1 accept_memory= [MM] 2 Format: { eager | lazy } 3 default: lazy 4 By default, unaccepted memory is accepted lazily to 5 avoid prolonged boot times. The lazy option will add 6 some runtime overhead until all memory is eventually 7 accepted. In most cases the overhead is negligible. 8 For some workloads or for debugging purposes 9 accept_memory=eager can be used to accept all memory 10 at once during boot. 11 12 acpi= [HW,ACPI,X86,ARM64,RISCV64,EARLY] 13 Advanced Configuration and Power Interface 14 Format: { force | on | off | strict | noirq | rsdt | 15 copy_dsdt | nospcr } 16 force -- enable ACPI if default was off 17 on -- enable ACPI but allow fallback to DT [arm64,riscv64] 18 off -- disable ACPI if default was on 19 noirq -- do not use ACPI for IRQ routing 20 strict -- Be less tolerant of platforms that are not 21 strictly ACPI specification compliant. 22 rsdt -- prefer RSDT over (default) XSDT 23 copy_dsdt -- copy DSDT to memory 24 nocmcff -- Disable firmware first mode for corrected 25 errors. This disables parsing the HEST CMC error 26 source to check if firmware has set the FF flag. This 27 may result in duplicate corrected error reports. 28 nospcr -- disable console in ACPI SPCR table as 29 default _serial_ console on ARM64 30 For ARM64, ONLY "acpi=off", "acpi=on", "acpi=force" or 31 "acpi=nospcr" are available 32 For RISCV64, ONLY "acpi=off", "acpi=on" or "acpi=force" 33 are available 34 35 See also Documentation/power/runtime_pm.rst, pci=noacpi 36 37 acpi_apic_instance= [ACPI,IOAPIC,EARLY] 38 Format: <int> 39 2: use 2nd APIC table, if available 40 1,0: use 1st APIC table 41 default: 0 42 43 acpi_backlight= [HW,ACPI] 44 { vendor | video | native | none } 45 If set to vendor, prefer vendor-specific driver 46 (e.g. thinkpad_acpi, sony_acpi, etc.) instead 47 of the ACPI video.ko driver. 48 If set to video, use the ACPI video.ko driver. 49 If set to native, use the device's native backlight mode. 50 If set to none, disable the ACPI backlight interface. 51 52 acpi_force_32bit_fadt_addr [ACPI,EARLY] 53 force FADT to use 32 bit addresses rather than the 54 64 bit X_* addresses. Some firmware have broken 64 55 bit addresses for force ACPI ignore these and use 56 the older legacy 32 bit addresses. 57 58 acpica_no_return_repair [HW, ACPI] 59 Disable AML predefined validation mechanism 60 This mechanism can repair the evaluation result to make 61 the return objects more ACPI specification compliant. 62 This option is useful for developers to identify the 63 root cause of an AML interpreter issue when the issue 64 has something to do with the repair mechanism. 65 66 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG] 67 acpi.debug_level= [HW,ACPI,ACPI_DEBUG] 68 Format: <int> 69 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI 70 debug output. Bits in debug_layer correspond to a 71 _COMPONENT in an ACPI source file, e.g., 72 #define _COMPONENT ACPI_EVENTS 73 Bits in debug_level correspond to a level in 74 ACPI_DEBUG_PRINT statements, e.g., 75 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ... 76 The debug_level mask defaults to "info". See 77 Documentation/firmware-guide/acpi/debug.rst for more information about 78 debug layers and levels. 79 80 Enable processor driver info messages: 81 acpi.debug_layer=0x20000000 82 Enable AML "Debug" output, i.e., stores to the Debug 83 object while interpreting AML: 84 acpi.debug_layer=0xffffffff acpi.debug_level=0x2 85 Enable all messages related to ACPI hardware: 86 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff 87 88 Some values produce so much output that the system is 89 unusable. The "log_buf_len" parameter may be useful 90 if you need to capture more output. 91 92 acpi_enforce_resources= [ACPI] 93 { strict | lax | no } 94 Check for resource conflicts between native drivers 95 and ACPI OperationRegions (SystemIO and SystemMemory 96 only). IO ports and memory declared in ACPI might be 97 used by the ACPI subsystem in arbitrary AML code and 98 can interfere with legacy drivers. 99 strict (default): access to resources claimed by ACPI 100 is denied; legacy drivers trying to access reserved 101 resources will fail to bind to device using them. 102 lax: access to resources claimed by ACPI is allowed; 103 legacy drivers trying to access reserved resources 104 will bind successfully but a warning message is logged. 105 no: ACPI OperationRegions are not marked as reserved, 106 no further checks are performed. 107 108 acpi_force_table_verification [HW,ACPI,EARLY] 109 Enable table checksum verification during early stage. 110 By default, this is disabled due to x86 early mapping 111 size limitation. 112 113 acpi_irq_balance [HW,ACPI] 114 ACPI will balance active IRQs 115 default in APIC mode 116 117 acpi_irq_nobalance [HW,ACPI] 118 ACPI will not move active IRQs (default) 119 default in PIC mode 120 121 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 122 Format: <irq>,<irq>... 123 124 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 125 use by PCI 126 Format: <irq>,<irq>... 127 128 acpi_mask_gpe= [HW,ACPI] 129 Due to the existence of _Lxx/_Exx, some GPEs triggered 130 by unsupported hardware/firmware features can result in 131 GPE floodings that cannot be automatically disabled by 132 the GPE dispatcher. 133 This facility can be used to prevent such uncontrolled 134 GPE floodings. 135 Format: <byte> or <bitmap-list> 136 137 acpi_no_auto_serialize [HW,ACPI] 138 Disable auto-serialization of AML methods 139 AML control methods that contain the opcodes to create 140 named objects will be marked as "Serialized" by the 141 auto-serialization feature. 142 This feature is enabled by default. 143 This option allows to turn off the feature. 144 145 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 146 kernels. 147 148 acpi_no_static_ssdt [HW,ACPI,EARLY] 149 Disable installation of static SSDTs at early boot time 150 By default, SSDTs contained in the RSDT/XSDT will be 151 installed automatically and they will appear under 152 /sys/firmware/acpi/tables. 153 This option turns off this feature. 154 Note that specifying this option does not affect 155 dynamic table installation which will install SSDT 156 tables to /sys/firmware/acpi/tables/dynamic. 157 158 acpi_no_watchdog [HW,ACPI,WDT] 159 Ignore the ACPI-based watchdog interface (WDAT) and let 160 a native driver control the watchdog device instead. 161 162 acpi_rsdp= [ACPI,EFI,KEXEC,EARLY] 163 Pass the RSDP address to the kernel, mostly used 164 on machines running EFI runtime service to boot the 165 second kernel for kdump. 166 167 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 168 Format: To spoof as Windows 98: ="Microsoft Windows" 169 170 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 171 of 2 which is mandated by ACPI 6) as the supported ACPI 172 specification revision (when using this switch, it may 173 be necessary to carry out a cold reboot _twice_ in a 174 row to make it take effect on the platform firmware). 175 176 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 177 acpi_osi="string1" # add string1 178 acpi_osi="!string2" # remove string2 179 acpi_osi=!* # remove all strings 180 acpi_osi=! # disable all built-in OS vendor 181 strings 182 acpi_osi=!! # enable all built-in OS vendor 183 strings 184 acpi_osi= # disable all strings 185 186 'acpi_osi=!' can be used in combination with single or 187 multiple 'acpi_osi="string1"' to support specific OS 188 vendor string(s). Note that such command can only 189 affect the default state of the OS vendor strings, thus 190 it cannot affect the default state of the feature group 191 strings and the current state of the OS vendor strings, 192 specifying it multiple times through kernel command line 193 is meaningless. This command is useful when one do not 194 care about the state of the feature group strings which 195 should be controlled by the OSPM. 196 Examples: 197 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 198 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 199 can make '_OSI("Windows 2000")' TRUE. 200 201 'acpi_osi=' cannot be used in combination with other 202 'acpi_osi=' command lines, the _OSI method will not 203 exist in the ACPI namespace. NOTE that such command can 204 only affect the _OSI support state, thus specifying it 205 multiple times through kernel command line is also 206 meaningless. 207 Examples: 208 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 209 FALSE. 210 211 'acpi_osi=!*' can be used in combination with single or 212 multiple 'acpi_osi="string1"' to support specific 213 string(s). Note that such command can affect the 214 current state of both the OS vendor strings and the 215 feature group strings, thus specifying it multiple times 216 through kernel command line is meaningful. But it may 217 still not able to affect the final state of a string if 218 there are quirks related to this string. This command 219 is useful when one want to control the state of the 220 feature group strings to debug BIOS issues related to 221 the OSPM features. 222 Examples: 223 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 224 '_OSI("Module Device")' FALSE. 225 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 226 '_OSI("Module Device")' TRUE. 227 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 228 equivalent to 229 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 230 and 231 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 232 they all will make '_OSI("Windows 2000")' TRUE. 233 234 acpi_pm_good [X86] 235 Override the pmtimer bug detection: force the kernel 236 to assume that this machine's pmtimer latches its value 237 and always returns good values. 238 239 acpi_sci= [HW,ACPI,EARLY] ACPI System Control Interrupt trigger mode 240 Format: { level | edge | high | low } 241 242 acpi_skip_timer_override [HW,ACPI,EARLY] 243 Recognize and ignore IRQ0/pin2 Interrupt Override. 244 For broken nForce2 BIOS resulting in XT-PIC timer. 245 246 acpi_sleep= [HW,ACPI] Sleep options 247 Format: { s3_bios, s3_mode, s3_beep, s4_hwsig, 248 s4_nohwsig, old_ordering, nonvs, 249 sci_force_enable, nobl } 250 See Documentation/power/video.rst for information on 251 s3_bios and s3_mode. 252 s3_beep is for debugging; it makes the PC's speaker beep 253 as soon as the kernel's real-mode entry point is called. 254 s4_hwsig causes the kernel to check the ACPI hardware 255 signature during resume from hibernation, and gracefully 256 refuse to resume if it has changed. This complies with 257 the ACPI specification but not with reality, since 258 Windows does not do this and many laptops do change it 259 on docking. So the default behaviour is to allow resume 260 and simply warn when the signature changes, unless the 261 s4_hwsig option is enabled. 262 s4_nohwsig prevents ACPI hardware signature from being 263 used (or even warned about) during resume. 264 old_ordering causes the ACPI 1.0 ordering of the _PTS 265 control method, with respect to putting devices into 266 low power states, to be enforced (the ACPI 2.0 ordering 267 of _PTS is used by default). 268 nonvs prevents the kernel from saving/restoring the 269 ACPI NVS memory during suspend/hibernation and resume. 270 sci_force_enable causes the kernel to set SCI_EN directly 271 on resume from S1/S3 (which is against the ACPI spec, 272 but some broken systems don't work without it). 273 nobl causes the internal blacklist of systems known to 274 behave incorrectly in some ways with respect to system 275 suspend and resume to be ignored (use wisely). 276 277 acpi_use_timer_override [HW,ACPI,EARLY] 278 Use timer override. For some broken Nvidia NF5 boards 279 that require a timer override, but don't have HPET 280 281 add_efi_memmap [EFI,X86,EARLY] Include EFI memory map in 282 kernel's map of available physical RAM. 283 284 agp= [AGP] 285 { off | try_unsupported } 286 off: disable AGP support 287 try_unsupported: try to drive unsupported chipsets 288 (may crash computer or cause data corruption) 289 290 ALSA [HW,ALSA] 291 See Documentation/sound/alsa-configuration.rst 292 293 alignment= [KNL,ARM] 294 Allow the default userspace alignment fault handler 295 behaviour to be specified. Bit 0 enables warnings, 296 bit 1 enables fixups, and bit 2 sends a segfault. 297 298 align_va_addr= [X86-64] 299 Align virtual addresses by clearing slice [14:12] when 300 allocating a VMA at process creation time. This option 301 gives you up to 3% performance improvement on AMD F15h 302 machines (where it is enabled by default) for a 303 CPU-intensive style benchmark, and it can vary highly in 304 a microbenchmark depending on workload and compiler. 305 306 32: only for 32-bit processes 307 64: only for 64-bit processes 308 on: enable for both 32- and 64-bit processes 309 off: disable for both 32- and 64-bit processes 310 311 alloc_snapshot [FTRACE] 312 Allocate the ftrace snapshot buffer on boot up when the 313 main buffer is allocated. This is handy if debugging 314 and you need to use tracing_snapshot() on boot up, and 315 do not want to use tracing_snapshot_alloc() as it needs 316 to be done where GFP_KERNEL allocations are allowed. 317 318 allow_mismatched_32bit_el0 [ARM64,EARLY] 319 Allow execve() of 32-bit applications and setting of the 320 PER_LINUX32 personality on systems where only a strict 321 subset of the CPUs support 32-bit EL0. When this 322 parameter is present, the set of CPUs supporting 32-bit 323 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0 324 and hot-unplug operations may be restricted. 325 326 See Documentation/arch/arm64/asymmetric-32bit.rst for more 327 information. 328 329 amd_iommu= [HW,X86-64] 330 Pass parameters to the AMD IOMMU driver in the system. 331 Possible values are: 332 fullflush - Deprecated, equivalent to iommu.strict=1 333 off - do not initialize any AMD IOMMU found in 334 the system 335 force_isolation - Force device isolation for all 336 devices. The IOMMU driver is not 337 allowed anymore to lift isolation 338 requirements as needed. This option 339 does not override iommu=pt 340 force_enable - Force enable the IOMMU on platforms known 341 to be buggy with IOMMU enabled. Use this 342 option with care. 343 pgtbl_v1 - Use v1 page table for DMA-API (Default). 344 pgtbl_v2 - Use v2 page table for DMA-API. 345 irtcachedis - Disable Interrupt Remapping Table (IRT) caching. 346 nohugepages - Limit page-sizes used for v1 page-tables 347 to 4 KiB. 348 v2_pgsizes_only - Limit page-sizes used for v1 page-tables 349 to 4KiB/2Mib/1GiB. 350 351 352 amd_iommu_dump= [HW,X86-64] 353 Enable AMD IOMMU driver option to dump the ACPI table 354 for AMD IOMMU. With this option enabled, AMD IOMMU 355 driver will print ACPI tables for AMD IOMMU during 356 IOMMU initialization. 357 358 amd_iommu_intr= [HW,X86-64] 359 Specifies one of the following AMD IOMMU interrupt 360 remapping modes: 361 legacy - Use legacy interrupt remapping mode. 362 vapic - Use virtual APIC mode, which allows IOMMU 363 to inject interrupts directly into guest. 364 This mode requires kvm-amd.avic=1. 365 (Default when IOMMU HW support is present.) 366 367 amd_pstate= [X86,EARLY] 368 disable 369 Do not enable amd_pstate as the default 370 scaling driver for the supported processors 371 passive 372 Use amd_pstate with passive mode as a scaling driver. 373 In this mode autonomous selection is disabled. 374 Driver requests a desired performance level and platform 375 tries to match the same performance level if it is 376 satisfied by guaranteed performance level. 377 active 378 Use amd_pstate_epp driver instance as the scaling driver, 379 driver provides a hint to the hardware if software wants 380 to bias toward performance (0x0) or energy efficiency (0xff) 381 to the CPPC firmware. then CPPC power algorithm will 382 calculate the runtime workload and adjust the realtime cores 383 frequency. 384 guided 385 Activate guided autonomous mode. Driver requests minimum and 386 maximum performance level and the platform autonomously 387 selects a performance level in this range and appropriate 388 to the current workload. 389 390 amd_prefcore= 391 [X86] 392 disable 393 Disable amd-pstate preferred core. 394 395 amijoy.map= [HW,JOY] Amiga joystick support 396 Map of devices attached to JOY0DAT and JOY1DAT 397 Format: <a>,<b> 398 See also Documentation/input/joydev/joystick.rst 399 400 analog.map= [HW,JOY] Analog joystick and gamepad support 401 Specifies type or capabilities of an analog joystick 402 connected to one of 16 gameports 403 Format: <type1>,<type2>,..<type16> 404 405 apc= [HW,SPARC] 406 Power management functions (SPARCstation-4/5 + deriv.) 407 Format: noidle 408 Disable APC CPU standby support. SPARCstation-Fox does 409 not play well with APC CPU idle - disable it if you have 410 APC and your system crashes randomly. 411 412 apic [APIC,X86-64] Use IO-APIC. Default. 413 414 apic= [APIC,X86,EARLY] Advanced Programmable Interrupt Controller 415 Change the output verbosity while booting 416 Format: { quiet (default) | verbose | debug } 417 Change the amount of debugging information output 418 when initialising the APIC and IO-APIC components. 419 420 apic_extnmi= [APIC,X86,EARLY] External NMI delivery setting 421 Format: { bsp (default) | all | none } 422 bsp: External NMI is delivered only to CPU 0 423 all: External NMIs are broadcast to all CPUs as a 424 backup of CPU 0 425 none: External NMI is masked for all CPUs. This is 426 useful so that a dump capture kernel won't be 427 shot down by NMI 428 429 apicpmtimer Do APIC timer calibration using the pmtimer. Implies 430 apicmaintimer. Useful when your PIT timer is totally 431 broken. 432 433 autoconf= [IPV6] 434 See Documentation/networking/ipv6.rst. 435 436 apm= [APM] Advanced Power Management 437 See header of arch/x86/kernel/apm_32.c. 438 439 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 440 Format: { "0" | "1" } 441 See security/apparmor/Kconfig help text 442 0 -- disable. 443 1 -- enable. 444 Default value is set via kernel config option. 445 446 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards 447 Format: <io>,<irq>,<nodeID> 448 449 arm64.no32bit_el0 [ARM64] Unconditionally disable the execution of 450 32 bit applications. 451 452 arm64.nobti [ARM64] Unconditionally disable Branch Target 453 Identification support 454 455 arm64.nogcs [ARM64] Unconditionally disable Guarded Control Stack 456 support 457 458 arm64.nomops [ARM64] Unconditionally disable Memory Copy and Memory 459 Set instructions support 460 461 arm64.nompam [ARM64] Unconditionally disable Memory Partitioning And 462 Monitoring support 463 464 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension 465 support 466 467 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 468 support 469 470 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix 471 Extension support 472 473 arm64.nosve [ARM64] Unconditionally disable Scalable Vector 474 Extension support 475 476 ataflop= [HW,M68k] 477 478 atarimouse= [HW,MOUSE] Atari Mouse 479 480 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 481 EzKey and similar keyboards 482 483 atkbd.reset= [HW] Reset keyboard during initialization 484 485 atkbd.set= [HW] Select keyboard code set 486 Format: <int> (2 = AT (default), 3 = PS/2) 487 488 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 489 keyboards 490 491 atkbd.softraw= [HW] Choose between synthetic and real raw mode 492 Format: <bool> (0 = real, 1 = synthetic (default)) 493 494 atkbd.softrepeat= [HW] 495 Use software keyboard repeat 496 497 audit= [KNL] Enable the audit sub-system 498 Format: { "0" | "1" | "off" | "on" } 499 0 | off - kernel audit is disabled and can not be 500 enabled until the next reboot 501 unset - kernel audit is initialized but disabled and 502 will be fully enabled by the userspace auditd. 503 1 | on - kernel audit is initialized and partially 504 enabled, storing at most audit_backlog_limit 505 messages in RAM until it is fully enabled by the 506 userspace auditd. 