KNL,EARLY] Specify a physical address and size from which to 2432 load the initrd. If an initrd is compiled in or 2433 specified in the bootparams, it takes priority over this 2434 setting. 2435 Format: ss[KMG],nn[KMG] 2436 Default is 0, 0 2437 2438 init_on_alloc= [MM,EARLY] Fill newly allocated pages and heap objects with 2439 zeroes. 2440 Format: 0 | 1 2441 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 2442 2443 init_on_free= [MM,EARLY] Fill freed pages and heap objects with zeroes. 2444 Format: 0 | 1 2445 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 2446 2447 init_pkru= [X86] Specify the default memory protection keys rights 2448 register contents for all processes. 0x55555554 by 2449 default (disallow access to all but pkey 0). Can 2450 override in debugfs after boot. 2451 2452 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 2453 Format: <irq> 2454 2455 int_pln_enable [X86] Enable power limit notification interrupt 2456 2457 integrity_audit=[IMA] 2458 Format: { "0" | "1" } 2459 0 -- basic integrity auditing messages. (Default) 2460 1 -- additional integrity auditing messages. 2461 2462 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 2463 on 2464 Enable intel iommu driver. 2465 off 2466 Disable intel iommu driver. 2467 igfx_off [Default Off] 2468 By default, gfx is mapped as normal device. If a gfx 2469 device has a dedicated DMAR unit, the DMAR unit is 2470 bypassed by not enabling DMAR with this option. In 2471 this case, gfx device will use physical address for 2472 DMA. 2473 strict [Default Off] 2474 Deprecated, equivalent to iommu.strict=1. 2475 sp_off [Default Off] 2476 By default, super page will be supported if Intel IOMMU 2477 has the capability. With this option, super page will 2478 not be supported. 2479 sm_on 2480 Enable the Intel IOMMU scalable mode if the hardware 2481 advertises that it has support for the scalable mode 2482 translation. 2483 sm_off 2484 Disallow use of the Intel IOMMU scalable mode. 2485 tboot_noforce [Default Off] 2486 Do not force the Intel IOMMU enabled under tboot. 2487 By default, tboot will force Intel IOMMU on, which 2488 could harm performance of some high-throughput 2489 devices like 40GBit network cards, even if identity 2490 mapping is enabled. 2491 Note that using this option lowers the security 2492 provided by tboot because it makes the system 2493 vulnerable to DMA attacks. 2494 2495 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 2496 0 disables intel_idle and fall back on acpi_idle. 2497 1 to 9 specify maximum depth of C-state. 2498 2499 intel_pstate= [X86,EARLY] 2500 disable 2501 Do not enable intel_pstate as the default 2502 scaling driver for the supported processors 2503 active 2504 Use intel_pstate driver to bypass the scaling 2505 governors layer of cpufreq and provides it own 2506 algorithms for p-state selection. There are two 2507 P-state selection algorithms provided by 2508 intel_pstate in the active mode: powersave and 2509 performance. The way they both operate depends 2510 on whether or not the hardware managed P-states 2511 (HWP) feature has been enabled in the processor 2512 and possibly on the processor model. 2513 passive 2514 Use intel_pstate as a scaling driver, but configure it 2515 to work with generic cpufreq governors (instead of 2516 enabling its internal governor). This mode cannot be 2517 used along with the hardware-managed P-states (HWP) 2518 feature. 2519 force 2520 Enable intel_pstate on systems that prohibit it by default 2521 in favor of acpi-cpufreq. Forcing the intel_pstate driver 2522 instead of acpi-cpufreq may disable platform features, such 2523 as thermal controls and power capping, that rely on ACPI 2524 P-States information being indicated to OSPM and therefore 2525 should be used with caution. This option does not work with 2526 processors that aren't supported by the intel_pstate driver 2527 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 2528 no_hwp 2529 Do not enable hardware P state control (HWP) 2530 if available. 2531 hwp_only 2532 Only load intel_pstate on systems which support 2533 hardware P state control (HWP) if available. 2534 support_acpi_ppc 2535 Enforce ACPI _PPC performance limits. If the Fixed ACPI 2536 Description Table, specifies preferred power management 2537 profile as "Enterprise Server" or "Performance Server", 2538 then this feature is turned on by default. 2539 per_cpu_perf_limits 2540 Allow per-logical-CPU P-State performance control limits using 2541 cpufreq sysfs interface 2542 no_cas 2543 Do not enable capacity-aware scheduling (CAS) on 2544 hybrid systems 2545 2546 intremap= [X86-64,Intel-IOMMU,EARLY] 2547 on enable Interrupt Remapping (default) 2548 off disable Interrupt Remapping 2549 nosid disable Source ID checking 2550 no_x2apic_optout 2551 BIOS x2APIC opt-out request will be ignored 2552 nopost disable Interrupt Posting 2553 posted_msi 2554 enable MSIs delivered as posted interrupts 2555 2556 iomem= Disable strict checking of access to MMIO memory 2557 strict regions from userspace. 2558 relaxed 2559 2560 iommu= [X86,EARLY] 2561 2562 off 2563 Don't initialize and use any kind of IOMMU. 2564 2565 force 2566 Force the use of the hardware IOMMU even when 2567 it is not actually needed (e.g. because < 3 GB 2568 memory). 2569 2570 noforce 2571 Don't force hardware IOMMU usage when it is not 2572 needed. (default). 2573 2574 biomerge 2575 panic 2576 nopanic 2577 merge 2578 nomerge 2579 2580 soft 2581 Use software bounce buffering (SWIOTLB) (default for 2582 Intel machines). This can be used to prevent the usage 2583 of an available hardware IOMMU. 2584 2585 [X86] 2586 pt 2587 [X86] 2588 nopt 2589 [PPC/POWERNV] 2590 nobypass 2591 Disable IOMMU bypass, using IOMMU for PCI devices. 2592 2593 [X86] 2594 AMD Gart HW IOMMU-specific options: 2595 2596 <size> 2597 Set the size of the remapping area in bytes. 2598 2599 allowed 2600 Overwrite iommu off workarounds for specific chipsets 2601 2602 fullflush 2603 Flush IOMMU on each allocation (default). 2604 2605 nofullflush 2606 Don't use IOMMU fullflush. 2607 2608 memaper[=<order>] 2609 Allocate an own aperture over RAM with size 2610 32MB<<order. (default: order=1, i.e. 64MB) 2611 2612 merge 2613 Do scatter-gather (SG) merging. Implies "force" 2614 (experimental). 2615 2616 nomerge 2617 Don't do scatter-gather (SG) merging. 2618 2619 noaperture 2620 Ask the IOMMU not to touch the aperture for AGP. 2621 2622 noagp 2623 Don't initialize the AGP driver and use full aperture. 2624 2625 panic 2626 Always panic when IOMMU overflows. 2627 2628 iommu.forcedac= [ARM64,X86,EARLY] Control IOVA allocation for PCI devices. 2629 Format: { "0" | "1" } 2630 0 - Try to allocate a 32-bit DMA address first, before 2631 falling back to the full range if needed. 2632 1 - Allocate directly from the full usable range, 2633 forcing Dual Address Cycle for PCI cards supporting 2634 greater than 32-bit addressing. 2635 2636 iommu.strict= [ARM64,X86,S390,EARLY] Configure TLB invalidation behaviour 2637 Format: { "0" | "1" } 2638 0 - Lazy mode. 2639 Request that DMA unmap operations use deferred 2640 invalidation of hardware TLBs, for increased 2641 throughput at the cost of reduced device isolation. 2642 Will fall back to strict mode if not supported by 2643 the relevant IOMMU driver. 2644 1 - Strict mode. 2645 DMA unmap operations invalidate IOMMU hardware TLBs 2646 synchronously. 2647 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}. 2648 Note: on x86, strict mode specified via one of the 2649 legacy driver-specific options takes precedence. 2650 2651 iommu.passthrough= 2652 [ARM64,X86,EARLY] Configure DMA to bypass the IOMMU by default. 2653 Format: { "0" | "1" } 2654 0 - Use IOMMU translation for DMA. 2655 1 - Bypass the IOMMU for DMA. 2656 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2657 2658 io7= [HW] IO7 for Marvel-based Alpha systems 2659 See comment before marvel_specify_io7 in 2660 arch/alpha/kernel/core_marvel.c. 2661 2662 io_delay= [X86,EARLY] I/O delay method 2663 0x80 2664 Standard port 0x80 based delay 2665 0xed 2666 Alternate port 0xed based delay (needed on some systems) 2667 udelay 2668 Simple two microseconds delay 2669 none 2670 No delay 2671 2672 ip= [IP_PNP] 2673 See Documentation/admin-guide/nfs/nfsroot.rst. 2674 2675 ipcmni_extend [KNL,EARLY] Extend the maximum number of unique System V 2676 IPC identifiers from 32,768 to 16,777,216. 2677 2678 ipe.enforce= [IPE] 2679 Format: <bool> 2680 Determine whether IPE starts in permissive (0) or 2681 enforce (1) mode. The default is enforce. 2682 2683 ipe.success_audit= 2684 [IPE] 2685 Format: <bool> 2686 Start IPE with success auditing enabled, emitting 2687 an audit event when a binary is allowed. The default 2688 is 0. 2689 2690 irqaffinity= [SMP] Set the default irq affinity mask 2691 The argument is a cpu list, as described above. 2692 2693 irqchip.gicv2_force_probe= 2694 [ARM,ARM64,EARLY] 2695 Format: <bool> 2696 Force the kernel to look for the second 4kB page 2697 of a GICv2 controller even if the memory range 2698 exposed by the device tree is too small. 2699 2700 irqchip.gicv3_nolpi= 2701 [ARM,ARM64,EARLY] 2702 Force the kernel to ignore the availability of 2703 LPIs (and by consequence ITSs). Intended for system 2704 that use the kernel as a bootloader, and thus want 2705 to let secondary kernels in charge of setting up 2706 LPIs. 2707 2708 irqchip.gicv3_pseudo_nmi= [ARM64,EARLY] 2709 Enables support for pseudo-NMIs in the kernel. This 2710 requires the kernel to be built with 2711 CONFIG_ARM64_PSEUDO_NMI. 2712 2713 irqchip.riscv_imsic_noipi 2714 [RISC-V,EARLY] 2715 Force the kernel to not use IMSIC software injected MSIs 2716 as IPIs. Intended for system where IMSIC is trap-n-emulated, 2717 and thus want to reduce MMIO traps when triggering IPIs 2718 to multiple harts. 2719 2720 irqfixup [HW] 2721 When an interrupt is not handled search all handlers 2722 for it. Intended to get systems with badly broken 2723 firmware running. 2724 2725 irqhandler.duration_warn_us= [KNL] 2726 Warn if an IRQ handler exceeds the specified duration 2727 threshold in microseconds. Useful for identifying 2728 long-running IRQs in the system. 2729 2730 irqpoll [HW] 2731 When an interrupt is not handled search all handlers 2732 for it. Also check all handlers each timer 2733 interrupt. Intended to get systems with badly broken 2734 firmware running. 2735 2736 isapnp= [ISAPNP] 2737 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2738 2739 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2740 [Deprecated - use cpusets instead] 2741 Format: [flag-list,]<cpu-list> 2742 2743 Specify one or more CPUs to isolate from disturbances 2744 specified in the flag list (default: domain): 2745 2746 nohz 2747 Disable the tick when a single task runs as well as 2748 disabling other kernel noises like having RCU callbacks 2749 offloaded. This is equivalent to the nohz_full parameter. 2750 2751 A residual 1Hz tick is offloaded to workqueues, which you 2752 need to affine to housekeeping through the global 2753 workqueue's affinity configured via the 2754 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2755 by using the 'domain' flag described below. 2756 2757 NOTE: by default the global workqueue runs on all CPUs, 2758 so to protect individual CPUs the 'cpumask' file has to 2759 be configured manually after bootup. 2760 2761 domain 2762 Isolate from the general SMP balancing and scheduling 2763 algorithms. Note that performing domain isolation this way 2764 is irreversible: it's not possible to bring back a CPU to 2765 the domains once isolated through isolcpus. It's strongly 2766 advised to use cpusets instead to disable scheduler load 2767 balancing through the "cpuset.sched_load_balance" file. 2768 It offers a much more flexible interface where CPUs can 2769 move in and out of an isolated set anytime. 2770 2771 You can move a process onto or off an "isolated" CPU via 2772 the CPU affinity syscalls or cpuset. 2773 <cpu number> begins at 0 and the maximum value is 2774 "number of CPUs in system - 1". 2775 2776 managed_irq 2777 2778 Isolate from being targeted by managed interrupts 2779 which have an interrupt mask containing isolated 2780 CPUs. The affinity of managed interrupts is 2781 handled by the kernel and cannot be changed via 2782 the /proc/irq/* interfaces. 2783 2784 This isolation is best effort and only effective 2785 if the automatically assigned interrupt mask of a 2786 device queue contains isolated and housekeeping 2787 CPUs. If housekeeping CPUs are online then such 2788 interrupts are directed to the housekeeping CPU 2789 so that IO submitted on the housekeeping CPU 2790 cannot disturb the isolated CPU. 2791 2792 If a queue's affinity mask contains only isolated 2793 CPUs then this parameter has no effect on the 2794 interrupt routing decision, though interrupts are 2795 only delivered when tasks running on those 2796 isolated CPUs submit IO. IO submitted on 2797 housekeeping CPUs has no influence on those 2798 queues. 2799 2800 The format of <cpu-list> is described above. 2801 2802 iucv= [HW,NET] 2803 2804 ivrs_ioapic [HW,X86-64] 2805 Provide an override to the IOAPIC-ID<->DEVICE-ID 2806 mapping provided in the IVRS ACPI table. 2807 By default, PCI segment is 0, and can be omitted. 2808 2809 For example, to map IOAPIC-ID decimal 10 to 2810 PCI segment 0x1 and PCI device 00:14.0, 2811 write the parameter as: 2812 ivrs_ioapic=10@0001:00:14.0 2813 2814 Deprecated formats: 2815 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0 2816 write the parameter as: 2817 ivrs_ioapic[10]=00:14.0 2818 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and 2819 PCI device 00:14.0 write the parameter as: 2820 ivrs_ioapic[10]=0001:00:14.0 2821 2822 ivrs_hpet [HW,X86-64] 2823 Provide an override to the HPET-ID<->DEVICE-ID 2824 mapping provided in the IVRS ACPI table. 2825 By default, PCI segment is 0, and can be omitted. 2826 2827 For example, to map HPET-ID decimal 10 to 2828 PCI segment 0x1 and PCI device 00:14.0, 2829 write the parameter as: 2830 ivrs_hpet=10@0001:00:14.0 2831 2832 Deprecated formats: 2833 * To map HPET-ID decimal 0 to PCI device 00:14.0 2834 write the parameter as: 2835 ivrs_hpet[0]=00:14.0 2836 * To map HPET-ID decimal 10 to PCI segment 0x1 and 2837 PCI device 00:14.0 write the parameter as: 2838 ivrs_ioapic[10]=0001:00:14.0 2839 2840 ivrs_acpihid [HW,X86-64] 2841 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2842 mapping provided in the IVRS ACPI table. 2843 By default, PCI segment is 0, and can be omitted. 2844 2845 For example, to map UART-HID:UID AMD0020:0 to 2846 PCI segment 0x1 and PCI device ID 00:14.5, 2847 write the parameter as: 2848 ivrs_acpihid=AMD0020:0@0001:00:14.5 2849 2850 Deprecated formats: 2851 * To map UART-HID:UID AMD0020:0 to PCI segment is 0, 2852 PCI device ID 00:14.5, write the parameter as: 2853 ivrs_acpihid[00:14.5]=AMD0020:0 2854 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and 2855 PCI device ID 00:14.5, write the parameter as: 2856 ivrs_acpihid[0001:00:14.5]=AMD0020:0 2857 2858 js= [HW,JOY] Analog joystick 2859 See Documentation/input/joydev/joystick.rst. 2860 2861 kasan_multi_shot 2862 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2863 report on every invalid memory access. Without this 2864 parameter KASAN will print report only for the first 2865 invalid access. 2866 2867 keep_bootcon [KNL,EARLY] 2868 Do not unregister boot console at start. This is only 2869 useful for debugging when something happens in the window 2870 between unregistering the boot console and initializing 2871 the real console. 2872 2873 keepinitrd [HW,ARM] See retain_initrd. 2874 2875 kernelcore= [KNL,X86,PPC,EARLY] 2876 Format: nn[KMGTPE] | nn% | "mirror" 2877 This parameter specifies the amount of memory usable by 2878 the kernel for non-movable allocations. The requested 2879 amount is spread evenly throughout all nodes in the 2880 system as ZONE_NORMAL. The remaining memory is used for 2881 movable memory in its own zone, ZONE_MOVABLE. In the 2882 event, a node is too small to have both ZONE_NORMAL and 2883 ZONE_MOVABLE, kernelcore memory will take priority and 2884 other nodes will have a larger ZONE_MOVABLE. 2885 2886 ZONE_MOVABLE is used for the allocation of pages that 2887 may be reclaimed or moved by the page migration 2888 subsystem. Note that allocations like PTEs-from-HighMem 2889 still use the HighMem zone if it exists, and the Normal 2890 zone if it does not. 2891 2892 It is possible to specify the exact amount of memory in 2893 the form of "nn[KMGTPE]", a percentage of total system 2894 memory in the form of "nn%", or "mirror". If "mirror" 2895 option is specified, mirrored (reliable) memory is used 2896 for non-movable allocations and remaining memory is used 2897 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2898 are exclusive, so you cannot specify multiple forms. 2899 2900 kgdbdbgp= [KGDB,HW,EARLY] kgdb over EHCI usb debug port. 2901 Format: <Controller#>[,poll interval] 2902 The controller # is the number of the ehci usb debug 2903 port as it is probed via PCI. The poll interval is 2904 optional and is the number seconds in between 2905 each poll cycle to the debug port in case you need 2906 the functionality for interrupting the kernel with 2907 gdb or control-c on the dbgp connection. When 2908 not using this parameter you use sysrq-g to break into 2909 the kernel debugger. 2910 2911 kgdboc= [KGDB,HW] kgdb over consoles. 2912 Requires a tty driver that supports console polling, 2913 or a supported polling keyboard driver (non-usb). 2914 Serial only format: <serial_device>[,baud] 2915 keyboard only format: kbd 2916 keyboard and serial format: kbd,<serial_device>[,baud] 2917 Optional Kernel mode setting: 2918 kms, kbd format: kms,kbd 2919 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2920 2921 kgdboc_earlycon= [KGDB,HW,EARLY] 2922 If the boot console provides the ability to read 2923 characters and can work in polling mode, you can use 2924 this parameter to tell kgdb to use it as a backend 2925 until the normal console is registered. Intended to 2926 be used together with the kgdboc parameter which 2927 specifies the normal console to transition to. 2928 2929 The name of the early console should be specified 2930 as the value of this parameter. Note that the name of 2931 the early console might be different than the tty 2932 name passed to kgdboc. It's OK to leave the value 2933 blank and the first boot console that implements 2934 read() will be picked. 2935 2936 kgdbwait [KGDB,EARLY] Stop kernel execution and enter the 2937 kernel debugger at the earliest opportunity. 2938 2939 kho= [KEXEC,EARLY] 2940 Format: { "0" | "1" | "off" | "on" | "y" | "n" } 2941 Enables or disables Kexec HandOver. 2942 "0" | "off" | "n" - kexec handover is disabled 2943 "1" | "on" | "y" - kexec handover is enabled 2944 2945 kho_scratch= [KEXEC,EARLY] 2946 Format: ll[KMG],mm[KMG],nn[KMG] | nn% 2947 Defines the size of the KHO scratch region. The KHO 2948 scratch regions are physically contiguous memory 2949 ranges that can only be used for non-kernel 2950 allocations. That way, even when memory is heavily 2951 fragmented with handed over memory, the kexeced 2952 kernel will always have enough contiguous ranges to 2953 bootstrap itself. 2954 2955 It is possible to specify the exact amount of 2956 memory in the form of "ll[KMG],mm[KMG],nn[KMG]" 2957 where the first parameter defines the size of a low 2958 memory scratch area, the second parameter defines 2959 the size of a global scratch area and the third 2960 parameter defines the size of additional per-node 2961 scratch areas. The form "nn%" defines scale factor 2962 (in percents) of memory that was used during boot. 2963 2964 kmac= [MIPS] Korina ethernet MAC address. 2965 Configure the RouterBoard 532 series on-chip 2966 Ethernet adapter MAC address. 2967 2968 kmemleak= [KNL,EARLY] Boot-time kmemleak enable/disable 2969 Valid arguments: on, off 2970 Default: on 2971 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 2972 the default is off. 2973 2974 kprobe_event=[probe-list] 2975 [FTRACE] Add kprobe events and enable at boot time. 2976 The probe-list is a semicolon delimited list of probe 2977 definitions. Each definition is same as kprobe_events 2978 interface, but the parameters are comma delimited. 2979 For example, to add a kprobe event on vfs_read with 2980 arg1 and arg2, add to the command line; 2981 2982 kprobe_event=p,vfs_read,$arg1,$arg2 2983 2984 See also Documentation/trace/kprobetrace.rst "Kernel 2985 Boot Parameter" section. 2986 2987 kpti= [ARM64,EARLY] Control page table isolation of 2988 user and kernel address spaces. 2989 Default: enabled on cores which need mitigation. 2990 0: force disabled 2991 1: force enabled 2992 2993 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires 2994 CONFIG_KUNIT to be set to be fully enabled. The 2995 default value can be overridden via 2996 KUNIT_DEFAULT_ENABLED. 2997 Default is 1 (enabled) 2998 2999 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 3000 Default is 0 (don't ignore, but inject #GP) 3001 3002 kvm.eager_page_split= 3003 [KVM,X86] Controls whether or not KVM will try to 3004 proactively split all huge pages during dirty logging. 3005 Eager page splitting reduces interruptions to vCPU 3006 execution by eliminating the write-protection faults 3007 and MMU lock contention that would otherwise be 3008 required to split huge pages lazily. 3009 3010 VM workloads that rarely perform writes or that write 3011 only to a small region of VM memory may benefit from 3012 disabling eager page splitting to allow huge pages to 3013 still be used for reads. 3014 3015 The behavior of eager page splitting depends on whether 3016 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If 3017 disabled, all huge pages in a memslot will be eagerly 3018 split when dirty logging is enabled on that memslot. If 3019 enabled, eager page splitting will be performed during 3020 the KVM_CLEAR_DIRTY ioctl, and only for the pages being 3021 cleared. 3022 3023 Eager page splitting is only supported when kvm.tdp_mmu=Y. 3024 3025 Default is Y (on). 3026 3027 kvm.enable_virt_at_load=[KVM,ARM64,LOONGARCH,MIPS,RISCV,X86] 3028 If enabled, KVM will enable virtualization in hardware 3029 when KVM is loaded, and disable virtualization when KVM 3030 is unloaded (if KVM is built as a module). 3031 3032 If disabled, KVM will dynamically enable and disable 3033 virtualization on-demand when creating and destroying 3034 VMs, i.e. on the 0=>1 and 1=>0 transitions of the 3035 number of VMs. 3036 3037 Enabling virtualization at module load avoids potential 3038 latency for creation of the 0=>1 VM, as KVM serializes 3039 virtualization enabling across all online CPUs. The 3040 "cost" of enabling virtualization when KVM is loaded, 3041 is that doing so may interfere with using out-of-tree 3042 hypervisors that want to "own" virtualization hardware. 3043 3044 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 3045 Default is false (don't support). 3046 3047 kvm.nx_huge_pages= 3048 [KVM] Controls the software workaround for the 3049 X86_BUG_ITLB_MULTIHIT bug. 3050 force : Always deploy workaround. 3051 off : Never deploy workaround. 3052 auto : Deploy workaround based on the presence of 3053 X86_BUG_ITLB_MULTIHIT. 3054 3055 Default is 'auto'. 3056 3057 If the software workaround is enabled for the host, 3058 guests do need not to enable it for nested guests. 3059 3060 kvm.nx_huge_pages_recovery_ratio= 3061 [KVM] Controls how many 4KiB pages are periodically zapped 3062 back to huge pages. 0 disables the recovery, otherwise if 3063 the value is N KVM will zap 1/Nth of the 4KiB pages every 3064 period (see below). The default is 60. 3065 3066 kvm.nx_huge_pages_recovery_period_ms= 3067 [KVM] Controls the time period at which KVM zaps 4KiB pages 3068 back to huge pages. If the value is a non-zero N, KVM will 3069 zap a portion (see ratio above) of the pages every N msecs. 3070 If the value is 0 (the default), KVM will pick a period based 3071 on the ratio, such that a page is zapped after 1 hour on average. 3072 3073 kvm-amd.nested= [KVM,AMD] Control nested virtualization feature in 3074 KVM/SVM. Default is 1 (enabled). 3075 3076 kvm-amd.npt= [KVM,AMD] Control KVM's use of Nested Page Tables, 3077 a.k.a. Two-Dimensional Page Tables. Default is 1 3078 (enabled). Disable by KVM if hardware lacks support 3079 for NPT. 3080 3081 kvm-amd.ciphertext_hiding_asids= 3082 [KVM,AMD] Ciphertext hiding prevents disallowed accesses 3083 to SNP private memory from reading ciphertext. Instead, 3084 reads will see constant default values (0xff). 3085 3086 If ciphertext hiding is enabled, the joint SEV-ES and 3087 SEV-SNP ASID space is partitioned into separate SEV-ES 3088 and SEV-SNP ASID ranges, with the SEV-SNP range being 3089 [1..max_snp_asid] and the SEV-ES range being 3090 (max_snp_asid..min_sev_asid), where min_sev_asid is 3091 enumerated by CPUID.0x.8000_001F[EDX]. 3092 3093 A non-zero value enables SEV-SNP ciphertext hiding and 3094 adjusts the ASID ranges for SEV-ES and SEV-SNP guests. 3095 KVM caps the number of SEV-SNP ASIDs at the maximum 3096 possible value, e.g. specifying -1u will assign all 3097 joint SEV-ES and SEV-SNP ASIDs to SEV-SNP. Note, 3098 assigning all joint ASIDs to SEV-SNP, i.e. configuring 3099 max_snp_asid == min_sev_asid-1, will effectively make 3100 SEV-ES unusable. 3101 3102 kvm-arm.mode= 3103 [KVM,ARM,EARLY] Select one of KVM/arm64's modes of 3104 operation. 3105 3106 none: Forcefully disable KVM. 3107 3108 nvhe: Standard nVHE-based mode, without support for 3109 protected guests. 3110 3111 protected: Mode with support for guests whose state is 3112 kept private from the host, using VHE or 3113 nVHE depending on HW support. 3114 3115 nested: VHE-based mode with support for nested 3116 virtualization. Requires at least ARMv8.4 3117 hardware (with FEAT_NV2). 3118 3119 Defaults to VHE/nVHE based on hardware support. Setting 3120 mode to "protected" will disable kexec and hibernation 3121 for the host. To force nVHE on VHE hardware, add 3122 "arm64_sw.hvhe=0 id_aa64mmfr1.vh=0" to the 3123 command-line. 3124 "nested" is experimental and should be used with 3125 extreme caution. 3126 3127 kvm-arm.vgic_v3_group0_trap= 3128 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-0 3129 system registers 3130 3131 kvm-arm.vgic_v3_group1_trap= 3132 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-1 3133 system registers 3134 3135 kvm-arm.vgic_v3_common_trap= 3136 [KVM,ARM,EARLY] Trap guest accesses to GICv3 common 3137 system registers 3138 3139 kvm-arm.vgic_v4_enable= 3140 [KVM,ARM,EARLY] Allow use of GICv4 for direct 3141 injection of LPIs. 3142 3143 kvm-arm.wfe_trap_policy= 3144 [KVM,ARM] Control when to set WFE instruction trap for 3145 KVM VMs. Traps are allowed but not guaranteed by the 3146 CPU architecture. 3147 3148 trap: set WFE instruction trap 3149 3150 notrap: clear WFE instruction trap 3151 3152 kvm-arm.wfi_trap_policy= 3153 [KVM,ARM] Control when to set WFI instruction trap for 3154 KVM VMs. Traps are allowed but not guaranteed by the 3155 CPU architecture. 3156 3157 trap: set WFI instruction trap 3158 3159 notrap: clear WFI instruction trap 3160 3161 kvm_cma_resv_ratio=n [PPC,EARLY] 3162 Reserves given percentage from system memory area for 3163 contiguous memory allocation for KVM hash pagetable 3164 allocation. 3165 By default it reserves 5% of total system memory. 3166 Format: <integer> 3167 Default: 5 3168 3169 kvm-intel.ept= [KVM,Intel] Control KVM's use of Extended Page Tables, 3170 a.k.a. Two-Dimensional Page Tables. Default is 1 3171 (enabled). Disable by KVM if hardware lacks support 3172 for EPT. 3173 3174 kvm-intel.emulate_invalid_guest_state= 3175 [KVM,Intel] Control whether to emulate invalid guest 3176 state. Ignored if kvm-intel.enable_unrestricted_guest=1, 3177 as guest state is never invalid for unrestricted 3178 guests. This param doesn't apply to nested guests (L2), 3179 as KVM never emulates invalid L2 guest state. 3180 Default is 1 (enabled). 3181 3182 kvm-intel.flexpriority= 3183 [KVM,Intel] Control KVM's use of FlexPriority feature 3184 (TPR shadow). Default is 1 (enabled). Disable by KVM if 3185 hardware lacks support for it. 3186 3187 kvm-intel.nested= 3188 [KVM,Intel] Control nested virtualization feature in 3189 KVM/VMX. Default is 1 (enabled). 3190 3191 kvm-intel.unrestricted_guest= 3192 [KVM,Intel] Control KVM's use of unrestricted guest 3193 feature (virtualized real and unpaged mode). Default 3194 is 1 (enabled). Disable by KVM if EPT is disabled or 3195 hardware lacks support for it. 3196 3197 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 3198 CVE-2018-3620. 3199 3200 Valid arguments: never, cond, always 3201 3202 always: L1D cache flush on every VMENTER. 3203 cond: Flush L1D on VMENTER only when the code between 3204 VMEXIT and VMENTER can leak host memory. 