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