507 Default: unset 508 509 audit_backlog_limit= [KNL] Set the audit queue size limit. 510 Format: <int> (must be >=0) 511 Default: 64 512 513 bau= [X86_UV] Enable the BAU on SGI UV. The default 514 behavior is to disable the BAU (i.e. bau=0). 515 Format: { "0" | "1" } 516 0 - Disable the BAU. 517 1 - Enable the BAU. 518 unset - Disable the BAU. 519 520 baycom_epp= [HW,AX25] 521 Format: <io>,<mode> 522 523 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem 524 Format: <io>,<mode> 525 See header of drivers/net/hamradio/baycom_par.c. 526 527 baycom_ser_fdx= [HW,AX25] 528 BayCom Serial Port AX.25 Modem (Full Duplex Mode) 529 Format: <io>,<irq>,<mode>[,<baud>] 530 See header of drivers/net/hamradio/baycom_ser_fdx.c. 531 532 baycom_ser_hdx= [HW,AX25] 533 BayCom Serial Port AX.25 Modem (Half Duplex Mode) 534 Format: <io>,<irq>,<mode> 535 See header of drivers/net/hamradio/baycom_ser_hdx.c. 536 537 bdev_allow_write_mounted= 538 Format: <bool> 539 Control the ability to open a mounted block device 540 for writing, i.e., allow / disallow writes that bypass 541 the FS. This was implemented as a means to prevent 542 fuzzers from crashing the kernel by overwriting the 543 metadata underneath a mounted FS without its awareness. 544 This also prevents destructive formatting of mounted 545 filesystems by naive storage tooling that don't use 546 O_EXCL. Default is Y and can be changed through the 547 Kconfig option CONFIG_BLK_DEV_WRITE_MOUNTED. 548 549 bert_disable [ACPI] 550 Disable BERT OS support on buggy BIOSes. 551 552 bgrt_disable [ACPI,X86,EARLY] 553 Disable BGRT to avoid flickering OEM logo. 554 555 blkdevparts= Manual partition parsing of block device(s) for 556 embedded devices based on command line input. 557 See Documentation/block/cmdline-partition.rst 558 559 boot_delay= [KNL,EARLY] 560 Milliseconds to delay each printk during boot. 561 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled, 562 and you may also have to specify "lpj=". Boot_delay 563 values larger than 10 seconds (10000) are assumed 564 erroneous and ignored. 565 Format: integer 566 567 bootconfig [KNL,EARLY] 568 Extended command line options can be added to an initrd 569 and this will cause the kernel to look for it. 570 571 See Documentation/admin-guide/bootconfig.rst 572 573 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 574 bttv.radio= Most important insmod options are available as 575 kernel args too. 576 bttv.pll= See Documentation/admin-guide/media/bttv.rst 577 bttv.tuner= 578 579 bulk_remove=off [PPC] This parameter disables the use of the pSeries 580 firmware feature for flushing multiple hpte entries 581 at a time. 582 583 c101= [NET] Moxa C101 synchronous serial card 584 585 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 586 Sometimes CPU hardware bugs make them report the cache 587 size incorrectly. The kernel will attempt work arounds 588 to fix known problems, but for some CPUs it is not 589 possible to determine what the correct size should be. 590 This option provides an override for these situations. 591 592 carrier_timeout= 593 [NET] Specifies amount of time (in seconds) that 594 the kernel should wait for a network carrier. By default 595 it waits 120 seconds. 596 597 ca_keys= [KEYS] This parameter identifies a specific key(s) on 598 the system trusted keyring to be used for certificate 599 trust validation. 600 format: { id:<keyid> | builtin } 601 602 cca= [MIPS,EARLY] Override the kernel pages' cache coherency 603 algorithm. Accepted values range from 0 to 7 604 inclusive. See arch/mips/include/asm/pgtable-bits.h 605 for platform specific values (SB1, Loongson3 and 606 others). 607 608 ccw_timeout_log [S390] 609 See Documentation/arch/s390/common_io.rst for details. 610 611 cfi= [X86-64] Set Control Flow Integrity checking features 612 when CONFIG_FINEIBT is enabled. 613 Format: feature[,feature...] 614 Default: auto 615 616 auto: Use FineIBT if IBT available, otherwise kCFI. 617 Under FineIBT, enable "paranoid" mode when 618 FRED is not available. 619 off: Turn off CFI checking. 620 kcfi: Use kCFI (disable FineIBT). 621 fineibt: Use FineIBT (even if IBT not available). 622 norand: Do not re-randomize CFI hashes. 623 paranoid: Add caller hash checking under FineIBT. 624 bhi: Enable register poisoning to stop speculation 625 across FineIBT. (Disabled by default.) 626 warn: Do not enforce CFI checking: warn only. 627 debug: Report CFI initialization details. 628 629 cgroup_disable= [KNL] Disable a particular controller or optional feature 630 Format: {name of the controller(s) or feature(s) to disable} 631 The effects of cgroup_disable=foo are: 632 - foo isn't auto-mounted if you mount all cgroups in 633 a single hierarchy 634 - foo isn't visible as an individually mountable 635 subsystem 636 - if foo is an optional feature then the feature is 637 disabled and corresponding cgroup files are not 638 created 639 {Currently only "memory" controller deal with this and 640 cut the overhead, others just disable the usage. So 641 only cgroup_disable=memory is actually worthy} 642 Specifying "pressure" disables per-cgroup pressure 643 stall information accounting feature 644 645 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 646 Format: { { controller | "all" | "named" } 647 [,{ controller | "all" | "named" }...] } 648 Like cgroup_disable, but only applies to cgroup v1; 649 the blacklisted controllers remain available in cgroup2. 650 "all" blacklists all controllers and "named" disables 651 named mounts. Specifying both "all" and "named" disables 652 all v1 hierarchies. 653 654 cgroup_v1_proc= [KNL] Show also missing controllers in /proc/cgroups 655 Format: { "true" | "false" } 656 /proc/cgroups lists only v1 controllers by default. 657 This compatibility option enables listing also v2 658 controllers (whose v1 code is not compiled!), so that 659 semi-legacy software can check this file to decide 660 about usage of v2 (sic) controllers. 661 662 cgroup_favordynmods= [KNL] Enable or Disable favordynmods. 663 Format: { "true" | "false" } 664 Defaults to the value of CONFIG_CGROUP_FAVOR_DYNMODS. 665 666 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 667 Format: <string> 668 nosocket -- Disable socket memory accounting. 669 nokmem -- Disable kernel memory accounting. 670 nobpf -- Disable BPF memory accounting. 671 672 checkreqprot= [SELINUX] Set initial checkreqprot flag value. 673 Format: { "0" | "1" } 674 See security/selinux/Kconfig help text. 675 0 -- check protection applied by kernel (includes 676 any implied execute protection). 677 1 -- check protection requested by application. 678 Default value is set via a kernel config option. 679 Value can be changed at runtime via 680 /sys/fs/selinux/checkreqprot. 681 Setting checkreqprot to 1 is deprecated. 682 683 cio_ignore= [S390] 684 See Documentation/arch/s390/common_io.rst for details. 685 686 clearcpuid=X[,X...] [X86] 687 Disable CPUID feature X for the kernel. See 688 arch/x86/include/asm/cpufeatures.h for the valid bit 689 numbers X. Note the Linux-specific bits are not necessarily 690 stable over kernel options, but the vendor-specific 691 ones should be. 692 X can also be a string as appearing in the flags: line 693 in /proc/cpuinfo which does not have the above 694 instability issue. However, not all features have names 695 in /proc/cpuinfo. 696 Note that using this option will taint your kernel. 697 Also note that user programs calling CPUID directly 698 or using the feature without checking anything 699 will still see it. This just prevents it from 700 being used by the kernel or shown in /proc/cpuinfo. 701 Also note the kernel might malfunction if you disable 702 some critical bits. 703 704 clk_ignore_unused 705 [CLK] 706 Prevents the clock framework from automatically gating 707 clocks that have not been explicitly enabled by a Linux 708 device driver but are enabled in hardware at reset or 709 by the bootloader/firmware. Note that this does not 710 force such clocks to be always-on nor does it reserve 711 those clocks in any way. This parameter is useful for 712 debug and development, but should not be needed on a 713 platform with proper driver support. For more 714 information, see Documentation/driver-api/clk.rst. 715 716 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 717 [Deprecated] 718 Forces specified clocksource (if available) to be used 719 when calculating gettimeofday(). If specified 720 clocksource is not available, it defaults to PIT. 721 Format: { pit | tsc | cyclone | pmtmr } 722 723 clocksource= Override the default clocksource 724 Format: <string> 725 Override the default clocksource and use the clocksource 726 with the name specified. 727 Some clocksource names to choose from, depending on 728 the platform: 729 [all] jiffies (this is the base, fallback clocksource) 730 [ACPI] acpi_pm 731 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 732 pxa_timer,timer3,32k_counter,timer0_1 733 [X86-32] pit,hpet,tsc; 734 scx200_hrt on Geode; cyclone on IBM x440 735 [MIPS] MIPS 736 [PARISC] cr16 737 [S390] tod 738 [SH] SuperH 739 [SPARC64] tick 740 [X86-64] hpet,tsc 741 742 clocksource.arm_arch_timer.evtstrm= 743 [ARM,ARM64,EARLY] 744 Format: <bool> 745 Enable/disable the eventstream feature of the ARM 746 architected timer so that code using WFE-based polling 747 loops can be debugged more effectively on production 748 systems. 749 750 clocksource.verify_n_cpus= [KNL] 751 Limit the number of CPUs checked for clocksources 752 marked with CLOCK_SOURCE_VERIFY_PERCPU that 753 are marked unstable due to excessive skew. 754 A negative value says to check all CPUs, while 755 zero says not to check any. Values larger than 756 nr_cpu_ids are silently truncated to nr_cpu_ids. 757 The actual CPUs are chosen randomly, with 758 no replacement if the same CPU is chosen twice. 759 760 clocksource-wdtest.holdoff= [KNL] 761 Set the time in seconds that the clocksource 762 watchdog test waits before commencing its tests. 763 Defaults to zero when built as a module and to 764 10 seconds when built into the kernel. 765 766 cma=nn[MG]@[start[MG][-end[MG]]] 767 [KNL,CMA,EARLY] 768 Sets the size of kernel global memory area for 769 contiguous memory allocations and optionally the 770 placement constraint by the physical address range of 771 memory allocations. A value of 0 disables CMA 772 altogether. For more information, see 773 kernel/dma/contiguous.c 774 775 cma_pernuma=nn[MG] 776 [KNL,CMA,EARLY] 777 Sets the size of kernel per-numa memory area for 778 contiguous memory allocations. A value of 0 disables 779 per-numa CMA altogether. And If this option is not 780 specified, the default value is 0. 781 With per-numa CMA enabled, DMA users on node nid will 782 first try to allocate buffer from the pernuma area 783 which is located in node nid, if the allocation fails, 784 they will fallback to the global default memory area. 785 786 numa_cma=<node>:nn[MG][,<node>:nn[MG]] 787 [KNL,CMA,EARLY] 788 Sets the size of kernel numa memory area for 789 contiguous memory allocations. It will reserve CMA 790 area for the specified node. 791 792 With numa CMA enabled, DMA users on node nid will 793 first try to allocate buffer from the numa area 794 which is located in node nid, if the allocation fails, 795 they will fallback to the global default memory area. 796 797 cmo_free_hint= [PPC] Format: { yes | no } 798 Specify whether pages are marked as being inactive 799 when they are freed. This is used in CMO environments 800 to determine OS memory pressure for page stealing by 801 a hypervisor. 802 Default: yes 803 804 coherent_pool=nn[KMG] [ARM,KNL,EARLY] 805 Sets the size of memory pool for coherent, atomic dma 806 allocations, by default set to 256K. 807 808 com20020= [HW,NET] ARCnet - COM20020 chipset 809 Format: 810 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 811 812 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 813 Format: <io>[,<irq>] 814 815 com90xx= [HW,NET] 816 ARCnet - COM90xx chipset (memory-mapped buffers) 817 Format: <io>[,<irq>[,<memstart>]] 818 819 condev= [HW,S390] console device 820 conmode= 821 822 con3215_drop= [S390,EARLY] 3215 console drop mode. 823 Format: y|n|Y|N|1|0 824 When set to true, drop data on the 3215 console when 825 the console buffer is full. In this case the 826 operator using a 3270 terminal emulator (for example 827 x3270) does not have to enter the clear key for the 828 console output to advance and the kernel to continue. 829 This leads to a much faster boot time when a 3270 830 terminal emulator is active. If no 3270 terminal 831 emulator is used, this parameter has no effect. 832 833 console= [KNL] Output console device and options. 834 835 tty<n> Use the virtual console device <n>. 836 837 ttyS<n>[,options] 838 ttyUSB0[,options] 839 Use the specified serial port. The options are of 840 the form "bbbbpnf", where "bbbb" is the baud rate, 841 "p" is parity ("n", "o", or "e"), "n" is number of 842 bits, and "f" is flow control ("r" for RTS or 843 omit it). Default is "9600n8". 844 845 See Documentation/admin-guide/serial-console.rst for more 846 information. See 847 Documentation/networking/netconsole.rst for an 848 alternative. 849 850 <DEVNAME>:<n>.<n>[,options] 851 Use the specified serial port on the serial core bus. 852 The addressing uses DEVNAME of the physical serial port 853 device, followed by the serial core controller instance, 854 and the serial port instance. The options are the same 855 as documented for the ttyS addressing above. 856 857 The mapping of the serial ports to the tty instances 858 can be viewed with: 859 860 $ ls -d /sys/bus/serial-base/devices/*:*.*/tty/* 861 /sys/bus/serial-base/devices/00:04:0.0/tty/ttyS0 862 863 In the above example, the console can be addressed with 864 console=00:04:0.0. Note that a console addressed this 865 way will only get added when the related device driver 866 is ready. The use of an earlycon parameter in addition to 867 the console may be desired for console output early on. 868 869 uart[8250],io,<addr>[,options] 870 uart[8250],mmio,<addr>[,options] 871 uart[8250],mmio16,<addr>[,options] 872 uart[8250],mmio32,<addr>[,options] 873 uart[8250],0x<addr>[,options] 874 Start an early, polled-mode console on the 8250/16550 875 UART at the specified I/O port or MMIO address, 876 switching to the matching ttyS device later. 877 MMIO inter-register address stride is either 8-bit 878 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 879 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 880 to be equivalent to 'mmio'. 'options' are specified in 881 the same format described for ttyS above; if unspecified, 882 the h/w is not re-initialized. 883 884 hvc<n> Use the hypervisor console device <n>. This is for 885 both Xen and PowerPC hypervisors. 886 887 { null | "" } 888 Use to disable console output, i.e., to have kernel 889 console messages discarded. 890 This must be the only console= parameter used on the 891 kernel command line. 892 893 If the device connected to the port is not a TTY but a braille 894 device, prepend "brl," before the device type, for instance 895 console=brl,ttyS0 896 For now, only VisioBraille is supported. 897 898 console_msg_format= 899 [KNL] Change console messages format 900 default 901 By default we print messages on consoles in 902 "[time stamp] text\n" format (time stamp may not be 903 printed, depending on CONFIG_PRINTK_TIME or 904 `printk_time' param). 905 syslog 906 Switch to syslog format: "<%u>[time stamp] text\n" 907 IOW, each message will have a facility and loglevel 908 prefix. The format is similar to one used by syslog() 909 syscall, or to executing "dmesg -S --raw" or to reading 910 from /proc/kmsg. 911 912 consoleblank= [KNL] The console blank (screen saver) timeout in 913 seconds. A value of 0 disables the blank timer. 914 Defaults to 0. 915 916 coredump_filter= 917 [KNL] Change the default value for 918 /proc/<pid>/coredump_filter. 919 See also Documentation/filesystems/proc.rst. 920 921 coresight_cpu_debug.enable 922 [ARM,ARM64] 923 Format: <bool> 924 Enable/disable the CPU sampling based debugging. 925 0: default value, disable debugging 926 1: enable debugging at boot time 927 928 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 929 Format: 930 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 931 932 cpuidle.off=1 [CPU_IDLE] 933 disable the cpuidle sub-system 934 935 cpuidle.governor= 936 [CPU_IDLE] Name of the cpuidle governor to use. 937 938 cpufreq.off=1 [CPU_FREQ] 939 disable the cpufreq sub-system 940 941 cpufreq.default_governor= 942 [CPU_FREQ] Name of the default cpufreq governor or 943 policy to use. This governor must be registered in the 944 kernel before the cpufreq driver probes. 945 946 cpu_init_udelay=N 947 [X86,EARLY] Delay for N microsec between assert and de-assert 948 of APIC INIT to start processors. This delay occurs 949 on every CPU online, such as boot, and resume from suspend. 950 Default: 10000 951 952 cpuhp.parallel= 953 [SMP] Enable/disable parallel bringup of secondary CPUs 954 Format: <bool> 955 Default is enabled if CONFIG_HOTPLUG_PARALLEL=y. Otherwise 956 the parameter has no effect. 957 958 crash_kexec_post_notifiers 959 Only jump to kdump kernel after running the panic 960 notifiers and dumping kmsg. This option increases 961 the risks of a kdump failure, since some panic 962 notifiers can make the crashed kernel more unstable. 963 In configurations where kdump may not be reliable, 964 running the panic notifiers could allow collecting 965 more data on dmesg, like stack traces from other CPUS 966 or extra data dumped by panic_print. Note that some 967 configurations enable this option unconditionally, 968 like Hyper-V, PowerPC (fadump) and AMD SEV-SNP. 969 970 crashkernel=size[KMG][@offset[KMG]] 971 [KNL,EARLY] Using kexec, Linux can switch to a 'crash kernel' 972 upon panic. This parameter reserves the physical 973 memory region [offset, offset + size] for that kernel 974 image. If '@offset' is omitted, then a suitable offset 975 is selected automatically. 976 [KNL, X86-64, ARM64, RISCV, LoongArch] Select a region 977 under 4G first, and fall back to reserve region above 978 4G when '@offset' hasn't been specified. 979 See Documentation/admin-guide/kdump/kdump.rst for further details. 980 981 crashkernel=range1:size1[,range2:size2,...][@offset] 982 [KNL] Same as above, but depends on the memory 983 in the running system. The syntax of range is 984 start-[end] where start and end are both 985 a memory unit (amount[KMG]). See also 986 Documentation/admin-guide/kdump/kdump.rst for an example. 987 988 crashkernel=size[KMG],high 989 [KNL, X86-64, ARM64, RISCV, LoongArch] range could be 990 above 4G. 991 Allow kernel to allocate physical memory region from top, 992 so could be above 4G if system have more than 4G ram 993 installed. Otherwise memory region will be allocated 994 below 4G, if available. 995 It will be ignored if crashkernel=X is specified. 996 crashkernel=size[KMG],low 997 [KNL, X86-64, ARM64, RISCV, LoongArch] range under 4G. 998 When crashkernel=X,high is passed, kernel could allocate 999 physical memory region above 4G, that cause second kernel 1000 crash on system that require some amount of low memory, 1001 e.g. swiotlb requires at least 64M+32K low memory, also 1002 enough extra low memory is needed to make sure DMA buffers 1003 for 32-bit devices won't run out. Kernel would try to allocate 1004 default size of memory below 4G automatically. The default 1005 size is platform dependent. 1006 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB) 1007 --> arm64: 128MiB 1008 --> riscv: 128MiB 1009 --> loongarch: 128MiB 1010 This one lets the user specify own low range under 4G 1011 for second kernel instead. 1012 0: to disable low allocation. 