3205 never: Disables the mitigation 3206 3207 Default is cond (do L1 cache flush in specific instances) 3208 3209 kvm-intel.vpid= [KVM,Intel] Control KVM's use of Virtual Processor 3210 Identification feature (tagged TLBs). Default is 1 3211 (enabled). Disable by KVM if hardware lacks support 3212 for it. 3213 3214 l1d_flush= [X86,INTEL,EARLY] 3215 Control mitigation for L1D based snooping vulnerability. 3216 3217 Certain CPUs are vulnerable to an exploit against CPU 3218 internal buffers which can forward information to a 3219 disclosure gadget under certain conditions. 3220 3221 In vulnerable processors, the speculatively 3222 forwarded data can be used in a cache side channel 3223 attack, to access data to which the attacker does 3224 not have direct access. 3225 3226 This parameter controls the mitigation. The 3227 options are: 3228 3229 on - enable the interface for the mitigation 3230 3231 l1tf= [X86,EARLY] Control mitigation of the L1TF vulnerability on 3232 affected CPUs 3233 3234 The kernel PTE inversion protection is unconditionally 3235 enabled and cannot be disabled. 3236 3237 full 3238 Provides all available mitigations for the 3239 L1TF vulnerability. Disables SMT and 3240 enables all mitigations in the 3241 hypervisors, i.e. unconditional L1D flush. 3242 3243 SMT control and L1D flush control via the 3244 sysfs interface is still possible after 3245 boot. Hypervisors will issue a warning 3246 when the first VM is started in a 3247 potentially insecure configuration, 3248 i.e. SMT enabled or L1D flush disabled. 3249 3250 full,force 3251 Same as 'full', but disables SMT and L1D 3252 flush runtime control. Implies the 3253 'nosmt=force' command line option. 3254 (i.e. sysfs control of SMT is disabled.) 3255 3256 flush 3257 Leaves SMT enabled and enables the default 3258 hypervisor mitigation, i.e. conditional 3259 L1D flush. 3260 3261 SMT control and L1D flush control via the 3262 sysfs interface is still possible after 3263 boot. Hypervisors will issue a warning 3264 when the first VM is started in a 3265 potentially insecure configuration, 3266 i.e. SMT enabled or L1D flush disabled. 3267 3268 flush,nosmt 3269 3270 Disables SMT and enables the default 3271 hypervisor mitigation. 3272 3273 SMT control and L1D flush control via the 3274 sysfs interface is still possible after 3275 boot. Hypervisors will issue a warning 3276 when the first VM is started in a 3277 potentially insecure configuration, 3278 i.e. SMT enabled or L1D flush disabled. 3279 3280 flush,nowarn 3281 Same as 'flush', but hypervisors will not 3282 warn when a VM is started in a potentially 3283 insecure configuration. 3284 3285 off 3286 Disables hypervisor mitigations and doesn't 3287 emit any warnings. 3288 It also drops the swap size and available 3289 RAM limit restriction on both hypervisor and 3290 bare metal. 3291 3292 Default is 'flush'. 3293 3294 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 3295 3296 l2cr= [PPC] 3297 3298 l3cr= [PPC] 3299 3300 lapic [X86-32,APIC,EARLY] Enable the local APIC even if BIOS 3301 disabled it. 3302 3303 lapic= [X86,APIC] Do not use TSC deadline 3304 value for LAPIC timer one-shot implementation. Default 3305 back to the programmable timer unit in the LAPIC. 3306 Format: notscdeadline 3307 3308 lapic_timer_c2_ok [X86,APIC,EARLY] trust the local apic timer 3309 in C2 power state. 3310 3311 libata.dma= [LIBATA] DMA control 3312 libata.dma=0 Disable all PATA and SATA DMA 3313 libata.dma=1 PATA and SATA Disk DMA only 3314 libata.dma=2 ATAPI (CDROM) DMA only 3315 libata.dma=4 Compact Flash DMA only 3316 Combinations also work, so libata.dma=3 enables DMA 3317 for disks and CDROMs, but not CFs. 3318 3319 libata.ignore_hpa= [LIBATA] Ignore HPA limit 3320 libata.ignore_hpa=0 keep BIOS limits (default) 3321 libata.ignore_hpa=1 ignore limits, using full disk 3322 3323 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 3324 when set. 3325 Format: <int> 3326 3327 libata.force= [LIBATA] Force configurations. The format is a comma- 3328 separated list of "[ID:]VAL" where ID is PORT[.DEVICE]. 3329 PORT and DEVICE are decimal numbers matching port, link 3330 or device. Basically, it matches the ATA ID string 3331 printed on console by libata. If the whole ID part is 3332 omitted, the last PORT and DEVICE values are used. If 3333 ID hasn't been specified yet, the configuration applies 3334 to all ports, links and devices. 3335 3336 If only DEVICE is omitted, the parameter applies to 3337 the port and all links and devices behind it. DEVICE 3338 number of 0 either selects the first device or the 3339 first fan-out link behind PMP device. It does not 3340 select the host link. DEVICE number of 15 selects the 3341 host link and device attached to it. 3342 3343 The VAL specifies the configuration to force. As long 3344 as there is no ambiguity, shortcut notation is allowed. 3345 For example, both 1.5 and 1.5G would work for 1.5Gbps. 3346 The following configurations can be forced. 3347 3348 * Cable type: 40c, 80c, short40c, unk, ign or sata. 3349 Any ID with matching PORT is used. 3350 3351 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 3352 3353 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 3354 udma[/][16,25,33,44,66,100,133] notation is also 3355 allowed. 3356 3357 * nohrst, nosrst, norst: suppress hard, soft and both 3358 resets. 3359 3360 * rstonce: only attempt one reset during hot-unplug 3361 link recovery. 3362 3363 * [no]dbdelay: Enable or disable the extra 200ms delay 3364 before debouncing a link PHY and device presence 3365 detection. 3366 3367 * [no]ncq: Turn on or off NCQ. 3368 3369 * [no]ncqtrim: Enable or disable queued DSM TRIM. 3370 3371 * [no]ncqati: Enable or disable NCQ trim on ATI chipset. 3372 3373 * [no]trim: Enable or disable (unqueued) TRIM. 3374 3375 * trim_zero: Indicate that TRIM command zeroes data. 3376 3377 * max_trim_128m: Set 128M maximum trim size limit. 3378 3379 * [no]dma: Turn on or off DMA transfers. 3380 3381 * atapi_dmadir: Enable ATAPI DMADIR bridge support. 3382 3383 * atapi_mod16_dma: Enable the use of ATAPI DMA for 3384 commands that are not a multiple of 16 bytes. 3385 3386 * [no]dmalog: Enable or disable the use of the 3387 READ LOG DMA EXT command to access logs. 3388 3389 * [no]iddevlog: Enable or disable access to the 3390 identify device data log. 3391 3392 * [no]logdir: Enable or disable access to the general 3393 purpose log directory. 3394 3395 * max_sec_128: Set transfer size limit to 128 sectors. 3396 3397 * max_sec_1024: Set or clear transfer size limit to 3398 1024 sectors. 3399 3400 * max_sec_lba48: Set or clear transfer size limit to 3401 65535 sectors. 3402 3403 * external: Mark port as external (hotplug-capable). 3404 3405 * [no]lpm: Enable or disable link power management. 3406 3407 * [no]setxfer: Indicate if transfer speed mode setting 3408 should be skipped. 3409 3410 * [no]fua: Disable or enable FUA (Force Unit Access) 3411 support for devices supporting this feature. 3412 3413 * dump_id: Dump IDENTIFY data. 3414 3415 * disable: Disable this device. 3416 3417 If there are multiple matching configurations changing 3418 the same attribute, the last one is used. 3419 3420 load_ramdisk= [RAM] [Deprecated] 3421 3422 lockd.nlm_grace_period=P [NFS] Assign grace period. 3423 Format: <integer> 3424 3425 lockd.nlm_tcpport=N [NFS] Assign TCP port. 3426 Format: <integer> 3427 3428 lockd.nlm_timeout=T [NFS] Assign timeout value. 3429 Format: <integer> 3430 3431 lockd.nlm_udpport=M [NFS] Assign UDP port. 3432 Format: <integer> 3433 3434 lockdown= [SECURITY,EARLY] 3435 { integrity | confidentiality } 3436 Enable the kernel lockdown feature. If set to 3437 integrity, kernel features that allow userland to 3438 modify the running kernel are disabled. If set to 3439 confidentiality, kernel features that allow userland 3440 to extract confidential information from the kernel 3441 are also disabled. 3442 3443 locktorture.acq_writer_lim= [KNL] 3444 Set the time limit in jiffies for a lock 3445 acquisition. Acquisitions exceeding this limit 3446 will result in a splat once they do complete. 3447 3448 locktorture.bind_readers= [KNL] 3449 Specify the list of CPUs to which the readers are 3450 to be bound. 3451 3452 locktorture.bind_writers= [KNL] 3453 Specify the list of CPUs to which the writers are 3454 to be bound. 3455 3456 locktorture.call_rcu_chains= [KNL] 3457 Specify the number of self-propagating call_rcu() 3458 chains to set up. These are used to ensure that 3459 there is a high probability of an RCU grace period 3460 in progress at any given time. Defaults to 0, 3461 which disables these call_rcu() chains. 3462 3463 locktorture.long_hold= [KNL] 3464 Specify the duration in milliseconds for the 3465 occasional long-duration lock hold time. Defaults 3466 to 100 milliseconds. Select 0 to disable. 3467 3468 locktorture.nested_locks= [KNL] 3469 Specify the maximum lock nesting depth that 3470 locktorture is to exercise, up to a limit of 8 3471 (MAX_NESTED_LOCKS). Specify zero to disable. 3472 Note that this parameter is ineffective on types 3473 of locks that do not support nested acquisition. 3474 3475 locktorture.nreaders_stress= [KNL] 3476 Set the number of locking read-acquisition kthreads. 3477 Defaults to being automatically set based on the 3478 number of online CPUs. 3479 3480 locktorture.nwriters_stress= [KNL] 3481 Set the number of locking write-acquisition kthreads. 3482 3483 locktorture.onoff_holdoff= [KNL] 3484 Set time (s) after boot for CPU-hotplug testing. 3485 3486 locktorture.onoff_interval= [KNL] 3487 Set time (s) between CPU-hotplug operations, or 3488 zero to disable CPU-hotplug testing. 3489 3490 locktorture.rt_boost= [KNL] 3491 Do periodic testing of real-time lock priority 3492 boosting. Select 0 to disable, 1 to boost 3493 only rt_mutex, and 2 to boost unconditionally. 3494 Defaults to 2, which might seem to be an 3495 odd choice, but which should be harmless for 3496 non-real-time spinlocks, due to their disabling 3497 of preemption. Note that non-realtime mutexes 3498 disable boosting. 3499 3500 locktorture.rt_boost_factor= [KNL] 3501 Number that determines how often and for how 3502 long priority boosting is exercised. This is 3503 scaled down by the number of writers, so that the 3504 number of boosts per unit time remains roughly 3505 constant as the number of writers increases. 3506 On the other hand, the duration of each boost 3507 increases with the number of writers. 3508 3509 locktorture.shuffle_interval= [KNL] 3510 Set task-shuffle interval (jiffies). Shuffling 3511 tasks allows some CPUs to go into dyntick-idle 3512 mode during the locktorture test. 3513 3514 locktorture.shutdown_secs= [KNL] 3515 Set time (s) after boot system shutdown. This 3516 is useful for hands-off automated testing. 3517 3518 locktorture.stat_interval= [KNL] 3519 Time (s) between statistics printk()s. 3520 3521 locktorture.stutter= [KNL] 3522 Time (s) to stutter testing, for example, 3523 specifying five seconds causes the test to run for 3524 five seconds, wait for five seconds, and so on. 3525 This tests the locking primitive's ability to 3526 transition abruptly to and from idle. 3527 3528 locktorture.torture_type= [KNL] 3529 Specify the locking implementation to test. 3530 3531 locktorture.verbose= [KNL] 3532 Enable additional printk() statements. 3533 3534 locktorture.writer_fifo= [KNL] 3535 Run the write-side locktorture kthreads at 3536 sched_set_fifo() real-time priority. 3537 3538 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 3539 Format: <irq> 3540 3541 loglevel= [KNL,EARLY] 3542 All Kernel Messages with a loglevel smaller than the 3543 console loglevel will be printed to the console. It can 3544 also be changed with klogd or other programs. The 3545 loglevels are defined as follows: 3546 3547 0 (KERN_EMERG) system is unusable 3548 1 (KERN_ALERT) action must be taken immediately 3549 2 (KERN_CRIT) critical conditions 3550 3 (KERN_ERR) error conditions 3551 4 (KERN_WARNING) warning conditions 3552 5 (KERN_NOTICE) normal but significant condition 3553 6 (KERN_INFO) informational 3554 7 (KERN_DEBUG) debug-level messages 3555 3556 log_buf_len=n[KMG] [KNL,EARLY] 3557 Sets the size of the printk ring buffer, in bytes. 3558 n must be a power of two and greater than the 3559 minimal size. The minimal size is defined by 3560 LOG_BUF_SHIFT kernel config parameter. There 3561 is also CONFIG_LOG_CPU_MAX_BUF_SHIFT config 3562 parameter that allows to increase the default size 3563 depending on the number of CPUs. See init/Kconfig 3564 for more details. 3565 3566 logo.nologo [FB] Disables display of the built-in Linux logo. 3567 This may be used to provide more screen space for 3568 kernel log messages and is useful when debugging 3569 kernel boot problems. 3570 3571 lp=0 [LP] Specify parallel ports to use, e.g, 3572 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 3573 lp=reset first parallel port). 'lp=0' disables the 3574 lp=auto printer driver. 'lp=reset' (which can be 3575 specified in addition to the ports) causes 3576 attached printers to be reset. Using 3577 lp=port1,port2,... specifies the parallel ports 3578 to associate lp devices with, starting with 3579 lp0. A port specification may be 'none' to skip 3580 that lp device, or a parport name such as 3581 'parport0'. Specifying 'lp=auto' instead of a 3582 port specification list means that device IDs 3583 from each port should be examined, to see if 3584 an IEEE 1284-compliant printer is attached; if 3585 so, the driver will manage that printer. 3586 See also header of drivers/char/lp.c. 3587 3588 lpj=n [KNL] 3589 Sets loops_per_jiffy to given constant, thus avoiding 3590 time-consuming boot-time autodetection (up to 250 ms per 3591 CPU). 0 enables autodetection (default). To determine 3592 the correct value for your kernel, boot with normal 3593 autodetection and see what value is printed. Note that 3594 on SMP systems the preset will be applied to all CPUs, 3595 which is likely to cause problems if your CPUs need 3596 significantly divergent settings. An incorrect value 3597 will cause delays in the kernel to be wrong, leading to 3598 unpredictable I/O errors and other breakage. Although 3599 unlikely, in the extreme case this might damage your 3600 hardware. 3601 3602 lsm.debug [SECURITY] Enable LSM initialization debugging output. 3603 3604 lsm=lsm1,...,lsmN 3605 [SECURITY] Choose order of LSM initialization. This 3606 overrides CONFIG_LSM, and the "security=" parameter. 3607 3608 machtype= [Loongson] Share the same kernel image file between 3609 different yeeloong laptops. 3610 Example: machtype=lemote-yeeloong-2f-7inch 3611 3612 maxcpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 3613 will bring up during bootup. maxcpus=n : n >= 0 limits 3614 the kernel to bring up 'n' processors. Surely after 3615 bootup you can bring up the other plugged cpu by executing 3616 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 3617 only takes effect during system bootup. 3618 While n=0 is a special case, it is equivalent to "nosmp", 3619 which also disables the IO APIC. 3620 3621 max_loop= [LOOP] The number of loop block devices that get 3622 (loop.max_loop) unconditionally pre-created at init time. The default 3623 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 3624 of statically allocating a predefined number, loop 3625 devices can be requested on-demand with the 3626 /dev/loop-control interface. 3627 3628 mce= [X86-{32,64}] 3629 3630 Please see Documentation/arch/x86/x86_64/machinecheck.rst for sysfs runtime tunables. 3631 3632 off 3633 disable machine check 3634 3635 no_cmci 3636 disable CMCI(Corrected Machine Check Interrupt) that 3637 Intel processor supports. Usually this disablement is 3638 not recommended, but it might be handy if your 3639 hardware is misbehaving. 3640 3641 Note that you'll get more problems without CMCI than 3642 with due to the shared banks, i.e. you might get 3643 duplicated error logs. 3644 3645 dont_log_ce 3646 don't make logs for corrected errors. All events 3647 reported as corrected are silently cleared by OS. This 3648 option will be useful if you have no interest in any 3649 of corrected errors. 3650 3651 ignore_ce 3652 disable features for corrected errors, e.g. 3653 polling timer and CMCI. All events reported as 3654 corrected are not cleared by OS and remained in its 3655 error banks. 3656 3657 Usually this disablement is not recommended, however 3658 if there is an agent checking/clearing corrected 3659 errors (e.g. BIOS or hardware monitoring 3660 applications), conflicting with OS's error handling, 3661 and you cannot deactivate the agent, then this option 3662 will be a help. 3663 3664 no_lmce 3665 do not opt-in to Local MCE delivery. Use legacy method 3666 to broadcast MCEs. 3667 3668 bootlog 3669 enable logging of machine checks left over from 3670 booting. Disabled by default on AMD Fam10h and older 3671 because some BIOS leave bogus ones. 3672 3673 If your BIOS doesn't do that it's a good idea to 3674 enable though to make sure you log even machine check 3675 events that result in a reboot. On Intel systems it is 3676 enabled by default. 3677 3678 nobootlog 3679 disable boot machine check logging. 3680 3681 monarchtimeout (number) 3682 sets the time in us to wait for other CPUs on machine 3683 checks. 0 to disable. 3684 3685 bios_cmci_threshold 3686 don't overwrite the bios-set CMCI threshold. This boot 3687 option prevents Linux from overwriting the CMCI 3688 threshold set by the bios. Without this option, Linux 3689 always sets the CMCI threshold to 1. Enabling this may 3690 make memory predictive failure analysis less effective 3691 if the bios sets thresholds for memory errors since we 3692 will not see details for all errors. 3693 3694 recovery 3695 force-enable recoverable machine check code paths 3696 3697 Everything else is in sysfs now. 3698 3699 3700 md= [HW] RAID subsystems devices and level 3701 See Documentation/admin-guide/md.rst. 3702 3703 mdacon= [MDA] 3704 Format: <first>,<last> 3705 Specifies range of consoles to be captured by the MDA. 3706 3707 mds= [X86,INTEL,EARLY] 3708 Control mitigation for the Micro-architectural Data 3709 Sampling (MDS) vulnerability. 3710 3711 Certain CPUs are vulnerable to an exploit against CPU 3712 internal buffers which can forward information to a 3713 disclosure gadget under certain conditions. 3714 3715 In vulnerable processors, the speculatively 3716 forwarded data can be used in a cache side channel 3717 attack, to access data to which the attacker does 3718 not have direct access. 3719 3720 This parameter controls the MDS mitigation. The 3721 options are: 3722 3723 full - Enable MDS mitigation on vulnerable CPUs 3724 full,nosmt - Enable MDS mitigation and disable 3725 SMT on vulnerable CPUs 3726 off - Unconditionally disable MDS mitigation 3727 3728 On TAA-affected machines, mds=off can be prevented by 3729 an active TAA mitigation as both vulnerabilities are 3730 mitigated with the same mechanism so in order to disable 3731 this mitigation, you need to specify tsx_async_abort=off 3732 too. 3733 3734 Not specifying this option is equivalent to 3735 mds=full. 3736 3737 For details see: Documentation/admin-guide/hw-vuln/mds.rst 3738 3739 mem=nn[KMG] [HEXAGON,EARLY] Set the memory size. 3740 Must be specified, otherwise memory size will be 0. 3741 3742 mem=nn[KMG] [KNL,BOOT,EARLY] Force usage of a specific amount 3743 of memory Amount of memory to be used in cases 3744 as follows: 3745 3746 1 for test; 3747 2 when the kernel is not able to see the whole system memory; 3748 3 memory that lies after 'mem=' boundary is excluded from 3749 the hypervisor, then assigned to KVM guests. 3750 4 to limit the memory available for kdump kernel. 3751 3752 [ARC,MICROBLAZE] - the limit applies only to low memory, 3753 high memory is not affected. 3754 3755 [ARM64] - only limits memory covered by the linear 3756 mapping. The NOMAP regions are not affected. 3757 3758 [X86] Work as limiting max address. Use together 3759 with memmap= to avoid physical address space collisions. 3760 Without memmap= PCI devices could be placed at addresses 3761 belonging to unused RAM. 3762 3763 Note that this only takes effects during boot time since 3764 in above case 3, memory may need be hot added after boot 3765 if system memory of hypervisor is not sufficient. 3766 3767 mem=nn[KMG]@ss[KMG] 3768 [ARM,MIPS,EARLY] - override the memory layout 3769 reported by firmware. 3770 Define a memory region of size nn[KMG] starting at 3771 ss[KMG]. 3772 Multiple different regions can be specified with 3773 multiple mem= parameters on the command line. 3774 3775 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 3776 memory. 3777 3778 memblock=debug [KNL,EARLY] Enable memblock debug messages. 3779 3780 memchunk=nn[KMG] 3781 [KNL,SH] Allow user to override the default size for 3782 per-device physically contiguous DMA buffers. 3783 3784 memhp_default_state=online/offline/online_kernel/online_movable 3785 [KNL] Set the initial state for the memory hotplug 3786 onlining policy. If not specified, the default value is 3787 set according to the 3788 CONFIG_MHP_DEFAULT_ONLINE_TYPE kernel config 3789 options. 3790 See Documentation/admin-guide/mm/memory-hotplug.rst. 3791 3792 memmap=exactmap [KNL,X86,EARLY] Enable setting of an exact 3793 E820 memory map, as specified by the user. 3794 Such memmap=exactmap lines can be constructed based on 3795 BIOS output or other requirements. See the memmap=nn@ss 3796 option description. 3797 3798 memmap=nn[KMG]@ss[KMG] 3799 [KNL, X86,MIPS,XTENSA,EARLY] Force usage of a specific region of memory. 3800 Region of memory to be used is from ss to ss+nn. 3801 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 3802 which limits max address to nn[KMG]. 3803 Multiple different regions can be specified, 3804 comma delimited. 3805 Example: 3806 memmap=100M@2G,100M#3G,1G!1024G 3807 3808 memmap=nn[KMG]#ss[KMG] 3809 [KNL,ACPI,EARLY] Mark specific memory as ACPI data. 3810 Region of memory to be marked is from ss to ss+nn. 3811 3812 memmap=nn[KMG]$ss[KMG] 3813 [KNL,ACPI,EARLY] Mark specific memory as reserved. 3814 Region of memory to be reserved is from ss to ss+nn. 3815 Example: Exclude memory from 0x18690000-0x1869ffff 3816 memmap=64K$0x18690000 3817 or 3818 memmap=0x10000$0x18690000 3819 Some bootloaders may need an escape character before '$', 3820 like Grub2, otherwise '$' and the following number 3821 will be eaten. 3822 3823 memmap=nn[KMG]!ss[KMG,EARLY] 3824 [KNL,X86] Mark specific memory as protected. 3825 Region of memory to be used, from ss to ss+nn. 3826 The memory region may be marked as e820 type 12 (0xc) 3827 and is NVDIMM or ADR memory. 3828 3829 memmap=<size>%<offset>-<oldtype>+<newtype> 3830 [KNL,ACPI,EARLY] Convert memory within the specified region 3831 from <oldtype> to <newtype>. If "-<oldtype>" is left 3832 out, the whole region will be marked as <newtype>, 3833 even if previously unavailable. If "+<newtype>" is left 3834 out, matching memory will be removed. Types are 3835 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 3836 3 = ACPI, 12 = PRAM. 3837 3838 memory_corruption_check=0/1 [X86,EARLY] 3839 Some BIOSes seem to corrupt the first 64k of 3840 memory when doing things like suspend/resume. 3841 Setting this option will scan the memory 3842 looking for corruption. Enabling this will 3843 both detect corruption and prevent the kernel 3844 from using the memory being corrupted. 3845 However, it's intended as a diagnostic tool; if 3846 repeatable BIOS-originated corruption always 3847 affects the same memory, you can use memmap= 3848 to prevent the kernel from using that memory. 3849 3850 memory_corruption_check_size=size [X86,EARLY] 3851 By default it checks for corruption in the low 3852 64k, making this memory unavailable for normal 3853 use. Use this parameter to scan for 3854 corruption in more or less memory. 3855 3856 memory_corruption_check_period=seconds [X86,EARLY] 3857 By default it checks for corruption every 60 3858 seconds. Use this parameter to check at some 3859 other rate. 0 disables periodic checking. 3860 3861 memory_hotplug.memmap_on_memory 3862 [KNL,X86,ARM] Boolean flag to enable this feature. 3863 Format: {on | off (default)} 3864 When enabled, runtime hotplugged memory will 3865 allocate its internal metadata (struct pages, 3866 those vmemmap pages cannot be optimized even 3867 if hugetlb_free_vmemmap is enabled) from the 3868 hotadded memory which will allow to hotadd a 3869 lot of memory without requiring additional 3870 memory to do so. 3871 This feature is disabled by default because it 3872 has some implication on large (e.g. GB) 3873 allocations in some configurations (e.g. small 3874 memory blocks). 3875 The state of the flag can be read in 3876 /sys/module/memory_hotplug/parameters/memmap_on_memory. 3877 Note that even when enabled, there are a few cases where 3878 the feature is not effective. 3879 3880 memtest= [KNL,X86,ARM,M68K,PPC,RISCV,EARLY] Enable memtest 3881 Format: <integer> 3882 default : 0 <disable> 3883 Specifies the number of memtest passes to be 3884 performed. Each pass selects another test 3885 pattern from a given set of patterns. Memtest 3886 fills the memory with this pattern, validates 3887 memory contents and reserves bad memory 3888 regions that are detected. 3889 3890 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 3891 Valid arguments: on, off 3892 Default: off 3893 mem_encrypt=on: Activate SME 3894 mem_encrypt=off: Do not activate SME 3895 3896 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst 3897 for details on when memory encryption can be activated. 3898 3899 mem_sleep_default= [SUSPEND] Default system suspend mode: 3900 s2idle - Suspend-To-Idle 3901 shallow - Power-On Suspend or equivalent (if supported) 3902 deep - Suspend-To-RAM or equivalent (if supported) 3903 See Documentation/admin-guide/pm/sleep-states.rst. 3904 3905 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 3906 the BIOS has incorrectly applied a workaround. TinyBIOS 3907 version 0.98 is known to be affected, 0.99 fixes the 3908 problem by letting the user disable the workaround. 3909 3910 mga= [HW,DRM] 3911 3912 microcode= [X86] Control the behavior of the microcode loader. 3913 Available options, comma separated: 3914 3915 base_rev=X - with <X> with format: <u32> 3916 Set the base microcode revision of each thread when in 3917 debug mode. 3918 3919 dis_ucode_ldr: disable the microcode loader 3920 3921 force_minrev: 3922 Enable or disable the microcode minimal revision 3923 enforcement for the runtime microcode loader. 3924 3925 mini2440= [ARM,HW,KNL] 3926 Format:[0..2][b][c][t] 3927 Default: "0tb" 3928 MINI2440 configuration specification: 3929 0 - The attached screen is the 3.5" TFT 3930 1 - The attached screen is the 7" TFT 3931 2 - The VGA Shield is attached (1024x768) 3932 Leaving out the screen size parameter will not load 3933 the TFT driver, and the framebuffer will be left 3934 unconfigured. 3935 b - Enable backlight. The TFT backlight pin will be 3936 linked to the kernel VESA blanking code and a GPIO 3937 LED. This parameter is not necessary when using the 3938 VGA shield. 3939 c - Enable the s3c camera interface. 3940 t - Reserved for enabling touchscreen support. The 3941 touchscreen support is not enabled in the mainstream 3942 kernel as of 2.6.30, a preliminary port can be found 3943 in the "bleeding edge" mini2440 support kernel at 3944 https://repo.or.cz/w/linux-2.6/mini2440.git 3945 3946 mitigations= 3947 [X86,PPC,S390,ARM64,EARLY] Control optional mitigations for 3948 CPU vulnerabilities. This is a set of curated, 3949 arch-independent options, each of which is an 3950 aggregation of existing arch-specific options. 3951 3952 Note, "mitigations" is supported if and only if the 3953 kernel was built with CPU_MITIGATIONS=y. 3954 3955 off 3956 Disable all optional CPU mitigations. This 3957 improves system performance, but it may also 3958 expose users to several CPU vulnerabilities. 3959 Equivalent to: if nokaslr then kpti=0 [ARM64] 3960 gather_data_sampling=off [X86] 3961 indirect_target_selection=off [X86] 3962 kvm.nx_huge_pages=off [X86] 3963 l1tf=off [X86] 3964 mds=off [X86] 3965 mmio_stale_data=off [X86] 3966 no_entry_flush [PPC] 3967 no_uaccess_flush [PPC] 3968 nobp=0 [S390] 3969 nopti [X86,PPC] 3970 nospectre_bhb [ARM64] 3971 nospectre_v1 [X86,PPC] 3972 nospectre_v2 [X86,PPC,S390,ARM64] 3973 reg_file_data_sampling=off [X86] 3974 retbleed=off [X86] 3975 spec_rstack_overflow=off [X86] 3976 spec_store_bypass_disable=off [X86,PPC] 3977 spectre_bhi=off [X86] 3978 spectre_v2_user=off [X86] 3979 srbds=off [X86,INTEL] 3980 ssbd=force-off [ARM64] 3981 tsx_async_abort=off [X86] 3982 vmscape=off [X86] 3983 3984 Exceptions: 3985 This does not have any effect on 3986 kvm.nx_huge_pages when 3987 kvm.nx_huge_pages=force. 3988 3989 auto (default) 3990 Mitigate all CPU vulnerabilities, but leave SMT 3991 enabled, even if it's vulnerable. This is for 3992 users who don't want to be surprised by SMT 3993 getting disabled across kernel upgrades, or who 3994 have other ways of avoiding SMT-based attacks. 