1013 It will be ignored when crashkernel=X,high is not used 1014 or memory reserved is below 4G. 1015 crashkernel=size[KMG],cma 1016 [KNL, X86] Reserve additional crash kernel memory from 1017 CMA. This reservation is usable by the first system's 1018 userspace memory and kernel movable allocations (memory 1019 balloon, zswap). Pages allocated from this memory range 1020 will not be included in the vmcore so this should not 1021 be used if dumping of userspace memory is intended and 1022 it has to be expected that some movable kernel pages 1023 may be missing from the dump. 1024 1025 A standard crashkernel reservation, as described above, 1026 is still needed to hold the crash kernel and initrd. 1027 1028 This option increases the risk of a kdump failure: DMA 1029 transfers configured by the first kernel may end up 1030 corrupting the second kernel's memory. 1031 1032 This reservation method is intended for systems that 1033 can't afford to sacrifice enough memory for standard 1034 crashkernel reservation and where less reliable and 1035 possibly incomplete kdump is preferable to no kdump at 1036 all. 1037 1038 cryptomgr.notests 1039 [KNL] Disable crypto self-tests 1040 1041 cs89x0_dma= [HW,NET] 1042 Format: <dma> 1043 1044 cs89x0_media= [HW,NET] 1045 Format: { rj45 | aui | bnc } 1046 1047 csdlock_debug= [KNL] Enable or disable debug add-ons of cross-CPU 1048 function call handling. When switched on, 1049 additional debug data is printed to the console 1050 in case a hanging CPU is detected, and that 1051 CPU is pinged again in order to try to resolve 1052 the hang situation. The default value of this 1053 option depends on the CSD_LOCK_WAIT_DEBUG_DEFAULT 1054 Kconfig option. 1055 1056 dasd= [HW,NET] 1057 See header of drivers/s390/block/dasd_devmap.c. 1058 1059 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 1060 (one device per port) 1061 Format: <port#>,<type> 1062 See also Documentation/input/devices/joystick-parport.rst 1063 1064 debug [KNL,EARLY] Enable kernel debugging (events log level). 1065 1066 debug_boot_weak_hash 1067 [KNL,EARLY] Enable printing [hashed] pointers early in the 1068 boot sequence. If enabled, we use a weak hash instead 1069 of siphash to hash pointers. Use this option if you are 1070 seeing instances of '(___ptrval___)') and need to see a 1071 value (hashed pointer) instead. Cryptographically 1072 insecure, please do not use on production kernels. 1073 1074 debug_locks_verbose= 1075 [KNL] verbose locking self-tests 1076 Format: <int> 1077 Print debugging info while doing the locking API 1078 self-tests. 1079 Bitmask for the various LOCKTYPE_ tests. Defaults to 0 1080 (no extra messages), setting it to -1 (all bits set) 1081 will print _a_lot_ more information - normally only 1082 useful to lockdep developers. 1083 1084 debug_objects [KNL,EARLY] Enable object debugging 1085 1086 debug_guardpage_minorder= 1087 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this 1088 parameter allows control of the order of pages that will 1089 be intentionally kept free (and hence protected) by the 1090 buddy allocator. Bigger value increase the probability 1091 of catching random memory corruption, but reduce the 1092 amount of memory for normal system use. The maximum 1093 possible value is MAX_PAGE_ORDER/2. Setting this 1094 parameter to 1 or 2 should be enough to identify most 1095 random memory corruption problems caused by bugs in 1096 kernel or driver code when a CPU writes to (or reads 1097 from) a random memory location. Note that there exists 1098 a class of memory corruptions problems caused by buggy 1099 H/W or F/W or by drivers badly programming DMA 1100 (basically when memory is written at bus level and the 1101 CPU MMU is bypassed) which are not detectable by 1102 CONFIG_DEBUG_PAGEALLOC, hence this option will not 1103 help tracking down these problems. 1104 1105 debug_pagealloc= 1106 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 1107 enables the feature at boot time. By default, it is 1108 disabled and the system will work mostly the same as a 1109 kernel built without CONFIG_DEBUG_PAGEALLOC. 1110 Note: to get most of debug_pagealloc error reports, it's 1111 useful to also enable the page_owner functionality. 1112 on: enable the feature 1113 1114 debugfs= [KNL,EARLY] This parameter enables what is exposed to 1115 userspace and debugfs internal clients. 1116 Format: { on, no-mount, off } 1117 on: All functions are enabled. 1118 no-mount: 1119 Filesystem is not registered but kernel clients can 1120 access APIs and a crashkernel can be used to read 1121 its content. There is nothing to mount. 1122 off: Filesystem is not registered and clients 1123 get a -EPERM as result when trying to register files 1124 or directories within debugfs. 1125 This is equivalent of the runtime functionality if 1126 debugfs was not enabled in the kernel at all. 1127 Default value is set in build-time with a kernel configuration. 1128 1129 debugpat [X86] Enable PAT debugging 1130 1131 default_hugepagesz= 1132 [HW] The size of the default HugeTLB page. This is 1133 the size represented by the legacy /proc/ hugepages 1134 APIs. In addition, this is the default hugetlb size 1135 used for shmget(), mmap() and mounting hugetlbfs 1136 filesystems. If not specified, defaults to the 1137 architecture's default huge page size. Huge page 1138 sizes are architecture dependent. See also 1139 Documentation/admin-guide/mm/hugetlbpage.rst. 1140 Format: size[KMG] 1141 1142 deferred_probe_timeout= 1143 [KNL] Debugging option to set a timeout in seconds for 1144 deferred probe to give up waiting on dependencies to 1145 probe. Only specific dependencies (subsystems or 1146 drivers) that have opted in will be ignored. A timeout 1147 of 0 will timeout at the end of initcalls. If the time 1148 out hasn't expired, it'll be restarted by each 1149 successful driver registration. This option will also 1150 dump out devices still on the deferred probe list after 1151 retrying. 1152 1153 delayacct [KNL] Enable per-task delay accounting 1154 1155 dell_smm_hwmon.ignore_dmi= 1156 [HW] Continue probing hardware even if DMI data 1157 indicates that the driver is running on unsupported 1158 hardware. 1159 1160 dell_smm_hwmon.force= 1161 [HW] Activate driver even if SMM BIOS signature does 1162 not match list of supported models and enable otherwise 1163 blacklisted features. 1164 1165 dell_smm_hwmon.power_status= 1166 [HW] Report power status in /proc/i8k 1167 (disabled by default). 1168 1169 dell_smm_hwmon.restricted= 1170 [HW] Allow controlling fans only if SYS_ADMIN 1171 capability is set. 1172 1173 dell_smm_hwmon.fan_mult= 1174 [HW] Factor to multiply fan speed with. 1175 1176 dell_smm_hwmon.fan_max= 1177 [HW] Maximum configurable fan speed. 1178 1179 dfltcc= [HW,S390] 1180 Format: { on | off | def_only | inf_only | always } 1181 on: s390 zlib hardware support for compression on 1182 level 1 and decompression (default) 1183 off: No s390 zlib hardware support 1184 def_only: s390 zlib hardware support for deflate 1185 only (compression on level 1) 1186 inf_only: s390 zlib hardware support for inflate 1187 only (decompression) 1188 always: Same as 'on' but ignores the selected compression 1189 level always using hardware support (used for debugging) 1190 1191 dhash_entries= [KNL] 1192 Set number of hash buckets for dentry cache. 1193 1194 disable_1tb_segments [PPC,EARLY] 1195 Disables the use of 1TB hash page table segments. This 1196 causes the kernel to fall back to 256MB segments which 1197 can be useful when debugging issues that require an SLB 1198 miss to occur. 1199 1200 disable= [IPV6] 1201 See Documentation/networking/ipv6.rst. 1202 1203 disable_radix [PPC,EARLY] 1204 Disable RADIX MMU mode on POWER9 1205 1206 disable_tlbie [PPC] 1207 Disable TLBIE instruction. Currently does not work 1208 with KVM, with HASH MMU, or with coherent accelerators. 1209 1210 disable_ddw [PPC/PSERIES,EARLY] 1211 Disable Dynamic DMA Window support. Use this 1212 to workaround buggy firmware. 1213 1214 disable_ipv6= [IPV6] 1215 See Documentation/networking/ipv6.rst. 1216 1217 disable_mtrr_cleanup [X86,EARLY] 1218 The kernel tries to adjust MTRR layout from continuous 1219 to discrete, to make X server driver able to add WB 1220 entry later. This parameter disables that. 1221 1222 disable_mtrr_trim [X86, Intel and AMD only,EARLY] 1223 By default the kernel will trim any uncacheable 1224 memory out of your available memory pool based on 1225 MTRR settings. This parameter disables that behavior, 1226 possibly causing your machine to run very slowly. 1227 1228 disable_timer_pin_1 [X86,EARLY] 1229 Disable PIN 1 of APIC timer 1230 Can be useful to work around chipset bugs. 1231 1232 dis_ucode_ldr [X86] Disable the microcode loader. 1233 1234 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 1235 this option disables the debugging code at boot. 1236 1237 dma_debug_entries=<number> 1238 This option allows to tune the number of preallocated 1239 entries for DMA-API debugging code. One entry is 1240 required per DMA-API allocation. Use this if the 1241 DMA-API debugging code disables itself because the 1242 architectural default is too low. 1243 1244 dma_debug_driver=<driver_name> 1245 With this option the DMA-API debugging driver 1246 filter feature can be enabled at boot time. Just 1247 pass the driver to filter for as the parameter. 1248 The filter can be disabled or changed to another 1249 driver later using sysfs. 1250 1251 reg_file_data_sampling= 1252 [X86] Controls mitigation for Register File Data 1253 Sampling (RFDS) vulnerability. RFDS is a CPU 1254 vulnerability which may allow userspace to infer 1255 kernel data values previously stored in floating point 1256 registers, vector registers, or integer registers. 1257 RFDS only affects Intel Atom processors. 1258 1259 on: Turns ON the mitigation. 1260 off: Turns OFF the mitigation. 1261 1262 This parameter overrides the compile time default set 1263 by CONFIG_MITIGATION_RFDS. Mitigation cannot be 1264 disabled when other VERW based mitigations (like MDS) 1265 are enabled. In order to disable RFDS mitigation all 1266 VERW based mitigations need to be disabled. 1267 1268 For details see: 1269 Documentation/admin-guide/hw-vuln/reg-file-data-sampling.rst 1270 1271 driver_async_probe= [KNL] 1272 List of driver names to be probed asynchronously. * 1273 matches with all driver names. If * is specified, the 1274 rest of the listed driver names are those that will NOT 1275 match the *. 1276 Format: <driver_name1>,<driver_name2>... 1277 1278 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1279 Broken monitors, graphic adapters, KVMs and EDIDless 1280 panels may send no or incorrect EDID data sets. 1281 This parameter allows to specify an EDID data sets 1282 in the /lib/firmware directory that are used instead. 1283 An EDID data set will only be used for a particular 1284 connector, if its name and a colon are prepended to 1285 the EDID name. Each connector may use a unique EDID 1286 data set by separating the files with a comma. An EDID 1287 data set with no connector name will be used for 1288 any connectors not explicitly specified. 1289 1290 dscc4.setup= [NET] 1291 1292 dt_cpu_ftrs= [PPC,EARLY] 1293 Format: {"off" | "known"} 1294 Control how the dt_cpu_ftrs device-tree binding is 1295 used for CPU feature discovery and setup (if it 1296 exists). 1297 off: Do not use it, fall back to legacy cpu table. 1298 known: Do not pass through unknown features to guests 1299 or userspace, only those that the kernel is aware of. 1300 1301 dump_apple_properties [X86] 1302 Dump name and content of EFI device properties on 1303 x86 Macs. Useful for driver authors to determine 1304 what data is available or for reverse-engineering. 1305 1306 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1307 <module>.dyndbg[="val"] 1308 Enable debug messages at boot time. See 1309 Documentation/admin-guide/dynamic-debug-howto.rst 1310 for details. 1311 1312 early_ioremap_debug [KNL,EARLY] 1313 Enable debug messages in early_ioremap support. This 1314 is useful for tracking down temporary early mappings 1315 which are not unmapped. 1316 1317 earlycon= [KNL,EARLY] Output early console device and options. 1318 1319 When used with no options, the early console is 1320 determined by stdout-path property in device tree's 1321 chosen node or the ACPI SPCR table if supported by 1322 the platform. 1323 1324 cdns,<addr>[,options] 1325 Start an early, polled-mode console on a Cadence 1326 (xuartps) serial port at the specified address. Only 1327 supported option is baud rate. If baud rate is not 1328 specified, the serial port must already be setup and 1329 configured. 1330 1331 uart[8250],io,<addr>[,options[,uartclk]] 1332 uart[8250],mmio,<addr>[,options[,uartclk]] 1333 uart[8250],mmio32,<addr>[,options[,uartclk]] 1334 uart[8250],mmio32be,<addr>[,options[,uartclk]] 1335 uart[8250],0x<addr>[,options] 1336 Start an early, polled-mode console on the 8250/16550 1337 UART at the specified I/O port or MMIO address. 1338 MMIO inter-register address stride is either 8-bit 1339 (mmio) or 32-bit (mmio32 or mmio32be). 1340 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1341 to be equivalent to 'mmio'. 'options' are specified 1342 in the same format described for "console=ttyS<n>"; if 1343 unspecified, the h/w is not initialized. 'uartclk' is 1344 the uart clock frequency; if unspecified, it is set 1345 to 'BASE_BAUD' * 16. 1346 1347 pl011,<addr> 1348 pl011,mmio32,<addr> 1349 Start an early, polled-mode console on a pl011 serial 1350 port at the specified address. The pl011 serial port 1351 must already be setup and configured. Options are not 1352 yet supported. If 'mmio32' is specified, then only 1353 the driver will use only 32-bit accessors to read/write 1354 the device registers. 1355 1356 liteuart,<addr> 1357 Start an early console on a litex serial port at the 1358 specified address. The serial port must already be 1359 setup and configured. Options are not yet supported. 1360 1361 meson,<addr> 1362 Start an early, polled-mode console on a meson serial 1363 port at the specified address. The serial port must 1364 already be setup and configured. Options are not yet 1365 supported. 1366 1367 msm_serial,<addr> 1368 Start an early, polled-mode console on an msm serial 1369 port at the specified address. The serial port 1370 must already be setup and configured. Options are not 1371 yet supported. 1372 1373 msm_serial_dm,<addr> 1374 Start an early, polled-mode console on an msm serial 1375 dm port at the specified address. The serial port 1376 must already be setup and configured. Options are not 1377 yet supported. 1378 1379 owl,<addr> 1380 Start an early, polled-mode console on a serial port 1381 of an Actions Semi SoC, such as S500 or S900, at the 1382 specified address. The serial port must already be 1383 setup and configured. Options are not yet supported. 1384 1385 rda,<addr> 1386 Start an early, polled-mode console on a serial port 1387 of an RDA Micro SoC, such as RDA8810PL, at the 1388 specified address. The serial port must already be 1389 setup and configured. Options are not yet supported. 1390 1391 sbi 1392 Use RISC-V SBI (Supervisor Binary Interface) for early 1393 console. 1394 1395 smh Use ARM semihosting calls for early console. 1396 1397 s3c2410,<addr> 1398 s3c2412,<addr> 1399 s3c2440,<addr> 1400 s3c6400,<addr> 1401 s5pv210,<addr> 1402 exynos4210,<addr> 1403 Use early console provided by serial driver available 1404 on Samsung SoCs, requires selecting proper type and 1405 a correct base address of the selected UART port. The 1406 serial port must already be setup and configured. 1407 Options are not yet supported. 1408 1409 lantiq,<addr> 1410 Start an early, polled-mode console on a lantiq serial 1411 (lqasc) port at the specified address. The serial port 1412 must already be setup and configured. Options are not 1413 yet supported. 1414 1415 lpuart,<addr> 1416 lpuart32,<addr> 1417 Use early console provided by Freescale LP UART driver 1418 found on Freescale Vybrid and QorIQ LS1021A processors. 1419 A valid base address must be provided, and the serial 1420 port must already be setup and configured. 1421 1422 ec_imx21,<addr> 1423 ec_imx6q,<addr> 1424 Start an early, polled-mode, output-only console on the 1425 Freescale i.MX UART at the specified address. The UART 1426 must already be setup and configured. 1427 1428 ar3700_uart,<addr> 1429 Start an early, polled-mode console on the 1430 Armada 3700 serial port at the specified 1431 address. The serial port must already be setup 1432 and configured. Options are not yet supported. 1433 1434 qcom_geni,<addr> 1435 Start an early, polled-mode console on a Qualcomm 1436 Generic Interface (GENI) based serial port at the 1437 specified address. The serial port must already be 1438 setup and configured. Options are not yet supported. 1439 1440 efifb,[options] 1441 Start an early, unaccelerated console on the EFI 1442 memory mapped framebuffer (if available). On cache 1443 coherent non-x86 systems that use system memory for 1444 the framebuffer, pass the 'ram' option so that it is 1445 mapped with the correct attributes. 1446 1447 linflex,<addr> 1448 Use early console provided by Freescale LINFlexD UART 1449 serial driver for NXP S32V234 SoCs. A valid base 1450 address must be provided, and the serial port must 1451 already be setup and configured. 1452 1453 earlyprintk= [X86,SH,ARM,M68k,S390,UM,EARLY] 1454 earlyprintk=vga 1455 earlyprintk=sclp 1456 earlyprintk=xen 1457 earlyprintk=serial[,ttySn[,baudrate]] 1458 earlyprintk=serial[,0x...[,baudrate]] 1459 earlyprintk=ttySn[,baudrate] 1460 earlyprintk=dbgp[debugController#] 1461 earlyprintk=mmio32,membase[,{nocfg|baudrate}] 1462 earlyprintk=pciserial[,force],bus:device.function[,{nocfg|baudrate}] 1463 earlyprintk=xdbc[xhciController#] 1464 earlyprintk=bios 1465 1466 earlyprintk is useful when the kernel crashes before 1467 the normal console is initialized. It is not enabled by 1468 default because it has some cosmetic problems. 1469 1470 Use "nocfg" to skip UART configuration, assume 1471 BIOS/firmware has configured UART correctly. 1472 1473 Append ",keep" to not disable it when the real console 1474 takes over. 1475 1476 Only one of vga, serial, or usb debug port can 1477 be used at a time. 1478 1479 Currently only ttyS0 and ttyS1 may be specified by 1480 name. Other I/O ports may be explicitly specified 1481 on some architectures (x86 and arm at least) by 1482 replacing ttySn with an I/O port address, like this: 1483 earlyprintk=serial,0x1008,115200 1484 You can find the port for a given device in 1485 /proc/tty/driver/serial: 1486 2: uart:ST16650V2 port:00001008 irq:18 ... 1487 1488 Interaction with the standard serial driver is not 1489 very good. 1490 1491 The VGA output is eventually overwritten by 1492 the real console. 1493 1494 The xen option can only be used in Xen domains. 1495 1496 The sclp output can only be used on s390. 1497 1498 The bios output can only be used on SuperH. 1499 1500 The optional "force" to "pciserial" enables use of a 1501 PCI device even when its classcode is not of the 1502 UART class. 1503 1504 edac_report= [HW,EDAC] Control how to report EDAC event 1505 Format: {"on" | "off" | "force"} 1506 on: enable EDAC to report H/W event. May be overridden 1507 by other higher priority error reporting module. 1508 off: disable H/W event reporting through EDAC. 1509 force: enforce the use of EDAC to report H/W event. 1510 default: on. 1511 1512 edd= [EDD] 1513 Format: {"off" | "on" | "skip[mbr]"} 1514 1515 efi= [EFI,EARLY] 1516 Format: { "debug", "disable_early_pci_dma", 1517 "nochunk", "noruntime", "nosoftreserve", 1518 "novamap", "no_disable_early_pci_dma" } 1519 debug: enable misc debug output. 1520 disable_early_pci_dma: disable the busmaster bit on all 1521 PCI bridges while in the EFI boot stub. 1522 nochunk: disable reading files in "chunks" in the EFI 1523 boot stub, as chunking can cause problems with some 1524 firmware implementations. 