3995 Equivalent to: (default behavior) 3996 3997 auto,nosmt 3998 Mitigate all CPU vulnerabilities, disabling SMT 3999 if needed. This is for users who always want to 4000 be fully mitigated, even if it means losing SMT. 4001 Equivalent to: l1tf=flush,nosmt [X86] 4002 mds=full,nosmt [X86] 4003 tsx_async_abort=full,nosmt [X86] 4004 mmio_stale_data=full,nosmt [X86] 4005 retbleed=auto,nosmt [X86] 4006 4007 [X86] After one of the above options, additionally 4008 supports attack-vector based controls as documented in 4009 Documentation/admin-guide/hw-vuln/attack_vector_controls.rst 4010 4011 mminit_loglevel= 4012 [KNL,EARLY] When CONFIG_DEBUG_MEMORY_INIT is set, this 4013 parameter allows control of the logging verbosity for 4014 the additional memory initialisation checks. A value 4015 of 0 disables mminit logging and a level of 4 will 4016 log everything. Information is printed at KERN_DEBUG 4017 so loglevel=8 may also need to be specified. 4018 4019 mmio_stale_data= 4020 [X86,INTEL,EARLY] Control mitigation for the Processor 4021 MMIO Stale Data vulnerabilities. 4022 4023 Processor MMIO Stale Data is a class of 4024 vulnerabilities that may expose data after an MMIO 4025 operation. Exposed data could originate or end in 4026 the same CPU buffers as affected by MDS and TAA. 4027 Therefore, similar to MDS and TAA, the mitigation 4028 is to clear the affected CPU buffers. 4029 4030 This parameter controls the mitigation. The 4031 options are: 4032 4033 full - Enable mitigation on vulnerable CPUs 4034 4035 full,nosmt - Enable mitigation and disable SMT on 4036 vulnerable CPUs. 4037 4038 off - Unconditionally disable mitigation 4039 4040 On MDS or TAA affected machines, 4041 mmio_stale_data=off can be prevented by an active 4042 MDS or TAA mitigation as these vulnerabilities are 4043 mitigated with the same mechanism so in order to 4044 disable this mitigation, you need to specify 4045 mds=off and tsx_async_abort=off too. 4046 4047 Not specifying this option is equivalent to 4048 mmio_stale_data=full. 4049 4050 For details see: 4051 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst 4052 4053 <module>.async_probe[=<bool>] [KNL] 4054 If no <bool> value is specified or if the value 4055 specified is not a valid <bool>, enable asynchronous 4056 probe on this module. Otherwise, enable/disable 4057 asynchronous probe on this module as indicated by the 4058 <bool> value. See also: module.async_probe 4059 4060 module.async_probe=<bool> 4061 [KNL] When set to true, modules will use async probing 4062 by default. To enable/disable async probing for a 4063 specific module, use the module specific control that 4064 is documented under <module>.async_probe. When both 4065 module.async_probe and <module>.async_probe are 4066 specified, <module>.async_probe takes precedence for 4067 the specific module. 4068 4069 module.enable_dups_trace 4070 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set, 4071 this means that duplicate request_module() calls will 4072 trigger a WARN_ON() instead of a pr_warn(). Note that 4073 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s 4074 will always be issued and this option does nothing. 4075 module.sig_enforce 4076 [KNL] When CONFIG_MODULE_SIG is set, this means that 4077 modules without (valid) signatures will fail to load. 4078 Note that if CONFIG_MODULE_SIG_FORCE is set, that 4079 is always true, so this option does nothing. 4080 4081 module_blacklist= [KNL] Do not load a comma-separated list of 4082 modules. Useful for debugging problem modules. 4083 4084 mousedev.tap_time= 4085 [MOUSE] Maximum time between finger touching and 4086 leaving touchpad surface for touch to be considered 4087 a tap and be reported as a left button click (for 4088 touchpads working in absolute mode only). 4089 Format: <msecs> 4090 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 4091 reporting absolute coordinates, such as tablets 4092 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 4093 reporting absolute coordinates, such as tablets 4094 4095 movablecore= [KNL,X86,PPC,EARLY] 4096 Format: nn[KMGTPE] | nn% 4097 This parameter is the complement to kernelcore=, it 4098 specifies the amount of memory used for migratable 4099 allocations. If both kernelcore and movablecore is 4100 specified, then kernelcore will be at *least* the 4101 specified value but may be more. If movablecore on its 4102 own is specified, the administrator must be careful 4103 that the amount of memory usable for all allocations 4104 is not too small. 4105 4106 movable_node [KNL,EARLY] Boot-time switch to make hotplugable memory 4107 NUMA nodes to be movable. This means that the memory 4108 of such nodes will be usable only for movable 4109 allocations which rules out almost all kernel 4110 allocations. Use with caution! 4111 4112 MTD_Partition= [MTD] 4113 Format: <name>,<region-number>,<size>,<offset> 4114 4115 MTD_Region= [MTD] Format: 4116 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 4117 4118 mtdparts= [MTD] 4119 See drivers/mtd/parsers/cmdlinepart.c 4120 4121 mtouchusb.raw_coordinates= 4122 [HW] Make the MicroTouch USB driver use raw coordinates 4123 ('y', default) or cooked coordinates ('n') 4124 4125 mtrr=debug [X86,EARLY] 4126 Enable printing debug information related to MTRR 4127 registers at boot time. 4128 4129 mtrr_chunk_size=nn[KMG,X86,EARLY] 4130 used for mtrr cleanup. It is largest continuous chunk 4131 that could hold holes aka. UC entries. 4132 4133 mtrr_gran_size=nn[KMG,X86,EARLY] 4134 Used for mtrr cleanup. It is granularity of mtrr block. 4135 Default is 1. 4136 Large value could prevent small alignment from 4137 using up MTRRs. 4138 4139 mtrr_spare_reg_nr=n [X86,EARLY] 4140 Format: <integer> 4141 Range: 0,7 : spare reg number 4142 Default : 1 4143 Used for mtrr cleanup. It is spare mtrr entries number. 4144 Set to 2 or more if your graphical card needs more. 4145 4146 multitce=off [PPC] This parameter disables the use of the pSeries 4147 firmware feature for updating multiple TCE entries 4148 at a time. 4149 4150 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 4151 4152 netdev= [NET] Network devices parameters 4153 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 4154 Note that mem_start is often overloaded to mean 4155 something different and driver-specific. 4156 This usage is only documented in each driver source 4157 file if at all. 4158 4159 netpoll.carrier_timeout= 4160 [NET] Specifies amount of time (in seconds) that 4161 netpoll should wait for a carrier. By default netpoll 4162 waits 4 seconds. 4163 4164 nf_conntrack.acct= 4165 [NETFILTER] Enable connection tracking flow accounting 4166 0 to disable accounting 4167 1 to enable accounting 4168 Default value is 0. 4169 4170 nfs.cache_getent= 4171 [NFS] sets the pathname to the program which is used 4172 to update the NFS client cache entries. 4173 4174 nfs.cache_getent_timeout= 4175 [NFS] sets the timeout after which an attempt to 4176 update a cache entry is deemed to have failed. 4177 4178 nfs.callback_nr_threads= 4179 [NFSv4] set the total number of threads that the 4180 NFS client will assign to service NFSv4 callback 4181 requests. 4182 4183 nfs.callback_tcpport= 4184 [NFS] set the TCP port on which the NFSv4 callback 4185 channel should listen. 4186 4187 nfs.delay_retrans= 4188 [NFS] specifies the number of times the NFSv4 client 4189 retries the request before returning an EAGAIN error, 4190 after a reply of NFS4ERR_DELAY from the server. 4191 Only applies if the softerr mount option is enabled, 4192 and the specified value is >= 0. 4193 4194 nfs.enable_ino64= 4195 [NFS] enable 64-bit inode numbers. 4196 If zero, the NFS client will fake up a 32-bit inode 4197 number for the readdir() and stat() syscalls instead 4198 of returning the full 64-bit number. 4199 The default is to return 64-bit inode numbers. 4200 4201 nfs.idmap_cache_timeout= 4202 [NFS] set the maximum lifetime for idmapper cache 4203 entries. 4204 4205 nfs.max_session_cb_slots= 4206 [NFSv4.1] Sets the maximum number of session 4207 slots the client will assign to the callback 4208 channel. This determines the maximum number of 4209 callbacks the client will process in parallel for 4210 a particular server. 4211 4212 nfs.max_session_slots= 4213 [NFSv4.1] Sets the maximum number of session slots 4214 the client will attempt to negotiate with the server. 4215 This limits the number of simultaneous RPC requests 4216 that the client can send to the NFSv4.1 server. 4217 Note that there is little point in setting this 4218 value higher than the max_tcp_slot_table_limit. 4219 4220 nfs.nfs4_disable_idmapping= 4221 [NFSv4] When set to the default of '1', this option 4222 ensures that both the RPC level authentication 4223 scheme and the NFS level operations agree to use 4224 numeric uids/gids if the mount is using the 4225 'sec=sys' security flavour. In effect it is 4226 disabling idmapping, which can make migration from 4227 legacy NFSv2/v3 systems to NFSv4 easier. 4228 Servers that do not support this mode of operation 4229 will be autodetected by the client, and it will fall 4230 back to using the idmapper. 4231 To turn off this behaviour, set the value to '0'. 4232 4233 nfs.nfs4_unique_id= 4234 [NFS4] Specify an additional fixed unique ident- 4235 ification string that NFSv4 clients can insert into 4236 their nfs_client_id4 string. This is typically a 4237 UUID that is generated at system install time. 4238 4239 nfs.recover_lost_locks= 4240 [NFSv4] Attempt to recover locks that were lost due 4241 to a lease timeout on the server. Please note that 4242 doing this risks data corruption, since there are 4243 no guarantees that the file will remain unchanged 4244 after the locks are lost. 4245 If you want to enable the kernel legacy behaviour of 4246 attempting to recover these locks, then set this 4247 parameter to '1'. 4248 The default parameter value of '0' causes the kernel 4249 not to attempt recovery of lost locks. 4250 4251 nfs.send_implementation_id= 4252 [NFSv4.1] Send client implementation identification 4253 information in exchange_id requests. 4254 If zero, no implementation identification information 4255 will be sent. 4256 The default is to send the implementation identification 4257 information. 4258 4259 nfs4.layoutstats_timer= 4260 [NFSv4.2] Change the rate at which the kernel sends 4261 layoutstats to the pNFS metadata server. 4262 4263 Setting this to value to 0 causes the kernel to use 4264 whatever value is the default set by the layout 4265 driver. A non-zero value sets the minimum interval 4266 in seconds between layoutstats transmissions. 4267 4268 nfsd.inter_copy_offload_enable= 4269 [NFSv4.2] When set to 1, the server will support 4270 server-to-server copies for which this server is 4271 the destination of the copy. 4272 4273 nfsd.nfs4_disable_idmapping= 4274 [NFSv4] When set to the default of '1', the NFSv4 4275 server will return only numeric uids and gids to 4276 clients using auth_sys, and will accept numeric uids 4277 and gids from such clients. This is intended to ease 4278 migration from NFSv2/v3. 4279 4280 nfsd.nfsd4_ssc_umount_timeout= 4281 [NFSv4.2] When used as the destination of a 4282 server-to-server copy, knfsd temporarily mounts 4283 the source server. It caches the mount in case 4284 it will be needed again, and discards it if not 4285 used for the number of milliseconds specified by 4286 this parameter. 4287 4288 nfsaddrs= [NFS] Deprecated. Use ip= instead. 4289 See Documentation/admin-guide/nfs/nfsroot.rst. 4290 4291 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 4292 See Documentation/admin-guide/nfs/nfsroot.rst. 4293 4294 nfsrootdebug [NFS] enable nfsroot debugging messages. 4295 See Documentation/admin-guide/nfs/nfsroot.rst. 4296 4297 nmi_backtrace.backtrace_idle [KNL] 4298 Dump stacks even of idle CPUs in response to an 4299 NMI stack-backtrace request. 4300 4301 nmi_debug= [KNL,SH] Specify one or more actions to take 4302 when a NMI is triggered. 4303 Format: [state][,regs][,debounce][,die] 4304 4305 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 4306 Format: [panic,][nopanic,][rNNN,][num] 4307 Valid num: 0 or 1 4308 0 - turn hardlockup detector in nmi_watchdog off 4309 1 - turn hardlockup detector in nmi_watchdog on 4310 rNNN - configure the watchdog with raw perf event 0xNNN 4311 4312 When panic is specified, panic when an NMI watchdog 4313 timeout occurs (or 'nopanic' to not panic on an NMI 4314 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 4315 To disable both hard and soft lockup detectors, 4316 please see 'nowatchdog'. 4317 This is useful when you use a panic=... timeout and 4318 need the box quickly up again. 4319 4320 These settings can be accessed at runtime via 4321 the nmi_watchdog and hardlockup_panic sysctls. 4322 4323 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 4324 emulation library even if a 387 maths coprocessor 4325 is present. 4326 4327 no4lvl [RISCV,EARLY] Disable 4-level and 5-level paging modes. 4328 Forces kernel to use 3-level paging instead. 4329 4330 no5lvl [X86-64,RISCV,EARLY] Disable 5-level paging mode. Forces 4331 kernel to use 4-level paging instead. 4332 4333 noalign [KNL,ARM] 4334 4335 noapic [SMP,APIC,EARLY] Tells the kernel to not make use of any 4336 IOAPICs that may be present in the system. 4337 4338 noapictimer [APIC,X86] Don't set up the APIC timer 4339 4340 noautogroup Disable scheduler automatic task group creation. 4341 4342 nocache [ARM,EARLY] 4343 4344 no_console_suspend 4345 [HW] Never suspend the console 4346 Disable suspending of consoles during suspend and 4347 hibernate operations. Once disabled, debugging 4348 messages can reach various consoles while the rest 4349 of the system is being put to sleep (ie, while 4350 debugging driver suspend/resume hooks). This may 4351 not work reliably with all consoles, but is known 4352 to work with serial and VGA consoles. 4353 To facilitate more flexible debugging, we also add 4354 console_suspend, a printk module parameter to control 4355 it. Users could use console_suspend (usually 4356 /sys/module/printk/parameters/console_suspend) to 4357 turn on/off it dynamically. 4358 4359 no_debug_objects 4360 [KNL,EARLY] Disable object debugging 4361 4362 nodsp [SH] Disable hardware DSP at boot time. 4363 4364 noefi [EFI,EARLY] Disable EFI runtime services support. 4365 4366 no_entry_flush [PPC,EARLY] Don't flush the L1-D cache when entering the kernel. 4367 4368 noexec32 [X86-64] 4369 This affects only 32-bit executables. 4370 noexec32=on: enable non-executable mappings (default) 4371 read doesn't imply executable mappings 4372 noexec32=off: disable non-executable mappings 4373 read implies executable mappings 4374 4375 no_file_caps Tells the kernel not to honor file capabilities. The 4376 only way then for a file to be executed with privilege 4377 is to be setuid root or executed by root. 4378 4379 nofpu [MIPS,SH] Disable hardware FPU at boot time. 4380 4381 nofsgsbase [X86] Disables FSGSBASE instructions. 4382 4383 nofxsr [BUGS=X86-32] Disables x86 floating point extended 4384 register save and restore. The kernel will only save 4385 legacy floating-point registers on task switch. 4386 4387 nogbpages [X86] Do not use GB pages for kernel direct mappings. 4388 4389 no_hash_pointers 4390 [KNL,EARLY] 4391 Alias for "hash_pointers=never". 4392 4393 nohibernate [HIBERNATION] Disable hibernation and resume. 4394 4395 nohlt [ARM,ARM64,MICROBLAZE,MIPS,PPC,RISCV,SH] Forces the kernel to 4396 busy wait in do_idle() and not use the arch_cpu_idle() 4397 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP 4398 to be effective. This is useful on platforms where the 4399 sleep(SH) or wfi(ARM,ARM64) instructions do not work 4400 correctly or when doing power measurements to evaluate 4401 the impact of the sleep instructions. This is also 4402 useful when using JTAG debugger. 4403 4404 nohpet [X86] Don't use the HPET timer. 4405 4406 nohugeiomap [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge I/O mappings. 4407 4408 nohugevmalloc [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge vmalloc mappings. 4409 4410 nohz= [KNL] Boottime enable/disable dynamic ticks 4411 Valid arguments: on, off 4412 Default: on 4413 4414 nohz_full= [KNL,BOOT,SMP,ISOL] 4415 The argument is a cpu list, as described above. 4416 In kernels built with CONFIG_NO_HZ_FULL=y, set 4417 the specified list of CPUs whose tick will be stopped 4418 whenever possible. The boot CPU will be forced outside 4419 the range to maintain the timekeeping. Any CPUs 4420 in this list will have their RCU callbacks offloaded, 4421 just as if they had also been called out in the 4422 rcu_nocbs= boot parameter. 4423 4424 Note that this argument takes precedence over 4425 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 4426 4427 noinitrd [RAM] Tells the kernel not to load any configured 4428 initial RAM disk. 4429 4430 nointremap [X86-64,Intel-IOMMU,EARLY] Do not enable interrupt 4431 remapping. 4432 [Deprecated - use intremap=off] 4433 4434 noinvpcid [X86,EARLY] Disable the INVPCID cpu feature. 4435 4436 noiotrap [SH] Disables trapped I/O port accesses. 4437 4438 noirqdebug [X86-32] Disables the code which attempts to detect and 4439 disable unhandled interrupt sources. 4440 4441 noisapnp [ISAPNP] Disables ISA PnP code. 4442 4443 nokaslr [KNL,EARLY] 4444 When CONFIG_RANDOMIZE_BASE is set, this disables 4445 kernel and module base offset ASLR (Address Space 4446 Layout Randomization). 4447 4448 no-kvmapf [X86,KVM,EARLY] Disable paravirtualized asynchronous page 4449 fault handling. 4450 4451 no-kvmclock [X86,KVM,EARLY] Disable paravirtualized KVM clock driver 4452 4453 nolapic [X86-32,APIC,EARLY] Do not enable or use the local APIC. 4454 4455 nolapic_timer [X86-32,APIC,EARLY] Do not use the local APIC timer. 4456 4457 nomce [X86-32] Disable Machine Check Exception 4458 4459 nomfgpt [X86-32] Disable Multi-Function General Purpose 4460 Timer usage (for AMD Geode machines). 4461 4462 nomodeset Disable kernel modesetting. Most systems' firmware 4463 sets up a display mode and provides framebuffer memory 4464 for output. With nomodeset, DRM and fbdev drivers will 4465 not load if they could possibly displace the pre- 4466 initialized output. Only the system framebuffer will 4467 be available for use. The respective drivers will not 4468 perform display-mode changes or accelerated rendering. 4469 4470 Useful as error fallback, or for testing and debugging. 4471 4472 nomodule Disable module load 4473 4474 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 4475 shutdown the other cpus. Instead use the REBOOT_VECTOR 4476 irq. 4477 4478 nopat [X86,EARLY] Disable PAT (page attribute table extension of 4479 pagetables) support. 4480 4481 nopcid [X86-64,EARLY] Disable the PCID cpu feature. 4482 4483 nopku [X86] Disable Memory Protection Keys CPU feature found 4484 in some Intel CPUs. 4485 4486 nopti [X86-64,EARLY] 4487 Equivalent to pti=off 4488 4489 nopv= [X86,XEN,KVM,HYPER_V,VMWARE,EARLY] 4490 Disables the PV optimizations forcing the guest to run 4491 as generic guest with no PV drivers. Currently support 4492 XEN HVM, KVM, HYPER_V and VMWARE guest. 4493 4494 nopvspin [X86,XEN,KVM,EARLY] 4495 Disables the qspinlock slow path using PV optimizations 4496 which allow the hypervisor to 'idle' the guest on lock 4497 contention. 4498 4499 norandmaps Don't use address space randomization. Equivalent to 4500 echo 0 > /proc/sys/kernel/randomize_va_space 4501 4502 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 4503 with UP alternatives 4504 4505 noresume [SWSUSP] Disables resume and restores original swap 4506 space. 4507 4508 no-scroll [VGA] Disables scrollback. 4509 This is required for the Braillex ib80-piezo Braille 4510 reader made by F.H. Papenmeier (Germany). 4511 4512 nosgx [X86-64,SGX,EARLY] Disables Intel SGX kernel support. 4513 4514 nosmap [PPC,EARLY] 4515 Disable SMAP (Supervisor Mode Access Prevention) 4516 even if it is supported by processor. 4517 4518 nosmep [PPC64s,EARLY] 4519 Disable SMEP (Supervisor Mode Execution Prevention) 4520 even if it is supported by processor. 4521 4522 nosmp [SMP,EARLY] Tells an SMP kernel to act as a UP kernel, 4523 and disable the IO APIC. legacy for "maxcpus=0". 4524 4525 nosmt [KNL,MIPS,PPC,EARLY] Disable symmetric multithreading (SMT). 4526 Equivalent to smt=1. 4527 4528 [KNL,X86,PPC,S390] Disable symmetric multithreading (SMT). 4529 nosmt=force: Force disable SMT, cannot be undone 4530 via the sysfs control file. 4531 4532 nosoftlockup [KNL] Disable the soft-lockup detector. 4533 4534 nospec_store_bypass_disable 4535 [HW,EARLY] Disable all mitigations for the Speculative 4536 Store Bypass vulnerability 4537 4538 nospectre_bhb [ARM64,EARLY] Disable all mitigations for Spectre-BHB (branch 4539 history injection) vulnerability. System may allow data leaks 4540 with this option. 4541 4542 nospectre_v1 [X86,PPC,EARLY] Disable mitigations for Spectre Variant 1 4543 (bounds check bypass). With this option data leaks are 4544 possible in the system. 4545 4546 nospectre_v2 [X86,PPC_E500,ARM64,EARLY] Disable all mitigations 4547 for the Spectre variant 2 (indirect branch 4548 prediction) vulnerability. System may allow data 4549 leaks with this option. 4550 4551 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES,RISCV,LOONGARCH,EARLY] 4552 Disable paravirtualized steal time accounting. steal time 4553 is computed, but won't influence scheduler behaviour 4554 4555 nosync [HW,M68K] Disables sync negotiation for all devices. 4556 4557 no_timer_check [X86,APIC] Disables the code which tests for broken 4558 timer IRQ sources, i.e., the IO-APIC timer. This can 4559 work around problems with incorrect timer 4560 initialization on some boards. 4561 4562 no_uaccess_flush 4563 [PPC,EARLY] Don't flush the L1-D cache after accessing user data. 4564 4565 novmcoredd [KNL,KDUMP] 4566 Disable device dump. Device dump allows drivers to 4567 append dump data to vmcore so you can collect driver 4568 specified debug info. Drivers can append the data 4569 without any limit and this data is stored in memory, 4570 so this may cause significant memory stress. Disabling 4571 device dump can help save memory but the driver debug 4572 data will be no longer available. This parameter 4573 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 4574 is set. 4575 4576 no-vmw-sched-clock 4577 [X86,PV_OPS,EARLY] Disable paravirtualized VMware 4578 scheduler clock and use the default one. 4579 4580 nowatchdog [KNL] Disable both lockup detectors, i.e. 4581 soft-lockup and NMI watchdog (hard-lockup). 4582 4583 nowb [ARM,EARLY] 4584 4585 nox2apic [X86-64,APIC,EARLY] Do not enable x2APIC mode. 4586 4587 NOTE: this parameter will be ignored on systems with the 4588 LEGACY_XAPIC_DISABLED bit set in the 4589 IA32_XAPIC_DISABLE_STATUS MSR. 4590 4591 noxsave [BUGS=X86] Disables x86 extended register state save 4592 and restore using xsave. The kernel will fallback to 4593 enabling legacy floating-point and sse state. 4594 4595 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 4596 register states. The kernel will fall back to use 4597 xsave to save the states. By using this parameter, 4598 performance of saving the states is degraded because 4599 xsave doesn't support modified optimization while 4600 xsaveopt supports it on xsaveopt enabled systems. 4601 4602 noxsaves [X86] Disables xsaves and xrstors used in saving and 4603 restoring x86 extended register state in compacted 4604 form of xsave area. The kernel will fall back to use 4605 xsaveopt and xrstor to save and restore the states 4606 in standard form of xsave area. By using this 4607 parameter, xsave area per process might occupy more 4608 memory on xsaves enabled systems. 4609 4610 nr_cpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 4611 could support. nr_cpus=n : n >= 1 limits the kernel to 4612 support 'n' processors. It could be larger than the 4613 number of already plugged CPU during bootup, later in 4614 runtime you can physically add extra cpu until it reaches 4615 n. So during boot up some boot time memory for per-cpu 4616 variables need be pre-allocated for later physical cpu 4617 hot plugging. 4618 4619 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 4620 4621 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86, EARLY] 4622 Disable NUMA, Only set up a single NUMA node 4623 spanning all memory. 4624 4625 numa=fake=<size>[MG] 4626 [KNL, ARM64, RISCV, X86, EARLY] 4627 If given as a memory unit, fills all system RAM with 4628 nodes of size interleaved over physical nodes. 4629 4630 numa=fake=<N> 4631 [KNL, ARM64, RISCV, X86, EARLY] 4632 If given as an integer, fills all system RAM with N 4633 fake nodes interleaved over physical nodes. 4634 4635 numa=fake=<N>U 4636 [KNL, ARM64, RISCV, X86, EARLY] 4637 If given as an integer followed by 'U', it will 4638 divide each physical node into N emulated nodes. 4639 4640 numa=noacpi [X86] Don't parse the SRAT table for NUMA setup 4641 4642 numa=nohmat [X86] Don't parse the HMAT table for NUMA setup, or 4643 soft-reserved memory partitioning. 4644 4645 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic 4646 NUMA balancing. 4647 Allowed values are enable and disable 4648 4649 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 4650 'node', 'default' can be specified 4651 This can be set from sysctl after boot. 4652 See Documentation/admin-guide/sysctl/vm.rst for details. 4653 4654 ohci1394_dma=early [HW,EARLY] enable debugging via the ohci1394 driver. 4655 See Documentation/core-api/debugging-via-ohci1394.rst for more 4656 info. 4657 4658 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 4659 Rather than timing out after 20 ms if an EC 4660 command is not properly ACKed, override the length 4661 of the timeout. We have interrupts disabled while 4662 waiting for the ACK, so if this is set too high 4663 interrupts *may* be lost! 4664 4665 omap_mux= [OMAP] Override bootloader pin multiplexing. 4666 Format: <mux_mode0.mode_name=value>... 4667 For example, to override I2C bus2: 4668 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 4669 4670 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 4671 4672 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 4673 4674 boundary - index of last SLC block on Flex-OneNAND. 4675 The remaining blocks are configured as MLC blocks. 4676 lock - Configure if Flex-OneNAND boundary should be locked. 4677 Once locked, the boundary cannot be changed. 4678 1 indicates lock status, 0 indicates unlock status. 4679 4680 oops=panic [KNL,EARLY] 4681 Always panic on oopses. Default is to just kill the 4682 process, but there is a small probability of 4683 deadlocking the machine. 4684 This will also cause panics on machine check exceptions. 4685 Useful together with panic=30 to trigger a reboot. 4686 4687 page_alloc.shuffle= 4688 [KNL] Boolean flag to control whether the page allocator 4689 should randomize its free lists. This parameter can be 4690 used to enable/disable page randomization. The state of 4691 the flag can be read from sysfs at: 4692 /sys/module/page_alloc/parameters/shuffle. 4693 This parameter is only available if CONFIG_SHUFFLE_PAGE_ALLOCATOR=y. 4694 4695 page_owner= [KNL,EARLY] Boot-time page_owner enabling option. 4696 Storage of the information about who allocated 4697 each page is disabled in default. With this switch, 4698 we can turn it on. 4699 on: enable the feature 4700 4701 page_poison= [KNL,EARLY] Boot-time parameter changing the state of 4702 poisoning on the buddy allocator, available with 4703 CONFIG_PAGE_POISONING=y. 4704 off: turn off poisoning (default) 4705 on: turn on poisoning 4706 4707 page_reporting.page_reporting_order= 4708 [KNL] Minimal page reporting order 4709 Format: <integer> 4710 Adjust the minimal page reporting order. The page 4711 reporting is disabled when it exceeds MAX_PAGE_ORDER. 4712 4713 panic= [KNL] Kernel behaviour on panic: delay <timeout> 4714 timeout > 0: seconds before rebooting 4715 timeout = 0: wait forever 4716 timeout < 0: reboot immediately 4717 Format: <timeout> 4718 4719 panic_on_taint= [KNL,EARLY] 4720 Bitmask for conditionally calling panic() in add_taint() 4721 Format: <hex>[,nousertaint] 4722 Hexadecimal bitmask representing the set of TAINT flags 4723 that will cause the kernel to panic when add_taint() is 4724 called with any of the flags in this set. 4725 The optional switch "nousertaint" can be utilized to 4726 prevent userspace forced crashes by writing to sysctl 4727 /proc/sys/kernel/tainted any flagset matching with the 4728 bitmask set on panic_on_taint. 4729 See Documentation/admin-guide/tainted-kernels.rst for 4730 extra details on the taint flags that users can pick 4731 to compose the bitmask to assign to panic_on_taint. 4732 4733 panic_on_warn=1 panic() instead of WARN(). Useful to cause kdump 4734 on a WARN(). 4735 4736 panic_print= Bitmask for printing system info when panic happens. 4737 User can chose combination of the following bits: 4738 bit 0: print all tasks info 4739 bit 1: print system memory info 4740 bit 2: print timer info 4741 bit 3: print locks info if CONFIG_LOCKDEP is on 4742 bit 4: print ftrace buffer 4743 bit 5: replay all kernel messages on consoles at the end of panic 4744 bit 6: print all CPUs backtrace (if available in the arch) 4745 bit 7: print only tasks in uninterruptible (blocked) state 4746 *Be aware* that this option may print a _lot_ of lines, 4747 so there are risks of losing older messages in the log. 4748 Use this option carefully, maybe worth to setup a 4749 bigger log buffer with "log_buf_len" along with this. 4750 4751 panic_sys_info= A comma separated list of extra information to be dumped 4752 on panic. 