1525 noruntime : disable EFI runtime services support 1526 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1527 attribute may cause the kernel to reserve the 1528 memory range for a memory mapping driver to 1529 claim. Specify efi=nosoftreserve to disable this 1530 reservation and treat the memory by its base type 1531 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1532 novamap: do not call SetVirtualAddressMap(). 1533 no_disable_early_pci_dma: Leave the busmaster bit set 1534 on all PCI bridges while in the EFI boot stub 1535 1536 efi_no_storage_paranoia [EFI,X86,EARLY] 1537 Using this parameter you can use more than 50% of 1538 your efi variable storage. Use this parameter only if 1539 you are really sure that your UEFI does sane gc and 1540 fulfills the spec otherwise your board may brick. 1541 1542 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1543 that is to be dynamically loaded by Linux. If there are 1544 multiple variables with the same name but with different 1545 vendor GUIDs, all of them will be loaded. See 1546 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1547 1548 1549 eisa_irq_edge= [PARISC,HW] 1550 See header of drivers/parisc/eisa.c. 1551 1552 ekgdboc= [X86,KGDB,EARLY] Allow early kernel console debugging 1553 Format: ekgdboc=kbd 1554 1555 This is designed to be used in conjunction with 1556 the boot argument: earlyprintk=vga 1557 1558 This parameter works in place of the kgdboc parameter 1559 but can only be used if the backing tty is available 1560 very early in the boot process. For early debugging 1561 via a serial port see kgdboc_earlycon instead. 1562 1563 elanfreq= [X86-32] 1564 See comment before function elanfreq_setup() in 1565 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1566 1567 elfcorehdr=[size[KMG]@]offset[KMG] [PPC,SH,X86,S390,EARLY] 1568 Specifies physical address of start of kernel core 1569 image elf header and optionally the size. Generally 1570 kexec loader will pass this option to capture kernel. 1571 See Documentation/admin-guide/kdump/kdump.rst for details. 1572 1573 enable_mtrr_cleanup [X86,EARLY] 1574 The kernel tries to adjust MTRR layout from continuous 1575 to discrete, to make X server driver able to add WB 1576 entry later. This parameter enables that. 1577 1578 enable_timer_pin_1 [X86] 1579 Enable PIN 1 of APIC timer 1580 Can be useful to work around chipset bugs 1581 (in particular on some ATI chipsets). 1582 The kernel tries to set a reasonable default. 1583 1584 enforcing= [SELINUX] Set initial enforcing status. 1585 Format: {"0" | "1"} 1586 See security/selinux/Kconfig help text. 1587 0 -- permissive (log only, no denials). 1588 1 -- enforcing (deny and log). 1589 Default value is 0. 1590 Value can be changed at runtime via 1591 /sys/fs/selinux/enforce. 1592 1593 erst_disable [ACPI] 1594 Disable Error Record Serialization Table (ERST) 1595 support. 1596 1597 ether= [HW,NET] Ethernet cards parameters 1598 This option is obsoleted by the "netdev=" option, which 1599 has equivalent usage. See its documentation for details. 1600 1601 evm= [EVM] 1602 Format: { "fix" } 1603 Permit 'security.evm' to be updated regardless of 1604 current integrity status. 1605 1606 early_page_ext [KNL,EARLY] Enforces page_ext initialization to earlier 1607 stages so cover more early boot allocations. 1608 Please note that as side effect some optimizations 1609 might be disabled to achieve that (e.g. parallelized 1610 memory initialization is disabled) so the boot process 1611 might take longer, especially on systems with a lot of 1612 memory. Available with CONFIG_PAGE_EXTENSION=y. 1613 1614 failslab= 1615 fail_usercopy= 1616 fail_page_alloc= 1617 fail_skb_realloc= 1618 fail_make_request=[KNL] 1619 General fault injection mechanism. 1620 Format: <interval>,<probability>,<space>,<times> 1621 See also Documentation/fault-injection/. 1622 1623 fb_tunnels= [NET] 1624 Format: { initns | none } 1625 See Documentation/admin-guide/sysctl/net.rst for 1626 fb_tunnels_only_for_init_ns 1627 1628 floppy= [HW] 1629 See Documentation/admin-guide/blockdev/floppy.rst. 1630 1631 forcepae [X86-32] 1632 Forcefully enable Physical Address Extension (PAE). 1633 Many Pentium M systems disable PAE but may have a 1634 functionally usable PAE implementation. 1635 Warning: use of this parameter will taint the kernel 1636 and may cause unknown problems. 1637 1638 fred= [X86-64] 1639 Enable/disable Flexible Return and Event Delivery. 1640 Format: { on | off } 1641 on: enable FRED when it's present. 1642 off: disable FRED, the default setting. 1643 1644 ftrace=[tracer] 1645 [FTRACE] will set and start the specified tracer 1646 as early as possible in order to facilitate early 1647 boot debugging. 1648 1649 ftrace_boot_snapshot 1650 [FTRACE] On boot up, a snapshot will be taken of the 1651 ftrace ring buffer that can be read at: 1652 /sys/kernel/tracing/snapshot. 1653 This is useful if you need tracing information from kernel 1654 boot up that is likely to be overridden by user space 1655 start up functionality. 1656 1657 Optionally, the snapshot can also be defined for a tracing 1658 instance that was created by the trace_instance= command 1659 line parameter. 1660 1661 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo 1662 1663 The above will cause the "foo" tracing instance to trigger 1664 a snapshot at the end of boot up. 1665 1666 ftrace_dump_on_oops[=2(orig_cpu) | =<instance>][,<instance> | 1667 ,<instance>=2(orig_cpu)] 1668 [FTRACE] will dump the trace buffers on oops. 1669 If no parameter is passed, ftrace will dump global 1670 buffers of all CPUs, if you pass 2 or orig_cpu, it 1671 will dump only the buffer of the CPU that triggered 1672 the oops, or the specific instance will be dumped if 1673 its name is passed. Multiple instance dump is also 1674 supported, and instances are separated by commas. Each 1675 instance supports only dump on CPU that triggered the 1676 oops by passing 2 or orig_cpu to it. 1677 1678 ftrace_dump_on_oops=foo=orig_cpu 1679 1680 The above will dump only the buffer of "foo" instance 1681 on CPU that triggered the oops. 1682 1683 ftrace_dump_on_oops,foo,bar=orig_cpu 1684 1685 The above will dump global buffer on all CPUs, the 1686 buffer of "foo" instance on all CPUs and the buffer 1687 of "bar" instance on CPU that triggered the oops. 1688 1689 ftrace_filter=[function-list] 1690 [FTRACE] Limit the functions traced by the function 1691 tracer at boot up. function-list is a comma-separated 1692 list of functions. This list can be changed at run 1693 time by the set_ftrace_filter file in the debugfs 1694 tracing directory. 1695 1696 ftrace_notrace=[function-list] 1697 [FTRACE] Do not trace the functions specified in 1698 function-list. This list can be changed at run time 1699 by the set_ftrace_notrace file in the debugfs 1700 tracing directory. 1701 1702 ftrace_graph_filter=[function-list] 1703 [FTRACE] Limit the top level callers functions traced 1704 by the function graph tracer at boot up. 1705 function-list is a comma-separated list of functions 1706 that can be changed at run time by the 1707 set_graph_function file in the debugfs tracing directory. 1708 1709 ftrace_graph_notrace=[function-list] 1710 [FTRACE] Do not trace from the functions specified in 1711 function-list. This list is a comma-separated list of 1712 functions that can be changed at run time by the 1713 set_graph_notrace file in the debugfs tracing directory. 1714 1715 ftrace_graph_max_depth=<uint> 1716 [FTRACE] Used with the function graph tracer. This is 1717 the max depth it will trace into a function. This value 1718 can be changed at run time by the max_graph_depth file 1719 in the tracefs tracing directory. default: 0 (no limit) 1720 1721 fw_devlink= [KNL,EARLY] Create device links between consumer and supplier 1722 devices by scanning the firmware to infer the 1723 consumer/supplier relationships. This feature is 1724 especially useful when drivers are loaded as modules as 1725 it ensures proper ordering of tasks like device probing 1726 (suppliers first, then consumers), supplier boot state 1727 clean up (only after all consumers have probed), 1728 suspend/resume & runtime PM (consumers first, then 1729 suppliers). 1730 Format: { off | permissive | on | rpm } 1731 off -- Don't create device links from firmware info. 1732 permissive -- Create device links from firmware info 1733 but use it only for ordering boot state clean 1734 up (sync_state() calls). 1735 on -- Create device links from firmware info and use it 1736 to enforce probe and suspend/resume ordering. 1737 rpm -- Like "on", but also use to order runtime PM. 1738 1739 fw_devlink.strict=<bool> 1740 [KNL,EARLY] Treat all inferred dependencies as mandatory 1741 dependencies. This only applies for fw_devlink=on|rpm. 1742 Format: <bool> 1743 1744 fw_devlink.sync_state = 1745 [KNL,EARLY] When all devices that could probe have finished 1746 probing, this parameter controls what to do with 1747 devices that haven't yet received their sync_state() 1748 calls. 1749 Format: { strict | timeout } 1750 strict -- Default. Continue waiting on consumers to 1751 probe successfully. 1752 timeout -- Give up waiting on consumers and call 1753 sync_state() on any devices that haven't yet 1754 received their sync_state() calls after 1755 deferred_probe_timeout has expired or by 1756 late_initcall() if !CONFIG_MODULES. 1757 1758 gamecon.map[2|3]= 1759 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1760 support via parallel port (up to 5 devices per port) 1761 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1762 See also Documentation/input/devices/joystick-parport.rst 1763 1764 gamma= [HW,DRM] 1765 1766 gart_fix_e820= [X86-64,EARLY] disable the fix e820 for K8 GART 1767 Format: off | on 1768 default: on 1769 1770 gather_data_sampling= 1771 [X86,INTEL,EARLY] Control the Gather Data Sampling (GDS) 1772 mitigation. 1773 1774 Gather Data Sampling is a hardware vulnerability which 1775 allows unprivileged speculative access to data which was 1776 previously stored in vector registers. 1777 1778 This issue is mitigated by default in updated microcode. 1779 The mitigation may have a performance impact but can be 1780 disabled. On systems without the microcode mitigation 1781 disabling AVX serves as a mitigation. 1782 1783 force: Disable AVX to mitigate systems without 1784 microcode mitigation. No effect if the microcode 1785 mitigation is present. Known to cause crashes in 1786 userspace with buggy AVX enumeration. 1787 1788 off: Disable GDS mitigation. 1789 1790 gbpages [X86] Use GB pages for kernel direct mappings. 1791 1792 gcov_persist= [GCOV] When non-zero (default), profiling data for 1793 kernel modules is saved and remains accessible via 1794 debugfs, even when the module is unloaded/reloaded. 1795 When zero, profiling data is discarded and associated 1796 debugfs files are removed at module unload time. 1797 1798 goldfish [X86] Enable the goldfish android emulator platform. 1799 Don't use this when you are not running on the 1800 android emulator 1801 1802 gpio-mockup.gpio_mockup_ranges 1803 [HW] Sets the ranges of gpiochip of for this device. 1804 Format: <start1>,<end1>,<start2>,<end2>... 1805 gpio-mockup.gpio_mockup_named_lines 1806 [HW] Let the driver know GPIO lines should be named. 1807 1808 gpt [EFI] Forces disk with valid GPT signature but 1809 invalid Protective MBR to be treated as GPT. If the 1810 primary GPT is corrupted, it enables the backup/alternate 1811 GPT to be used instead. 1812 1813 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1814 the "Enable 0" bit of the configuration register. 1815 Format: 0 | 1 1816 Default: 0 1817 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1818 the "Enable 0" bit of the configuration register. 1819 Format: 0 | 1 1820 Default: 0 1821 grcan.select= [HW] Select which physical interface to use. 1822 Format: 0 | 1 1823 Default: 0 1824 grcan.txsize= [HW] Sets the size of the tx buffer. 1825 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1826 Default: 1024 1827 grcan.rxsize= [HW] Sets the size of the rx buffer. 1828 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1829 Default: 1024 1830 1831 hardened_usercopy= 1832 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 1833 hardening is enabled for this boot. Hardened 1834 usercopy checking is used to protect the kernel 1835 from reading or writing beyond known memory 1836 allocation boundaries as a proactive defense 1837 against bounds-checking flaws in the kernel's 1838 copy_to_user()/copy_from_user() interface. 1839 The default is determined by 1840 CONFIG_HARDENED_USERCOPY_DEFAULT_ON. 1841 on Perform hardened usercopy checks. 1842 off Disable hardened usercopy checks. 1843 1844 hardlockup_all_cpu_backtrace= 1845 [KNL] Should the hard-lockup detector generate 1846 backtraces on all cpus. 1847 Format: 0 | 1 1848 1849 hash_pointers= 1850 [KNL,EARLY] 1851 By default, when pointers are printed to the console 1852 or buffers via the %p format string, that pointer is 1853 "hashed", i.e. obscured by hashing the pointer value. 1854 This is a security feature that hides actual kernel 1855 addresses from unprivileged users, but it also makes 1856 debugging the kernel more difficult since unequal 1857 pointers can no longer be compared. The choices are: 1858 Format: { auto | always | never } 1859 Default: auto 1860 1861 auto - Hash pointers unless slab_debug is enabled. 1862 always - Always hash pointers (even if slab_debug is 1863 enabled). 1864 never - Never hash pointers. This option should only 1865 be specified when debugging the kernel. Do 1866 not use on production kernels. The boot 1867 param "no_hash_pointers" is an alias for 1868 this mode. 1869 1870 hashdist= [KNL,NUMA] Large hashes allocated during boot 1871 are distributed across NUMA nodes. Defaults on 1872 for 64-bit NUMA, off otherwise. 1873 Format: 0 | 1 (for off | on) 1874 1875 hd= [EIDE] (E)IDE hard drive subsystem geometry 1876 Format: <cyl>,<head>,<sect> 1877 1878 hest_disable [ACPI] 1879 Disable Hardware Error Source Table (HEST) support; 1880 corresponding firmware-first mode error processing 1881 logic will be disabled. 1882 1883 hibernate= [HIBERNATION] 1884 noresume Don't check if there's a hibernation image 1885 present during boot. 1886 nocompress Don't compress/decompress hibernation images. 1887 no Disable hibernation and resume. 1888 protect_image Turn on image protection during restoration 1889 (that will set all pages holding image data 1890 during restoration read-only). 1891 1892 hibernate.compressor= [HIBERNATION] Compression algorithm to be 1893 used with hibernation. 1894 Format: { lzo | lz4 } 1895 Default: lzo 1896 1897 lzo: Select LZO compression algorithm to 1898 compress/decompress hibernation image. 1899 1900 lz4: Select LZ4 compression algorithm to 1901 compress/decompress hibernation image. 1902 1903 hibernate.pm_test_delay= 1904 [HIBERNATION] 1905 Sets the number of seconds to remain in a hibernation test 1906 mode before resuming the system (see 1907 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 1908 is set. Default value is 5. 1909 1910 highmem=nn[KMG] [KNL,BOOT,EARLY] forces the highmem zone to have an exact 1911 size of <nn>. This works even on boxes that have no 1912 highmem otherwise. This also works to reduce highmem 1913 size on bigger boxes. 1914 1915 highres= [KNL] Enable/disable high resolution timer mode. 1916 Valid parameters: "on", "off" 1917 Default: "on" 1918 1919 hlt [BUGS=ARM,SH] 1920 1921 hostname= [KNL,EARLY] Set the hostname (aka UTS nodename). 1922 Format: <string> 1923 This allows setting the system's hostname during early 1924 startup. This sets the name returned by gethostname. 1925 Using this parameter to set the hostname makes it 1926 possible to ensure the hostname is correctly set before 1927 any userspace processes run, avoiding the possibility 1928 that a process may call gethostname before the hostname 1929 has been explicitly set, resulting in the calling 1930 process getting an incorrect result. The string must 1931 not exceed the maximum allowed hostname length (usually 1932 64 characters) and will be truncated otherwise. 1933 1934 hpet= [X86-32,HPET] option to control HPET usage 1935 Format: { enable (default) | disable | force | 1936 verbose } 1937 disable: disable HPET and use PIT instead 1938 force: allow force enabled of undocumented chips (ICH4, 1939 VIA, nVidia) 1940 verbose: show contents of HPET registers during setup 1941 1942 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 1943 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 1944 1945 hugepages= [HW,EARLY] Number of HugeTLB pages to allocate at boot. 1946 If this follows hugepagesz (below), it specifies 1947 the number of pages of hugepagesz to be allocated. 1948 If this is the first HugeTLB parameter on the command 1949 line, it specifies the number of pages to allocate for 1950 the default huge page size. If using node format, the 1951 number of pages to allocate per-node can be specified. 1952 See also Documentation/admin-guide/mm/hugetlbpage.rst. 1953 Format: <integer> or (node format) 1954 <node>:<integer>[,<node>:<integer>] 1955 1956 hugepagesz= 1957 [HW,EARLY] The size of the HugeTLB pages. This is 1958 used in conjunction with hugepages (above) to 1959 allocate huge pages of a specific size at boot. The 1960 pair hugepagesz=X hugepages=Y can be specified once 1961 for each supported huge page size. Huge page sizes 1962 are architecture dependent. See also 1963 Documentation/admin-guide/mm/hugetlbpage.rst. 1964 Format: size[KMG] 1965 1966 hugepage_alloc_threads= 1967 [HW] The number of threads that should be used to 1968 allocate hugepages during boot. This option can be 1969 used to improve system bootup time when allocating 1970 a large amount of huge pages. 1971 The default value is 25% of the available hardware threads. 1972 1973 Note that this parameter only applies to non-gigantic huge pages. 1974 1975 hugetlb_cma= [HW,CMA,EARLY] The size of a CMA area used for allocation 1976 of gigantic hugepages. Or using node format, the size 1977 of a CMA area per node can be specified. 1978 Format: nn[KMGTPE] or (node format) 1979 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]] 1980 1981 Reserve a CMA area of given size and allocate gigantic 1982 hugepages using the CMA allocator. If enabled, the 1983 boot-time allocation of gigantic hugepages is skipped. 1984 1985 hugetlb_cma_only= 1986 [HW,CMA,EARLY] When allocating new HugeTLB pages, only 1987 try to allocate from the CMA areas. 1988 1989 This option does nothing if hugetlb_cma= is not also 1990 specified. 1991 1992 hugetlb_free_vmemmap= 1993 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP 1994 enabled. 1995 Control if HugeTLB Vmemmap Optimization (HVO) is enabled. 1996 Allows heavy hugetlb users to free up some more 1997 memory (7 * PAGE_SIZE for each 2MB hugetlb page). 1998 Format: { on | off (default) } 1999 2000 on: enable HVO 2001 off: disable HVO 2002 2003 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y, 2004 the default is on. 2005 2006 Note that the vmemmap pages may be allocated from the added 2007 memory block itself when memory_hotplug.memmap_on_memory is 2008 enabled, those vmemmap pages cannot be optimized even if this 2009 feature is enabled. Other vmemmap pages not allocated from 2010 the added memory block itself do not be affected. 2011 2012 hung_task_panic= 2013 [KNL] Number of hung tasks to trigger kernel panic. 2014 Format: <int> 2015 2016 When set to a non-zero value, a kernel panic will be triggered if 2017 the number of detected hung tasks reaches this value. 2018 2019 0: don't panic 2020 1: panic immediately on first hung task 2021 N: panic after N hung tasks are detected in a single scan 2022 2023 The default value is controlled by the 2024 CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time option. The value 2025 selected by this boot parameter can be changed later by the 2026 kernel.hung_task_panic sysctl. 