4753 Format: val[,val...] 4754 Where @val can be any of the following: 4755 4756 tasks: print all tasks info 4757 mem: print system memory info 4758 timers: print timers info 4759 locks: print locks info if CONFIG_LOCKDEP is on 4760 ftrace: print ftrace buffer 4761 all_bt: print all CPUs backtrace (if available in the arch) 4762 blocked_tasks: print only tasks in uninterruptible (blocked) state 4763 4764 This is a human readable alternative to the 'panic_print' option. 4765 4766 panic_console_replay 4767 When panic happens, replay all kernel messages on 4768 consoles at the end of panic. 4769 4770 parkbd.port= [HW] Parallel port number the keyboard adapter is 4771 connected to, default is 0. 4772 Format: <parport#> 4773 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 4774 0 for XT, 1 for AT (default is AT). 4775 Format: <mode> 4776 4777 parport= [HW,PPT] Specify parallel ports. 0 disables. 4778 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 4779 Use 'auto' to force the driver to use any 4780 IRQ/DMA settings detected (the default is to 4781 ignore detected IRQ/DMA settings because of 4782 possible conflicts). You can specify the base 4783 address, IRQ, and DMA settings; IRQ and DMA 4784 should be numbers, or 'auto' (for using detected 4785 settings on that particular port), or 'nofifo' 4786 (to avoid using a FIFO even if it is detected). 4787 Parallel ports are assigned in the order they 4788 are specified on the command line, starting 4789 with parport0. 4790 4791 parport_init_mode= [HW,PPT] 4792 Configure VIA parallel port to operate in 4793 a specific mode. This is necessary on Pegasos 4794 computer where firmware has no options for setting 4795 up parallel port mode and sets it to spp. 4796 Currently this function knows 686a and 8231 chips. 4797 Format: [spp|ps2|epp|ecp|ecpepp] 4798 4799 pata_legacy.all= [HW,LIBATA] 4800 Format: <int> 4801 Set to non-zero to probe primary and secondary ISA 4802 port ranges on PCI systems where no PCI PATA device 4803 has been found at either range. Disabled by default. 4804 4805 pata_legacy.autospeed= [HW,LIBATA] 4806 Format: <int> 4807 Set to non-zero if a chip is present that snoops speed 4808 changes. Disabled by default. 4809 4810 pata_legacy.ht6560a= [HW,LIBATA] 4811 Format: <int> 4812 Set to 1, 2, or 3 for HT 6560A on the primary channel, 4813 the secondary channel, or both channels respectively. 4814 Disabled by default. 4815 4816 pata_legacy.ht6560b= [HW,LIBATA] 4817 Format: <int> 4818 Set to 1, 2, or 3 for HT 6560B on the primary channel, 4819 the secondary channel, or both channels respectively. 4820 Disabled by default. 4821 4822 pata_legacy.iordy_mask= [HW,LIBATA] 4823 Format: <int> 4824 IORDY enable mask. Set individual bits to allow IORDY 4825 for the respective channel. Bit 0 is for the first 4826 legacy channel handled by this driver, bit 1 is for 4827 the second channel, and so on. The sequence will often 4828 correspond to the primary legacy channel, the secondary 4829 legacy channel, and so on, but the handling of a PCI 4830 bus and the use of other driver options may interfere 4831 with the sequence. By default IORDY is allowed across 4832 all channels. 4833 4834 pata_legacy.opti82c46x= [HW,LIBATA] 4835 Format: <int> 4836 Set to 1, 2, or 3 for Opti 82c611A on the primary 4837 channel, the secondary channel, or both channels 4838 respectively. Disabled by default. 4839 4840 pata_legacy.opti82c611a= [HW,LIBATA] 4841 Format: <int> 4842 Set to 1, 2, or 3 for Opti 82c465MV on the primary 4843 channel, the secondary channel, or both channels 4844 respectively. Disabled by default. 4845 4846 pata_legacy.pio_mask= [HW,LIBATA] 4847 Format: <int> 4848 PIO mode mask for autospeed devices. Set individual 4849 bits to allow the use of the respective PIO modes. 4850 Bit 0 is for mode 0, bit 1 is for mode 1, and so on. 4851 All modes allowed by default. 4852 4853 pata_legacy.probe_all= [HW,LIBATA] 4854 Format: <int> 4855 Set to non-zero to probe tertiary and further ISA 4856 port ranges on PCI systems. Disabled by default. 4857 4858 pata_legacy.probe_mask= [HW,LIBATA] 4859 Format: <int> 4860 Probe mask for legacy ISA PATA ports. Depending on 4861 platform configuration and the use of other driver 4862 options up to 6 legacy ports are supported: 0x1f0, 4863 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing 4864 of individual ports can be disabled by setting the 4865 corresponding bits in the mask to 1. Bit 0 is for 4866 the first port in the list above (0x1f0), and so on. 4867 By default all supported ports are probed. 4868 4869 pata_legacy.qdi= [HW,LIBATA] 4870 Format: <int> 4871 Set to non-zero to probe QDI controllers. By default 4872 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise. 4873 4874 pata_legacy.winbond= [HW,LIBATA] 4875 Format: <int> 4876 Set to non-zero to probe Winbond controllers. Use 4877 the standard I/O port (0x130) if 1, otherwise the 4878 value given is the I/O port to use (typically 0x1b0). 4879 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE, 4880 0 otherwise. 4881 4882 pata_platform.pio_mask= [HW,LIBATA] 4883 Format: <int> 4884 Supported PIO mode mask. Set individual bits to allow 4885 the use of the respective PIO modes. Bit 0 is for 4886 mode 0, bit 1 is for mode 1, and so on. Mode 0 only 4887 allowed by default. 4888 4889 pause_on_oops=<int> 4890 Halt all CPUs after the first oops has been printed for 4891 the specified number of seconds. This is to be used if 4892 your oopses keep scrolling off the screen. 4893 4894 pcbit= [HW,ISDN] 4895 4896 pci=option[,option...] [PCI,EARLY] various PCI subsystem options. 4897 4898 Some options herein operate on a specific device 4899 or a set of devices (<pci_dev>). These are 4900 specified in one of the following formats: 4901 4902 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 4903 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 4904 4905 Note: the first format specifies a PCI 4906 bus/device/function address which may change 4907 if new hardware is inserted, if motherboard 4908 firmware changes, or due to changes caused 4909 by other kernel parameters. If the 4910 domain is left unspecified, it is 4911 taken to be zero. Optionally, a path 4912 to a device through multiple device/function 4913 addresses can be specified after the base 4914 address (this is more robust against 4915 renumbering issues). The second format 4916 selects devices using IDs from the 4917 configuration space which may match multiple 4918 devices in the system. 4919 4920 earlydump dump PCI config space before the kernel 4921 changes anything 4922 off [X86] don't probe for the PCI bus 4923 bios [X86-32] force use of PCI BIOS, don't access 4924 the hardware directly. Use this if your machine 4925 has a non-standard PCI host bridge. 4926 nobios [X86-32] disallow use of PCI BIOS, only direct 4927 hardware access methods are allowed. Use this 4928 if you experience crashes upon bootup and you 4929 suspect they are caused by the BIOS. 4930 conf1 [X86] Force use of PCI Configuration Access 4931 Mechanism 1 (config address in IO port 0xCF8, 4932 data in IO port 0xCFC, both 32-bit). 4933 conf2 [X86] Force use of PCI Configuration Access 4934 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 4935 the function, IO port 0xCFA, also 8-bit, sets 4936 bus number. The config space is then accessed 4937 through ports 0xC000-0xCFFF). 4938 See http://wiki.osdev.org/PCI for more info 4939 on the configuration access mechanisms. 4940 noaer [PCIE] If the PCIEAER kernel config parameter is 4941 enabled, this kernel boot option can be used to 4942 disable the use of PCIE advanced error reporting. 4943 nodomains [PCI] Disable support for multiple PCI 4944 root domains (aka PCI segments, in ACPI-speak). 4945 nommconf [X86] Disable use of MMCONFIG for PCI 4946 Configuration 4947 check_enable_amd_mmconf [X86] check for and enable 4948 properly configured MMIO access to PCI 4949 config space on AMD family 10h CPU 4950 nomsi [MSI] If the PCI_MSI kernel config parameter is 4951 enabled, this kernel boot option can be used to 4952 disable the use of MSI interrupts system-wide. 4953 noioapicquirk [APIC] Disable all boot interrupt quirks. 4954 Safety option to keep boot IRQs enabled. This 4955 should never be necessary. 4956 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 4957 primary IO-APIC for bridges that cannot disable 4958 boot IRQs. This fixes a source of spurious IRQs 4959 when the system masks IRQs. 4960 noioapicreroute [APIC] Disable workaround that uses the 4961 boot IRQ equivalent of an IRQ that connects to 4962 a chipset where boot IRQs cannot be disabled. 4963 The opposite of ioapicreroute. 4964 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 4965 routing table. These calls are known to be buggy 4966 on several machines and they hang the machine 4967 when used, but on other computers it's the only 4968 way to get the interrupt routing table. Try 4969 this option if the kernel is unable to allocate 4970 IRQs or discover secondary PCI buses on your 4971 motherboard. 4972 rom [X86] Assign address space to expansion ROMs. 4973 Use with caution as certain devices share 4974 address decoders between ROMs and other 4975 resources. 4976 norom [X86] Do not assign address space to 4977 expansion ROMs that do not already have 4978 BIOS assigned address ranges. 4979 nobar [X86] Do not assign address space to the 4980 BARs that weren't assigned by the BIOS. 4981 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 4982 assigned automatically to PCI devices. You can 4983 make the kernel exclude IRQs of your ISA cards 4984 this way. 4985 pirqaddr=0xAAAAA [X86] Specify the physical address 4986 of the PIRQ table (normally generated 4987 by the BIOS) if it is outside the 4988 F0000h-100000h range. 4989 lastbus=N [X86] Scan all buses thru bus #N. Can be 4990 useful if the kernel is unable to find your 4991 secondary buses and you want to tell it 4992 explicitly which ones they are. 4993 assign-busses [X86] Always assign all PCI bus 4994 numbers ourselves, overriding 4995 whatever the firmware may have done. 4996 usepirqmask [X86] Honor the possible IRQ mask stored 4997 in the BIOS $PIR table. This is needed on 4998 some systems with broken BIOSes, notably 4999 some HP Pavilion N5400 and Omnibook XE3 5000 notebooks. This will have no effect if ACPI 5001 IRQ routing is enabled. 5002 noacpi [X86] Do not use ACPI for IRQ routing 5003 or for PCI scanning. 5004 use_crs [X86] Use PCI host bridge window information 5005 from ACPI. On BIOSes from 2008 or later, this 5006 is enabled by default. If you need to use this, 5007 please report a bug. 5008 nocrs [X86] Ignore PCI host bridge windows from ACPI. 5009 If you need to use this, please report a bug. 5010 use_e820 [X86] Use E820 reservations to exclude parts of 5011 PCI host bridge windows. This is a workaround 5012 for BIOS defects in host bridge _CRS methods. 5013 If you need to use this, please report a bug to 5014 <linux-pci@vger.kernel.org>. 5015 no_e820 [X86] Ignore E820 reservations for PCI host 5016 bridge windows. This is the default on modern 5017 hardware. If you need to use this, please report 5018 a bug to <linux-pci@vger.kernel.org>. 5019 routeirq Do IRQ routing for all PCI devices. 5020 This is normally done in pci_enable_device(), 5021 so this option is a temporary workaround 5022 for broken drivers that don't call it. 5023 skip_isa_align [X86] do not align io start addr, so can 5024 handle more pci cards 5025 noearly [X86] Don't do any early type 1 scanning. 5026 This might help on some broken boards which 5027 machine check when some devices' config space 5028 is read. But various workarounds are disabled 5029 and some IOMMU drivers will not work. 5030 bfsort Sort PCI devices into breadth-first order. 5031 This sorting is done to get a device 5032 order compatible with older (<= 2.4) kernels. 5033 nobfsort Don't sort PCI devices into breadth-first order. 5034 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 5035 tuning and use the BIOS-configured MPS defaults. 5036 pcie_bus_safe Set every device's MPS to the largest value 5037 supported by all devices below the root complex. 5038 pcie_bus_perf Set device MPS to the largest allowable MPS 5039 based on its parent bus. Also set MRRS (Max 5040 Read Request Size) to the largest supported 5041 value (no larger than the MPS that the device 5042 or bus can support) for best performance. 5043 pcie_bus_peer2peer Set every device's MPS to 128B, which 5044 every device is guaranteed to support. This 5045 configuration allows peer-to-peer DMA between 5046 any pair of devices, possibly at the cost of 5047 reduced performance. This also guarantees 5048 that hot-added devices will work. 5049 cbiosize=nn[KMG] The fixed amount of bus space which is 5050 reserved for the CardBus bridge's IO window. 5051 The default value is 256 bytes. 5052 cbmemsize=nn[KMG] The fixed amount of bus space which is 5053 reserved for the CardBus bridge's memory 5054 window. The default value is 64 megabytes. 5055 resource_alignment= 5056 Format: 5057 [<order of align>@]<pci_dev>[; ...] 5058 Specifies alignment and device to reassign 5059 aligned memory resources. How to 5060 specify the device is described above. 5061 If <order of align> is not specified, 5062 PAGE_SIZE is used as alignment. 5063 A PCI-PCI bridge can be specified if resource 5064 windows need to be expanded. 5065 To specify the alignment for several 5066 instances of a device, the PCI vendor, 5067 device, subvendor, and subdevice may be 5068 specified, e.g., 12@pci:8086:9c22:103c:198f 5069 for 4096-byte alignment. 5070 ecrc= Enable/disable PCIe ECRC (transaction layer 5071 end-to-end CRC checking). Only effective if 5072 OS has native AER control (either granted by 5073 ACPI _OSC or forced via "pcie_ports=native") 5074 bios: Use BIOS/firmware settings. This is the 5075 the default. 5076 off: Turn ECRC off 5077 on: Turn ECRC on. 5078 hpiosize=nn[KMG] The fixed amount of bus space which is 5079 reserved for hotplug bridge's IO window. 5080 Default size is 256 bytes. 5081 hpmmiosize=nn[KMG] The fixed amount of bus space which is 5082 reserved for hotplug bridge's MMIO window. 5083 Default size is 2 megabytes. 5084 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 5085 reserved for hotplug bridge's MMIO_PREF window. 5086 Default size is 2 megabytes. 5087 hpmemsize=nn[KMG] The fixed amount of bus space which is 5088 reserved for hotplug bridge's MMIO and 5089 MMIO_PREF window. 5090 Default size is 2 megabytes. 5091 hpbussize=nn The minimum amount of additional bus numbers 5092 reserved for buses below a hotplug bridge. 5093 Default is 1. 5094 realloc= Enable/disable reallocating PCI bridge resources 5095 if allocations done by BIOS are too small to 5096 accommodate resources required by all child 5097 devices. 5098 off: Turn realloc off 5099 on: Turn realloc on 5100 realloc same as realloc=on 5101 noari do not use PCIe ARI. 5102 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 5103 do not use PCIe ATS (and IOMMU device IOTLB). 5104 pcie_scan_all Scan all possible PCIe devices. Otherwise we 5105 only look for one device below a PCIe downstream 5106 port. 5107 big_root_window Try to add a big 64bit memory window to the PCIe 5108 root complex on AMD CPUs. Some GFX hardware 5109 can resize a BAR to allow access to all VRAM. 5110 Adding the window is slightly risky (it may 5111 conflict with unreported devices), so this 5112 taints the kernel. 5113 disable_acs_redir=<pci_dev>[; ...] 5114 Specify one or more PCI devices (in the format 5115 specified above) separated by semicolons. 5116 Each device specified will have the PCI ACS 5117 redirect capabilities forced off which will 5118 allow P2P traffic between devices through 5119 bridges without forcing it upstream. Note: 5120 this removes isolation between devices and 5121 may put more devices in an IOMMU group. 5122 config_acs= 5123 Format: 5124 <ACS flags>@<pci_dev>[; ...] 5125 Specify one or more PCI devices (in the format 5126 specified above) optionally prepended with flags 5127 and separated by semicolons. The respective 5128 capabilities will be enabled, disabled or 5129 unchanged based on what is specified in 5130 flags. 5131 5132 ACS Flags is defined as follows: 5133 bit-0 : ACS Source Validation 5134 bit-1 : ACS Translation Blocking 5135 bit-2 : ACS P2P Request Redirect 5136 bit-3 : ACS P2P Completion Redirect 5137 bit-4 : ACS Upstream Forwarding 5138 bit-5 : ACS P2P Egress Control 5139 bit-6 : ACS Direct Translated P2P 5140 Each bit can be marked as: 5141 '0' – force disabled 5142 '1' – force enabled 5143 'x' – unchanged 5144 For example, 5145 pci=config_acs=10x@pci:0:0 5146 would configure all devices that support 5147 ACS to enable P2P Request Redirect, disable 5148 Translation Blocking, and leave Source 5149 Validation unchanged from whatever power-up 5150 or firmware set it to. 5151 5152 Note: this may remove isolation between devices 5153 and may put more devices in an IOMMU group. 5154 force_floating [S390] Force usage of floating interrupts. 5155 nomio [S390] Do not use MIO instructions. 5156 norid [S390] ignore the RID field and force use of 5157 one PCI domain per PCI function 5158 notph [PCIE] If the PCIE_TPH kernel config parameter 5159 is enabled, this kernel boot option can be used 5160 to disable PCIe TLP Processing Hints support 5161 system-wide. 5162 5163 pcie_aspm= [PCIE] Forcibly enable or ignore PCIe Active State Power 5164 Management. 5165 off Don't touch ASPM configuration at all. Leave any 5166 configuration done by firmware unchanged. 5167 force Enable ASPM even on devices that claim not to support it. 5168 WARNING: Forcing ASPM on may cause system lockups. 5169 5170 pcie_ports= [PCIE] PCIe port services handling: 5171 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 5172 even if the platform doesn't give the OS permission to 5173 use them. This may cause conflicts if the platform 5174 also tries to use these services. 5175 dpc-native Use native PCIe service for DPC only. May 5176 cause conflicts if firmware uses AER or DPC. 5177 compat Disable native PCIe services (PME, AER, DPC, PCIe 5178 hotplug). 5179 5180 pcie_port_pm= [PCIE] PCIe port power management handling: 5181 off Disable power management of all PCIe ports 5182 force Forcibly enable power management of all PCIe ports 5183 5184 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 5185 nomsi Do not use MSI for native PCIe PME signaling (this makes 5186 all PCIe root ports use INTx for all services). 5187 5188 pcmv= [HW,PCMCIA] BadgePAD 4 5189 5190 pd_ignore_unused 5191 [PM] 5192 Keep all power-domains already enabled by bootloader on, 5193 even if no driver has claimed them. This is useful 5194 for debug and development, but should not be 5195 needed on a platform with proper driver support. 5196 5197 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 5198 boot time. 5199 Format: { 0 | 1 } 5200 See arch/parisc/kernel/pdc_chassis.c 5201 5202 percpu_alloc= [MM,EARLY] 5203 Select which percpu first chunk allocator to use. 5204 Currently supported values are "embed" and "page". 5205 Archs may support subset or none of the selections. 5206 See comments in mm/percpu.c for details on each 5207 allocator. This parameter is primarily for debugging 5208 and performance comparison. 5209 5210 pirq= [SMP,APIC] Manual mp-table setup 5211 See Documentation/arch/x86/i386/IO-APIC.rst. 5212 5213 plip= [PPT,NET] Parallel port network link 5214 Format: { parport<nr> | timid | 0 } 5215 See also Documentation/admin-guide/parport.rst. 5216 5217 pmtmr= [X86] Manual setup of pmtmr I/O Port. 5218 Override pmtimer IOPort with a hex value. 5219 e.g. pmtmr=0x508 5220 5221 pmu_override= [PPC] Override the PMU. 5222 This option takes over the PMU facility, so it is no 5223 longer usable by perf. Setting this option starts the 5224 PMU counters by setting MMCR0 to 0 (the FC bit is 5225 cleared). If a number is given, then MMCR1 is set to 5226 that number, otherwise (e.g., 'pmu_override=on'), MMCR1 5227 remains 0. 5228 5229 pm_async= [PM] 5230 Format: off 5231 This parameter sets the initial value of the 5232 /sys/power/pm_async sysfs knob at boot time. 5233 If set to "off", disables asynchronous suspend and 5234 resume of devices during system-wide power transitions. 5235 This can be useful on platforms where device 5236 dependencies are not well-defined, or for debugging 5237 power management issues. Asynchronous operations are 5238 enabled by default. 5239 5240 5241 pm_debug_messages [SUSPEND,KNL] 5242 Enable suspend/resume debug messages during boot up. 5243 5244 pnp.debug=1 [PNP] 5245 Enable PNP debug messages (depends on the 5246 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 5247 via /sys/module/pnp/parameters/debug. We always show 5248 current resource usage; turning this on also shows 5249 possible settings and some assignment information. 5250 5251 pnpacpi= [ACPI] 5252 { off } 5253 5254 pnpbios= [ISAPNP] 5255 { on | off | curr | res | no-curr | no-res } 5256 5257 pnp_reserve_irq= 5258 [ISAPNP] Exclude IRQs for the autoconfiguration 5259 5260 pnp_reserve_dma= 5261 [ISAPNP] Exclude DMAs for the autoconfiguration 5262 5263 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 5264 Ranges are in pairs (I/O port base and size). 5265 5266 pnp_reserve_mem= 5267 [ISAPNP] Exclude memory regions for the 5268 autoconfiguration. 5269 Ranges are in pairs (memory base and size). 5270 5271 ports= [IP_VS_FTP] IPVS ftp helper module 5272 Default is 21. 5273 Up to 8 (IP_VS_APP_MAX_PORTS) ports 5274 may be specified. 5275 Format: <port>,<port>.... 5276 5277 possible_cpus= [SMP,S390,X86] 5278 Format: <unsigned int> 5279 Set the number of possible CPUs, overriding the 5280 regular discovery mechanisms (such as ACPI/FW, etc). 5281 5282 powersave=off [PPC] This option disables power saving features. 5283 It specifically disables cpuidle and sets the 5284 platform machine description specific power_save 5285 function to NULL. On Idle the CPU just reduces 5286 execution priority. 5287 5288 ppc_strict_facility_enable 5289 [PPC,ENABLE] This option catches any kernel floating point, 5290 Altivec, VSX and SPE outside of regions specifically 5291 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 5292 There is some performance impact when enabling this. 5293 5294 ppc_tm= [PPC,EARLY] 5295 Format: {"off"} 5296 Disable Hardware Transactional Memory 5297 5298 preempt= [KNL] 5299 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC 5300 none - Limited to cond_resched() calls 5301 voluntary - Limited to cond_resched() and might_sleep() calls 5302 full - Any section that isn't explicitly preempt disabled 5303 can be preempted anytime. Tasks will also yield 5304 contended spinlocks (if the critical section isn't 5305 explicitly preempt disabled beyond the lock itself). 5306 lazy - Scheduler controlled. Similar to full but instead 5307 of preempting the task immediately, the task gets 5308 one HZ tick time to yield itself before the 5309 preemption will be forced. One preemption is when the 5310 task returns to user space. 5311 5312 print-fatal-signals= 5313 [KNL] debug: print fatal signals 5314 5315 If enabled, warn about various signal handling 5316 related application anomalies: too many signals, 5317 too many POSIX.1 timers, fatal signals causing a 5318 coredump - etc. 5319 5320 If you hit the warning due to signal overflow, 5321 you might want to try "ulimit -i unlimited". 5322 5323 default: off. 5324 5325 printk.always_kmsg_dump= 5326 Trigger kmsg_dump for cases other than kernel oops or 5327 panics 5328 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5329 default: disabled 5330 5331 printk.console_no_auto_verbose= 5332 Disable console loglevel raise on oops, panic 5333 or lockdep-detected issues (only if lock debug is on). 5334 With an exception to setups with low baudrate on 5335 serial console, keeping this 0 is a good choice 5336 in order to provide more debug information. 5337 Format: <bool> 5338 default: 0 (auto_verbose is enabled) 5339 5340 printk.debug_non_panic_cpus= 5341 Allows storing messages from non-panic CPUs into 5342 the printk log buffer during panic(). They are 5343 flushed to consoles by the panic-CPU on 5344 a best-effort basis. 5345 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5346 Default: disabled 5347 5348 printk.devkmsg={on,off,ratelimit} 5349 Control writing to /dev/kmsg. 5350 on - unlimited logging to /dev/kmsg from userspace 5351 off - logging to /dev/kmsg disabled 5352 ratelimit - ratelimit the logging 5353 Default: ratelimit 5354 5355 printk.time= Show timing data prefixed to each printk message line 5356 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5357 5358 proc_mem.force_override= [KNL] 5359 Format: {always | ptrace | never} 5360 Traditionally /proc/pid/mem allows memory permissions to be 5361 overridden without restrictions. This option may be set to 5362 restrict that. Can be one of: 5363 - 'always': traditional behavior always allows mem overrides. 5364 - 'ptrace': only allow mem overrides for active ptracers. 5365 - 'never': never allow mem overrides. 5366 If not specified, default is the CONFIG_PROC_MEM_* choice. 5367 5368 processor.max_cstate= [HW,ACPI] 5369 Limit processor to maximum C-state 5370 max_cstate=9 overrides any DMI blacklist limit. 5371 5372 processor.nocst [HW,ACPI] 5373 Ignore the _CST method to determine C-states, 5374 instead using the legacy FADT method 5375 5376 profile= [KNL] Enable kernel profiling via /proc/profile 5377 Format: [<profiletype>,]<number> 5378 Param: <profiletype>: "schedule" or "kvm" 5379 [defaults to kernel profiling] 5380 Param: "schedule" - profile schedule points. 5381 Param: "kvm" - profile VM exits. 5382 Param: <number> - step/bucket size as a power of 2 for 5383 statistical time based profiling. 5384 5385 prompt_ramdisk= [RAM] [Deprecated] 5386 5387 prot_virt= [S390] enable hosting protected virtual machines 5388 isolated from the hypervisor (if hardware supports 5389 that). If enabled, the default kernel base address 5390 might be overridden even when Kernel Address Space 5391 Layout Randomization is disabled. 5392 Format: <bool> 5393 5394 psi= [KNL] Enable or disable pressure stall information 5395 tracking. 5396 Format: <bool> 5397 5398 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 5399 probe for; one of (bare|imps|exps|lifebook|any). 5400 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 5401 per second. 5402 psmouse.resetafter= [HW,MOUSE] 5403 Try to reset the device after so many bad packets 5404 (0 = never). 5405 psmouse.resolution= 5406 [HW,MOUSE] Set desired mouse resolution, in dpi. 5407 psmouse.smartscroll= 5408 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 5409 0 = disabled, 1 = enabled (default). 5410 5411 pstore.backend= Specify the name of the pstore backend to use 5412 5413 pti= [X86-64] Control Page Table Isolation of user and 5414 kernel address spaces. Disabling this feature 5415 removes hardening, but improves performance of 5416 system calls and interrupts. 5417 5418 on - unconditionally enable 5419 off - unconditionally disable 5420 auto - kernel detects whether your CPU model is 5421 vulnerable to issues that PTI mitigates 5422 5423 Not specifying this option is equivalent to pti=auto. 5424 5425 pty.legacy_count= 5426 [KNL] Number of legacy pty's. Overwrites compiled-in 5427 default number. 5428 5429 quiet [KNL,EARLY] Disable most log messages 5430 5431 r128= [HW,DRM] 5432 5433 radix_hcall_invalidate=on [PPC/PSERIES] 5434 Disable RADIX GTSE feature and use hcall for TLB 5435 invalidate. 5436 5437 raid= [HW,RAID] 5438 See Documentation/admin-guide/md.rst. 5439 5440 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 5441 See Documentation/admin-guide/blockdev/ramdisk.rst. 5442 5443 ramdisk_start= [RAM] RAM disk image start address 5444 5445 random.trust_cpu=off 5446 [KNL,EARLY] Disable trusting the use of the CPU's 5447 random number generator (if available) to 5448 initialize the kernel's RNG. 5449 5450 random.trust_bootloader=off 5451 [KNL,EARLY] Disable trusting the use of the a seed 5452 passed by the bootloader (if available) to 5453 initialize the kernel's RNG. 5454 5455 randomize_kstack_offset= 5456 [KNL,EARLY] Enable or disable kernel stack offset 5457 randomization, which provides roughly 5 bits of 5458 entropy, frustrating memory corruption attacks 5459 that depend on stack address determinism or 5460 cross-syscall address exposures. This is only 5461 available on architectures that have defined 5462 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET. 5463 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5464 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT. 5465 5466 ras=option[,option,...] [KNL] RAS-specific options 5467 5468 cec_disable [X86] 5469 Disable the Correctable Errors Collector, 5470 see CONFIG_RAS_CEC help text. 5471 5472 rcu_nocbs[=cpu-list] 5473 [KNL] The optional argument is a cpu list, 5474 as described above. 5475 5476 In kernels built with CONFIG_RCU_NOCB_CPU=y, 5477 enable the no-callback CPU mode, which prevents 5478 such CPUs' callbacks from being invoked in 5479 softirq context. Invocation of such CPUs' RCU 5480 callbacks will instead be offloaded to "rcuox/N" 5481 kthreads created for that purpose, where "x" is 5482 "p" for RCU-preempt, "s" for RCU-sched, and "g" 5483 for the kthreads that mediate grace periods; and 5484 "N" is the CPU number. This reduces OS jitter on 5485 the offloaded CPUs, which can be useful for HPC 5486 and real-time workloads. It can also improve 5487 energy efficiency for asymmetric multiprocessors. 5488 5489 If a cpulist is passed as an argument, the specified 5490 list of CPUs is set to no-callback mode from boot. 5491 5492 Otherwise, if the '=' sign and the cpulist 5493 arguments are omitted, no CPU will be set to 5494 no-callback mode from boot but the mode may be 5495 toggled at runtime via cpusets. 5496 5497 Note that this argument takes precedence over 5498 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 5499 5500 rcu_nocb_poll [KNL] 5501 Rather than requiring that offloaded CPUs 5502 (specified by rcu_nocbs= above) explicitly 5503 awaken the corresponding "rcuoN" kthreads, 5504 make these kthreads poll for callbacks. 5505 This improves the real-time response for the 5506 offloaded CPUs by relieving them of the need to 5507 wake up the corresponding kthread, but degrades 5508 energy efficiency by requiring that the kthreads 5509 periodically wake up to do the polling. 5510 5511 rcutree.blimit= [KNL] 5512 Set maximum number of finished RCU callbacks to 5513 process in one batch. 5514 5515 rcutree.csd_lock_suppress_rcu_stall= [KNL] 5516 Do only a one-line RCU CPU stall warning when 5517 there is an ongoing too-long CSD-lock wait. 5518 5519 rcutree.do_rcu_barrier= [KNL] 5520 Request a call to rcu_barrier(). This is 5521 throttled so that userspace tests can safely 5522 hammer on the sysfs variable if they so choose. 5523 If triggered before the RCU grace-period machinery 5524 is fully active, this will error out with EAGAIN. 5525 5526 rcutree.dump_tree= [KNL] 5527 Dump the structure of the rcu_node combining tree 5528 out at early boot. This is used for diagnostic 5529 purposes, to verify correct tree setup. 5530 5531 rcutree.gp_cleanup_delay= [KNL] 5532 Set the number of jiffies to delay each step of 5533 RCU grace-period cleanup. 5534 5535 rcutree.gp_init_delay= [KNL] 5536 Set the number of jiffies to delay each step of 5537 RCU grace-period initialization. 5538 5539 rcutree.gp_preinit_delay= [KNL] 5540 Set the number of jiffies to delay each step of 5541 RCU grace-period pre-initialization, that is, 5542 the propagation of recent CPU-hotplug changes up 5543 the rcu_node combining tree. 5544 5545 rcutree.jiffies_till_first_fqs= [KNL] 5546 Set delay from grace-period initialization to 5547 first attempt to force quiescent states. 5548 Units are jiffies, minimum value is zero, 5549 and maximum value is HZ. 5550 5551 rcutree.jiffies_till_next_fqs= [KNL] 5552 Set delay between subsequent attempts to force 5553 quiescent states. Units are jiffies, minimum 5554 value is one, and maximum value is HZ. 5555 5556 rcutree.jiffies_till_sched_qs= [KNL] 5557 Set required age in jiffies for a 5558 given grace period before RCU starts 5559 soliciting quiescent-state help from 5560 rcu_note_context_switch() and cond_resched(). 5561 If not specified, the kernel will calculate 5562 a value based on the most recent settings 5563 of rcutree.jiffies_till_first_fqs 5564 and rcutree.jiffies_till_next_fqs. 5565 This calculated value may be viewed in 5566 rcutree.jiffies_to_sched_qs. Any attempt to set 5567 rcutree.jiffies_to_sched_qs will be cheerfully 5568 overwritten. 5569 5570 rcutree.kthread_prio= [KNL,BOOT] 5571 Set the SCHED_FIFO priority of the RCU per-CPU 5572 kthreads (rcuc/N). This value is also used for 5573 the priority of the RCU boost threads (rcub/N) 5574 and for the RCU grace-period kthreads (rcu_bh, 5575 rcu_preempt, and rcu_sched). If RCU_BOOST is 5576 set, valid values are 1-99 and the default is 1 5577 (the least-favored priority). Otherwise, when 5578 RCU_BOOST is not set, valid values are 0-99 and 5579 the default is zero (non-realtime operation). 5580 When RCU_NOCB_CPU is set, also adjust the 5581 priority of NOCB callback kthreads. 5582 5583 rcutree.nocb_nobypass_lim_per_jiffy= [KNL] 5584 On callback-offloaded (rcu_nocbs) CPUs, 5585 RCU reduces the lock contention that would 5586 otherwise be caused by callback floods through 5587 use of the ->nocb_bypass list. However, in the 5588 common non-flooded case, RCU queues directly to 5589 the main ->cblist in order to avoid the extra 5590 overhead of the ->nocb_bypass list and its lock. 5591 But if there are too many callbacks queued during 5592 a single jiffy, RCU pre-queues the callbacks into 5593 the ->nocb_bypass queue. The definition of "too 5594 many" is supplied by this kernel boot parameter. 5595 5596 rcutree.nohz_full_patience_delay= [KNL] 5597 On callback-offloaded (rcu_nocbs) CPUs, avoid 5598 disturbing RCU unless the grace period has 5599 reached the specified age in milliseconds. 5600 Defaults to zero. Large values will be capped 5601 at five seconds. All values will be rounded down 5602 to the nearest value representable by jiffies. 5603 5604 rcutree.qhimark= [KNL] 5605 Set threshold of queued RCU callbacks beyond which 5606 batch limiting is disabled. 5607 5608 rcutree.qlowmark= [KNL] 5609 Set threshold of queued RCU callbacks below which 5610 batch limiting is re-enabled. 5611 5612 rcutree.qovld= [KNL] 5613 Set threshold of queued RCU callbacks beyond which 5614 RCU's force-quiescent-state scan will aggressively 5615 enlist help from cond_resched() and sched IPIs to 5616 help CPUs more quickly reach quiescent states. 5617 Set to less than zero to make this be set based 5618 on rcutree.qhimark at boot time and to zero to 5619 disable more aggressive help enlistment. 5620 5621 rcutree.rcu_delay_page_cache_fill_msec= [KNL] 5622 Set the page-cache refill delay (in milliseconds) 5623 in response to low-memory conditions. The range 5624 of permitted values is in the range 0:100000. 5625 5626 rcutree.rcu_divisor= [KNL] 5627 Set the shift-right count to use to compute 5628 the callback-invocation batch limit bl from 5629 the number of callbacks queued on this CPU. 5630 The result will be bounded below by the value of 5631 the rcutree.blimit kernel parameter. Every bl 5632 callbacks, the softirq handler will exit in 5633 order to allow the CPU to do other work. 5634 5635 Please note that this callback-invocation batch 5636 limit applies only to non-offloaded callback 5637 invocation. Offloaded callbacks are instead 5638 invoked in the context of an rcuoc kthread, which 5639 scheduler will preempt as it does any other task. 5640 5641 rcutree.rcu_fanout_exact= [KNL] 5642 Disable autobalancing of the rcu_node combining 5643 tree. This is used by rcutorture, and might 5644 possibly be useful for architectures having high 5645 cache-to-cache transfer latencies. 5646 5647 rcutree.rcu_fanout_leaf= [KNL] 5648 Change the number of CPUs assigned to each 5649 leaf rcu_node structure. Useful for very 5650 large systems, which will choose the value 64, 5651 and for NUMA systems with large remote-access 5652 latencies, which will choose a value aligned 5653 with the appropriate hardware boundaries. 5654 5655 rcutree.rcu_min_cached_objs= [KNL] 5656 Minimum number of objects which are cached and 5657 maintained per one CPU. Object size is equal 5658 to PAGE_SIZE. The cache allows to reduce the 5659 pressure to page allocator, also it makes the 5660 whole algorithm to behave better in low memory 5661 condition. 5662 5663 rcutree.rcu_nocb_gp_stride= [KNL] 5664 Set the number of NOCB callback kthreads in 5665 each group, which defaults to the square root 5666 of the number of CPUs. Larger numbers reduce 5667 the wakeup overhead on the global grace-period 5668 kthread, but increases that same overhead on 5669 each group's NOCB grace-period kthread. 5670 5671 rcutree.rcu_kick_kthreads= [KNL] 5672 Cause the grace-period kthread to get an extra 5673 wake_up() if it sleeps three times longer than 5674 it should at force-quiescent-state time. 5675 This wake_up() will be accompanied by a 5676 WARN_ONCE() splat and an ftrace_dump(). 5677 5678 rcutree.rcu_resched_ns= [KNL] 5679 Limit the time spend invoking a batch of RCU 5680 callbacks to the specified number of nanoseconds. 5681 By default, this limit is checked only once 5682 every 32 callbacks in order to limit the pain 5683 inflicted by local_clock() overhead. 5684 5685 rcutree.rcu_unlock_delay= [KNL] 5686 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 5687 this specifies an rcu_read_unlock()-time delay 5688 in microseconds. This defaults to zero. 5689 Larger delays increase the probability of 5690 catching RCU pointer leaks, that is, buggy use 5691 of RCU-protected pointers after the relevant 5692 rcu_read_unlock() has completed. 5693 5694 rcutree.sysrq_rcu= [KNL] 5695 Commandeer a sysrq key to dump out Tree RCU's 5696 rcu_node tree with an eye towards determining 5697 why a new grace period has not yet started. 5698 5699 rcutree.use_softirq= [KNL] 5700 If set to zero, move all RCU_SOFTIRQ processing to 5701 per-CPU rcuc kthreads. Defaults to a non-zero 5702 value, meaning that RCU_SOFTIRQ is used by default. 5703 Specify rcutree.use_softirq=0 to use rcuc kthreads. 5704 5705 But note that CONFIG_PREEMPT_RT=y kernels disable 5706 this kernel boot parameter, forcibly setting it 5707 to zero. 5708 5709 rcutree.enable_rcu_lazy= [KNL] 5710 To save power, batch RCU callbacks and flush after 5711 delay, memory pressure or callback list growing too 5712 big. 5713 5714 rcutree.rcu_normal_wake_from_gp= [KNL] 5715 Reduces a latency of synchronize_rcu() call. This approach 5716 maintains its own track of synchronize_rcu() callers, so it 5717 does not interact with regular callbacks because it does not 5718 use a call_rcu[_hurry]() path. Please note, this is for a 5719 normal grace period. 5720 5721 How to enable it: 5722 5723 echo 1 > /sys/module/rcutree/parameters/rcu_normal_wake_from_gp 5724 or pass a boot parameter "rcutree.rcu_normal_wake_from_gp=1" 5725 5726 Default is 1 if num_possible_cpus() <= 16 and it is not explicitly 5727 disabled by the boot parameter passing 0. 5728 5729 rcuscale.gp_async= [KNL] 5730 Measure performance of asynchronous 5731 grace-period primitives such as call_rcu(). 5732 5733 rcuscale.gp_async_max= [KNL] 5734 Specify the maximum number of outstanding 5735 callbacks per writer thread. When a writer 5736 thread exceeds this limit, it invokes the 5737 corresponding flavor of rcu_barrier() to allow 5738 previously posted callbacks to drain. 5739 5740 rcuscale.gp_exp= [KNL] 5741 Measure performance of expedited synchronous 5742 grace-period primitives. 5743 5744 rcuscale.holdoff= [KNL] 5745 Set test-start holdoff period. The purpose of 5746 this parameter is to delay the start of the 5747 test until boot completes in order to avoid 5748 interference. 5749 5750 rcuscale.kfree_by_call_rcu= [KNL] 5751 In kernels built with CONFIG_RCU_LAZY=y, test 5752 call_rcu() instead of kfree_rcu(). 5753 5754 rcuscale.kfree_mult= [KNL] 5755 Instead of allocating an object of size kfree_obj, 5756 allocate one of kfree_mult * sizeof(kfree_obj). 5757 Defaults to 1. 5758 5759 rcuscale.kfree_rcu_test= [KNL] 5760 Set to measure performance of kfree_rcu() flooding. 5761 5762 rcuscale.kfree_rcu_test_double= [KNL] 5763 Test the double-argument variant of kfree_rcu(). 5764 If this parameter has the same value as 5765 rcuscale.kfree_rcu_test_single, both the single- 5766 and double-argument variants are tested. 5767 5768 rcuscale.kfree_rcu_test_single= [KNL] 5769 Test the single-argument variant of kfree_rcu(). 5770 If this parameter has the same value as 5771 rcuscale.kfree_rcu_test_double, both the single- 5772 and double-argument variants are tested. 5773 5774 rcuscale.kfree_nthreads= [KNL] 5775 The number of threads running loops of kfree_rcu(). 5776 5777 rcuscale.kfree_alloc_num= [KNL] 5778 Number of allocations and frees done in an iteration. 5779 5780 rcuscale.kfree_loops= [KNL] 5781 Number of loops doing rcuscale.kfree_alloc_num number 5782 of allocations and frees. 5783 5784 rcuscale.minruntime= [KNL] 5785 Set the minimum test run time in seconds. This 5786 does not affect the data-collection interval, 5787 but instead allows better measurement of things 5788 like CPU consumption. 5789 5790 rcuscale.nreaders= [KNL] 5791 Set number of RCU readers. The value -1 selects 5792 N, where N is the number of CPUs. A value 5793 "n" less than -1 selects N-n+1, where N is again 5794 the number of CPUs. For example, -2 selects N 5795 (the number of CPUs), -3 selects N+1, and so on. 5796 A value of "n" less than or equal to -N selects 5797 a single reader. 5798 5799 rcuscale.nwriters= [KNL] 5800 Set number of RCU writers. The values operate 5801 the same as for rcuscale.nreaders. 5802 N, where N is the number of CPUs 5803 5804 rcuscale.scale_type= [KNL] 5805 Specify the RCU implementation to test. 5806 5807 rcuscale.shutdown= [KNL] 5808 Shut the system down after performance tests 5809 complete. This is useful for hands-off automated 5810 testing. 5811 5812 rcuscale.verbose= [KNL] 5813 Enable additional printk() statements. 5814 5815 rcuscale.writer_holdoff= [KNL] 5816 Write-side holdoff between grace periods, 5817 in microseconds. The default of zero says 5818 no holdoff. 5819 5820 rcuscale.writer_holdoff_jiffies= [KNL] 5821 Additional write-side holdoff between grace 5822 periods, but in jiffies. The default of zero 5823 says no holdoff. 5824 5825 rcutorture.fqs_duration= [KNL] 5826 Set duration of force_quiescent_state bursts 5827 in microseconds. 5828 5829 rcutorture.fqs_holdoff= [KNL] 5830 Set holdoff time within force_quiescent_state bursts 5831 in microseconds. 5832 5833 rcutorture.fqs_stutter= [KNL] 5834 Set wait time between force_quiescent_state bursts 5835 in seconds. 5836 5837 rcutorture.fwd_progress= [KNL] 5838 Specifies the number of kthreads to be used 5839 for RCU grace-period forward-progress testing 5840 for the types of RCU supporting this notion. 5841 Defaults to 1 kthread, values less than zero or 5842 greater than the number of CPUs cause the number 5843 of CPUs to be used. 5844 5845 rcutorture.fwd_progress_div= [KNL] 5846 Specify the fraction of a CPU-stall-warning 5847 period to do tight-loop forward-progress testing. 5848 5849 rcutorture.fwd_progress_holdoff= [KNL] 5850 Number of seconds to wait between successive 5851 forward-progress tests. 5852 5853 rcutorture.fwd_progress_need_resched= [KNL] 5854 Enclose cond_resched() calls within checks for 5855 need_resched() during tight-loop forward-progress 5856 testing. 5857 5858 rcutorture.gp_cond= [KNL] 5859 Use conditional/asynchronous update-side 5860 normal-grace-period primitives, if available. 5861 5862 rcutorture.gp_cond_exp= [KNL] 5863 Use conditional/asynchronous update-side 5864 expedited-grace-period primitives, if available. 5865 5866 rcutorture.gp_cond_full= [KNL] 5867 Use conditional/asynchronous update-side 5868 normal-grace-period primitives that also take 5869 concurrent expedited grace periods into account, 5870 if available. 5871 5872 rcutorture.gp_cond_exp_full= [KNL] 5873 Use conditional/asynchronous update-side 5874 expedited-grace-period primitives that also take 5875 concurrent normal grace periods into account, 5876 if available. 5877 5878 rcutorture.gp_cond_wi= [KNL] 5879 Nominal wait interval for normal conditional 5880 grace periods (specified by rcutorture's 5881 gp_cond and gp_cond_full module parameters), 5882 in microseconds. The actual wait interval will 5883 be randomly selected to nanosecond granularity up 5884 to this wait interval. Defaults to 16 jiffies, 5885 for example, 16,000 microseconds on a system 5886 with HZ=1000. 5887 5888 rcutorture.gp_cond_wi_exp= [KNL] 5889 Nominal wait interval for expedited conditional 5890 grace periods (specified by rcutorture's 5891 gp_cond_exp and gp_cond_exp_full module 5892 parameters), in microseconds. The actual wait 5893 interval will be randomly selected to nanosecond 5894 granularity up to this wait interval. Defaults to 5895 128 microseconds. 5896 5897 rcutorture.gp_exp= [KNL] 5898 Use expedited update-side primitives, if available. 5899 5900 rcutorture.gp_normal= [KNL] 5901 Use normal (non-expedited) asynchronous 5902 update-side primitives, if available. 5903 5904 rcutorture.gp_poll= [KNL] 5905 Use polled update-side normal-grace-period 5906 primitives, if available. 5907 5908 rcutorture.gp_poll_exp= [KNL] 5909 Use polled update-side expedited-grace-period 5910 primitives, if available. 5911 5912 rcutorture.gp_poll_full= [KNL] 5913 Use polled update-side normal-grace-period 5914 primitives that also take concurrent expedited 5915 grace periods into account, if available. 5916 5917 rcutorture.gp_poll_exp_full= [KNL] 5918 Use polled update-side expedited-grace-period 5919 primitives that also take concurrent normal 5920 grace periods into account, if available. 5921 5922 rcutorture.gp_poll_wi= [KNL] 5923 Nominal wait interval for normal conditional 5924 grace periods (specified by rcutorture's 5925 gp_poll and gp_poll_full module parameters), 5926 in microseconds. The actual wait interval will 5927 be randomly selected to nanosecond granularity up 5928 to this wait interval. Defaults to 16 jiffies, 5929 for example, 16,000 microseconds on a system 5930 with HZ=1000. 5931 5932 rcutorture.gp_poll_wi_exp= [KNL] 5933 Nominal wait interval for expedited conditional 5934 grace periods (specified by rcutorture's 5935 gp_poll_exp and gp_poll_exp_full module 5936 parameters), in microseconds. The actual wait 5937 interval will be randomly selected to nanosecond 5938 granularity up to this wait interval. Defaults to 5939 128 microseconds. 5940 5941 rcutorture.gp_sync= [KNL] 5942 Use normal (non-expedited) synchronous 5943 update-side primitives, if available. If all 5944 of rcutorture.gp_cond=, rcutorture.gp_exp=, 5945 rcutorture.gp_normal=, and rcutorture.gp_sync= 5946 are zero, rcutorture acts as if is interpreted 5947 they are all non-zero. 5948 5949 rcutorture.gpwrap_lag= [KNL] 5950 Enable grace-period wrap lag testing. Setting 5951 to false prevents the gpwrap lag test from 5952 running. Default is true. 5953 5954 rcutorture.gpwrap_lag_gps= [KNL] 5955 Set the value for grace-period wrap lag during 5956 active lag testing periods. This controls how many 5957 grace periods differences we tolerate between 5958 rdp and rnp's gp_seq before setting overflow flag. 5959 The default is always set to 8. 5960 5961 rcutorture.gpwrap_lag_cycle_mins= [KNL] 5962 Set the total cycle duration for gpwrap lag 5963 testing in minutes. This is the total time for 5964 one complete cycle of active and inactive 5965 testing periods. Default is 30 minutes. 5966 5967 rcutorture.gpwrap_lag_active_mins= [KNL] 5968 Set the duration for which gpwrap lag is active 5969 within each cycle, in minutes. During this time, 5970 the grace-period wrap lag will be set to the 5971 value specified by gpwrap_lag_gps. Default is 5972 5 minutes. 5973 5974 rcutorture.irqreader= [KNL] 5975 Run RCU readers from irq handlers, or, more 5976 accurately, from a timer handler. Not all RCU 5977 flavors take kindly to this sort of thing. 5978 5979 rcutorture.leakpointer= [KNL] 5980 Leak an RCU-protected pointer out of the reader. 5981 This can of course result in splats, and is 5982 intended to test the ability of things like 5983 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 5984 such leaks. 5985 5986 rcutorture.n_barrier_cbs= [KNL] 5987 Set callbacks/threads for rcu_barrier() testing. 5988 5989 rcutorture.nfakewriters= [KNL] 5990 Set number of concurrent RCU writers. These just 5991 stress RCU, they don't participate in the actual 5992 test, hence the "fake". 5993 5994 rcutorture.nocbs_nthreads= [KNL] 5995 Set number of RCU callback-offload togglers. 5996 Zero (the default) disables toggling. 5997 5998 rcutorture.nocbs_toggle= [KNL] 5999 Set the delay in milliseconds between successive 6000 callback-offload toggling attempts. 6001 6002 rcutorture.nreaders= [KNL] 6003 Set number of RCU readers. The value -1 selects 6004 N-1, where N is the number of CPUs. A value 6005 "n" less than -1 selects N-n-2, where N is again 6006 the number of CPUs. For example, -2 selects N 6007 (the number of CPUs), -3 selects N+1, and so on. 6008 6009 rcutorture.object_debug= [KNL] 6010 Enable debug-object double-call_rcu() testing. 6011 6012 rcutorture.onoff_holdoff= [KNL] 6013 Set time (s) after boot for CPU-hotplug testing. 6014 6015 rcutorture.onoff_interval= [KNL] 6016 Set time (jiffies) between CPU-hotplug operations, 6017 or zero to disable CPU-hotplug testing. 6018 6019 rcutorture.preempt_duration= [KNL] 6020 Set duration (in milliseconds) of preemptions 6021 by a high-priority FIFO real-time task. Set to 6022 zero (the default) to disable. The CPUs to 6023 preempt are selected randomly from the set that 6024 are online at a given point in time. Races with 6025 CPUs going offline are ignored, with that attempt 6026 at preemption skipped. 6027 6028 rcutorture.preempt_interval= [KNL] 6029 Set interval (in milliseconds, defaulting to one 6030 second) between preemptions by a high-priority 6031 FIFO real-time task. This delay is mediated 6032 by an hrtimer and is further fuzzed to avoid 6033 inadvertent synchronizations. 6034 6035 rcutorture.read_exit_burst= [KNL] 6036 The number of times in a given read-then-exit 6037 episode that a set of read-then-exit kthreads 6038 is spawned. 6039 6040 rcutorture.read_exit_delay= [KNL] 6041 The delay, in seconds, between successive 6042 read-then-exit testing episodes. 6043 6044 rcutorture.reader_flavor= [KNL] 6045 A bit mask indicating which readers to use. 6046 If there is more than one bit set, the readers 6047 are entered from low-order bit up, and are 6048 exited in the opposite order. For SRCU, the 6049 0x1 bit is normal readers, 0x2 NMI-safe readers, 6050 and 0x4 light-weight readers. 6051 6052 rcutorture.shuffle_interval= [KNL] 6053 Set task-shuffle interval (s). Shuffling tasks 6054 allows some CPUs to go into dyntick-idle mode 6055 during the rcutorture test. 6056 6057 rcutorture.shutdown_secs= [KNL] 6058 Set time (s) after boot system shutdown. This 6059 is useful for hands-off automated testing. 6060 6061 rcutorture.stall_cpu= [KNL] 6062 Duration of CPU stall (s) to test RCU CPU stall 6063 warnings, zero to disable. 6064 6065 rcutorture.stall_cpu_block= [KNL] 6066 Sleep while stalling if set. This will result 6067 in warnings from preemptible RCU in addition to 6068 any other stall-related activity. Note that 6069 in kernels built with CONFIG_PREEMPTION=n and 6070 CONFIG_PREEMPT_COUNT=y, this parameter will 6071 cause the CPU to pass through a quiescent state. 6072 Given CONFIG_PREEMPTION=n, this will suppress 6073 RCU CPU stall warnings, but will instead result 6074 in scheduling-while-atomic splats. 6075 6076 Use of this module parameter results in splats. 6077 6078 6079 rcutorture.stall_cpu_holdoff= [KNL] 6080 Time to wait (s) after boot before inducing stall. 6081 6082 rcutorture.stall_cpu_irqsoff= [KNL] 6083 Disable interrupts while stalling if set, but only 6084 on the first stall in the set. 6085 6086 rcutorture.stall_cpu_repeat= [KNL] 6087 Number of times to repeat the stall sequence, 6088 so that rcutorture.stall_cpu_repeat=3 will result 6089 in four stall sequences. 6090 6091 rcutorture.stall_gp_kthread= [KNL] 6092 Duration (s) of forced sleep within RCU 6093 grace-period kthread to test RCU CPU stall 6094 warnings, zero to disable. If both stall_cpu 6095 and stall_gp_kthread are specified, the 6096 kthread is starved first, then the CPU. 6097 6098 rcutorture.stat_interval= [KNL] 6099 Time (s) between statistics printk()s. 6100 6101 rcutorture.stutter= [KNL] 6102 Time (s) to stutter testing, for example, specifying 6103 five seconds causes the test to run for five seconds, 6104 wait for five seconds, and so on. This tests RCU's 6105 ability to transition abruptly to and from idle. 6106 6107 rcutorture.test_boost= [KNL] 6108 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 6109 "Maybe" means test if the RCU implementation 6110 under test support RCU priority boosting. 6111 6112 rcutorture.test_boost_duration= [KNL] 6113 Duration (s) of each individual boost test. 6114 6115 rcutorture.test_boost_holdoff= [KNL] 6116 Holdoff time (s) from start of test to the start 6117 of RCU priority-boost testing. Defaults to zero, 6118 that is, no holdoff. 6119 6120 rcutorture.test_boost_interval= [KNL] 6121 Interval (s) between each boost test. 6122 6123 rcutorture.test_no_idle_hz= [KNL] 6124 Test RCU's dyntick-idle handling. See also the 6125 rcutorture.shuffle_interval parameter. 6126 6127 rcutorture.torture_type= [KNL] 6128 Specify the RCU implementation to test. 6129 6130 rcutorture.verbose= [KNL] 6131 Enable additional printk() statements. 6132 6133 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 6134 Dump ftrace buffer after reporting RCU CPU 6135 stall warning. 6136 6137 rcupdate.rcu_cpu_stall_notifiers= [KNL] 6138 Provide RCU CPU stall notifiers, but see the 6139 warnings in the RCU_CPU_STALL_NOTIFIER Kconfig 6140 option's help text. TL;DR: You almost certainly 6141 do not want rcupdate.rcu_cpu_stall_notifiers. 6142 6143 rcupdate.rcu_cpu_stall_suppress= [KNL] 6144 Suppress RCU CPU stall warning messages. 6145 6146 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 6147 Suppress RCU CPU stall warning messages and 6148 rcutorture writer stall warnings that occur 6149 during early boot, that is, during the time 6150 before the init task is spawned. 6151 6152 rcupdate.rcu_cpu_stall_timeout= [KNL] 6153 Set timeout for RCU CPU stall warning messages. 6154 The value is in seconds and the maximum allowed 6155 value is 300 seconds. 6156 6157 rcupdate.rcu_exp_cpu_stall_timeout= [KNL] 6158 Set timeout for expedited RCU CPU stall warning 6159 messages. The value is in milliseconds 6160 and the maximum allowed value is 21000 6161 milliseconds. Please note that this value is 6162 adjusted to an arch timer tick resolution. 6163 Setting this to zero causes the value from 6164 rcupdate.rcu_cpu_stall_timeout to be used (after 6165 conversion from seconds to milliseconds). 6166 6167 rcupdate.rcu_cpu_stall_cputime= [KNL] 6168 Provide statistics on the cputime and count of 6169 interrupts and tasks during the sampling period. For 6170 multiple continuous RCU stalls, all sampling periods 6171 begin at half of the first RCU stall timeout. 6172 6173 rcupdate.rcu_exp_stall_task_details= [KNL] 6174 Print stack dumps of any tasks blocking the 6175 current expedited RCU grace period during an 6176 expedited RCU CPU stall warning. 6177 6178 rcupdate.rcu_expedited= [KNL] 6179 Use expedited grace-period primitives, for 6180 example, synchronize_rcu_expedited() instead 6181 of synchronize_rcu(). This reduces latency, 6182 but can increase CPU utilization, degrade 6183 real-time latency, and degrade energy efficiency. 6184 No effect on CONFIG_TINY_RCU kernels. 6185 6186 rcupdate.rcu_normal= [KNL] 6187 Use only normal grace-period primitives, 6188 for example, synchronize_rcu() instead of 6189 synchronize_rcu_expedited(). This improves 6190 real-time latency, CPU utilization, and 6191 energy efficiency, but can expose users to 6192 increased grace-period latency. This parameter 6193 overrides rcupdate.rcu_expedited. No effect on 6194 CONFIG_TINY_RCU kernels. 6195 6196 rcupdate.rcu_normal_after_boot= [KNL] 6197 Once boot has completed (that is, after 6198 rcu_end_inkernel_boot() has been invoked), use 6199 only normal grace-period primitives. No effect 6200 on CONFIG_TINY_RCU kernels. 6201 6202 But note that CONFIG_PREEMPT_RT=y kernels enables 6203 this kernel boot parameter, forcibly setting 6204 it to the value one, that is, converting any 6205 post-boot attempt at an expedited RCU grace 6206 period to instead use normal non-expedited 6207 grace-period processing. 6208 6209 rcupdate.rcu_task_collapse_lim= [KNL] 6210 Set the maximum number of callbacks present 6211 at the beginning of a grace period that allows 6212 the RCU Tasks flavors to collapse back to using 6213 a single callback queue. This switching only 6214 occurs when rcupdate.rcu_task_enqueue_lim is 6215 set to the default value of -1. 6216 6217 rcupdate.rcu_task_contend_lim= [KNL] 6218 Set the minimum number of callback-queuing-time 6219 lock-contention events per jiffy required to 6220 cause the RCU Tasks flavors to switch to per-CPU 6221 callback queuing. This switching only occurs 6222 when rcupdate.rcu_task_enqueue_lim is set to 6223 the default value of -1. 