2027 2028 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 2029 terminal devices. Valid values: 0..8 2030 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 2031 If specified, z/VM IUCV HVC accepts connections 2032 from listed z/VM user IDs only. 2033 2034 hv_nopvspin [X86,HYPER_V,EARLY] 2035 Disables the paravirt spinlock optimizations 2036 which allow the hypervisor to 'idle' the guest 2037 on lock contention. 2038 2039 hw_protection= [HW] 2040 Format: reboot | shutdown 2041 2042 Hardware protection action taken on critical events like 2043 overtemperature or imminent voltage loss. 2044 2045 i2c_bus= [HW] Override the default board specific I2C bus speed 2046 or register an additional I2C bus that is not 2047 registered from board initialization code. 2048 Format: 2049 <bus_id>,<clkrate> 2050 2051 i2c_touchscreen_props= [HW,ACPI,X86] 2052 Set device-properties for ACPI-enumerated I2C-attached 2053 touchscreen, to e.g. fix coordinates of upside-down 2054 mounted touchscreens. If you need this option please 2055 submit a drivers/platform/x86/touchscreen_dmi.c patch 2056 adding a DMI quirk for this. 2057 2058 Format: 2059 <ACPI_HW_ID>:<prop_name>=<val>[:prop_name=val][:...] 2060 Where <val> is one of: 2061 Omit "=<val>" entirely Set a boolean device-property 2062 Unsigned number Set a u32 device-property 2063 Anything else Set a string device-property 2064 2065 Examples (split over multiple lines): 2066 i2c_touchscreen_props=GDIX1001:touchscreen-inverted-x: 2067 touchscreen-inverted-y 2068 2069 i2c_touchscreen_props=MSSL1680:touchscreen-size-x=1920: 2070 touchscreen-size-y=1080:touchscreen-inverted-y: 2071 firmware-name=gsl1680-vendor-model.fw:silead,home-button 2072 2073 i8042.debug [HW] Toggle i8042 debug mode 2074 i8042.unmask_kbd_data 2075 [HW] Enable printing of interrupt data from the KBD port 2076 (disabled by default, and as a pre-condition 2077 requires that i8042.debug=1 be enabled) 2078 i8042.direct [HW] Put keyboard port into non-translated mode 2079 i8042.dumbkbd [HW] Pretend that controller can only read data from 2080 keyboard and cannot control its state 2081 (Don't attempt to blink the leds) 2082 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 2083 i8042.nokbd [HW] Don't check/create keyboard port 2084 i8042.noloop [HW] Disable the AUX Loopback command while probing 2085 for the AUX port 2086 i8042.nomux [HW] Don't check presence of an active multiplexing 2087 controller 2088 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 2089 controllers 2090 i8042.notimeout [HW] Ignore timeout condition signalled by controller 2091 i8042.reset [HW] Reset the controller during init, cleanup and 2092 suspend-to-ram transitions, only during s2r 2093 transitions, or never reset 2094 Format: { 1 | Y | y | 0 | N | n } 2095 1, Y, y: always reset controller 2096 0, N, n: don't ever reset controller 2097 Default: only on s2r transitions on x86; most other 2098 architectures force reset to be always executed 2099 i8042.unlock [HW] Unlock (ignore) the keylock 2100 i8042.kbdreset [HW] Reset device connected to KBD port 2101 i8042.probe_defer 2102 [HW] Allow deferred probing upon i8042 probe errors 2103 2104 i810= [HW,DRM] 2105 2106 i915.invert_brightness= 2107 [DRM] Invert the sense of the variable that is used to 2108 set the brightness of the panel backlight. Normally a 2109 brightness value of 0 indicates backlight switched off, 2110 and the maximum of the brightness value sets the backlight 2111 to maximum brightness. If this parameter is set to 0 2112 (default) and the machine requires it, or this parameter 2113 is set to 1, a brightness value of 0 sets the backlight 2114 to maximum brightness, and the maximum of the brightness 2115 value switches the backlight off. 2116 -1 -- never invert brightness 2117 0 -- machine default 2118 1 -- force brightness inversion 2119 2120 ia32_emulation= [X86-64] 2121 Format: <bool> 2122 When true, allows loading 32-bit programs and executing 32-bit 2123 syscalls, essentially overriding IA32_EMULATION_DEFAULT_DISABLED at 2124 boot time. When false, unconditionally disables IA32 emulation. 2125 2126 icn= [HW,ISDN] 2127 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 2128 2129 2130 idle= [X86,EARLY] 2131 Format: idle=poll, idle=halt, idle=nomwait 2132 2133 idle=poll: Don't do power saving in the idle loop 2134 using HLT, but poll for rescheduling event. This will 2135 make the CPUs eat a lot more power, but may be useful 2136 to get slightly better performance in multiprocessor 2137 benchmarks. It also makes some profiling using 2138 performance counters more accurate. Please note that 2139 on systems with MONITOR/MWAIT support (like Intel 2140 EM64T CPUs) this option has no performance advantage 2141 over the normal idle loop. It may also interact badly 2142 with hyperthreading. 2143 2144 idle=halt: Halt is forced to be used for CPU idle. 2145 In such case C2/C3 won't be used again. 2146 2147 idle=nomwait: Disable mwait for CPU C-states 2148 2149 idxd.sva= [HW] 2150 Format: <bool> 2151 Allow force disabling of Shared Virtual Memory (SVA) 2152 support for the idxd driver. By default it is set to 2153 true (1). 2154 2155 idxd.tc_override= [HW] 2156 Format: <bool> 2157 Allow override of default traffic class configuration 2158 for the device. By default it is set to false (0). 2159 2160 ieee754= [MIPS] Select IEEE Std 754 conformance mode 2161 Format: { strict | legacy | 2008 | relaxed | emulated } 2162 Default: strict 2163 2164 Choose which programs will be accepted for execution 2165 based on the IEEE 754 NaN encoding(s) supported by 2166 the FPU and the NaN encoding requested with the value 2167 of an ELF file header flag individually set by each 2168 binary. Hardware implementations are permitted to 2169 support either or both of the legacy and the 2008 NaN 2170 encoding mode. 2171 2172 Available settings are as follows: 2173 strict accept binaries that request a NaN encoding 2174 supported by the FPU 2175 legacy only accept legacy-NaN binaries, if supported 2176 by the FPU 2177 2008 only accept 2008-NaN binaries, if supported 2178 by the FPU 2179 relaxed accept any binaries regardless of whether 2180 supported by the FPU 2181 emulated accept any binaries but enable FPU emulator 2182 if binary mode is unsupported by the FPU. 2183 2184 The FPU emulator is always able to support both NaN 2185 encodings, so if no FPU hardware is present or it has 2186 been disabled with 'nofpu', then the settings of 2187 'legacy' and '2008' strap the emulator accordingly, 2188 'relaxed' straps the emulator for both legacy-NaN and 2189 2008-NaN, whereas 'strict' enables legacy-NaN only on 2190 legacy processors and both NaN encodings on MIPS32 or 2191 MIPS64 CPUs. 2192 2193 The setting for ABS.fmt/NEG.fmt instruction execution 2194 mode generally follows that for the NaN encoding, 2195 except where unsupported by hardware. 2196 2197 ignore_loglevel [KNL,EARLY] 2198 Ignore loglevel setting - this will print /all/ 2199 kernel messages to the console. Useful for debugging. 2200 We also add it as printk module parameter, so users 2201 could change it dynamically, usually by 2202 /sys/module/printk/parameters/ignore_loglevel. 2203 2204 ignore_rlimit_data 2205 Ignore RLIMIT_DATA setting for data mappings, 2206 print warning at first misuse. Can be changed via 2207 /sys/module/kernel/parameters/ignore_rlimit_data. 2208 2209 ihash_entries= [KNL] 2210 Set number of hash buckets for inode cache. 2211 2212 ima_appraise= [IMA] appraise integrity measurements 2213 Format: { "off" | "enforce" | "fix" | "log" } 2214 default: "enforce" 2215 2216 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 2217 The builtin appraise policy appraises all files 2218 owned by uid=0. 2219 2220 ima_canonical_fmt [IMA] 2221 Use the canonical format for the binary runtime 2222 measurements, instead of host native format. 2223 2224 ima_hash= [IMA] 2225 Format: { md5 | sha1 | rmd160 | sha256 | sha384 2226 | sha512 | ... } 2227 default: "sha1" 2228 2229 The list of supported hash algorithms is defined 2230 in crypto/hash_info.h. 2231 2232 ima_policy= [IMA] 2233 The builtin policies to load during IMA setup. 2234 Format: "tcb | appraise_tcb | secure_boot | 2235 fail_securely | critical_data" 2236 2237 The "tcb" policy measures all programs exec'd, files 2238 mmap'd for exec, and all files opened with the read 2239 mode bit set by either the effective uid (euid=0) or 2240 uid=0. 2241 2242 The "appraise_tcb" policy appraises the integrity of 2243 all files owned by root. 2244 2245 The "secure_boot" policy appraises the integrity 2246 of files (eg. kexec kernel image, kernel modules, 2247 firmware, policy, etc) based on file signatures. 2248 2249 The "fail_securely" policy forces file signature 2250 verification failure also on privileged mounted 2251 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 2252 flag. 2253 2254 The "critical_data" policy measures kernel integrity 2255 critical data. 2256 2257 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 2258 Load a policy which meets the needs of the Trusted 2259 Computing Base. This means IMA will measure all 2260 programs exec'd, files mmap'd for exec, and all files 2261 opened for read by uid=0. 2262 2263 ima_template= [IMA] 2264 Select one of defined IMA measurements template formats. 2265 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" | 2266 "ima-sigv2" } 2267 Default: "ima-ng" 2268 2269 ima_template_fmt= 2270 [IMA] Define a custom template format. 2271 Format: { "field1|...|fieldN" } 2272 2273 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 2274 Format: <min_file_size> 2275 Set the minimal file size for using asynchronous hash. 2276 If left unspecified, ahash usage is disabled. 2277 2278 ahash performance varies for different data sizes on 2279 different crypto accelerators. This option can be used 2280 to achieve the best performance for a particular HW. 2281 2282 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size 2283 Format: <bufsize> 2284 Set hashing buffer size. Default: 4k. 2285 2286 ahash performance varies for different chunk sizes on 2287 different crypto accelerators. This option can be used 2288 to achieve best performance for particular HW. 2289 2290 ima= [IMA] Enable or disable IMA 2291 Format: { "off" | "on" } 2292 Default: "on" 2293 Note that disabling IMA is limited to kdump kernel. 2294 2295 indirect_target_selection= [X86,Intel] Mitigation control for Indirect 2296 Target Selection(ITS) bug in Intel CPUs. Updated 2297 microcode is also required for a fix in IBPB. 2298 2299 on: Enable mitigation (default). 2300 off: Disable mitigation. 2301 force: Force the ITS bug and deploy default 2302 mitigation. 2303 vmexit: Only deploy mitigation if CPU is affected by 2304 guest/host isolation part of ITS. 2305 stuff: Deploy RSB-fill mitigation when retpoline is 2306 also deployed. Otherwise, deploy the default 2307 mitigation. 2308 2309 For details see: 2310 Documentation/admin-guide/hw-vuln/indirect-target-selection.rst 2311 2312 init= [KNL] 2313 Format: <full_path> 2314 Run specified binary instead of /sbin/init as init 2315 process. 2316 2317 initcall_debug [KNL] Trace initcalls as they are executed. Useful 2318 for working out where the kernel is dying during 2319 startup. 2320 2321 initcall_blacklist= [KNL] Do not execute a comma-separated list of 2322 initcall functions. Useful for debugging built-in 2323 modules and initcalls. 2324 2325 initramfs_async= [KNL] 2326 Format: <bool> 2327 Default: 1 2328 This parameter controls whether the initramfs 2329 image is unpacked asynchronously, concurrently 2330 with devices being probed and 2331 initialized. This should normally just work, 2332 but as a debugging aid, one can get the 2333 historical behaviour of the initramfs 2334 unpacking being completed before device_ and 2335 late_ initcalls. 2336 2337 initrd= [BOOT,EARLY] Specify the location of the initial ramdisk 2338 2339 initrdmem= [KNL,EARLY] Specify a physical address and size from which to 2340 load the initrd. If an initrd is compiled in or 2341 specified in the bootparams, it takes priority over this 2342 setting. 2343 Format: ss[KMG],nn[KMG] 2344 Default is 0, 0 2345 2346 init_on_alloc= [MM,EARLY] Fill newly allocated pages and heap objects with 2347 zeroes. 2348 Format: 0 | 1 2349 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 2350 2351 init_on_free= [MM,EARLY] Fill freed pages and heap objects with zeroes. 2352 Format: 0 | 1 2353 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 2354 2355 init_pkru= [X86] Specify the default memory protection keys rights 2356 register contents for all processes. 0x55555554 by 2357 default (disallow access to all but pkey 0). Can 2358 override in debugfs after boot. 2359 2360 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 2361 Format: <irq> 2362 2363 int_pln_enable [X86] Enable power limit notification interrupt 2364 2365 integrity_audit=[IMA] 2366 Format: { "0" | "1" } 2367 0 -- basic integrity auditing messages. (Default) 2368 1 -- additional integrity auditing messages. 2369 2370 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 2371 on 2372 Enable intel iommu driver. 2373 off 2374 Disable intel iommu driver. 2375 igfx_off [Default Off] 2376 By default, gfx is mapped as normal device. If a gfx 2377 device has a dedicated DMAR unit, the DMAR unit is 2378 bypassed by not enabling DMAR with this option. In 2379 this case, gfx device will use physical address for 2380 DMA. 2381 strict [Default Off] 2382 Deprecated, equivalent to iommu.strict=1. 2383 sp_off [Default Off] 2384 By default, super page will be supported if Intel IOMMU 2385 has the capability. With this option, super page will 2386 not be supported. 2387 sm_on 2388 Enable the Intel IOMMU scalable mode if the hardware 2389 advertises that it has support for the scalable mode 2390 translation. 2391 sm_off 2392 Disallow use of the Intel IOMMU scalable mode. 2393 tboot_noforce [Default Off] 2394 Do not force the Intel IOMMU enabled under tboot. 2395 By default, tboot will force Intel IOMMU on, which 2396 could harm performance of some high-throughput 2397 devices like 40GBit network cards, even if identity 2398 mapping is enabled. 2399 Note that using this option lowers the security 2400 provided by tboot because it makes the system 2401 vulnerable to DMA attacks. 2402 2403 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 2404 0 disables intel_idle and fall back on acpi_idle. 2405 1 to 9 specify maximum depth of C-state. 2406 2407 intel_pstate= [X86,EARLY] 2408 disable 2409 Do not enable intel_pstate as the default 2410 scaling driver for the supported processors 2411 active 2412 Use intel_pstate driver to bypass the scaling 2413 governors layer of cpufreq and provides it own 2414 algorithms for p-state selection. There are two 2415 P-state selection algorithms provided by 2416 intel_pstate in the active mode: powersave and 2417 performance. The way they both operate depends 2418 on whether or not the hardware managed P-states 2419 (HWP) feature has been enabled in the processor 2420 and possibly on the processor model. 2421 passive 2422 Use intel_pstate as a scaling driver, but configure it 2423 to work with generic cpufreq governors (instead of 2424 enabling its internal governor). This mode cannot be 2425 used along with the hardware-managed P-states (HWP) 2426 feature. 2427 force 2428 Enable intel_pstate on systems that prohibit it by default 2429 in favor of acpi-cpufreq. Forcing the intel_pstate driver 2430 instead of acpi-cpufreq may disable platform features, such 2431 as thermal controls and power capping, that rely on ACPI 2432 P-States information being indicated to OSPM and therefore 2433 should be used with caution. This option does not work with 2434 processors that aren't supported by the intel_pstate driver 2435 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 2436 no_hwp 2437 Do not enable hardware P state control (HWP) 2438 if available. 2439 hwp_only 2440 Only load intel_pstate on systems which support 2441 hardware P state control (HWP) if available. 2442 support_acpi_ppc 2443 Enforce ACPI _PPC performance limits. If the Fixed ACPI 2444 Description Table, specifies preferred power management 2445 profile as "Enterprise Server" or "Performance Server", 2446 then this feature is turned on by default. 2447 per_cpu_perf_limits 2448 Allow per-logical-CPU P-State performance control limits using 2449 cpufreq sysfs interface 2450 no_cas 2451 Do not enable capacity-aware scheduling (CAS) on 2452 hybrid systems 2453 2454 intremap= [X86-64,Intel-IOMMU,EARLY] 2455 on enable Interrupt Remapping (default) 2456 off disable Interrupt Remapping 2457 nosid disable Source ID checking 2458 no_x2apic_optout 2459 BIOS x2APIC opt-out request will be ignored 2460 nopost disable Interrupt Posting 2461 posted_msi 2462 enable MSIs delivered as posted interrupts 2463 2464 iomem= Disable strict checking of access to MMIO memory 2465 strict regions from userspace. 2466 relaxed 2467 2468 iommu= [X86,EARLY] 2469 2470 off 2471 Don't initialize and use any kind of IOMMU. 2472 2473 force 2474 Force the use of the hardware IOMMU even when 2475 it is not actually needed (e.g. because < 3 GB 2476 memory). 2477 2478 noforce 2479 Don't force hardware IOMMU usage when it is not 2480 needed. (default). 2481 2482 biomerge 2483 panic 2484 nopanic 2485 merge 2486 nomerge 2487 2488 soft 2489 Use software bounce buffering (SWIOTLB) (default for 2490 Intel machines). This can be used to prevent the usage 2491 of an available hardware IOMMU. 2492 2493 [X86] 2494 pt 2495 [X86] 2496 nopt 2497 [PPC/POWERNV] 2498 nobypass 2499 Disable IOMMU bypass, using IOMMU for PCI devices. 2500 2501 [X86] 2502 AMD Gart HW IOMMU-specific options: 2503 2504 <size> 2505 Set the size of the remapping area in bytes. 2506 2507 allowed 2508 Overwrite iommu off workarounds for specific chipsets 2509 2510 fullflush 2511 Flush IOMMU on each allocation (default). 2512 2513 nofullflush 2514 Don't use IOMMU fullflush. 2515 2516 memaper[=<order>] 2517 Allocate an own aperture over RAM with size 2518 32MB<<order. (default: order=1, i.e. 64MB) 2519 2520 merge 2521 Do scatter-gather (SG) merging. Implies "force" 2522 (experimental). 2523 2524 nomerge 2525 Don't do scatter-gather (SG) merging. 2526 2527 noaperture 2528 Ask the IOMMU not to touch the aperture for AGP. 2529 2530 noagp 2531 Don't initialize the AGP driver and use full aperture. 2532 2533 panic 2534 Always panic when IOMMU overflows. 2535 2536 iommu.forcedac= [ARM64,X86,EARLY] Control IOVA allocation for PCI devices. 2537 Format: { "0" | "1" } 2538 0 - Try to allocate a 32-bit DMA address first, before 2539 falling back to the full range if needed. 2540 1 - Allocate directly from the full usable range, 2541 forcing Dual Address Cycle for PCI cards supporting 2542 greater than 32-bit addressing. 2543 2544 iommu.strict= [ARM64,X86,S390,EARLY] Configure TLB invalidation behaviour 2545 Format: { "0" | "1" } 2546 0 - Lazy mode. 2547 Request that DMA unmap operations use deferred 2548 invalidation of hardware TLBs, for increased 2549 throughput at the cost of reduced device isolation. 2550 Will fall back to strict mode if not supported by 2551 the relevant IOMMU driver. 2552 1 - Strict mode. 2553 DMA unmap operations invalidate IOMMU hardware TLBs 2554 synchronously. 2555 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}. 2556 Note: on x86, strict mode specified via one of the 2557 legacy driver-specific options takes precedence. 2558 2559 iommu.passthrough= 2560 [ARM64,X86,EARLY] Configure DMA to bypass the IOMMU by default. 2561 Format: { "0" | "1" } 2562 0 - Use IOMMU translation for DMA. 2563 1 - Bypass the IOMMU for DMA. 2564 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2565 2566 io7= [HW] IO7 for Marvel-based Alpha systems 2567 See comment before marvel_specify_io7 in 2568 arch/alpha/kernel/core_marvel.c. 2569 2570 io_delay= [X86,EARLY] I/O delay method 2571 0x80 2572 Standard port 0x80 based delay 2573 0xed 2574 Alternate port 0xed based delay (needed on some systems) 2575 udelay 2576 Simple two microseconds delay 2577 none 2578 No delay 2579 2580 ip= [IP_PNP] 2581 See Documentation/admin-guide/nfs/nfsroot.rst. 2582 2583 ipcmni_extend [KNL,EARLY] Extend the maximum number of unique System V 2584 IPC identifiers from 32,768 to 16,777,216. 