6224 6225 rcupdate.rcu_task_enqueue_lim= [KNL] 6226 Set the number of callback queues to use for the 6227 RCU Tasks family of RCU flavors. The default 6228 of -1 allows this to be automatically (and 6229 dynamically) adjusted. This parameter is intended 6230 for use in testing. 6231 6232 rcupdate.rcu_task_ipi_delay= [KNL] 6233 Set time in jiffies during which RCU tasks will 6234 avoid sending IPIs, starting with the beginning 6235 of a given grace period. Setting a large 6236 number avoids disturbing real-time workloads, 6237 but lengthens grace periods. 6238 6239 rcupdate.rcu_task_lazy_lim= [KNL] 6240 Number of callbacks on a given CPU that will 6241 cancel laziness on that CPU. Use -1 to disable 6242 cancellation of laziness, but be advised that 6243 doing so increases the danger of OOM due to 6244 callback flooding. 6245 6246 rcupdate.rcu_task_stall_info= [KNL] 6247 Set initial timeout in jiffies for RCU task stall 6248 informational messages, which give some indication 6249 of the problem for those not patient enough to 6250 wait for ten minutes. Informational messages are 6251 only printed prior to the stall-warning message 6252 for a given grace period. Disable with a value 6253 less than or equal to zero. Defaults to ten 6254 seconds. A change in value does not take effect 6255 until the beginning of the next grace period. 6256 6257 rcupdate.rcu_task_stall_info_mult= [KNL] 6258 Multiplier for time interval between successive 6259 RCU task stall informational messages for a given 6260 RCU tasks grace period. This value is clamped 6261 to one through ten, inclusive. It defaults to 6262 the value three, so that the first informational 6263 message is printed 10 seconds into the grace 6264 period, the second at 40 seconds, the third at 6265 160 seconds, and then the stall warning at 600 6266 seconds would prevent a fourth at 640 seconds. 6267 6268 rcupdate.rcu_task_stall_timeout= [KNL] 6269 Set timeout in jiffies for RCU task stall 6270 warning messages. Disable with a value less 6271 than or equal to zero. Defaults to ten minutes. 6272 A change in value does not take effect until 6273 the beginning of the next grace period. 6274 6275 rcupdate.rcu_tasks_lazy_ms= [KNL] 6276 Set timeout in milliseconds RCU Tasks asynchronous 6277 callback batching for call_rcu_tasks(). 6278 A negative value will take the default. A value 6279 of zero will disable batching. Batching is 6280 always disabled for synchronize_rcu_tasks(). 6281 6282 rcupdate.rcu_tasks_trace_lazy_ms= [KNL] 6283 Set timeout in milliseconds RCU Tasks 6284 Trace asynchronous callback batching for 6285 call_rcu_tasks_trace(). A negative value 6286 will take the default. A value of zero will 6287 disable batching. Batching is always disabled 6288 for synchronize_rcu_tasks_trace(). 6289 6290 rcupdate.rcu_self_test= [KNL] 6291 Run the RCU early boot self tests 6292 6293 rdinit= [KNL] 6294 Format: <full_path> 6295 Run specified binary instead of /init from the ramdisk, 6296 used for early userspace startup. See initrd. 6297 6298 rdrand= [X86,EARLY] 6299 force - Override the decision by the kernel to hide the 6300 advertisement of RDRAND support (this affects 6301 certain AMD processors because of buggy BIOS 6302 support, specifically around the suspend/resume 6303 path). 6304 6305 rdt= [HW,X86,RDT] 6306 Turn on/off individual RDT features. List is: 6307 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 6308 mba, smba, bmec, abmc. 6309 E.g. to turn on cmt and turn off mba use: 6310 rdt=cmt,!mba 6311 6312 reboot= [KNL] 6313 Format (x86 or x86_64): 6314 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \ 6315 [[,]s[mp]#### \ 6316 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 6317 [[,]f[orce] 6318 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 6319 (prefix with 'panic_' to set mode for panic 6320 reboot only), 6321 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 6322 reboot_force is either force or not specified, 6323 reboot_cpu is s[mp]#### with #### being the processor 6324 to be used for rebooting. 6325 6326 acpi 6327 Use the ACPI RESET_REG in the FADT. If ACPI is not 6328 configured or the ACPI reset does not work, the reboot 6329 path attempts the reset using the keyboard controller. 6330 6331 bios 6332 Use the CPU reboot vector for warm reset 6333 6334 cold 6335 Set the cold reboot flag 6336 6337 default 6338 There are some built-in platform specific "quirks" 6339 - you may see: "reboot: <name> series board detected. 6340 Selecting <type> for reboots." In the case where you 6341 think the quirk is in error (e.g. you have newer BIOS, 6342 or newer board) using this option will ignore the 6343 built-in quirk table, and use the generic default 6344 reboot actions. 6345 6346 efi 6347 Use efi reset_system runtime service. If EFI is not 6348 configured or the EFI reset does not work, the reboot 6349 path attempts the reset using the keyboard controller. 6350 6351 force 6352 Don't stop other CPUs on reboot. This can make reboot 6353 more reliable in some cases. 6354 6355 kbd 6356 Use the keyboard controller. cold reset (default) 6357 6358 pci 6359 Use a write to the PCI config space register 0xcf9 to 6360 trigger reboot. 6361 6362 triple 6363 Force a triple fault (init) 6364 6365 warm 6366 Don't set the cold reboot flag 6367 6368 Using warm reset will be much faster especially on big 6369 memory systems because the BIOS will not go through 6370 the memory check. Disadvantage is that not all 6371 hardware will be completely reinitialized on reboot so 6372 there may be boot problems on some systems. 6373 6374 6375 refscale.holdoff= [KNL] 6376 Set test-start holdoff period. The purpose of 6377 this parameter is to delay the start of the 6378 test until boot completes in order to avoid 6379 interference. 6380 6381 refscale.lookup_instances= [KNL] 6382 Number of data elements to use for the forms of 6383 SLAB_TYPESAFE_BY_RCU testing. A negative number 6384 is negated and multiplied by nr_cpu_ids, while 6385 zero specifies nr_cpu_ids. 6386 6387 refscale.loops= [KNL] 6388 Set the number of loops over the synchronization 6389 primitive under test. Increasing this number 6390 reduces noise due to loop start/end overhead, 6391 but the default has already reduced the per-pass 6392 noise to a handful of picoseconds on ca. 2020 6393 x86 laptops. 6394 6395 refscale.nreaders= [KNL] 6396 Set number of readers. The default value of -1 6397 selects N, where N is roughly 75% of the number 6398 of CPUs. A value of zero is an interesting choice. 6399 6400 refscale.nruns= [KNL] 6401 Set number of runs, each of which is dumped onto 6402 the console log. 6403 6404 refscale.readdelay= [KNL] 6405 Set the read-side critical-section duration, 6406 measured in microseconds. 6407 6408 refscale.scale_type= [KNL] 6409 Specify the read-protection implementation to test. 6410 6411 refscale.shutdown= [KNL] 6412 Shut down the system at the end of the performance 6413 test. This defaults to 1 (shut it down) when 6414 refscale is built into the kernel and to 0 (leave 6415 it running) when refscale is built as a module. 6416 6417 refscale.verbose= [KNL] 6418 Enable additional printk() statements. 6419 6420 refscale.verbose_batched= [KNL] 6421 Batch the additional printk() statements. If zero 6422 (the default) or negative, print everything. Otherwise, 6423 print every Nth verbose statement, where N is the value 6424 specified. 6425 6426 regulator_ignore_unused 6427 [REGULATOR] 6428 Prevents regulator framework from disabling regulators 6429 that are unused, due no driver claiming them. This may 6430 be useful for debug and development, but should not be 6431 needed on a platform with proper driver support. 6432 6433 relax_domain_level= 6434 [KNL, SMP] Set scheduler's default relax_domain_level. 6435 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 6436 6437 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 6438 Format: <base1>,<size1>[,<base2>,<size2>,...] 6439 Reserve I/O ports or memory so the kernel won't use 6440 them. If <base> is less than 0x10000, the region 6441 is assumed to be I/O ports; otherwise it is memory. 6442 6443 reserve_mem= [RAM] 6444 Format: nn[KMG]:<align>:<label> 6445 Reserve physical memory and label it with a name that 6446 other subsystems can use to access it. This is typically 6447 used for systems that do not wipe the RAM, and this command 6448 line will try to reserve the same physical memory on 6449 soft reboots. Note, it is not guaranteed to be the same 6450 location. For example, if anything about the system changes 6451 or if booting a different kernel. It can also fail if KASLR 6452 places the kernel at the location of where the RAM reservation 6453 was from a previous boot, the new reservation will be at a 6454 different location. 6455 Any subsystem using this feature must add a way to verify 6456 that the contents of the physical memory is from a previous 6457 boot, as there may be cases where the memory will not be 6458 located at the same location. 6459 6460 The format is size:align:label for example, to request 6461 12 megabytes of 4096 alignment for ramoops: 6462 6463 reserve_mem=12M:4096:oops ramoops.mem_name=oops 6464 6465 reservetop= [X86-32,EARLY] 6466 Format: nn[KMG] 6467 Reserves a hole at the top of the kernel virtual 6468 address space. 6469 6470 reset_devices [KNL] Force drivers to reset the underlying device 6471 during initialization. 6472 6473 resume= [SWSUSP] 6474 Specify the partition device for software suspend 6475 Format: 6476 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 6477 6478 resume_offset= [SWSUSP] 6479 Specify the offset from the beginning of the partition 6480 given by "resume=" at which the swap header is located, 6481 in <PAGE_SIZE> units (needed only for swap files). 6482 See Documentation/power/swsusp-and-swap-files.rst 6483 6484 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 6485 read the resume files 6486 6487 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 6488 Useful for devices that are detected asynchronously 6489 (e.g. USB and MMC devices). 6490 6491 retain_initrd [RAM] Keep initrd memory after extraction. After boot, it will 6492 be accessible via /sys/firmware/initrd. 6493 6494 retbleed= [X86] Control mitigation of RETBleed (Arbitrary 6495 Speculative Code Execution with Return Instructions) 6496 vulnerability. 6497 6498 AMD-based UNRET and IBPB mitigations alone do not stop 6499 sibling threads from influencing the predictions of other 6500 sibling threads. For that reason, STIBP is used on pro- 6501 cessors that support it, and mitigate SMT on processors 6502 that don't. 6503 6504 off - no mitigation 6505 auto - automatically select a mitigation 6506 auto,nosmt - automatically select a mitigation, 6507 disabling SMT if necessary for 6508 the full mitigation (only on Zen1 6509 and older without STIBP). 6510 ibpb - On AMD, mitigate short speculation 6511 windows on basic block boundaries too. 6512 Safe, highest perf impact. It also 6513 enables STIBP if present. Not suitable 6514 on Intel. 6515 ibpb,nosmt - Like "ibpb" above but will disable SMT 6516 when STIBP is not available. This is 6517 the alternative for systems which do not 6518 have STIBP. 6519 unret - Force enable untrained return thunks, 6520 only effective on AMD f15h-f17h based 6521 systems. 6522 unret,nosmt - Like unret, but will disable SMT when STIBP 6523 is not available. This is the alternative for 6524 systems which do not have STIBP. 6525 6526 Selecting 'auto' will choose a mitigation method at run 6527 time according to the CPU. 6528 6529 Not specifying this option is equivalent to retbleed=auto. 6530 6531 rfkill.default_state= 6532 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 6533 etc. communication is blocked by default. 6534 1 Unblocked. 6535 6536 rfkill.master_switch_mode= 6537 0 The "airplane mode" button does nothing. 6538 1 The "airplane mode" button toggles between everything 6539 blocked and the previous configuration. 6540 2 The "airplane mode" button toggles between everything 6541 blocked and everything unblocked. 6542 6543 ring3mwait=disable 6544 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 6545 CPUs. 6546 6547 riscv_isa_fallback [RISCV,EARLY] 6548 When CONFIG_RISCV_ISA_FALLBACK is not enabled, permit 6549 falling back to detecting extension support by parsing 6550 "riscv,isa" property on devicetree systems when the 6551 replacement properties are not found. See the Kconfig 6552 entry for RISCV_ISA_FALLBACK. 6553 6554 ro [KNL] Mount root device read-only on boot 6555 6556 rodata= [KNL,EARLY] 6557 on Mark read-only kernel memory as read-only (default). 6558 off Leave read-only kernel memory writable for debugging. 6559 noalias Mark read-only kernel memory as read-only but retain 6560 writable aliases in the direct map for regions outside 6561 of the kernel image. [arm64] 6562 6563 rockchip.usb_uart 6564 [EARLY] 6565 Enable the uart passthrough on the designated usb port 6566 on Rockchip SoCs. When active, the signals of the 6567 debug-uart get routed to the D+ and D- pins of the usb 6568 port and the regular usb controller gets disabled. 6569 6570 root= [KNL] Root filesystem 6571 Usually this is a block device specifier of some kind, 6572 see the early_lookup_bdev comment in 6573 block/early-lookup.c for details. 6574 Alternatively this can be "ram" for the legacy initial 6575 ramdisk, "nfs" and "cifs" for root on a network file 6576 system, or "mtd" and "ubi" for mounting from raw flash. 6577 6578 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 6579 mount the root filesystem 6580 6581 rootflags= [KNL] Set root filesystem mount option string 6582 6583 initramfs_options= [KNL] 6584 Specify mount options for for the initramfs mount. 6585 6586 rootfstype= [KNL] Set root filesystem type 6587 6588 rootwait [KNL] Wait (indefinitely) for root device to show up. 6589 Useful for devices that are detected asynchronously 6590 (e.g. USB and MMC devices). 6591 6592 rootwait= [KNL] Maximum time (in seconds) to wait for root device 6593 to show up before attempting to mount the root 6594 filesystem. 6595 6596 rproc_mem=nn[KMG][@address] 6597 [KNL,ARM,CMA] Remoteproc physical memory block. 6598 Memory area to be used by remote processor image, 6599 managed by CMA. 6600 6601 rt_group_sched= [KNL] Enable or disable SCHED_RR/FIFO group scheduling 6602 when CONFIG_RT_GROUP_SCHED=y. Defaults to 6603 !CONFIG_RT_GROUP_SCHED_DEFAULT_DISABLED. 6604 Format: <bool> 6605 6606 rw [KNL] Mount root device read-write on boot 6607 6608 S [KNL] Run init in single mode 6609 6610 s390_iommu= [HW,S390] 6611 Set s390 IOTLB flushing mode 6612 strict 6613 With strict flushing every unmap operation will result 6614 in an IOTLB flush. Default is lazy flushing before 6615 reuse, which is faster. Deprecated, equivalent to 6616 iommu.strict=1. 6617 6618 s390_iommu_aperture= [KNL,S390] 6619 Specifies the size of the per device DMA address space 6620 accessible through the DMA and IOMMU APIs as a decimal 6621 factor of the size of main memory. 6622 The default is 1 meaning that one can concurrently use 6623 as many DMA addresses as physical memory is installed, 6624 if supported by hardware, and thus map all of memory 6625 once. With a value of 2 one can map all of memory twice 6626 and so on. As a special case a factor of 0 imposes no 6627 restrictions other than those given by hardware at the 6628 cost of significant additional memory use for tables. 6629 6630 sa1100ir [NET] 6631 See drivers/net/irda/sa1100_ir.c. 6632 6633 sched_proxy_exec= [KNL] 6634 Enables or disables "proxy execution" style 6635 solution to mutex-based priority inversion. 6636 Format: <bool> 6637 6638 sched_verbose [KNL,EARLY] Enables verbose scheduler debug messages. 6639 6640 schedstats= [KNL,X86] Enable or disable scheduled statistics. 6641 Allowed values are enable and disable. This feature 6642 incurs a small amount of overhead in the scheduler 6643 but is useful for debugging and performance tuning. 6644 6645 sched_thermal_decay_shift= 6646 [Deprecated] 6647 [KNL, SMP] Set a decay shift for scheduler thermal 6648 pressure signal. Thermal pressure signal follows the 6649 default decay period of other scheduler pelt 6650 signals(usually 32 ms but configurable). Setting 6651 sched_thermal_decay_shift will left shift the decay 6652 period for the thermal pressure signal by the shift 6653 value. 6654 i.e. with the default pelt decay period of 32 ms 6655 sched_thermal_decay_shift thermal pressure decay pr 6656 1 64 ms 6657 2 128 ms 6658 and so on. 6659 Format: integer between 0 and 10 6660 Default is 0. 6661 6662 scftorture.holdoff= [KNL] 6663 Number of seconds to hold off before starting 6664 test. Defaults to zero for module insertion and 6665 to 10 seconds for built-in smp_call_function() 6666 tests. 6667 6668 scftorture.longwait= [KNL] 6669 Request ridiculously long waits randomly selected 6670 up to the chosen limit in seconds. Zero (the 6671 default) disables this feature. Please note 6672 that requesting even small non-zero numbers of 6673 seconds can result in RCU CPU stall warnings, 6674 softlockup complaints, and so on. 6675 6676 scftorture.nthreads= [KNL] 6677 Number of kthreads to spawn to invoke the 6678 smp_call_function() family of functions. 6679 The default of -1 specifies a number of kthreads 6680 equal to the number of CPUs. 6681 6682 scftorture.onoff_holdoff= [KNL] 6683 Number seconds to wait after the start of the 6684 test before initiating CPU-hotplug operations. 6685 6686 scftorture.onoff_interval= [KNL] 6687 Number seconds to wait between successive 6688 CPU-hotplug operations. Specifying zero (which 6689 is the default) disables CPU-hotplug operations. 6690 6691 scftorture.shutdown_secs= [KNL] 6692 The number of seconds following the start of the 6693 test after which to shut down the system. The 6694 default of zero avoids shutting down the system. 6695 Non-zero values are useful for automated tests. 6696 6697 scftorture.stat_interval= [KNL] 6698 The number of seconds between outputting the 6699 current test statistics to the console. A value 6700 of zero disables statistics output. 6701 6702 scftorture.stutter_cpus= [KNL] 6703 The number of jiffies to wait between each change 6704 to the set of CPUs under test. 6705 6706 scftorture.use_cpus_read_lock= [KNL] 6707 Use use_cpus_read_lock() instead of the default 6708 preempt_disable() to disable CPU hotplug 6709 while invoking one of the smp_call_function*() 6710 functions. 6711 6712 scftorture.verbose= [KNL] 6713 Enable additional printk() statements. 6714 6715 scftorture.weight_single= [KNL] 6716 The probability weighting to use for the 6717 smp_call_function_single() function with a zero 6718 "wait" parameter. A value of -1 selects the 6719 default if all other weights are -1. However, 6720 if at least one weight has some other value, a 6721 value of -1 will instead select a weight of zero. 6722 6723 scftorture.weight_single_wait= [KNL] 6724 The probability weighting to use for the 6725 smp_call_function_single() function with a 6726 non-zero "wait" parameter. See weight_single. 6727 6728 scftorture.weight_many= [KNL] 6729 The probability weighting to use for the 6730 smp_call_function_many() function with a zero 6731 "wait" parameter. See weight_single. 6732 Note well that setting a high probability for 6733 this weighting can place serious IPI load 6734 on the system. 6735 6736 scftorture.weight_many_wait= [KNL] 6737 The probability weighting to use for the 6738 smp_call_function_many() function with a 6739 non-zero "wait" parameter. See weight_single 6740 and weight_many. 6741 6742 scftorture.weight_all= [KNL] 6743 The probability weighting to use for the 6744 smp_call_function_all() function with a zero 6745 "wait" parameter. See weight_single and 6746 weight_many. 6747 6748 scftorture.weight_all_wait= [KNL] 6749 The probability weighting to use for the 6750 smp_call_function_all() function with a 6751 non-zero "wait" parameter. See weight_single 6752 and weight_many. 6753 6754 sdw_mclk_divider=[SDW] 6755 Specify the MCLK divider for Intel SoundWire buses in 6756 case the BIOS does not provide the clock rate properly. 6757 6758 skew_tick= [KNL,EARLY] Offset the periodic timer tick per cpu to mitigate 6759 xtime_lock contention on larger systems, and/or RCU lock 6760 contention on all systems with CONFIG_MAXSMP set. 6761 Format: { "0" | "1" } 6762 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 6763 1 -- enable. 6764 Note: increases power consumption, thus should only be 6765 enabled if running jitter sensitive (HPC/RT) workloads. 6766 6767 security= [SECURITY] Choose a legacy "major" security module to 6768 enable at boot. This has been deprecated by the 6769 "lsm=" parameter. 6770 6771 selinux= [SELINUX] Disable or enable SELinux at boot time. 6772 Format: { "0" | "1" } 6773 See security/selinux/Kconfig help text. 6774 0 -- disable. 6775 1 -- enable. 6776 Default value is 1. 6777 6778 serialnumber [BUGS=X86-32] 6779 6780 sev=option[,option...] [X86-64] 6781 6782 debug 6783 Enable debug messages. 6784 6785 nosnp 6786 Do not enable SEV-SNP (applies to host/hypervisor 6787 only). Setting 'nosnp' avoids the RMP check overhead 6788 in memory accesses when users do not want to run 6789 SEV-SNP guests. 6790 6791 shapers= [NET] 6792 Maximal number of shapers. 6793 6794 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 6795 Limit apic dumping. The parameter defines the maximal 6796 number of local apics being dumped. Also it is possible 6797 to set it to "all" by meaning -- no limit here. 6798 Format: { 1 (default) | 2 | ... | all }. 6799 The parameter valid if only apic=debug or 6800 apic=verbose is specified. 6801 Example: apic=debug show_lapic=all 6802 6803 slab_debug[=options[,slabs][;[options[,slabs]]...] [MM] 6804 Enabling slab_debug allows one to determine the 6805 culprit if slab objects become corrupted. Enabling 6806 slab_debug can create guard zones around objects and 6807 may poison objects when not in use. Also tracks the 6808 last alloc / free. For more information see 6809 Documentation/admin-guide/mm/slab.rst. 6810 (slub_debug legacy name also accepted for now) 6811 6812 Using this option implies the "no_hash_pointers" 6813 option which can be undone by adding the 6814 "hash_pointers=always" option. 6815 6816 slab_max_order= [MM] 6817 Determines the maximum allowed order for slabs. 6818 A high setting may cause OOMs due to memory 6819 fragmentation. For more information see 6820 Documentation/admin-guide/mm/slab.rst. 6821 (slub_max_order legacy name also accepted for now) 6822 6823 slab_merge [MM] 6824 Enable merging of slabs with similar size when the 6825 kernel is built without CONFIG_SLAB_MERGE_DEFAULT. 6826 (slub_merge legacy name also accepted for now) 6827 6828 slab_min_objects= [MM] 6829 The minimum number of objects per slab. SLUB will 6830 increase the slab order up to slab_max_order to 6831 generate a sufficiently large slab able to contain 6832 the number of objects indicated. The higher the number 6833 of objects the smaller the overhead of tracking slabs 6834 and the less frequently locks need to be acquired. 6835 For more information see 6836 Documentation/admin-guide/mm/slab.rst. 6837 (slub_min_objects legacy name also accepted for now) 6838 6839 slab_min_order= [MM] 6840 Determines the minimum page order for slabs. Must be 6841 lower or equal to slab_max_order. For more information see 6842 Documentation/admin-guide/mm/slab.rst. 6843 (slub_min_order legacy name also accepted for now) 6844 6845 slab_nomerge [MM] 6846 Disable merging of slabs with similar size. May be 6847 necessary if there is some reason to distinguish 6848 allocs to different slabs, especially in hardened 6849 environments where the risk of heap overflows and 6850 layout control by attackers can usually be 6851 frustrated by disabling merging. This will reduce 6852 most of the exposure of a heap attack to a single 6853 cache (risks via metadata attacks are mostly 6854 unchanged). Debug options disable merging on their 6855 own. 6856 For more information see 6857 Documentation/admin-guide/mm/slab.rst. 6858 (slub_nomerge legacy name also accepted for now) 6859 6860 slab_strict_numa [MM] 6861 Support memory policies on a per object level 6862 in the slab allocator. The default is for memory 6863 policies to be applied at the folio level when 6864 a new folio is needed or a partial folio is 6865 retrieved from the lists. Increases overhead 6866 in the slab fastpaths but gains more accurate 6867 NUMA kernel object placement which helps with slow 6868 interconnects in NUMA systems. 6869 6870 slram= [HW,MTD] 6871 6872 smart2= [HW] 6873 Format: <io1>[,<io2>[,...,<io8>]] 6874 6875 smp.csd_lock_timeout= [KNL] 6876 Specify the period of time in milliseconds 6877 that smp_call_function() and friends will wait 6878 for a CPU to release the CSD lock. This is 6879 useful when diagnosing bugs involving CPUs 6880 disabling interrupts for extended periods 6881 of time. Defaults to 5,000 milliseconds, and 6882 setting a value of zero disables this feature. 6883 This feature may be more efficiently disabled 6884 using the csdlock_debug- kernel parameter. 6885 6886 smp.panic_on_ipistall= [KNL] 6887 If a csd_lock_timeout extends for more than 6888 the specified number of milliseconds, panic the 6889 system. By default, let CSD-lock acquisition 6890 take as long as they take. Specifying 300,000 6891 for this value provides a 5-minute timeout. 6892 6893 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 6894 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 6895 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 6896 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 6897 smsc-ircc2.ircc_irq= [HW] IRQ line 6898 smsc-ircc2.ircc_dma= [HW] DMA channel 6899 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 6900 0: Toshiba Satellite 1800 (GP data pin select) 6901 1: Fast pin select (default) 6902 2: ATC IRMode 6903 6904 smt= [KNL,MIPS,S390,EARLY] Set the maximum number of threads 6905 (logical CPUs) to use per physical CPU on systems 6906 capable of symmetric multithreading (SMT). Will 6907 be capped to the actual hardware limit. 6908 Format: <integer> 6909 Default: -1 (no limit) 6910 6911 softlockup_panic= 6912 [KNL] Should the soft-lockup detector generate panics. 6913 Format: 0 | 1 6914 6915 A value of 1 instructs the soft-lockup detector 6916 to panic the machine when a soft-lockup occurs. It is 6917 also controlled by the kernel.softlockup_panic sysctl 6918 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 6919 respective build-time switch to that functionality. 6920 6921 softlockup_all_cpu_backtrace= 6922 [KNL] Should the soft-lockup detector generate 6923 backtraces on all cpus. 6924 Format: 0 | 1 6925 6926 sonypi.*= [HW] Sony Programmable I/O Control Device driver 6927 See Documentation/admin-guide/laptops/sonypi.rst 6928 6929 spectre_bhi= [X86] Control mitigation of Branch History Injection 6930 (BHI) vulnerability. This setting affects the 6931 deployment of the HW BHI control and the SW BHB 6932 clearing sequence. 6933 6934 on - (default) Enable the HW or SW mitigation as 6935 needed. This protects the kernel from 6936 both syscalls and VMs. 6937 vmexit - On systems which don't have the HW mitigation 6938 available, enable the SW mitigation on vmexit 6939 ONLY. On such systems, the host kernel is 6940 protected from VM-originated BHI attacks, but 6941 may still be vulnerable to syscall attacks. 6942 off - Disable the mitigation. 6943 6944 spectre_v2= [X86,EARLY] Control mitigation of Spectre variant 2 6945 (indirect branch speculation) vulnerability. 6946 The default operation protects the kernel from 6947 user space attacks. 6948 6949 on - unconditionally enable, implies 6950 spectre_v2_user=on 6951 off - unconditionally disable, implies 6952 spectre_v2_user=off 6953 auto - kernel detects whether your CPU model is 6954 vulnerable 6955 6956 Selecting 'on' will, and 'auto' may, choose a 6957 mitigation method at run time according to the 6958 CPU, the available microcode, the setting of the 6959 CONFIG_MITIGATION_RETPOLINE configuration option, 6960 and the compiler with which the kernel was built. 6961 6962 Selecting 'on' will also enable the mitigation 6963 against user space to user space task attacks. 6964 Selecting specific mitigation does not force enable 6965 user mitigations. 6966 6967 Selecting 'off' will disable both the kernel and 6968 the user space protections. 6969 6970 Specific mitigations can also be selected manually: 6971 6972 retpoline - replace indirect branches 6973 retpoline,generic - Retpolines 6974 retpoline,lfence - LFENCE; indirect branch 6975 retpoline,amd - alias for retpoline,lfence 6976 eibrs - Enhanced/Auto IBRS 6977 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines 6978 eibrs,lfence - Enhanced/Auto IBRS + LFENCE 6979 ibrs - use IBRS to protect kernel 6980 6981 Not specifying this option is equivalent to 6982 spectre_v2=auto. 6983 6984 spectre_v2_user= 6985 [X86] Control mitigation of Spectre variant 2 6986 (indirect branch speculation) vulnerability between 6987 user space tasks 6988 6989 on - Unconditionally enable mitigations. Is 6990 enforced by spectre_v2=on 6991 6992 off - Unconditionally disable mitigations. Is 6993 enforced by spectre_v2=off 6994 6995 prctl - Indirect branch speculation is enabled, 6996 but mitigation can be enabled via prctl 6997 per thread. The mitigation control state 6998 is inherited on fork. 6999 7000 prctl,ibpb 7001 - Like "prctl" above, but only STIBP is 7002 controlled per thread. IBPB is issued 7003 always when switching between different user 7004 space processes. 7005 7006 seccomp 7007 - Same as "prctl" above, but all seccomp 7008 threads will enable the mitigation unless 7009 they explicitly opt out. 7010 7011 seccomp,ibpb 7012 - Like "seccomp" above, but only STIBP is 7013 controlled per thread. IBPB is issued 7014 always when switching between different 7015 user space processes. 7016 7017 auto - Kernel selects the mitigation depending on 7018 the available CPU features and vulnerability. 7019 7020 Default mitigation: "prctl" 7021 7022 Not specifying this option is equivalent to 7023 spectre_v2_user=auto. 7024 7025 spec_rstack_overflow= 7026 [X86,EARLY] Control RAS overflow mitigation on AMD Zen CPUs 7027 7028 off - Disable mitigation 7029 microcode - Enable microcode mitigation only 7030 safe-ret - Enable sw-only safe RET mitigation (default) 7031 ibpb - Enable mitigation by issuing IBPB on 7032 kernel entry 7033 ibpb-vmexit - Issue IBPB only on VMEXIT 7034 (cloud-specific mitigation) 7035 7036 spec_store_bypass_disable= 7037 [HW,EARLY] Control Speculative Store Bypass (SSB) Disable mitigation 7038 (Speculative Store Bypass vulnerability) 7039 7040 Certain CPUs are vulnerable to an exploit against a 7041 a common industry wide performance optimization known 7042 as "Speculative Store Bypass" in which recent stores 7043 to the same memory location may not be observed by 7044 later loads during speculative execution. The idea 7045 is that such stores are unlikely and that they can 7046 be detected prior to instruction retirement at the 7047 end of a particular speculation execution window. 7048 7049 In vulnerable processors, the speculatively forwarded 7050 store can be used in a cache side channel attack, for 7051 example to read memory to which the attacker does not 7052 directly have access (e.g. inside sandboxed code). 7053 7054 This parameter controls whether the Speculative Store 7055 Bypass optimization is used. 7056 7057 On x86 the options are: 7058 7059 on - Unconditionally disable Speculative Store Bypass 7060 off - Unconditionally enable Speculative Store Bypass 7061 auto - Kernel detects whether the CPU model contains an 7062 implementation of Speculative Store Bypass and 7063 picks the most appropriate mitigation. If the 7064 CPU is not vulnerable, "off" is selected. If the 7065 CPU is vulnerable the default mitigation is 7066 architecture and Kconfig dependent. See below. 7067 prctl - Control Speculative Store Bypass per thread 7068 via prctl. Speculative Store Bypass is enabled 7069 for a process by default. The state of the control 7070 is inherited on fork. 7071 seccomp - Same as "prctl" above, but all seccomp threads 7072 will disable SSB unless they explicitly opt out. 7073 7074 Default mitigations: 7075 X86: "prctl" 7076 7077 On powerpc the options are: 7078 7079 on,auto - On Power8 and Power9 insert a store-forwarding 7080 barrier on kernel entry and exit. On Power7 7081 perform a software flush on kernel entry and 7082 exit. 7083 off - No action. 7084 7085 Not specifying this option is equivalent to 7086 spec_store_bypass_disable=auto. 7087 7088 split_lock_detect= 7089 [X86] Enable split lock detection or bus lock detection 7090 7091 When enabled (and if hardware support is present), atomic 7092 instructions that access data across cache line 7093 boundaries will result in an alignment check exception 7094 for split lock detection or a debug exception for 7095 bus lock detection. 7096 7097 off - not enabled 7098 7099 warn - the kernel will emit rate-limited warnings 7100 about applications triggering the #AC 7101 exception or the #DB exception. This mode is 7102 the default on CPUs that support split lock 7103 detection or bus lock detection. Default 7104 behavior is by #AC if both features are 7105 enabled in hardware. 7106 7107 fatal - the kernel will send SIGBUS to applications 7108 that trigger the #AC exception or the #DB 7109 exception. Default behavior is by #AC if 7110 both features are enabled in hardware. 7111 7112 ratelimit:N - 7113 Set system wide rate limit to N bus locks 7114 per second for bus lock detection. 7115 0 < N <= 1000. 7116 7117 N/A for split lock detection. 7118 7119 7120 If an #AC exception is hit in the kernel or in 7121 firmware (i.e. not while executing in user mode) 7122 the kernel will oops in either "warn" or "fatal" 7123 mode. 7124 7125 #DB exception for bus lock is triggered only when 7126 CPL > 0. 7127 7128 srbds= [X86,INTEL,EARLY] 7129 Control the Special Register Buffer Data Sampling 7130 (SRBDS) mitigation. 7131 7132 Certain CPUs are vulnerable to an MDS-like 7133 exploit which can leak bits from the random 7134 number generator. 7135 7136 By default, this issue is mitigated by 7137 microcode. However, the microcode fix can cause 7138 the RDRAND and RDSEED instructions to become 7139 much slower. Among other effects, this will 7140 result in reduced throughput from /dev/urandom. 7141 7142 The microcode mitigation can be disabled with 7143 the following option: 7144 7145 off: Disable mitigation and remove 7146 performance impact to RDRAND and RDSEED 7147 7148 srcutree.big_cpu_lim [KNL] 7149 Specifies the number of CPUs constituting a 7150 large system, such that srcu_struct structures 7151 should immediately allocate an srcu_node array. 7152 This kernel-boot parameter defaults to 128, 7153 but takes effect only when the low-order four 7154 bits of srcutree.convert_to_big is equal to 3 7155 (decide at boot). 7156 7157 srcutree.convert_to_big [KNL] 7158 Specifies under what conditions an SRCU tree 7159 srcu_struct structure will be converted to big 7160 form, that is, with an rcu_node tree: 7161 7162 0: Never. 7163 1: At init_srcu_struct() time. 7164 2: When rcutorture decides to. 7165 3: Decide at boot time (default). 7166 0x1X: Above plus if high contention. 7167 7168 Either way, the srcu_node tree will be sized based 7169 on the actual runtime number of CPUs (nr_cpu_ids) 7170 instead of the compile-time CONFIG_NR_CPUS. 7171 7172 srcutree.counter_wrap_check [KNL] 7173 Specifies how frequently to check for 7174 grace-period sequence counter wrap for the 7175 srcu_data structure's ->srcu_gp_seq_needed field. 7176 The greater the number of bits set in this kernel 7177 parameter, the less frequently counter wrap will 7178 be checked for. Note that the bottom two bits 7179 are ignored. 7180 7181 srcutree.exp_holdoff [KNL] 7182 Specifies how many nanoseconds must elapse 7183 since the end of the last SRCU grace period for 7184 a given srcu_struct until the next normal SRCU 7185 grace period will be considered for automatic 7186 expediting. Set to zero to disable automatic 7187 expediting. 7188 7189 srcutree.srcu_max_nodelay [KNL] 7190 Specifies the number of no-delay instances 7191 per jiffy for which the SRCU grace period 7192 worker thread will be rescheduled with zero 7193 delay. Beyond this limit, worker thread will 7194 be rescheduled with a sleep delay of one jiffy. 7195 7196 srcutree.srcu_max_nodelay_phase [KNL] 7197 Specifies the per-grace-period phase, number of 7198 non-sleeping polls of readers. Beyond this limit, 7199 grace period worker thread will be rescheduled 7200 with a sleep delay of one jiffy, between each 7201 rescan of the readers, for a grace period phase. 7202 7203 srcutree.srcu_retry_check_delay [KNL] 7204 Specifies number of microseconds of non-sleeping 7205 delay between each non-sleeping poll of readers. 7206 7207 srcutree.small_contention_lim [KNL] 7208 Specifies the number of update-side contention 7209 events per jiffy will be tolerated before 7210 initiating a conversion of an srcu_struct 7211 structure to big form. Note that the value of 7212 srcutree.convert_to_big must have the 0x10 bit 7213 set for contention-based conversions to occur. 7214 7215 ssbd= [ARM64,HW,EARLY] 7216 Speculative Store Bypass Disable control 7217 7218 On CPUs that are vulnerable to the Speculative 7219 Store Bypass vulnerability and offer a 7220 firmware based mitigation, this parameter 7221 indicates how the mitigation should be used: 7222 7223 force-on: Unconditionally enable mitigation for 7224 for both kernel and userspace 7225 force-off: Unconditionally disable mitigation for 7226 for both kernel and userspace 7227 kernel: Always enable mitigation in the 7228 kernel, and offer a prctl interface 7229 to allow userspace to register its 7230 interest in being mitigated too. 7231 7232 stack_guard_gap= [MM] 7233 override the default stack gap protection. The value 7234 is in page units and it defines how many pages prior 7235 to (for stacks growing down) resp. after (for stacks 7236 growing up) the main stack are reserved for no other 7237 mapping. Default value is 256 pages. 7238 7239 stack_depot_disable= [KNL,EARLY] 7240 Setting this to true through kernel command line will 7241 disable the stack depot thereby saving the static memory 7242 consumed by the stack hash table. By default this is set 7243 to false. 7244 7245 stack_depot_max_pools= [KNL,EARLY] 7246 Specify the maximum number of pools to use for storing 7247 stack traces. Pools are allocated on-demand up to this 7248 limit. Default value is 8191 pools. 7249 7250 stacktrace [FTRACE] 7251 Enable the stack tracer on boot up. 7252 7253 stacktrace_filter=[function-list] 7254 [FTRACE] Limit the functions that the stack tracer 7255 will trace at boot up. function-list is a comma-separated 7256 list of functions. This list can be changed at run 7257 time by the stack_trace_filter file in the debugfs 7258 tracing directory. Note, this enables stack tracing 7259 and the stacktrace above is not needed. 7260 7261 sti= [PARISC,HW] 7262 Format: <num> 7263 Set the STI (builtin display/keyboard on the HP-PARISC 7264 machines) console (graphic card) which should be used 7265 as the initial boot-console. 7266 See also comment in drivers/video/console/sticore.c. 7267 7268 sti_font= [HW] 7269 See comment in drivers/video/console/sticore.c. 7270 7271 stifb= [HW] 7272 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 7273 7274 strict_sas_size= 7275 [X86] 7276 Format: <bool> 7277 Enable or disable strict sigaltstack size checks 7278 against the required signal frame size which 7279 depends on the supported FPU features. This can 7280 be used to filter out binaries which have 7281 not yet been made aware of AT_MINSIGSTKSZ. 7282 7283 stress_hpt [PPC,EARLY] 7284 Limits the number of kernel HPT entries in the hash 7285 page table to increase the rate of hash page table 7286 faults on kernel addresses. 7287 7288 stress_slb [PPC,EARLY] 7289 Limits the number of kernel SLB entries, and flushes 7290 them frequently to increase the rate of SLB faults 7291 on kernel addresses. 7292 7293 sunrpc.min_resvport= 7294 sunrpc.max_resvport= 7295 [NFS,SUNRPC] 7296 SunRPC servers often require that client requests 7297 originate from a privileged port (i.e. a port in the 7298 range 0 < portnr < 1024). 7299 An administrator who wishes to reserve some of these 7300 ports for other uses may adjust the range that the 7301 kernel's sunrpc client considers to be privileged 7302 using these two parameters to set the minimum and 7303 maximum port values. 7304 7305 sunrpc.svc_rpc_per_connection_limit= 7306 [NFS,SUNRPC] 7307 Limit the number of requests that the server will 7308 process in parallel from a single connection. 7309 The default value is 0 (no limit). 7310 7311 sunrpc.pool_mode= 7312 [NFS] 7313 Control how the NFS server code allocates CPUs to 7314 service thread pools. Depending on how many NICs 7315 you have and where their interrupts are bound, this 7316 option will affect which CPUs will do NFS serving. 7317 Note: this parameter cannot be changed while the 7318 NFS server is running. 7319 7320 auto the server chooses an appropriate mode 7321 automatically using heuristics 7322 global a single global pool contains all CPUs 7323 percpu one pool for each CPU 7324 pernode one pool for each NUMA node (equivalent 7325 to global on non-NUMA machines) 7326 7327 sunrpc.tcp_slot_table_entries= 7328 sunrpc.udp_slot_table_entries= 7329 [NFS,SUNRPC] 7330 Sets the upper limit on the number of simultaneous 7331 RPC calls that can be sent from the client to a 7332 server. Increasing these values may allow you to 7333 improve throughput, but will also increase the 7334 amount of memory reserved for use by the client. 7335 7336 suspend.pm_test_delay= 7337 [SUSPEND] 7338 Sets the number of seconds to remain in a suspend test 7339 mode before resuming the system (see 7340 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 7341 is set. Default value is 5. 7342 7343 svm= [PPC] 7344 Format: { on | off | y | n | 1 | 0 } 7345 This parameter controls use of the Protected 7346 Execution Facility on pSeries. 7347 7348 swiotlb= [ARM,PPC,MIPS,X86,S390,EARLY] 7349 Format: { <int> [,<int>] | force | noforce } 7350 <int> -- Number of I/O TLB slabs 7351 <int> -- Second integer after comma. Number of swiotlb 7352 areas with their own lock. Will be rounded up 7353 to a power of 2. 7354 force -- force using of bounce buffers even if they 7355 wouldn't be automatically used by the kernel 7356 noforce -- Never use bounce buffers (for debugging) 7357 7358 switches= [HW,M68k,EARLY] 7359 7360 sysctl.*= [KNL] 7361 Set a sysctl parameter, right before loading the init 7362 process, as if the value was written to the respective 7363 /proc/sys/... file. Both '.' and '/' are recognized as 7364 separators. Unrecognized parameters and invalid values 7365 are reported in the kernel log. Sysctls registered 7366 later by a loaded module cannot be set this way. 7367 Example: sysctl.vm.swappiness=40 7368 7369 sysrq_always_enabled 7370 [KNL] 7371 Ignore sysrq setting - this boot parameter will 7372 neutralize any effect of /proc/sys/kernel/sysrq. 7373 Useful for debugging. 7374 7375 tcpmhash_entries= [KNL,NET] 7376 Set the number of tcp_metrics_hash slots. 7377 Default value is 8192 or 16384 depending on total 7378 ram pages. This is used to specify the TCP metrics 7379 cache size. See Documentation/networking/ip-sysctl.rst 7380 "tcp_no_metrics_save" section for more details. 7381 7382 tdfx= [HW,DRM] 7383 7384 test_suspend= [SUSPEND] 7385 Format: { "mem" | "standby" | "freeze" }[,N] 7386 Specify "mem" (for Suspend-to-RAM) or "standby" (for 7387 standby suspend) or "freeze" (for suspend type freeze) 7388 as the system sleep state during system startup with 7389 the optional capability to repeat N number of times. 7390 The system is woken from this state using a 7391 wakeup-capable RTC alarm. 7392 7393 thash_entries= [KNL,NET] 7394 Set number of hash buckets for TCP connection 7395 7396 thermal.act= [HW,ACPI] 7397 -1: disable all active trip points in all thermal zones 7398 <degrees C>: override all lowest active trip points 7399 7400 thermal.crt= [HW,ACPI] 7401 -1: disable all critical trip points in all thermal zones 7402 <degrees C>: override all critical trip points 7403 7404 thermal.off= [HW,ACPI] 7405 1: disable ACPI thermal control 7406 7407 thermal.psv= [HW,ACPI] 7408 -1: disable all passive trip points 7409 <degrees C>: override all passive trip points to this 7410 value 7411 7412 thermal.tzp= [HW,ACPI] 7413 Specify global default ACPI thermal zone polling rate 7414 <deci-seconds>: poll all this frequency 7415 0: no polling (default) 7416 7417 thp_anon= [KNL] 7418 Format: <size>[KMG],<size>[KMG]:<state>;<size>[KMG]-<size>[KMG]:<state> 7419 state is one of "always", "madvise", "never" or "inherit". 7420 Control the default behavior of the system with respect 7421 to anonymous transparent hugepages. 7422 Can be used multiple times for multiple anon THP sizes. 7423 See Documentation/admin-guide/mm/transhuge.rst for more 7424 details. 7425 7426 threadirqs [KNL,EARLY] 7427 Force threading of all interrupt handlers except those 7428 marked explicitly IRQF_NO_THREAD. 7429 7430 thp_shmem= [KNL] 7431 Format: <size>[KMG],<size>[KMG]:<policy>;<size>[KMG]-<size>[KMG]:<policy> 7432 Control the default policy of each hugepage size for the 7433 internal shmem mount. <policy> is one of policies available 7434 for the shmem mount ("always", "inherit", "never", "within_size", 7435 and "advise"). 7436 It can be used multiple times for multiple shmem THP sizes. 7437 See Documentation/admin-guide/mm/transhuge.rst for more 7438 details. 7439 7440 topology= [S390,EARLY] 7441 Format: {off | on} 7442 Specify if the kernel should make use of the cpu 7443 topology information if the hardware supports this. 7444 The scheduler will make use of this information and 7445 e.g. base its process migration decisions on it. 7446 Default is on. 7447 7448 torture.disable_onoff_at_boot= [KNL] 7449 Prevent the CPU-hotplug component of torturing 7450 until after init has spawned. 7451 7452 torture.ftrace_dump_at_shutdown= [KNL] 7453 Dump the ftrace buffer at torture-test shutdown, 7454 even if there were no errors. This can be a 7455 very costly operation when many torture tests 7456 are running concurrently, especially on systems 7457 with rotating-rust storage. 7458 7459 torture.verbose_sleep_frequency= [KNL] 7460 Specifies how many verbose printk()s should be 7461 emitted between each sleep. The default of zero 7462 disables verbose-printk() sleeping. 7463 7464 torture.verbose_sleep_duration= [KNL] 7465 Duration of each verbose-printk() sleep in jiffies. 7466 7467 tpm.disable_pcr_integrity= [HW,TPM] 7468 Do not protect PCR registers from unintended physical 7469 access, or interposers in the bus by the means of 7470 having an integrity protected session wrapped around 7471 TPM2_PCR_Extend command. Consider this in a situation 7472 where TPM is heavily utilized by IMA, thus protection 7473 causing a major performance hit, and the space where 7474 machines are deployed is by other means guarded. 7475 7476 tpm_crb_ffa.busy_timeout_ms= [ARM64,TPM] 7477 Maximum time in milliseconds to retry sending a message 7478 to the TPM service before giving up. This parameter controls 7479 how long the system will continue retrying when the TPM 7480 service is busy. 7481 Format: <unsigned int> 7482 Default: 2000 (2 seconds) 7483 7484 tpm_suspend_pcr=[HW,TPM] 7485 Format: integer pcr id 7486 Specify that at suspend time, the tpm driver 7487 should extend the specified pcr with zeros, 7488 as a workaround for some chips which fail to 7489 flush the last written pcr on TPM_SaveState. 7490 This will guarantee that all the other pcrs 7491 are saved. 7492 7493 tpm_tis.interrupts= [HW,TPM] 7494 Enable interrupts for the MMIO based physical layer 7495 for the FIFO interface. By default it is set to false 7496 (0). For more information about TPM hardware interfaces 7497 defined by Trusted Computing Group (TCG) see 7498 https://trustedcomputinggroup.org/resource/pc-client-platform-tpm-profile-ptp-specification/ 7499 7500 tp_printk [FTRACE] 7501 Have the tracepoints sent to printk as well as the 7502 tracing ring buffer. This is useful for early boot up 7503 where the system hangs or reboots and does not give the 7504 option for reading the tracing buffer or performing a 7505 ftrace_dump_on_oops. 7506 7507 To turn off having tracepoints sent to printk, 7508 echo 0 > /proc/sys/kernel/tracepoint_printk 7509 Note, echoing 1 into this file without the 7510 tp_printk kernel cmdline option has no effect. 7511 7512 The tp_printk_stop_on_boot (see below) can also be used 7513 to stop the printing of events to console at 7514 late_initcall_sync. 7515 7516 ** CAUTION ** 7517 7518 Having tracepoints sent to printk() and activating high 7519 frequency tracepoints such as irq or sched, can cause 7520 the system to live lock. 7521 7522 tp_printk_stop_on_boot [FTRACE] 7523 When tp_printk (above) is set, it can cause a lot of noise 7524 on the console. It may be useful to only include the 7525 printing of events during boot up, as user space may 7526 make the system inoperable. 7527 7528 This command line option will stop the printing of events 7529 to console at the late_initcall_sync() time frame. 7530 7531 trace_buf_size=nn[KMG] 7532 [FTRACE] will set tracing buffer size on each cpu. 7533 7534 trace_clock= [FTRACE] Set the clock used for tracing events 7535 at boot up. 7536 local - Use the per CPU time stamp counter 7537 (converted into nanoseconds). Fast, but 7538 depending on the architecture, may not be 7539 in sync between CPUs. 7540 global - Event time stamps are synchronized across 7541 CPUs. May be slower than the local clock, 7542 but better for some race conditions. 7543 counter - Simple counting of events (1, 2, ..) 7544 note, some counts may be skipped due to the 7545 infrastructure grabbing the clock more than 7546 once per event. 7547 uptime - Use jiffies as the time stamp. 7548 perf - Use the same clock that perf uses. 7549 mono - Use ktime_get_mono_fast_ns() for time stamps. 7550 mono_raw - Use ktime_get_raw_fast_ns() for time 7551 stamps. 7552 boot - Use ktime_get_boot_fast_ns() for time stamps. 7553 Architectures may add more clocks. See 7554 Documentation/trace/ftrace.rst for more details. 7555 7556 trace_event=[event-list] 7557 [FTRACE] Set and start specified trace events in order 7558 to facilitate early boot debugging. The event-list is a 7559 comma-separated list of trace events to enable. See 7560 also Documentation/trace/events.rst 7561 7562 To enable modules, use :mod: keyword: 7563 7564 trace_event=:mod:<module> 7565 7566 The value before :mod: will only enable specific events 7567 that are part of the module. See the above mentioned 7568 document for more information. 7569 7570 trace_instance=[instance-info] 7571 [FTRACE] Create a ring buffer instance early in boot up. 7572 This will be listed in: 7573 7574 /sys/kernel/tracing/instances 7575 7576 Events can be enabled at the time the instance is created 7577 via: 7578 7579 trace_instance=<name>,<system1>:<event1>,<system2>:<event2> 7580 7581 Note, the "<system*>:" portion is optional if the event is 7582 unique. 7583 7584 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall 7585 7586 will enable the "sched_switch" event (note, the "sched:" is optional, and 7587 the same thing would happen if it was left off). The irq_handler_entry 7588 event, and all events under the "initcall" system. 7589 7590 Flags can be added to the instance to modify its behavior when it is 7591 created. The flags are separated by '^'. 7592 7593 The available flags are: 7594 7595 traceoff - Have the tracing instance tracing disabled after it is created. 7596 traceprintk - Have trace_printk() write into this trace instance 7597 (note, "printk" and "trace_printk" can also be used) 7598 7599 trace_instance=foo^traceoff^traceprintk,sched,irq 7600 7601 The flags must come before the defined events. 7602 7603 If memory has been reserved (see memmap for x86), the instance 7604 can use that memory: 7605 7606 memmap=12M$0x284500000 trace_instance=boot_map@0x284500000:12M 7607 7608 The above will create a "boot_map" instance that uses the physical 7609 memory at 0x284500000 that is 12Megs. The per CPU buffers of that 7610 instance will be split up accordingly. 7611 7612 Alternatively, the memory can be reserved by the reserve_mem option: 7613 7614 reserve_mem=12M:4096:trace trace_instance=boot_map@trace 7615 7616 This will reserve 12 megabytes at boot up with a 4096 byte alignment 7617 and place the ring buffer in this memory. Note that due to KASLR, the 7618 memory may not be the same location each time, which will not preserve 7619 the buffer content. 7620 7621 Also note that the layout of the ring buffer data may change between 7622 kernel versions where the validator will fail and reset the ring buffer 7623 if the layout is not the same as the previous kernel. 7624 7625 If the ring buffer is used for persistent bootups and has events enabled, 7626 it is recommend to disable tracing so that events from a previous boot do not 7627 mix with events of the current boot (unless you are debugging a random crash 7628 at boot up). 7629 7630 reserve_mem=12M:4096:trace trace_instance=boot_map^traceoff^traceprintk@trace,sched,irq 7631 7632 Note, saving the trace buffer across reboots does require that the system 7633 is set up to not wipe memory. For instance, CONFIG_RESET_ATTACK_MITIGATION 7634 can force a memory reset on boot which will clear any trace that was stored. 7635 This is just one of many ways that can clear memory. Make sure your system 7636 keeps the content of memory across reboots before relying on this option. 7637 7638 NB: Both the mapped address and size must be page aligned for the architecture. 7639 7640 See also Documentation/trace/debugging.rst 7641 7642 7643 trace_options=[option-list] 7644 [FTRACE] Enable or disable tracer options at boot. 7645 The option-list is a comma delimited list of options 7646 that can be enabled or disabled just as if you were 7647 to echo the option name into 7648 7649 /sys/kernel/tracing/trace_options 7650 7651 For example, to enable stacktrace option (to dump the 7652 stack trace of each event), add to the command line: 7653 7654 trace_options=stacktrace 7655 7656 See also Documentation/trace/ftrace.rst "trace options" 7657 section. 7658 7659 trace_trigger=[trigger-list] 7660 [FTRACE] Add an event trigger on specific events. 7661 Set a trigger on top of a specific event, with an optional 7662 filter. 7663 7664 The format is "trace_trigger=<event>.<trigger>[ if <filter>],..." 7665 Where more than one trigger may be specified that are comma delimited. 7666 7667 For example: 7668 7669 trace_trigger="sched_switch.stacktrace if prev_state == 2" 7670 7671 The above will enable the "stacktrace" trigger on the "sched_switch" 7672 event but only trigger it if the "prev_state" of the "sched_switch" 7673 event is "2" (TASK_UNINTERRUPTIBLE). 7674 7675 See also "Event triggers" in Documentation/trace/events.rst 7676 7677 7678 traceoff_after_boot 7679 [FTRACE] Sometimes tracing is used to debug issues 7680 during the boot process. Since the trace buffer has a 7681 limited amount of storage, it may be prudent to 7682 disable tracing after the boot is finished, otherwise 7683 the critical information may be overwritten. With this 7684 option, the main tracing buffer will be turned off at 7685 the end of the boot process. 7686 7687 traceoff_on_warning 7688 [FTRACE] enable this option to disable tracing when a 7689 warning is hit. This turns off "tracing_on". Tracing can 7690 be enabled again by echoing '1' into the "tracing_on" 7691 file located in /sys/kernel/tracing/ 7692 7693 This option is useful, as it disables the trace before 7694 the WARNING dump is called, which prevents the trace to 7695 be filled with content caused by the warning output. 7696 7697 This option can also be set at run time via the sysctl 7698 option: kernel/traceoff_on_warning 7699 7700 transparent_hugepage= 7701 [KNL] 7702 Format: [always|madvise|never] 7703 Can be used to control the default behavior of the system 7704 with respect to transparent hugepages. 7705 See Documentation/admin-guide/mm/transhuge.rst 7706 for more details. 7707 7708 transparent_hugepage_shmem= [KNL] 7709 Format: [always|within_size|advise|never|deny|force] 7710 Can be used to control the hugepage allocation policy for 7711 the internal shmem mount. 7712 See Documentation/admin-guide/mm/transhuge.rst 7713 for more details. 7714 7715 transparent_hugepage_tmpfs= [KNL] 7716 Format: [always|within_size|advise|never] 7717 Can be used to control the default hugepage allocation policy 7718 for the tmpfs mount. 7719 See Documentation/admin-guide/mm/transhuge.rst 7720 for more details. 7721 7722 trusted.source= [KEYS] 7723 Format: <string> 7724 This parameter identifies the trust source as a backend 7725 for trusted keys implementation. Supported trust 7726 sources: 7727 - "tpm" 7728 - "tee" 7729 - "caam" 7730 - "dcp" 7731 If not specified then it defaults to iterating through 7732 the trust source list starting with TPM and assigns the 7733 first trust source as a backend which is initialized 7734 successfully during iteration. 7735 7736 trusted.rng= [KEYS] 7737 Format: <string> 7738 The RNG used to generate key material for trusted keys. 7739 Can be one of: 7740 - "kernel" 7741 - the same value as trusted.source: "tpm" or "tee" 7742 - "default" 7743 If not specified, "default" is used. In this case, 7744 the RNG's choice is left to each individual trust source. 7745 7746 trusted.dcp_use_otp_key 7747 This is intended to be used in combination with 7748 trusted.source=dcp and will select the DCP OTP key 7749 instead of the DCP UNIQUE key blob encryption. 7750 7751 trusted.dcp_skip_zk_test 7752 This is intended to be used in combination with 7753 trusted.source=dcp and will disable the check if the 7754 blob key is all zeros. This is helpful for situations where 7755 having this key zero'ed is acceptable. E.g. in testing 7756 scenarios. 