2585 2586 ipe.enforce= [IPE] 2587 Format: <bool> 2588 Determine whether IPE starts in permissive (0) or 2589 enforce (1) mode. The default is enforce. 2590 2591 ipe.success_audit= 2592 [IPE] 2593 Format: <bool> 2594 Start IPE with success auditing enabled, emitting 2595 an audit event when a binary is allowed. The default 2596 is 0. 2597 2598 irqaffinity= [SMP] Set the default irq affinity mask 2599 The argument is a cpu list, as described above. 2600 2601 irqchip.gicv2_force_probe= 2602 [ARM,ARM64,EARLY] 2603 Format: <bool> 2604 Force the kernel to look for the second 4kB page 2605 of a GICv2 controller even if the memory range 2606 exposed by the device tree is too small. 2607 2608 irqchip.gicv3_nolpi= 2609 [ARM,ARM64,EARLY] 2610 Force the kernel to ignore the availability of 2611 LPIs (and by consequence ITSs). Intended for system 2612 that use the kernel as a bootloader, and thus want 2613 to let secondary kernels in charge of setting up 2614 LPIs. 2615 2616 irqchip.gicv3_pseudo_nmi= [ARM64,EARLY] 2617 Enables support for pseudo-NMIs in the kernel. This 2618 requires the kernel to be built with 2619 CONFIG_ARM64_PSEUDO_NMI. 2620 2621 irqchip.riscv_imsic_noipi 2622 [RISC-V,EARLY] 2623 Force the kernel to not use IMSIC software injected MSIs 2624 as IPIs. Intended for system where IMSIC is trap-n-emulated, 2625 and thus want to reduce MMIO traps when triggering IPIs 2626 to multiple harts. 2627 2628 irqfixup [HW] 2629 When an interrupt is not handled search all handlers 2630 for it. Intended to get systems with badly broken 2631 firmware running. 2632 2633 irqhandler.duration_warn_us= [KNL] 2634 Warn if an IRQ handler exceeds the specified duration 2635 threshold in microseconds. Useful for identifying 2636 long-running IRQs in the system. 2637 2638 irqpoll [HW] 2639 When an interrupt is not handled search all handlers 2640 for it. Also check all handlers each timer 2641 interrupt. Intended to get systems with badly broken 2642 firmware running. 2643 2644 isapnp= [ISAPNP] 2645 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2646 2647 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2648 [Deprecated - use cpusets instead] 2649 Format: [flag-list,]<cpu-list> 2650 2651 Specify one or more CPUs to isolate from disturbances 2652 specified in the flag list (default: domain): 2653 2654 nohz 2655 Disable the tick when a single task runs as well as 2656 disabling other kernel noises like having RCU callbacks 2657 offloaded. This is equivalent to the nohz_full parameter. 2658 2659 A residual 1Hz tick is offloaded to workqueues, which you 2660 need to affine to housekeeping through the global 2661 workqueue's affinity configured via the 2662 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2663 by using the 'domain' flag described below. 2664 2665 NOTE: by default the global workqueue runs on all CPUs, 2666 so to protect individual CPUs the 'cpumask' file has to 2667 be configured manually after bootup. 2668 2669 domain 2670 Isolate from the general SMP balancing and scheduling 2671 algorithms. Note that performing domain isolation this way 2672 is irreversible: it's not possible to bring back a CPU to 2673 the domains once isolated through isolcpus. It's strongly 2674 advised to use cpusets instead to disable scheduler load 2675 balancing through the "cpuset.sched_load_balance" file. 2676 It offers a much more flexible interface where CPUs can 2677 move in and out of an isolated set anytime. 2678 2679 You can move a process onto or off an "isolated" CPU via 2680 the CPU affinity syscalls or cpuset. 2681 <cpu number> begins at 0 and the maximum value is 2682 "number of CPUs in system - 1". 2683 2684 managed_irq 2685 2686 Isolate from being targeted by managed interrupts 2687 which have an interrupt mask containing isolated 2688 CPUs. The affinity of managed interrupts is 2689 handled by the kernel and cannot be changed via 2690 the /proc/irq/* interfaces. 2691 2692 This isolation is best effort and only effective 2693 if the automatically assigned interrupt mask of a 2694 device queue contains isolated and housekeeping 2695 CPUs. If housekeeping CPUs are online then such 2696 interrupts are directed to the housekeeping CPU 2697 so that IO submitted on the housekeeping CPU 2698 cannot disturb the isolated CPU. 2699 2700 If a queue's affinity mask contains only isolated 2701 CPUs then this parameter has no effect on the 2702 interrupt routing decision, though interrupts are 2703 only delivered when tasks running on those 2704 isolated CPUs submit IO. IO submitted on 2705 housekeeping CPUs has no influence on those 2706 queues. 2707 2708 The format of <cpu-list> is described above. 2709 2710 iucv= [HW,NET] 2711 2712 ivrs_ioapic [HW,X86-64] 2713 Provide an override to the IOAPIC-ID<->DEVICE-ID 2714 mapping provided in the IVRS ACPI table. 2715 By default, PCI segment is 0, and can be omitted. 2716 2717 For example, to map IOAPIC-ID decimal 10 to 2718 PCI segment 0x1 and PCI device 00:14.0, 2719 write the parameter as: 2720 ivrs_ioapic=10@0001:00:14.0 2721 2722 Deprecated formats: 2723 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0 2724 write the parameter as: 2725 ivrs_ioapic[10]=00:14.0 2726 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and 2727 PCI device 00:14.0 write the parameter as: 2728 ivrs_ioapic[10]=0001:00:14.0 2729 2730 ivrs_hpet [HW,X86-64] 2731 Provide an override to the HPET-ID<->DEVICE-ID 2732 mapping provided in the IVRS ACPI table. 2733 By default, PCI segment is 0, and can be omitted. 2734 2735 For example, to map HPET-ID decimal 10 to 2736 PCI segment 0x1 and PCI device 00:14.0, 2737 write the parameter as: 2738 ivrs_hpet=10@0001:00:14.0 2739 2740 Deprecated formats: 2741 * To map HPET-ID decimal 0 to PCI device 00:14.0 2742 write the parameter as: 2743 ivrs_hpet[0]=00:14.0 2744 * To map HPET-ID decimal 10 to PCI segment 0x1 and 2745 PCI device 00:14.0 write the parameter as: 2746 ivrs_ioapic[10]=0001:00:14.0 2747 2748 ivrs_acpihid [HW,X86-64] 2749 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2750 mapping provided in the IVRS ACPI table. 2751 By default, PCI segment is 0, and can be omitted. 2752 2753 For example, to map UART-HID:UID AMD0020:0 to 2754 PCI segment 0x1 and PCI device ID 00:14.5, 2755 write the parameter as: 2756 ivrs_acpihid=AMD0020:0@0001:00:14.5 2757 2758 Deprecated formats: 2759 * To map UART-HID:UID AMD0020:0 to PCI segment is 0, 2760 PCI device ID 00:14.5, write the parameter as: 2761 ivrs_acpihid[00:14.5]=AMD0020:0 2762 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and 2763 PCI device ID 00:14.5, write the parameter as: 2764 ivrs_acpihid[0001:00:14.5]=AMD0020:0 2765 2766 js= [HW,JOY] Analog joystick 2767 See Documentation/input/joydev/joystick.rst. 2768 2769 kasan_multi_shot 2770 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2771 report on every invalid memory access. Without this 2772 parameter KASAN will print report only for the first 2773 invalid access. 2774 2775 keep_bootcon [KNL,EARLY] 2776 Do not unregister boot console at start. This is only 2777 useful for debugging when something happens in the window 2778 between unregistering the boot console and initializing 2779 the real console. 2780 2781 keepinitrd [HW,ARM] See retain_initrd. 2782 2783 kernelcore= [KNL,X86,PPC,EARLY] 2784 Format: nn[KMGTPE] | nn% | "mirror" 2785 This parameter specifies the amount of memory usable by 2786 the kernel for non-movable allocations. The requested 2787 amount is spread evenly throughout all nodes in the 2788 system as ZONE_NORMAL. The remaining memory is used for 2789 movable memory in its own zone, ZONE_MOVABLE. In the 2790 event, a node is too small to have both ZONE_NORMAL and 2791 ZONE_MOVABLE, kernelcore memory will take priority and 2792 other nodes will have a larger ZONE_MOVABLE. 2793 2794 ZONE_MOVABLE is used for the allocation of pages that 2795 may be reclaimed or moved by the page migration 2796 subsystem. Note that allocations like PTEs-from-HighMem 2797 still use the HighMem zone if it exists, and the Normal 2798 zone if it does not. 2799 2800 It is possible to specify the exact amount of memory in 2801 the form of "nn[KMGTPE]", a percentage of total system 2802 memory in the form of "nn%", or "mirror". If "mirror" 2803 option is specified, mirrored (reliable) memory is used 2804 for non-movable allocations and remaining memory is used 2805 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2806 are exclusive, so you cannot specify multiple forms. 2807 2808 kgdbdbgp= [KGDB,HW,EARLY] kgdb over EHCI usb debug port. 2809 Format: <Controller#>[,poll interval] 2810 The controller # is the number of the ehci usb debug 2811 port as it is probed via PCI. The poll interval is 2812 optional and is the number seconds in between 2813 each poll cycle to the debug port in case you need 2814 the functionality for interrupting the kernel with 2815 gdb or control-c on the dbgp connection. When 2816 not using this parameter you use sysrq-g to break into 2817 the kernel debugger. 2818 2819 kgdboc= [KGDB,HW] kgdb over consoles. 2820 Requires a tty driver that supports console polling, 2821 or a supported polling keyboard driver (non-usb). 2822 Serial only format: <serial_device>[,baud] 2823 keyboard only format: kbd 2824 keyboard and serial format: kbd,<serial_device>[,baud] 2825 Optional Kernel mode setting: 2826 kms, kbd format: kms,kbd 2827 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2828 2829 kgdboc_earlycon= [KGDB,HW,EARLY] 2830 If the boot console provides the ability to read 2831 characters and can work in polling mode, you can use 2832 this parameter to tell kgdb to use it as a backend 2833 until the normal console is registered. Intended to 2834 be used together with the kgdboc parameter which 2835 specifies the normal console to transition to. 2836 2837 The name of the early console should be specified 2838 as the value of this parameter. Note that the name of 2839 the early console might be different than the tty 2840 name passed to kgdboc. It's OK to leave the value 2841 blank and the first boot console that implements 2842 read() will be picked. 2843 2844 kgdbwait [KGDB,EARLY] Stop kernel execution and enter the 2845 kernel debugger at the earliest opportunity. 2846 2847 kho= [KEXEC,EARLY] 2848 Format: { "0" | "1" | "off" | "on" | "y" | "n" } 2849 Enables or disables Kexec HandOver. 2850 "0" | "off" | "n" - kexec handover is disabled 2851 "1" | "on" | "y" - kexec handover is enabled 2852 2853 kho_scratch= [KEXEC,EARLY] 2854 Format: ll[KMG],mm[KMG],nn[KMG] | nn% 2855 Defines the size of the KHO scratch region. The KHO 2856 scratch regions are physically contiguous memory 2857 ranges that can only be used for non-kernel 2858 allocations. That way, even when memory is heavily 2859 fragmented with handed over memory, the kexeced 2860 kernel will always have enough contiguous ranges to 2861 bootstrap itself. 2862 2863 It is possible to specify the exact amount of 2864 memory in the form of "ll[KMG],mm[KMG],nn[KMG]" 2865 where the first parameter defines the size of a low 2866 memory scratch area, the second parameter defines 2867 the size of a global scratch area and the third 2868 parameter defines the size of additional per-node 2869 scratch areas. The form "nn%" defines scale factor 2870 (in percents) of memory that was used during boot. 2871 2872 kmac= [MIPS] Korina ethernet MAC address. 2873 Configure the RouterBoard 532 series on-chip 2874 Ethernet adapter MAC address. 2875 2876 kmemleak= [KNL,EARLY] Boot-time kmemleak enable/disable 2877 Valid arguments: on, off 2878 Default: on 2879 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 2880 the default is off. 2881 2882 kprobe_event=[probe-list] 2883 [FTRACE] Add kprobe events and enable at boot time. 2884 The probe-list is a semicolon delimited list of probe 2885 definitions. Each definition is same as kprobe_events 2886 interface, but the parameters are comma delimited. 2887 For example, to add a kprobe event on vfs_read with 2888 arg1 and arg2, add to the command line; 2889 2890 kprobe_event=p,vfs_read,$arg1,$arg2 2891 2892 See also Documentation/trace/kprobetrace.rst "Kernel 2893 Boot Parameter" section. 2894 2895 kpti= [ARM64,EARLY] Control page table isolation of 2896 user and kernel address spaces. 2897 Default: enabled on cores which need mitigation. 2898 0: force disabled 2899 1: force enabled 2900 2901 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires 2902 CONFIG_KUNIT to be set to be fully enabled. The 2903 default value can be overridden via 2904 KUNIT_DEFAULT_ENABLED. 2905 Default is 1 (enabled) 2906 2907 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 2908 Default is 0 (don't ignore, but inject #GP) 2909 2910 kvm.eager_page_split= 2911 [KVM,X86] Controls whether or not KVM will try to 2912 proactively split all huge pages during dirty logging. 2913 Eager page splitting reduces interruptions to vCPU 2914 execution by eliminating the write-protection faults 2915 and MMU lock contention that would otherwise be 2916 required to split huge pages lazily. 2917 2918 VM workloads that rarely perform writes or that write 2919 only to a small region of VM memory may benefit from 2920 disabling eager page splitting to allow huge pages to 2921 still be used for reads. 2922 2923 The behavior of eager page splitting depends on whether 2924 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If 2925 disabled, all huge pages in a memslot will be eagerly 2926 split when dirty logging is enabled on that memslot. If 2927 enabled, eager page splitting will be performed during 2928 the KVM_CLEAR_DIRTY ioctl, and only for the pages being 2929 cleared. 2930 2931 Eager page splitting is only supported when kvm.tdp_mmu=Y. 2932 2933 Default is Y (on). 2934 2935 kvm.enable_virt_at_load=[KVM,ARM64,LOONGARCH,MIPS,RISCV,X86] 2936 If enabled, KVM will enable virtualization in hardware 2937 when KVM is loaded, and disable virtualization when KVM 2938 is unloaded (if KVM is built as a module). 2939 2940 If disabled, KVM will dynamically enable and disable 2941 virtualization on-demand when creating and destroying 2942 VMs, i.e. on the 0=>1 and 1=>0 transitions of the 2943 number of VMs. 2944 2945 Enabling virtualization at module load avoids potential 2946 latency for creation of the 0=>1 VM, as KVM serializes 2947 virtualization enabling across all online CPUs. The 2948 "cost" of enabling virtualization when KVM is loaded, 2949 is that doing so may interfere with using out-of-tree 2950 hypervisors that want to "own" virtualization hardware. 2951 2952 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 2953 Default is false (don't support). 2954 2955 kvm.nx_huge_pages= 2956 [KVM] Controls the software workaround for the 2957 X86_BUG_ITLB_MULTIHIT bug. 2958 force : Always deploy workaround. 2959 off : Never deploy workaround. 2960 auto : Deploy workaround based on the presence of 2961 X86_BUG_ITLB_MULTIHIT. 2962 2963 Default is 'auto'. 2964 2965 If the software workaround is enabled for the host, 2966 guests do need not to enable it for nested guests. 2967 2968 kvm.nx_huge_pages_recovery_ratio= 2969 [KVM] Controls how many 4KiB pages are periodically zapped 2970 back to huge pages. 0 disables the recovery, otherwise if 2971 the value is N KVM will zap 1/Nth of the 4KiB pages every 2972 period (see below). The default is 60. 2973 2974 kvm.nx_huge_pages_recovery_period_ms= 2975 [KVM] Controls the time period at which KVM zaps 4KiB pages 2976 back to huge pages. If the value is a non-zero N, KVM will 2977 zap a portion (see ratio above) of the pages every N msecs. 2978 If the value is 0 (the default), KVM will pick a period based 2979 on the ratio, such that a page is zapped after 1 hour on average. 2980 2981 kvm-amd.nested= [KVM,AMD] Control nested virtualization feature in 2982 KVM/SVM. Default is 1 (enabled). 2983 2984 kvm-amd.npt= [KVM,AMD] Control KVM's use of Nested Page Tables, 2985 a.k.a. Two-Dimensional Page Tables. Default is 1 2986 (enabled). Disable by KVM if hardware lacks support 2987 for NPT. 2988 2989 kvm-amd.ciphertext_hiding_asids= 2990 [KVM,AMD] Ciphertext hiding prevents disallowed accesses 2991 to SNP private memory from reading ciphertext. Instead, 2992 reads will see constant default values (0xff). 2993 2994 If ciphertext hiding is enabled, the joint SEV-ES and 2995 SEV-SNP ASID space is partitioned into separate SEV-ES 2996 and SEV-SNP ASID ranges, with the SEV-SNP range being 2997 [1..max_snp_asid] and the SEV-ES range being 2998 (max_snp_asid..min_sev_asid), where min_sev_asid is 2999 enumerated by CPUID.0x.8000_001F[EDX]. 3000 3001 A non-zero value enables SEV-SNP ciphertext hiding and 3002 adjusts the ASID ranges for SEV-ES and SEV-SNP guests. 3003 KVM caps the number of SEV-SNP ASIDs at the maximum 3004 possible value, e.g. specifying -1u will assign all 3005 joint SEV-ES and SEV-SNP ASIDs to SEV-SNP. Note, 3006 assigning all joint ASIDs to SEV-SNP, i.e. configuring 3007 max_snp_asid == min_sev_asid-1, will effectively make 3008 SEV-ES unusable. 3009 3010 kvm-arm.mode= 3011 [KVM,ARM,EARLY] Select one of KVM/arm64's modes of 3012 operation. 3013 3014 none: Forcefully disable KVM. 3015 3016 nvhe: Standard nVHE-based mode, without support for 3017 protected guests. 3018 3019 protected: Mode with support for guests whose state is 3020 kept private from the host, using VHE or 3021 nVHE depending on HW support. 3022 3023 nested: VHE-based mode with support for nested 3024 virtualization. Requires at least ARMv8.4 3025 hardware (with FEAT_NV2). 3026 3027 Defaults to VHE/nVHE based on hardware support. Setting 3028 mode to "protected" will disable kexec and hibernation 3029 for the host. To force nVHE on VHE hardware, add 3030 "arm64_sw.hvhe=0 id_aa64mmfr1.vh=0" to the 3031 command-line. 3032 "nested" is experimental and should be used with 3033 extreme caution. 3034 3035 kvm-arm.vgic_v3_group0_trap= 3036 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-0 3037 system registers 3038 3039 kvm-arm.vgic_v3_group1_trap= 3040 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-1 3041 system registers 3042 3043 kvm-arm.vgic_v3_common_trap= 3044 [KVM,ARM,EARLY] Trap guest accesses to GICv3 common 3045 system registers 3046 3047 kvm-arm.vgic_v4_enable= 3048 [KVM,ARM,EARLY] Allow use of GICv4 for direct 3049 injection of LPIs. 3050 3051 kvm-arm.wfe_trap_policy= 3052 [KVM,ARM] Control when to set WFE instruction trap for 3053 KVM VMs. Traps are allowed but not guaranteed by the 3054 CPU architecture. 3055 3056 trap: set WFE instruction trap 3057 3058 notrap: clear WFE instruction trap 3059 3060 kvm-arm.wfi_trap_policy= 3061 [KVM,ARM] Control when to set WFI instruction trap for 3062 KVM VMs. Traps are allowed but not guaranteed by the 3063 CPU architecture. 3064 3065 trap: set WFI instruction trap 3066 3067 notrap: clear WFI instruction trap 3068 3069 kvm_cma_resv_ratio=n [PPC,EARLY] 3070 Reserves given percentage from system memory area for 3071 contiguous memory allocation for KVM hash pagetable 3072 allocation. 3073 By default it reserves 5% of total system memory. 3074 Format: <integer> 3075 Default: 5 3076 3077 kvm-intel.ept= [KVM,Intel] Control KVM's use of Extended Page Tables, 3078 a.k.a. Two-Dimensional Page Tables. Default is 1 3079 (enabled). Disable by KVM if hardware lacks support 3080 for EPT. 3081 3082 kvm-intel.emulate_invalid_guest_state= 3083 [KVM,Intel] Control whether to emulate invalid guest 3084 state. Ignored if kvm-intel.enable_unrestricted_guest=1, 3085 as guest state is never invalid for unrestricted 3086 guests. This param doesn't apply to nested guests (L2), 3087 as KVM never emulates invalid L2 guest state. 