7757 7758 tsa= [X86] Control mitigation for Transient Scheduler 7759 Attacks on AMD CPUs. Search the following in your 7760 favourite search engine for more details: 7761 7762 "Technical guidance for mitigating transient scheduler 7763 attacks". 7764 7765 off - disable the mitigation 7766 on - enable the mitigation (default) 7767 user - mitigate only user/kernel transitions 7768 vm - mitigate only guest/host transitions 7769 7770 7771 tsc= Disable clocksource stability checks for TSC. 7772 Format: <string> 7773 [x86] reliable: mark tsc clocksource as reliable, this 7774 disables clocksource verification at runtime, as well 7775 as the stability checks done at bootup. Used to enable 7776 high-resolution timer mode on older hardware, and in 7777 virtualized environment. 7778 [x86] noirqtime: Do not use TSC to do irq accounting. 7779 Used to run time disable IRQ_TIME_ACCOUNTING on any 7780 platforms where RDTSC is slow and this accounting 7781 can add overhead. 7782 [x86] unstable: mark the TSC clocksource as unstable, this 7783 marks the TSC unconditionally unstable at bootup and 7784 avoids any further wobbles once the TSC watchdog notices. 7785 [x86] nowatchdog: disable clocksource watchdog. Used 7786 in situations with strict latency requirements (where 7787 interruptions from clocksource watchdog are not 7788 acceptable). 7789 [x86] recalibrate: force recalibration against a HW timer 7790 (HPET or PM timer) on systems whose TSC frequency was 7791 obtained from HW or FW using either an MSR or CPUID(0x15). 7792 Warn if the difference is more than 500 ppm. 7793 [x86] watchdog: Use TSC as the watchdog clocksource with 7794 which to check other HW timers (HPET or PM timer), but 7795 only on systems where TSC has been deemed trustworthy. 7796 This will be suppressed by an earlier tsc=nowatchdog and 7797 can be overridden by a later tsc=nowatchdog. A console 7798 message will flag any such suppression or overriding. 7799 7800 tsc_early_khz= [X86,EARLY] Skip early TSC calibration and use the given 7801 value instead. Useful when the early TSC frequency discovery 7802 procedure is not reliable, such as on overclocked systems 7803 with CPUID.16h support and partial CPUID.15h support. 7804 Format: <unsigned int> 7805 7806 tsx= [X86] Control Transactional Synchronization 7807 Extensions (TSX) feature in Intel processors that 7808 support TSX control. 7809 7810 This parameter controls the TSX feature. The options are: 7811 7812 on - Enable TSX on the system. Although there are 7813 mitigations for all known security vulnerabilities, 7814 TSX has been known to be an accelerator for 7815 several previous speculation-related CVEs, and 7816 so there may be unknown security risks associated 7817 with leaving it enabled. 7818 7819 off - Disable TSX on the system. (Note that this 7820 option takes effect only on newer CPUs which are 7821 not vulnerable to MDS, i.e., have 7822 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 7823 the new IA32_TSX_CTRL MSR through a microcode 7824 update. This new MSR allows for the reliable 7825 deactivation of the TSX functionality.) 7826 7827 auto - Disable TSX if X86_BUG_TAA is present, 7828 otherwise enable TSX on the system. 7829 7830 Not specifying this option is equivalent to tsx=off. 7831 7832 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 7833 for more details. 7834 7835 tsx_async_abort= [X86,INTEL,EARLY] Control mitigation for the TSX Async 7836 Abort (TAA) vulnerability. 7837 7838 Similar to Micro-architectural Data Sampling (MDS) 7839 certain CPUs that support Transactional 7840 Synchronization Extensions (TSX) are vulnerable to an 7841 exploit against CPU internal buffers which can forward 7842 information to a disclosure gadget under certain 7843 conditions. 7844 7845 In vulnerable processors, the speculatively forwarded 7846 data can be used in a cache side channel attack, to 7847 access data to which the attacker does not have direct 7848 access. 7849 7850 This parameter controls the TAA mitigation. The 7851 options are: 7852 7853 full - Enable TAA mitigation on vulnerable CPUs 7854 if TSX is enabled. 7855 7856 full,nosmt - Enable TAA mitigation and disable SMT on 7857 vulnerable CPUs. If TSX is disabled, SMT 7858 is not disabled because CPU is not 7859 vulnerable to cross-thread TAA attacks. 7860 off - Unconditionally disable TAA mitigation 7861 7862 On MDS-affected machines, tsx_async_abort=off can be 7863 prevented by an active MDS mitigation as both vulnerabilities 7864 are mitigated with the same mechanism so in order to disable 7865 this mitigation, you need to specify mds=off too. 7866 7867 Not specifying this option is equivalent to 7868 tsx_async_abort=full. On CPUs which are MDS affected 7869 and deploy MDS mitigation, TAA mitigation is not 7870 required and doesn't provide any additional 7871 mitigation. 7872 7873 For details see: 7874 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 7875 7876 turbografx.map[2|3]= [HW,JOY] 7877 TurboGraFX parallel port interface 7878 Format: 7879 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 7880 See also Documentation/input/devices/joystick-parport.rst 7881 7882 udbg-immortal [PPC] When debugging early kernel crashes that 7883 happen after console_init() and before a proper 7884 console driver takes over, this boot options might 7885 help "seeing" what's going on. 7886 7887 uhash_entries= [KNL,NET] 7888 Set number of hash buckets for UDP/UDP-Lite connections 7889 7890 uhci-hcd.ignore_oc= 7891 [USB] Ignore overcurrent events (default N). 7892 Some badly-designed motherboards generate lots of 7893 bogus events, for ports that aren't wired to 7894 anything. Set this parameter to avoid log spamming. 7895 Note that genuine overcurrent events won't be 7896 reported either. 7897 7898 unaligned_scalar_speed= 7899 [RISCV] 7900 Format: {slow | fast | unsupported} 7901 Allow skipping scalar unaligned access speed tests. This 7902 is useful for testing alternative code paths and to skip 7903 the tests in environments where they run too slowly. All 7904 CPUs must have the same scalar unaligned access speed. 7905 7906 unaligned_vector_speed= 7907 [RISCV] 7908 Format: {slow | fast | unsupported} 7909 Allow skipping vector unaligned access speed tests. This 7910 is useful for testing alternative code paths and to skip 7911 the tests in environments where they run too slowly. All 7912 CPUs must have the same vector unaligned access speed. 7913 7914 unknown_nmi_panic 7915 [X86] Cause panic on unknown NMI. 7916 7917 unwind_debug [X86-64,EARLY] 7918 Enable unwinder debug output. This can be 7919 useful for debugging certain unwinder error 7920 conditions, including corrupt stacks and 7921 bad/missing unwinder metadata. 7922 7923 usbcore.authorized_default= 7924 [USB] Default USB device authorization: 7925 (default -1 = authorized (same as 1), 7926 0 = not authorized, 1 = authorized, 2 = authorized 7927 if device connected to internal port) 7928 7929 usbcore.autosuspend= 7930 [USB] The autosuspend time delay (in seconds) used 7931 for newly-detected USB devices (default 2). This 7932 is the time required before an idle device will be 7933 autosuspended. Devices for which the delay is set 7934 to a negative value won't be autosuspended at all. 7935 7936 usbcore.usbfs_snoop= 7937 [USB] Set to log all usbfs traffic (default 0 = off). 7938 7939 usbcore.usbfs_snoop_max= 7940 [USB] Maximum number of bytes to snoop in each URB 7941 (default = 65536). 7942 7943 usbcore.blinkenlights= 7944 [USB] Set to cycle leds on hubs (default 0 = off). 7945 7946 usbcore.old_scheme_first= 7947 [USB] Start with the old device initialization 7948 scheme (default 0 = off). 7949 7950 usbcore.usbfs_memory_mb= 7951 [USB] Memory limit (in MB) for buffers allocated by 7952 usbfs (default = 16, 0 = max = 2047). 7953 7954 usbcore.use_both_schemes= 7955 [USB] Try the other device initialization scheme 7956 if the first one fails (default 1 = enabled). 7957 7958 usbcore.initial_descriptor_timeout= 7959 [USB] Specifies timeout for the initial 64-byte 7960 USB_REQ_GET_DESCRIPTOR request in milliseconds 7961 (default 5000 = 5.0 seconds). 7962 7963 usbcore.nousb [USB] Disable the USB subsystem 7964 7965 usbcore.quirks= 7966 [USB] A list of quirk entries to augment the built-in 7967 usb core quirk list. List entries are separated by 7968 commas. Each entry has the form 7969 VendorID:ProductID:Flags. The IDs are 4-digit hex 7970 numbers and Flags is a set of letters. Each letter 7971 will change the built-in quirk; setting it if it is 7972 clear and clearing it if it is set. The letters have 7973 the following meanings: 7974 a = USB_QUIRK_STRING_FETCH_255 (string 7975 descriptors must not be fetched using 7976 a 255-byte read); 7977 b = USB_QUIRK_RESET_RESUME (device can't resume 7978 correctly so reset it instead); 7979 c = USB_QUIRK_NO_SET_INTF (device can't handle 7980 Set-Interface requests); 7981 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 7982 handle its Configuration or Interface 7983 strings); 7984 e = USB_QUIRK_RESET (device can't be reset 7985 (e.g morph devices), don't use reset); 7986 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 7987 more interface descriptions than the 7988 bNumInterfaces count, and can't handle 7989 talking to these interfaces); 7990 g = USB_QUIRK_DELAY_INIT (device needs a pause 7991 during initialization, after we read 7992 the device descriptor); 7993 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 7994 high speed and super speed interrupt 7995 endpoints, the USB 2.0 and USB 3.0 spec 7996 require the interval in microframes (1 7997 microframe = 125 microseconds) to be 7998 calculated as interval = 2 ^ 7999 (bInterval-1). 8000 Devices with this quirk report their 8001 bInterval as the result of this 8002 calculation instead of the exponent 8003 variable used in the calculation); 8004 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 8005 handle device_qualifier descriptor 8006 requests); 8007 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 8008 generates spurious wakeup, ignore 8009 remote wakeup capability); 8010 k = USB_QUIRK_NO_LPM (device can't handle Link 8011 Power Management); 8012 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 8013 (Device reports its bInterval as linear 8014 frames instead of the USB 2.0 8015 calculation); 8016 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 8017 to be disconnected before suspend to 8018 prevent spurious wakeup); 8019 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 8020 pause after every control message); 8021 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 8022 delay after resetting its port); 8023 p = USB_QUIRK_SHORT_SET_ADDRESS_REQ_TIMEOUT 8024 (Reduce timeout of the SET_ADDRESS 8025 request from 5000 ms to 500 ms); 8026 Example: quirks=0781:5580:bk,0a5c:5834:gij 8027 8028 usbhid.mousepoll= 8029 [USBHID] The interval which mice are to be polled at. 8030 8031 usbhid.jspoll= 8032 [USBHID] The interval which joysticks are to be polled at. 8033 8034 usbhid.kbpoll= 8035 [USBHID] The interval which keyboards are to be polled at. 8036 8037 usb-storage.delay_use= 8038 [UMS] The delay in seconds before a new device is 8039 scanned for Logical Units (default 1). 8040 Optionally the delay in milliseconds if the value has 8041 suffix with "ms". 8042 Example: delay_use=2567ms 8043 8044 usb-storage.quirks= 8045 [UMS] A list of quirks entries to supplement or 8046 override the built-in unusual_devs list. List 8047 entries are separated by commas. Each entry has 8048 the form VID:PID:Flags where VID and PID are Vendor 8049 and Product ID values (4-digit hex numbers) and 8050 Flags is a set of characters, each corresponding 8051 to a common usb-storage quirk flag as follows: 8052 a = SANE_SENSE (collect more than 18 bytes 8053 of sense data, not on uas); 8054 b = BAD_SENSE (don't collect more than 18 8055 bytes of sense data, not on uas); 8056 c = FIX_CAPACITY (decrease the reported 8057 device capacity by one sector); 8058 d = NO_READ_DISC_INFO (don't use 8059 READ_DISC_INFO command, not on uas); 8060 e = NO_READ_CAPACITY_16 (don't use 8061 READ_CAPACITY_16 command); 8062 f = NO_REPORT_OPCODES (don't use report opcodes 8063 command, uas only); 8064 g = MAX_SECTORS_240 (don't transfer more than 8065 240 sectors at a time, uas only); 8066 h = CAPACITY_HEURISTICS (decrease the 8067 reported device capacity by one 8068 sector if the number is odd); 8069 i = IGNORE_DEVICE (don't bind to this 8070 device); 8071 j = NO_REPORT_LUNS (don't use report luns 8072 command, uas only); 8073 k = NO_SAME (do not use WRITE_SAME, uas only) 8074 l = NOT_LOCKABLE (don't try to lock and 8075 unlock ejectable media, not on uas); 8076 m = MAX_SECTORS_64 (don't transfer more 8077 than 64 sectors = 32 KB at a time, 8078 not on uas); 8079 n = INITIAL_READ10 (force a retry of the 8080 initial READ(10) command, not on uas); 8081 o = CAPACITY_OK (accept the capacity 8082 reported by the device, not on uas); 8083 p = WRITE_CACHE (the device cache is ON 8084 by default, not on uas); 8085 r = IGNORE_RESIDUE (the device reports 8086 bogus residue values, not on uas); 8087 s = SINGLE_LUN (the device has only one 8088 Logical Unit); 8089 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 8090 commands, uas only); 8091 u = IGNORE_UAS (don't bind to the uas driver); 8092 w = NO_WP_DETECT (don't test whether the 8093 medium is write-protected). 8094 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 8095 even if the device claims no cache, 8096 not on uas) 8097 Example: quirks=0419:aaf5:rl,0421:0433:rc 8098 8099 user_debug= [KNL,ARM] 8100 Format: <int> 8101 See arch/arm/Kconfig.debug help text. 8102 1 - undefined instruction events 8103 2 - system calls 8104 4 - invalid data aborts 8105 8 - SIGSEGV faults 8106 16 - SIGBUS faults 8107 Example: user_debug=31 8108 8109 vdso= [X86,SH,SPARC] 8110 On X86_32, this is an alias for vdso32=. Otherwise: 8111 8112 vdso=1: enable VDSO (the default) 8113 vdso=0: disable VDSO mapping 8114 8115 vdso32= [X86] Control the 32-bit vDSO 8116 vdso32=1: enable 32-bit VDSO 8117 vdso32=0 or vdso32=2: disable 32-bit VDSO 8118 8119 See the help text for CONFIG_COMPAT_VDSO for more 8120 details. If CONFIG_COMPAT_VDSO is set, the default is 8121 vdso32=0; otherwise, the default is vdso32=1. 8122 8123 For compatibility with older kernels, vdso32=2 is an 8124 alias for vdso32=0. 8125 8126 Try vdso32=0 if you encounter an error that says: 8127 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 8128 8129 video= [FB,EARLY] Frame buffer configuration 8130 See Documentation/fb/modedb.rst. 8131 8132 video.brightness_switch_enabled= [ACPI] 8133 Format: [0|1] 8134 If set to 1, on receiving an ACPI notify event 8135 generated by hotkey, video driver will adjust brightness 8136 level and then send out the event to user space through 8137 the allocated input device. If set to 0, video driver 8138 will only send out the event without touching backlight 8139 brightness level. 8140 default: 1 8141 8142 virtio_mmio.device= 8143 [VMMIO] Memory mapped virtio (platform) device. 8144 8145 <size>@<baseaddr>:<irq>[:<id>] 8146 where: 8147 <size> := size (can use standard suffixes 8148 like K, M and G) 8149 <baseaddr> := physical base address 8150 <irq> := interrupt number (as passed to 8151 request_irq()) 8152 <id> := (optional) platform device id 8153 example: 8154 virtio_mmio.device=1K@0x100b0000:48:7 8155 8156 Can be used multiple times for multiple devices. 8157 8158 vga= [BOOT,X86-32] Select a particular video mode 8159 See Documentation/arch/x86/boot.rst and 8160 Documentation/admin-guide/svga.rst. 8161 Use vga=ask for menu. 8162 This is actually a boot loader parameter; the value is 8163 passed to the kernel using a special protocol. 8164 8165 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 8166 May slow down system boot speed, especially when 8167 enabled on systems with a large amount of memory. 8168 All options are enabled by default, and this 8169 interface is meant to allow for selectively 8170 enabling or disabling specific virtual memory 8171 debugging features. 8172 8173 Available options are: 8174 P Enable page structure init time poisoning 8175 - Disable all of the above options 8176 8177 vmalloc=nn[KMG] [KNL,BOOT,EARLY] Forces the vmalloc area to have an 8178 exact size of <nn>. This can be used to increase 8179 the minimum size (128MB on x86, arm32 platforms). 8180 It can also be used to decrease the size and leave more room 8181 for directly mapped kernel RAM. Note that this parameter does 8182 not exist on many other platforms (including arm64, alpha, 8183 loongarch, arc, csky, hexagon, microblaze, mips, nios2, openrisc, 8184 parisc, m64k, powerpc, riscv, sh, um, xtensa, s390, sparc). 8185 8186 vmcp_cma=nn[MG] [KNL,S390,EARLY] 8187 Sets the memory size reserved for contiguous memory 8188 allocations for the vmcp device driver. 8189 8190 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 8191 Format: <command> 8192 8193 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 8194 Format: <command> 8195 8196 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 8197 Format: <command> 8198 8199 vmscape= [X86] Controls mitigation for VMscape attacks. 8200 VMscape attacks can leak information from a userspace 8201 hypervisor to a guest via speculative side-channels. 8202 8203 off - disable the mitigation 8204 ibpb - use Indirect Branch Prediction Barrier 8205 (IBPB) mitigation (default) 8206 force - force vulnerability detection even on 8207 unaffected processors 8208 8209 vsyscall= [X86-64,EARLY] 8210 Controls the behavior of vsyscalls (i.e. calls to 8211 fixed addresses of 0xffffffffff600x00 from legacy 8212 code). Most statically-linked binaries and older 8213 versions of glibc use these calls. Because these 8214 functions are at fixed addresses, they make nice 8215 targets for exploits that can control RIP. 8216 8217 emulate Vsyscalls turn into traps and are emulated 8218 reasonably safely. The vsyscall page is 8219 readable. 8220 8221 xonly [default] Vsyscalls turn into traps and are 8222 emulated reasonably safely. The vsyscall 8223 page is not readable. 8224 8225 none Vsyscalls don't work at all. This makes 8226 them quite hard to use for exploits but 8227 might break your system. 8228 8229 vt.color= [VT] Default text color. 8230 Format: 0xYX, X = foreground, Y = background. 8231 Default: 0x07 = light gray on black. 8232 8233 vt.cur_default= [VT] Default cursor shape. 8234 Format: 0xCCBBAA, where AA, BB, and CC are the same as 8235 the parameters of the <Esc>[?A;B;Cc escape sequence; 8236 see vga-softcursor.rst. Default: 2 = underline. 8237 8238 vt.default_blu= [VT] 8239 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 8240 Change the default blue palette of the console. 8241 This is a 16-member array composed of values 8242 ranging from 0-255. 8243 8244 vt.default_grn= [VT] 8245 Format: <green0>,<green1>,<green2>,...,<green15> 8246 Change the default green palette of the console. 8247 This is a 16-member array composed of values 8248 ranging from 0-255. 8249 8250 vt.default_red= [VT] 8251 Format: <red0>,<red1>,<red2>,...,<red15> 8252 Change the default red palette of the console. 8253 This is a 16-member array composed of values 8254 ranging from 0-255. 8255 8256 vt.default_utf8= 8257 [VT] 8258 Format=<0|1> 8259 Set system-wide default UTF-8 mode for all tty's. 8260 Default is 1, i.e. UTF-8 mode is enabled for all 8261 newly opened terminals. 8262 8263 vt.global_cursor_default= 8264 [VT] 8265 Format=<-1|0|1> 8266 Set system-wide default for whether a cursor 8267 is shown on new VTs. Default is -1, 8268 i.e. cursors will be created by default unless 8269 overridden by individual drivers. 0 will hide 8270 cursors, 1 will display them. 8271 8272 vt.italic= [VT] Default color for italic text; 0-15. 8273 Default: 2 = green. 8274 8275 vt.underline= [VT] Default color for underlined text; 0-15. 8276 Default: 3 = cyan. 8277 8278 watchdog timers [HW,WDT] For information on watchdog timers, 8279 see Documentation/watchdog/watchdog-parameters.rst 8280 or other driver-specific files in the 8281 Documentation/watchdog/ directory. 8282 8283 watchdog_thresh= 8284 [KNL] 8285 Set the hard lockup detector stall duration 8286 threshold in seconds. The soft lockup detector 8287 threshold is set to twice the value. A value of 0 8288 disables both lockup detectors. Default is 10 8289 seconds. 8290 8291 workqueue.unbound_cpus= 8292 [KNL,SMP] Specify to constrain one or some CPUs 8293 to use in unbound workqueues. 8294 Format: <cpu-list> 8295 By default, all online CPUs are available for 8296 unbound workqueues. 8297 8298 workqueue.watchdog_thresh= 8299 If CONFIG_WQ_WATCHDOG is configured, workqueue can 8300 warn stall conditions and dump internal state to 8301 help debugging. 0 disables workqueue stall 8302 detection; otherwise, it's the stall threshold 8303 duration in seconds. The default value is 30 and 8304 it can be updated at runtime by writing to the 8305 corresponding sysfs file. 8306 8307 workqueue.panic_on_stall=<uint> 8308 Panic when workqueue stall is detected by 8309 CONFIG_WQ_WATCHDOG. It sets the number times of the 8310 stall to trigger panic. 8311 8312 The default is 0, which disables the panic on stall. 8313 8314 workqueue.cpu_intensive_thresh_us= 8315 Per-cpu work items which run for longer than this 8316 threshold are automatically considered CPU intensive 8317 and excluded from concurrency management to prevent 8318 them from noticeably delaying other per-cpu work 8319 items. Default is 10000 (10ms). 8320 8321 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8322 will report the work functions which violate this 8323 threshold repeatedly. They are likely good 8324 candidates for using WQ_UNBOUND workqueues instead. 8325 8326 workqueue.cpu_intensive_warning_thresh=<uint> 8327 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8328 will report the work functions which violate the 8329 intensive_threshold_us repeatedly. In order to prevent 8330 spurious warnings, start printing only after a work 8331 function has violated this threshold number of times. 8332 8333 The default is 4 times. 0 disables the warning. 8334 8335 workqueue.power_efficient 8336 Per-cpu workqueues are generally preferred because 8337 they show better performance thanks to cache 8338 locality; unfortunately, per-cpu workqueues tend to 8339 be more power hungry than unbound workqueues. 8340 8341 Enabling this makes the per-cpu workqueues which 8342 were observed to contribute significantly to power 8343 consumption unbound, leading to measurably lower 8344 power usage at the cost of small performance 8345 overhead. 8346 8347 The default value of this parameter is determined by 8348 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 8349 8350 workqueue.default_affinity_scope= 8351 Select the default affinity scope to use for unbound 8352 workqueues. Can be one of "cpu", "smt", "cache", 8353 "numa" and "system". Default is "cache". For more 8354 information, see the Affinity Scopes section in 8355 Documentation/core-api/workqueue.rst. 8356 8357 This can be changed after boot by writing to the 8358 matching /sys/module/workqueue/parameters file. All 8359 workqueues with the "default" affinity scope will be 8360 updated accordingly. 8361 8362 workqueue.debug_force_rr_cpu 8363 Workqueue used to implicitly guarantee that work 8364 items queued without explicit CPU specified are put 8365 on the local CPU. This guarantee is no longer true 8366 and while local CPU is still preferred work items 8367 may be put on foreign CPUs. This debug option 8368 forces round-robin CPU selection to flush out 8369 usages which depend on the now broken guarantee. 8370 When enabled, memory and cache locality will be 8371 impacted. 8372 8373 writecombine= [LOONGARCH,EARLY] Control the MAT (Memory Access 8374 Type) of ioremap_wc(). 8375 8376 on - Enable writecombine, use WUC for ioremap_wc() 8377 off - Disable writecombine, use SUC for ioremap_wc() 8378 8379 x2apic_phys [X86-64,APIC,EARLY] Use x2apic physical mode instead of 8380 default x2apic cluster mode on platforms 8381 supporting x2apic. 8382 8383 xen_512gb_limit [KNL,X86-64,XEN] 8384 Restricts the kernel running paravirtualized under Xen 8385 to use only up to 512 GB of RAM. The reason to do so is 8386 crash analysis tools and Xen tools for doing domain 8387 save/restore/migration must be enabled to handle larger 8388 domains. 8389 8390 xen_emul_unplug= [HW,X86,XEN,EARLY] 8391 Unplug Xen emulated devices 8392 Format: [unplug0,][unplug1] 8393 ide-disks -- unplug primary master IDE devices 8394 aux-ide-disks -- unplug non-primary-master IDE devices 8395 nics -- unplug network devices 8396 all -- unplug all emulated devices (NICs and IDE disks) 8397 unnecessary -- unplugging emulated devices is 8398 unnecessary even if the host did not respond to 8399 the unplug protocol 8400 never -- do not unplug even if version check succeeds 8401 8402 xen_legacy_crash [X86,XEN,EARLY] 8403 Crash from Xen panic notifier, without executing late 8404 panic() code such as dumping handler. 8405 8406 xen_mc_debug [X86,XEN,EARLY] 8407 Enable multicall debugging when running as a Xen PV guest. 8408 Enabling this feature will reduce performance a little 8409 bit, so it should only be enabled for obtaining extended 8410 debug data in case of multicall errors. 8411 8412 xen_msr_safe= [X86,XEN,EARLY] 8413 Format: <bool> 8414 Select whether to always use non-faulting (safe) MSR 8415 access functions when running as Xen PV guest. The 8416 default value is controlled by CONFIG_XEN_PV_MSR_SAFE. 8417 8418 xen_nopv [X86] 8419 Disables the PV optimizations forcing the HVM guest to 8420 run as generic HVM guest with no PV drivers. 8421 This option is obsoleted by the "nopv" option, which 8422 has equivalent effect for XEN platform. 8423 8424 xen_no_vector_callback 8425 [KNL,X86,XEN,EARLY] Disable the vector callback for Xen 8426 event channel interrupts. 8427 8428 xen_scrub_pages= [XEN] 8429 Boolean option to control scrubbing pages before giving them back 8430 to Xen, for use by other domains. Can be also changed at runtime 8431 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 8432 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 8433 8434 xen_timer_slop= [X86-64,XEN,EARLY] 8435 Set the timer slop (in nanoseconds) for the virtual Xen 8436 timers (default is 100000). This adjusts the minimum 8437 delta of virtualized Xen timers, where lower values 8438 improve timer resolution at the expense of processing 8439 more timer interrupts. 8440 8441 xen.balloon_boot_timeout= [XEN] 8442 The time (in seconds) to wait before giving up to boot 8443 in case initial ballooning fails to free enough memory. 8444 Applies only when running as HVM or PVH guest and 8445 started with less memory configured than allowed at 8446 max. Default is 180. 8447 8448 xen.event_eoi_delay= [XEN] 8449 How long to delay EOI handling in case of event 8450 storms (jiffies). Default is 10. 8451 8452 xen.event_loop_timeout= [XEN] 8453 After which time (jiffies) the event handling loop 8454 should start to delay EOI handling. Default is 2. 8455 8456 xen.fifo_events= [XEN] 8457 Boolean parameter to disable using fifo event handling 8458 even if available. Normally fifo event handling is 8459 preferred over the 2-level event handling, as it is 8460 fairer and the number of possible event channels is 8461 much higher. Default is on (use fifo events). 8462 8463 xirc2ps_cs= [NET,PCMCIA] 8464 Format: 8465 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 8466 8467 xive= [PPC] 8468 By default on POWER9 and above, the kernel will 8469 natively use the XIVE interrupt controller. This option 8470 allows the fallback firmware mode to be used: 8471 8472 off Fallback to firmware control of XIVE interrupt 8473 controller on both pseries and powernv 8474 platforms. Only useful on POWER9 and above. 8475 8476 xive.store-eoi=off [PPC] 8477 By default on POWER10 and above, the kernel will use 8478 stores for EOI handling when the XIVE interrupt mode 8479 is active. This option allows the XIVE driver to use 8480 loads instead, as on POWER9. 8481 8482 xhci-hcd.quirks [USB,KNL] 8483 A hex value specifying bitmask with supplemental xhci 8484 host controller quirks. Meaning of each bit can be 8485 consulted in header drivers/usb/host/xhci.h. 8486 8487 xmon [PPC,EARLY] 8488 Format: { early | on | rw | ro | off } 8489 Controls if xmon debugger is enabled. Default is off. 8490 Passing only "xmon" is equivalent to "xmon=early". 8491 early Call xmon as early as possible on boot; xmon 8492 debugger is called from setup_arch(). 8493 on xmon debugger hooks will be installed so xmon 8494 is only called on a kernel crash. Default mode, 8495 i.e. either "ro" or "rw" mode, is controlled 8496 with CONFIG_XMON_DEFAULT_RO_MODE. 8497 rw xmon debugger hooks will be installed so xmon 8498 is called only on a kernel crash, mode is write, 8499 meaning SPR registers, memory and, other data 8500 can be written using xmon commands. 8501 ro same as "rw" option above but SPR registers, 8502 memory, and other data can't be written using 8503 xmon commands. 8504 off xmon is disabled. 8505