3088 Default is 1 (enabled). 3089 3090 kvm-intel.flexpriority= 3091 [KVM,Intel] Control KVM's use of FlexPriority feature 3092 (TPR shadow). Default is 1 (enabled). Disable by KVM if 3093 hardware lacks support for it. 3094 3095 kvm-intel.nested= 3096 [KVM,Intel] Control nested virtualization feature in 3097 KVM/VMX. Default is 1 (enabled). 3098 3099 kvm-intel.unrestricted_guest= 3100 [KVM,Intel] Control KVM's use of unrestricted guest 3101 feature (virtualized real and unpaged mode). Default 3102 is 1 (enabled). Disable by KVM if EPT is disabled or 3103 hardware lacks support for it. 3104 3105 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 3106 CVE-2018-3620. 3107 3108 Valid arguments: never, cond, always 3109 3110 always: L1D cache flush on every VMENTER. 3111 cond: Flush L1D on VMENTER only when the code between 3112 VMEXIT and VMENTER can leak host memory. 3113 never: Disables the mitigation 3114 3115 Default is cond (do L1 cache flush in specific instances) 3116 3117 kvm-intel.vpid= [KVM,Intel] Control KVM's use of Virtual Processor 3118 Identification feature (tagged TLBs). Default is 1 3119 (enabled). Disable by KVM if hardware lacks support 3120 for it. 3121 3122 l1d_flush= [X86,INTEL,EARLY] 3123 Control mitigation for L1D based snooping vulnerability. 3124 3125 Certain CPUs are vulnerable to an exploit against CPU 3126 internal buffers which can forward information to a 3127 disclosure gadget under certain conditions. 3128 3129 In vulnerable processors, the speculatively 3130 forwarded data can be used in a cache side channel 3131 attack, to access data to which the attacker does 3132 not have direct access. 3133 3134 This parameter controls the mitigation. The 3135 options are: 3136 3137 on - enable the interface for the mitigation 3138 3139 l1tf= [X86,EARLY] Control mitigation of the L1TF vulnerability on 3140 affected CPUs 3141 3142 The kernel PTE inversion protection is unconditionally 3143 enabled and cannot be disabled. 3144 3145 full 3146 Provides all available mitigations for the 3147 L1TF vulnerability. Disables SMT and 3148 enables all mitigations in the 3149 hypervisors, i.e. unconditional L1D flush. 3150 3151 SMT control and L1D flush control via the 3152 sysfs interface is still possible after 3153 boot. Hypervisors will issue a warning 3154 when the first VM is started in a 3155 potentially insecure configuration, 3156 i.e. SMT enabled or L1D flush disabled. 3157 3158 full,force 3159 Same as 'full', but disables SMT and L1D 3160 flush runtime control. Implies the 3161 'nosmt=force' command line option. 3162 (i.e. sysfs control of SMT is disabled.) 3163 3164 flush 3165 Leaves SMT enabled and enables the default 3166 hypervisor mitigation, i.e. conditional 3167 L1D flush. 3168 3169 SMT control and L1D flush control via the 3170 sysfs interface is still possible after 3171 boot. Hypervisors will issue a warning 3172 when the first VM is started in a 3173 potentially insecure configuration, 3174 i.e. SMT enabled or L1D flush disabled. 3175 3176 flush,nosmt 3177 3178 Disables SMT and enables the default 3179 hypervisor mitigation. 3180 3181 SMT control and L1D flush control via the 3182 sysfs interface is still possible after 3183 boot. Hypervisors will issue a warning 3184 when the first VM is started in a 3185 potentially insecure configuration, 3186 i.e. SMT enabled or L1D flush disabled. 3187 3188 flush,nowarn 3189 Same as 'flush', but hypervisors will not 3190 warn when a VM is started in a potentially 3191 insecure configuration. 3192 3193 off 3194 Disables hypervisor mitigations and doesn't 3195 emit any warnings. 3196 It also drops the swap size and available 3197 RAM limit restriction on both hypervisor and 3198 bare metal. 3199 3200 Default is 'flush'. 3201 3202 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 3203 3204 l2cr= [PPC] 3205 3206 l3cr= [PPC] 3207 3208 lapic [X86-32,APIC,EARLY] Enable the local APIC even if BIOS 3209 disabled it. 3210 3211 lapic= [X86,APIC] Do not use TSC deadline 3212 value for LAPIC timer one-shot implementation. Default 3213 back to the programmable timer unit in the LAPIC. 3214 Format: notscdeadline 3215 3216 lapic_timer_c2_ok [X86,APIC,EARLY] trust the local apic timer 3217 in C2 power state. 3218 3219 libata.dma= [LIBATA] DMA control 3220 libata.dma=0 Disable all PATA and SATA DMA 3221 libata.dma=1 PATA and SATA Disk DMA only 3222 libata.dma=2 ATAPI (CDROM) DMA only 3223 libata.dma=4 Compact Flash DMA only 3224 Combinations also work, so libata.dma=3 enables DMA 3225 for disks and CDROMs, but not CFs. 3226 3227 libata.ignore_hpa= [LIBATA] Ignore HPA limit 3228 libata.ignore_hpa=0 keep BIOS limits (default) 3229 libata.ignore_hpa=1 ignore limits, using full disk 3230 3231 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 3232 when set. 3233 Format: <int> 3234 3235 libata.force= [LIBATA] Force configurations. The format is a comma- 3236 separated list of "[ID:]VAL" where ID is PORT[.DEVICE]. 3237 PORT and DEVICE are decimal numbers matching port, link 3238 or device. Basically, it matches the ATA ID string 3239 printed on console by libata. If the whole ID part is 3240 omitted, the last PORT and DEVICE values are used. If 3241 ID hasn't been specified yet, the configuration applies 3242 to all ports, links and devices. 3243 3244 If only DEVICE is omitted, the parameter applies to 3245 the port and all links and devices behind it. DEVICE 3246 number of 0 either selects the first device or the 3247 first fan-out link behind PMP device. It does not 3248 select the host link. DEVICE number of 15 selects the 3249 host link and device attached to it. 3250 3251 The VAL specifies the configuration to force. As long 3252 as there is no ambiguity, shortcut notation is allowed. 3253 For example, both 1.5 and 1.5G would work for 1.5Gbps. 3254 The following configurations can be forced. 3255 3256 * Cable type: 40c, 80c, short40c, unk, ign or sata. 3257 Any ID with matching PORT is used. 3258 3259 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 3260 3261 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 3262 udma[/][16,25,33,44,66,100,133] notation is also 3263 allowed. 3264 3265 * nohrst, nosrst, norst: suppress hard, soft and both 3266 resets. 3267 3268 * rstonce: only attempt one reset during hot-unplug 3269 link recovery. 3270 3271 * [no]dbdelay: Enable or disable the extra 200ms delay 3272 before debouncing a link PHY and device presence 3273 detection. 3274 3275 * [no]ncq: Turn on or off NCQ. 3276 3277 * [no]ncqtrim: Enable or disable queued DSM TRIM. 3278 3279 * [no]ncqati: Enable or disable NCQ trim on ATI chipset. 3280 3281 * [no]trim: Enable or disable (unqueued) TRIM. 3282 3283 * trim_zero: Indicate that TRIM command zeroes data. 3284 3285 * max_trim_128m: Set 128M maximum trim size limit. 3286 3287 * [no]dma: Turn on or off DMA transfers. 3288 3289 * atapi_dmadir: Enable ATAPI DMADIR bridge support. 3290 3291 * atapi_mod16_dma: Enable the use of ATAPI DMA for 3292 commands that are not a multiple of 16 bytes. 3293 3294 * [no]dmalog: Enable or disable the use of the 3295 READ LOG DMA EXT command to access logs. 3296 3297 * [no]iddevlog: Enable or disable access to the 3298 identify device data log. 3299 3300 * [no]logdir: Enable or disable access to the general 3301 purpose log directory. 3302 3303 * max_sec_128: Set transfer size limit to 128 sectors. 3304 3305 * max_sec_1024: Set or clear transfer size limit to 3306 1024 sectors. 3307 3308 * max_sec_lba48: Set or clear transfer size limit to 3309 65535 sectors. 3310 3311 * external: Mark port as external (hotplug-capable). 3312 3313 * [no]lpm: Enable or disable link power management. 3314 3315 * [no]setxfer: Indicate if transfer speed mode setting 3316 should be skipped. 3317 3318 * [no]fua: Disable or enable FUA (Force Unit Access) 3319 support for devices supporting this feature. 3320 3321 * dump_id: Dump IDENTIFY data. 3322 3323 * disable: Disable this device. 3324 3325 If there are multiple matching configurations changing 3326 the same attribute, the last one is used. 3327 3328 load_ramdisk= [RAM] [Deprecated] 3329 3330 lockd.nlm_grace_period=P [NFS] Assign grace period. 3331 Format: <integer> 3332 3333 lockd.nlm_tcpport=N [NFS] Assign TCP port. 3334 Format: <integer> 3335 3336 lockd.nlm_timeout=T [NFS] Assign timeout value. 3337 Format: <integer> 3338 3339 lockd.nlm_udpport=M [NFS] Assign UDP port. 3340 Format: <integer> 3341 3342 lockdown= [SECURITY,EARLY] 3343 { integrity | confidentiality } 3344 Enable the kernel lockdown feature. If set to 3345 integrity, kernel features that allow userland to 3346 modify the running kernel are disabled. If set to 3347 confidentiality, kernel features that allow userland 3348 to extract confidential information from the kernel 3349 are also disabled. 3350 3351 locktorture.acq_writer_lim= [KNL] 3352 Set the time limit in jiffies for a lock 3353 acquisition. Acquisitions exceeding this limit 3354 will result in a splat once they do complete. 3355 3356 locktorture.bind_readers= [KNL] 3357 Specify the list of CPUs to which the readers are 3358 to be bound. 3359 3360 locktorture.bind_writers= [KNL] 3361 Specify the list of CPUs to which the writers are 3362 to be bound. 3363 3364 locktorture.call_rcu_chains= [KNL] 3365 Specify the number of self-propagating call_rcu() 3366 chains to set up. These are used to ensure that 3367 there is a high probability of an RCU grace period 3368 in progress at any given time. Defaults to 0, 3369 which disables these call_rcu() chains. 3370 3371 locktorture.long_hold= [KNL] 3372 Specify the duration in milliseconds for the 3373 occasional long-duration lock hold time. Defaults 3374 to 100 milliseconds. Select 0 to disable. 3375 3376 locktorture.nested_locks= [KNL] 3377 Specify the maximum lock nesting depth that 3378 locktorture is to exercise, up to a limit of 8 3379 (MAX_NESTED_LOCKS). Specify zero to disable. 3380 Note that this parameter is ineffective on types 3381 of locks that do not support nested acquisition. 3382 3383 locktorture.nreaders_stress= [KNL] 3384 Set the number of locking read-acquisition kthreads. 3385 Defaults to being automatically set based on the 3386 number of online CPUs. 3387 3388 locktorture.nwriters_stress= [KNL] 3389 Set the number of locking write-acquisition kthreads. 3390 3391 locktorture.onoff_holdoff= [KNL] 3392 Set time (s) after boot for CPU-hotplug testing. 3393 3394 locktorture.onoff_interval= [KNL] 3395 Set time (s) between CPU-hotplug operations, or 3396 zero to disable CPU-hotplug testing. 3397 3398 locktorture.rt_boost= [KNL] 3399 Do periodic testing of real-time lock priority 3400 boosting. Select 0 to disable, 1 to boost 3401 only rt_mutex, and 2 to boost unconditionally. 3402 Defaults to 2, which might seem to be an 3403 odd choice, but which should be harmless for 3404 non-real-time spinlocks, due to their disabling 3405 of preemption. Note that non-realtime mutexes 3406 disable boosting. 3407 3408 locktorture.rt_boost_factor= [KNL] 3409 Number that determines how often and for how 3410 long priority boosting is exercised. This is 3411 scaled down by the number of writers, so that the 3412 number of boosts per unit time remains roughly 3413 constant as the number of writers increases. 3414 On the other hand, the duration of each boost 3415 increases with the number of writers. 3416 3417 locktorture.shuffle_interval= [KNL] 3418 Set task-shuffle interval (jiffies). Shuffling 3419 tasks allows some CPUs to go into dyntick-idle 3420 mode during the locktorture test. 3421 3422 locktorture.shutdown_secs= [KNL] 3423 Set time (s) after boot system shutdown. This 3424 is useful for hands-off automated testing. 3425 3426 locktorture.stat_interval= [KNL] 3427 Time (s) between statistics printk()s. 3428 3429 locktorture.stutter= [KNL] 3430 Time (s) to stutter testing, for example, 3431 specifying five seconds causes the test to run for 3432 five seconds, wait for five seconds, and so on. 3433 This tests the locking primitive's ability to 3434 transition abruptly to and from idle. 3435 3436 locktorture.torture_type= [KNL] 3437 Specify the locking implementation to test. 3438 3439 locktorture.verbose= [KNL] 3440 Enable additional printk() statements. 3441 3442 locktorture.writer_fifo= [KNL] 3443 Run the write-side locktorture kthreads at 3444 sched_set_fifo() real-time priority. 3445 3446 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 3447 Format: <irq> 3448 3449 loglevel= [KNL,EARLY] 3450 All Kernel Messages with a loglevel smaller than the 3451 console loglevel will be printed to the console. It can 3452 also be changed with klogd or other programs. The 3453 loglevels are defined as follows: 3454 3455 0 (KERN_EMERG) system is unusable 3456 1 (KERN_ALERT) action must be taken immediately 3457 2 (KERN_CRIT) critical conditions 3458 3 (KERN_ERR) error conditions 3459 4 (KERN_WARNING) warning conditions 3460 5 (KERN_NOTICE) normal but significant condition 3461 6 (KERN_INFO) informational 3462 7 (KERN_DEBUG) debug-level messages 3463 3464 log_buf_len=n[KMG] [KNL,EARLY] 3465 Sets the size of the printk ring buffer, in bytes. 3466 n must be a power of two and greater than the 3467 minimal size. The minimal size is defined by 3468 LOG_BUF_SHIFT kernel config parameter. There 3469 is also CONFIG_LOG_CPU_MAX_BUF_SHIFT config 3470 parameter that allows to increase the default size 3471 depending on the number of CPUs. See init/Kconfig 3472 for more details. 3473 3474 logo.nologo [FB] Disables display of the built-in Linux logo. 3475 This may be used to provide more screen space for 3476 kernel log messages and is useful when debugging 3477 kernel boot problems. 3478 3479 lp=0 [LP] Specify parallel ports to use, e.g, 3480 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 3481 lp=reset first parallel port). 'lp=0' disables the 3482 lp=auto printer driver. 'lp=reset' (which can be 3483 specified in addition to the ports) causes 3484 attached printers to be reset. Using 3485 lp=port1,port2,... specifies the parallel ports 3486 to associate lp devices with, starting with 3487 lp0. A port specification may be 'none' to skip 3488 that lp device, or a parport name such as 3489 'parport0'. Specifying 'lp=auto' instead of a 3490 port specification list means that device IDs 3491 from each port should be examined, to see if 3492 an IEEE 1284-compliant printer is attached; if 3493 so, the driver will manage that printer. 3494 See also header of drivers/char/lp.c. 3495 3496 lpj=n [KNL] 3497 Sets loops_per_jiffy to given constant, thus avoiding 3498 time-consuming boot-time autodetection (up to 250 ms per 3499 CPU). 0 enables autodetection (default). To determine 3500 the correct value for your kernel, boot with normal 3501 autodetection and see what value is printed. Note that 3502 on SMP systems the preset will be applied to all CPUs, 3503 which is likely to cause problems if your CPUs need 3504 significantly divergent settings. An incorrect value 3505 will cause delays in the kernel to be wrong, leading to 3506 unpredictable I/O errors and other breakage. Although 3507 unlikely, in the extreme case this might damage your 3508 hardware. 3509 3510 lsm.debug [SECURITY] Enable LSM initialization debugging output. 3511 3512 lsm=lsm1,...,lsmN 3513 [SECURITY] Choose order of LSM initialization. This 3514 overrides CONFIG_LSM, and the "security=" parameter. 3515 3516 machtype= [Loongson] Share the same kernel image file between 3517 different yeeloong laptops. 3518 Example: machtype=lemote-yeeloong-2f-7inch 3519 3520 maxcpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 3521 will bring up during bootup. maxcpus=n : n >= 0 limits 3522 the kernel to bring up 'n' processors. Surely after 3523 bootup you can bring up the other plugged cpu by executing 3524 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 3525 only takes effect during system bootup. 3526 While n=0 is a special case, it is equivalent to "nosmp", 3527 which also disables the IO APIC. 3528 3529 max_loop= [LOOP] The number of loop block devices that get 3530 (loop.max_loop) unconditionally pre-created at init time. The default 3531 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 3532 of statically allocating a predefined number, loop 3533 devices can be requested on-demand with the 3534 /dev/loop-control interface. 3535 3536 mce= [X86-{32,64}] 3537 3538 Please see Documentation/arch/x86/x86_64/machinecheck.rst for sysfs runtime tunables. 3539 3540 off 3541 disable machine check 3542 3543 no_cmci 3544 disable CMCI(Corrected Machine Check Interrupt) that 3545 Intel processor supports. Usually this disablement is 3546 not recommended, but it might be handy if your 3547 hardware is misbehaving. 3548 3549 Note that you'll get more problems without CMCI than 3550 with due to the shared banks, i.e. you might get 3551 duplicated error logs. 3552 3553 dont_log_ce 3554 don't make logs for corrected errors. All events 3555 reported as corrected are silently cleared by OS. This 3556 option will be useful if you have no interest in any 3557 of corrected errors. 3558 3559 ignore_ce 3560 disable features for corrected errors, e.g. 3561 polling timer and CMCI. All events reported as 3562 corrected are not cleared by OS and remained in its 3563 error banks. 3564 3565 Usually this disablement is not recommended, however 3566 if there is an agent checking/clearing corrected 3567 errors (e.g. BIOS or hardware monitoring 3568 applications), conflicting with OS's error handling, 3569 and you cannot deactivate the agent, then this option 3570 will be a help. 3571 3572 no_lmce 3573 do not opt-in to Local MCE delivery. Use legacy method 3574 to broadcast MCEs. 3575 3576 bootlog 3577 enable logging of machine checks left over from 3578 booting. Disabled by default on AMD Fam10h and older 3579 because some BIOS leave bogus ones. 3580 3581 If your BIOS doesn't do that it's a good idea to 3582 enable though to make sure you log even machine check 3583 events that result in a reboot. On Intel systems it is 3584 enabled by default. 3585 3586 nobootlog 3587 disable boot machine check logging. 3588 3589 monarchtimeout (number) 3590 sets the time in us to wait for other CPUs on machine 3591 checks. 0 to disable. 3592 3593 bios_cmci_threshold 3594 don't overwrite the bios-set CMCI threshold. This boot 3595 option prevents Linux from overwriting the CMCI 3596 threshold set by the bios. Without this option, Linux 3597 always sets the CMCI threshold to 1. Enabling this may 3598 make memory predictive failure analysis less effective 3599 if the bios sets thresholds for memory errors since we 3600 will not see details for all errors. 3601 3602 recovery 3603 force-enable recoverable machine check code paths 3604 3605 Everything else is in sysfs now. 3606 3607 3608 md= [HW] RAID subsystems devices and level 3609 See Documentation/admin-guide/md.rst. 3610 3611 mdacon= [MDA] 3612 Format: <first>,<last> 3613 Specifies range of consoles to be captured by the MDA. 3614 3615 mds= [X86,INTEL,EARLY] 3616 Control mitigation for the Micro-architectural Data 3617 Sampling (MDS) vulnerability. 3618 3619 Certain CPUs are vulnerable to an exploit against CPU 3620 internal buffers which can forward information to a 3621 disclosure gadget under certain conditions. 3622 3623 In vulnerable processors, the speculatively 3624 forwarded data can be used in a cache side channel 3625 attack, to access data to which the attacker does 3626 not have direct access. 3627 3628 This parameter controls the MDS mitigation. The 3629 options are: 3630 3631 full - Enable MDS mitigation on vulnerable CPUs 3632 full,nosmt - Enable MDS mitigation and disable 3633 SMT on vulnerable CPUs 3634 off - Unconditionally disable MDS mitigation 3635 3636 On TAA-affected machines, mds=off can be prevented by 3637 an active TAA mitigation as both vulnerabilities are 3638 mitigated with the same mechanism so in order to disable 3639 this mitigation, you need to specify tsx_async_abort=off 3640 too. 3641 3642 Not specifying this option is equivalent to 3643 mds=full. 3644 3645 For details see: Documentation/admin-guide/hw-vuln/mds.rst 3646 3647 mem=nn[KMG] [HEXAGON,EARLY] Set the memory size. 3648 Must be specified, otherwise memory size will be 0. 3649 3650 mem=nn[KMG] [KNL,BOOT,EARLY] Force usage of a specific amount 3651 of memory Amount of memory to be used in cases 3652 as follows: 3653 3654 1 for test; 3655 2 when the kernel is not able to see the whole system memory; 3656 3 memory that lies after 'mem=' boundary is excluded from 3657 the hypervisor, then assigned to KVM guests. 3658 4 to limit the memory available for kdump kernel. 3659 3660 [ARC,MICROBLAZE] - the limit applies only to low memory, 3661 high memory is not affected. 3662 3663 [ARM64] - only limits memory covered by the linear 3664 mapping. The NOMAP regions are not affected. 3665 3666 [X86] Work as limiting max address. Use together 3667 with memmap= to avoid physical address space collisions. 3668 Without memmap= PCI devices could be placed at addresses 3669 belonging to unused RAM. 3670 3671 Note that this only takes effects during boot time since 3672 in above case 3, memory may need be hot added after boot 3673 if system memory of hypervisor is not sufficient. 3674 3675 mem=nn[KMG]@ss[KMG] 3676 [ARM,MIPS,EARLY] - override the memory layout 3677 reported by firmware. 3678 Define a memory region of size nn[KMG] starting at 3679 ss[KMG]. 3680 Multiple different regions can be specified with 3681 multiple mem= parameters on the command line. 3682 3683 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 3684 memory. 3685 3686 memblock=debug [KNL,EARLY] Enable memblock debug messages. 3687 3688 memchunk=nn[KMG] 3689 [KNL,SH] Allow user to override the default size for 3690 per-device physically contiguous DMA buffers. 3691 3692 memhp_default_state=online/offline/online_kernel/online_movable 3693 [KNL] Set the initial state for the memory hotplug 3694 onlining policy. If not specified, the default value is 3695 set according to the 3696 CONFIG_MHP_DEFAULT_ONLINE_TYPE kernel config 3697 options. 3698 See Documentation/admin-guide/mm/memory-hotplug.rst. 3699 3700 memmap=exactmap [KNL,X86,EARLY] Enable setting of an exact 3701 E820 memory map, as specified by the user. 3702 Such memmap=exactmap lines can be constructed based on 3703 BIOS output or other requirements. See the memmap=nn@ss 3704 option description. 3705 3706 memmap=nn[KMG]@ss[KMG] 3707 [KNL, X86,MIPS,XTENSA,EARLY] Force usage of a specific region of memory. 3708 Region of memory to be used is from ss to ss+nn. 3709 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 3710 which limits max address to nn[KMG]. 3711 Multiple different regions can be specified, 3712 comma delimited. 3713 Example: 3714 memmap=100M@2G,100M#3G,1G!1024G 3715 3716 memmap=nn[KMG]#ss[KMG] 3717 [KNL,ACPI,EARLY] Mark specific memory as ACPI data. 3718 Region of memory to be marked is from ss to ss+nn. 3719 3720 memmap=nn[KMG]$ss[KMG] 3721 [KNL,ACPI,EARLY] Mark specific memory as reserved. 3722 Region of memory to be reserved is from ss to ss+nn. 3723 Example: Exclude memory from 0x18690000-0x1869ffff 3724 memmap=64K$0x18690000 3725 or 3726 memmap=0x10000$0x18690000 3727 Some bootloaders may need an escape character before '$', 3728 like Grub2, otherwise '$' and the following number 3729 will be eaten. 3730 3731 memmap=nn[KMG]!ss[KMG,EARLY] 3732 [KNL,X86] Mark specific memory as protected. 3733 Region of memory to be used, from ss to ss+nn. 3734 The memory region may be marked as e820 type 12 (0xc) 3735 and is NVDIMM or ADR memory. 3736 3737 memmap=<size>%<offset>-<oldtype>+<newtype> 3738 [KNL,ACPI,EARLY] Convert memory within the specified region 3739 from <oldtype> to <newtype>. If "-<oldtype>" is left 3740 out, the whole region will be marked as <newtype>, 3741 even if previously unavailable. If "+<newtype>" is left 3742 out, matching memory will be removed. Types are 3743 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 3744 3 = ACPI, 12 = PRAM. 3745 3746 memory_corruption_check=0/1 [X86,EARLY] 3747 Some BIOSes seem to corrupt the first 64k of 3748 memory when doing things like suspend/resume. 3749 Setting this option will scan the memory 3750 looking for corruption. Enabling this will 3751 both detect corruption and prevent the kernel 3752 from using the memory being corrupted. 3753 However, it's intended as a diagnostic tool; if 3754 repeatable BIOS-originated corruption always 3755 affects the same memory, you can use memmap= 3756 to prevent the kernel from using that memory. 3757 3758 memory_corruption_check_size=size [X86,EARLY] 3759 By default it checks for corruption in the low 3760 64k, making this memory unavailable for normal 3761 use. Use this parameter to scan for 3762 corruption in more or less memory. 3763 3764 memory_corruption_check_period=seconds [X86,EARLY] 3765 By default it checks for corruption every 60 3766 seconds. Use this parameter to check at some 3767 other rate. 0 disables periodic checking. 3768 3769 memory_hotplug.memmap_on_memory 3770 [KNL,X86,ARM] Boolean flag to enable this feature. 3771 Format: {on | off (default)} 3772 When enabled, runtime hotplugged memory will 3773 allocate its internal metadata (struct pages, 3774 those vmemmap pages cannot be optimized even 3775 if hugetlb_free_vmemmap is enabled) from the 3776 hotadded memory which will allow to hotadd a 3777 lot of memory without requiring additional 3778 memory to do so. 3779 This feature is disabled by default because it 3780 has some implication on large (e.g. GB) 3781 allocations in some configurations (e.g. small 3782 memory blocks). 3783 The state of the flag can be read in 3784 /sys/module/memory_hotplug/parameters/memmap_on_memory. 3785 Note that even when enabled, there are a few cases where 3786 the feature is not effective. 3787 3788 memtest= [KNL,X86,ARM,M68K,PPC,RISCV,EARLY] Enable memtest 3789 Format: <integer> 3790 default : 0 <disable> 3791 Specifies the number of memtest passes to be 3792 performed. Each pass selects another test 3793 pattern from a given set of patterns. Memtest 3794 fills the memory with this pattern, validates 3795 memory contents and reserves bad memory 3796 regions that are detected. 3797 3798 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 3799 Valid arguments: on, off 3800 Default: off 3801 mem_encrypt=on: Activate SME 3802 mem_encrypt=off: Do not activate SME 3803 3804 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst 3805 for details on when memory encryption can be activated. 3806 3807 mem_sleep_default= [SUSPEND] Default system suspend mode: 3808 s2idle - Suspend-To-Idle 3809 shallow - Power-On Suspend or equivalent (if supported) 3810 deep - Suspend-To-RAM or equivalent (if supported) 3811 See Documentation/admin-guide/pm/sleep-states.rst. 3812 3813 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 3814 the BIOS has incorrectly applied a workaround. TinyBIOS 3815 version 0.98 is known to be affected, 0.99 fixes the 3816 problem by letting the user disable the workaround. 3817 3818 mga= [HW,DRM] 3819 3820 microcode= [X86] Control the behavior of the microcode loader. 3821 Available options, comma separated: 3822 3823 base_rev=X - with <X> with format: <u32> 3824 Set the base microcode revision of each thread when in 3825 debug mode. 3826 3827 dis_ucode_ldr: disable the microcode loader 3828 3829 force_minrev: 3830 Enable or disable the microcode minimal revision 3831 enforcement for the runtime microcode loader. 3832 3833 mini2440= [ARM,HW,KNL] 3834 Format:[0..2][b][c][t] 3835 Default: "0tb" 3836 MINI2440 configuration specification: 3837 0 - The attached screen is the 3.5" TFT 3838 1 - The attached screen is the 7" TFT 3839 2 - The VGA Shield is attached (1024x768) 3840 Leaving out the screen size parameter will not load 3841 the TFT driver, and the framebuffer will be left 3842 unconfigured. 3843 b - Enable backlight. The TFT backlight pin will be 3844 linked to the kernel VESA blanking code and a GPIO 3845 LED. This parameter is not necessary when using the 3846 VGA shield. 3847 c - Enable the s3c camera interface. 3848 t - Reserved for enabling touchscreen support. The 3849 touchscreen support is not enabled in the mainstream 3850 kernel as of 2.6.30, a preliminary port can be found 3851 in the "bleeding edge" mini2440 support kernel at 3852 https://repo.or.cz/w/linux-2.6/mini2440.git 3853 3854 mitigations= 3855 [X86,PPC,S390,ARM64,EARLY] Control optional mitigations for 3856 CPU vulnerabilities. This is a set of curated, 3857 arch-independent options, each of which is an 3858 aggregation of existing arch-specific options. 3859 3860 Note, "mitigations" is supported if and only if the 3861 kernel was built with CPU_MITIGATIONS=y. 3862 3863 off 3864 Disable all optional CPU mitigations. This 3865 improves system performance, but it may also 3866 expose users to several CPU vulnerabilities. 3867 Equivalent to: if nokaslr then kpti=0 [ARM64] 3868 gather_data_sampling=off [X86] 3869 indirect_target_selection=off [X86] 3870 kvm.nx_huge_pages=off [X86] 3871 l1tf=off [X86] 3872 mds=off [X86] 3873 mmio_stale_data=off [X86] 3874 no_entry_flush [PPC] 3875 no_uaccess_flush [PPC] 3876 nobp=0 [S390] 3877 nopti [X86,PPC] 3878 nospectre_bhb [ARM64] 3879 nospectre_v1 [X86,PPC] 3880 nospectre_v2 [X86,PPC,S390,ARM64] 3881 reg_file_data_sampling=off [X86] 3882 retbleed=off [X86] 3883 spec_rstack_overflow=off [X86] 3884 spec_store_bypass_disable=off [X86,PPC] 3885 spectre_bhi=off [X86] 3886 spectre_v2_user=off [X86] 3887 srbds=off [X86,INTEL] 3888 ssbd=force-off [ARM64] 3889 tsx_async_abort=off [X86] 3890 vmscape=off [X86] 3891 3892 Exceptions: 3893 This does not have any effect on 3894 kvm.nx_huge_pages when 3895 kvm.nx_huge_pages=force. 3896 3897 auto (default) 3898 Mitigate all CPU vulnerabilities, but leave SMT 3899 enabled, even if it's vulnerable. This is for 3900 users who don't want to be surprised by SMT 3901 getting disabled across kernel upgrades, or who 3902 have other ways of avoiding SMT-based attacks. 3903 Equivalent to: (default behavior) 3904 3905 auto,nosmt 3906 Mitigate all CPU vulnerabilities, disabling SMT 3907 if needed. This is for users who always want to 3908 be fully mitigated, even if it means losing SMT. 3909 Equivalent to: l1tf=flush,nosmt [X86] 3910 mds=full,nosmt [X86] 3911 tsx_async_abort=full,nosmt [X86] 3912 mmio_stale_data=full,nosmt [X86] 3913 retbleed=auto,nosmt [X86] 3914 3915 [X86] After one of the above options, additionally 3916 supports attack-vector based controls as documented in 3917 Documentation/admin-guide/hw-vuln/attack_vector_controls.rst 3918 3919 mminit_loglevel= 3920 [KNL,EARLY] When CONFIG_DEBUG_MEMORY_INIT is set, this 3921 parameter allows control of the logging verbosity for 3922 the additional memory initialisation checks. A value 3923 of 0 disables mminit logging and a level of 4 will 3924 log everything. Information is printed at KERN_DEBUG 3925 so loglevel=8 may also need to be specified. 3926 3927 mmio_stale_data= 3928 [X86,INTEL,EARLY] Control mitigation for the Processor 3929 MMIO Stale Data vulnerabilities. 3930 3931 Processor MMIO Stale Data is a class of 3932 vulnerabilities that may expose data after an MMIO 3933 operation. Exposed data could originate or end in 3934 the same CPU buffers as affected by MDS and TAA. 3935 Therefore, similar to MDS and TAA, the mitigation 3936 is to clear the affected CPU buffers. 3937 3938 This parameter controls the mitigation. The 3939 options are: 3940 3941 full - Enable mitigation on vulnerable CPUs 3942 3943 full,nosmt - Enable mitigation and disable SMT on 3944 vulnerable CPUs. 3945 3946 off - Unconditionally disable mitigation 3947 3948 On MDS or TAA affected machines, 3949 mmio_stale_data=off can be prevented by an active 3950 MDS or TAA mitigation as these vulnerabilities are 3951 mitigated with the same mechanism so in order to 3952 disable this mitigation, you need to specify 3953 mds=off and tsx_async_abort=off too. 3954 3955 Not specifying this option is equivalent to 3956 mmio_stale_data=full. 3957 3958 For details see: 3959 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst 3960 3961 <module>.async_probe[=<bool>] [KNL] 3962 If no <bool> value is specified or if the value 3963 specified is not a valid <bool>, enable asynchronous 3964 probe on this module. Otherwise, enable/disable 3965 asynchronous probe on this module as indicated by the 3966 <bool> value. See also: module.async_probe 3967 3968 module.async_probe=<bool> 3969 [KNL] When set to true, modules will use async probing 3970 by default. To enable/disable async probing for a 3971 specific module, use the module specific control that 3972 is documented under <module>.async_probe. When both 3973 module.async_probe and <module>.async_probe are 3974 specified, <module>.async_probe takes precedence for 3975 the specific module. 3976 3977 module.enable_dups_trace 3978 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set, 3979 this means that duplicate request_module() calls will 3980 trigger a WARN_ON() instead of a pr_warn(). Note that 3981 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s 3982 will always be issued and this option does nothing. 3983 module.sig_enforce 3984 [KNL] When CONFIG_MODULE_SIG is set, this means that 3985 modules without (valid) signatures will fail to load. 3986 Note that if CONFIG_MODULE_SIG_FORCE is set, that 3987 is always true, so this option does nothing. 3988 3989 module_blacklist= [KNL] Do not load a comma-separated list of 3990 modules. Useful for debugging problem modules. 3991 3992 mousedev.tap_time= 3993 [MOUSE] Maximum time between finger touching and 3994 leaving touchpad surface for touch to be considered 3995 a tap and be reported as a left button click (for 3996 touchpads working in absolute mode only). 3997 Format: <msecs> 3998 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 3999 reporting absolute coordinates, such as tablets 4000 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 4001 reporting absolute coordinates, such as tablets 4002 4003 movablecore= [KNL,X86,PPC,EARLY] 4004 Format: nn[KMGTPE] | nn% 4005 This parameter is the complement to kernelcore=, it 4006 specifies the amount of memory used for migratable 4007 allocations. If both kernelcore and movablecore is 4008 specified, then kernelcore will be at *least* the 4009 specified value but may be more. If movablecore on its 4010 own is specified, the administrator must be careful 4011 that the amount of memory usable for all allocations 4012 is not too small. 4013 4014 movable_node [KNL,EARLY] Boot-time switch to make hotplugable memory 4015 NUMA nodes to be movable. This means that the memory 4016 of such nodes will be usable only for movable 4017 allocations which rules out almost all kernel 4018 allocations. Use with caution! 4019 4020 MTD_Partition= [MTD] 4021 Format: <name>,<region-number>,<size>,<offset> 4022 4023 MTD_Region= [MTD] Format: 4024 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 4025 4026 mtdparts= [MTD] 4027 See drivers/mtd/parsers/cmdlinepart.c 4028 4029 mtouchusb.raw_coordinates= 4030 [HW] Make the MicroTouch USB driver use raw coordinates 4031 ('y', default) or cooked coordinates ('n') 4032 4033 mtrr=debug [X86,EARLY] 4034 Enable printing debug information related to MTRR 4035 registers at boot time. 4036 4037 mtrr_chunk_size=nn[KMG,X86,EARLY] 4038 used for mtrr cleanup. It is largest continuous chunk 4039 that could hold holes aka. UC entries. 4040 4041 mtrr_gran_size=nn[KMG,X86,EARLY] 4042 Used for mtrr cleanup. It is granularity of mtrr block. 4043 Default is 1. 4044 Large value could prevent small alignment from 4045 using up MTRRs. 4046 4047 mtrr_spare_reg_nr=n [X86,EARLY] 4048 Format: <integer> 4049 Range: 0,7 : spare reg number 4050 Default : 1 4051 Used for mtrr cleanup. It is spare mtrr entries number. 4052 Set to 2 or more if your graphical card needs more. 4053 4054 multitce=off [PPC] This parameter disables the use of the pSeries 4055 firmware feature for updating multiple TCE entries 4056 at a time. 4057 4058 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 4059 4060 netdev= [NET] Network devices parameters 4061 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 4062 Note that mem_start is often overloaded to mean 4063 something different and driver-specific. 4064 This usage is only documented in each driver source 4065 file if at all. 4066 4067 netpoll.carrier_timeout= 4068 [NET] Specifies amount of time (in seconds) that 4069 netpoll should wait for a carrier. By default netpoll 4070 waits 4 seconds. 4071 4072 nf_conntrack.acct= 4073 [NETFILTER] Enable connection tracking flow accounting 4074 0 to disable accounting 4075 1 to enable accounting 4076 Default value is 0. 4077 4078 nfs.cache_getent= 4079 [NFS] sets the pathname to the program which is used 4080 to update the NFS client cache entries. 4081 4082 nfs.cache_getent_timeout= 4083 [NFS] sets the timeout after which an attempt to 4084 update a cache entry is deemed to have failed. 4085 4086 nfs.callback_nr_threads= 4087 [NFSv4] set the total number of threads that the 4088 NFS client will assign to service NFSv4 callback 4089 requests. 4090 4091 nfs.callback_tcpport= 4092 [NFS] set the TCP port on which the NFSv4 callback 4093 channel should listen. 4094 4095 nfs.delay_retrans= 4096 [NFS] specifies the number of times the NFSv4 client 4097 retries the request before returning an EAGAIN error, 4098 after a reply of NFS4ERR_DELAY from the server. 4099 Only applies if the softerr mount option is enabled, 4100 and the specified value is >= 0. 4101 4102 nfs.enable_ino64= 4103 [NFS] enable 64-bit inode numbers. 4104 If zero, the NFS client will fake up a 32-bit inode 4105 number for the readdir() and stat() syscalls instead 4106 of returning the full 64-bit number. 4107 The default is to return 64-bit inode numbers. 4108 4109 nfs.idmap_cache_timeout= 4110 [NFS] set the maximum lifetime for idmapper cache 4111 entries. 4112 4113 nfs.max_session_cb_slots= 4114 [NFSv4.1] Sets the maximum number of session 4115 slots the client will assign to the callback 4116 channel. This determines the maximum number of 4117 callbacks the client will process in parallel for 4118 a particular server. 4119 4120 nfs.max_session_slots= 4121 [NFSv4.1] Sets the maximum number of session slots 4122 the client will attempt to negotiate with the server. 4123 This limits the number of simultaneous RPC requests 4124 that the client can send to the NFSv4.1 server. 4125 Note that there is little point in setting this 4126 value higher than the max_tcp_slot_table_limit. 4127 4128 nfs.nfs4_disable_idmapping= 4129 [NFSv4] When set to the default of '1', this option 4130 ensures that both the RPC level authentication 4131 scheme and the NFS level operations agree to use 4132 numeric uids/gids if the mount is using the 4133 'sec=sys' security flavour. In effect it is 4134 disabling idmapping, which can make migration from 4135 legacy NFSv2/v3 systems to NFSv4 easier. 4136 Servers that do not support this mode of operation 4137 will be autodetected by the client, and it will fall 4138 back to using the idmapper. 4139 To turn off this behaviour, set the value to '0'. 4140 4141 nfs.nfs4_unique_id= 4142 [NFS4] Specify an additional fixed unique ident- 4143 ification string that NFSv4 clients can insert into 4144 their nfs_client_id4 string. This is typically a 4145 UUID that is generated at system install time. 4146 4147 nfs.recover_lost_locks= 4148 [NFSv4] Attempt to recover locks that were lost due 4149 to a lease timeout on the server. Please note that 4150 doing this risks data corruption, since there are 4151 no guarantees that the file will remain unchanged 4152 after the locks are lost. 4153 If you want to enable the kernel legacy behaviour of 4154 attempting to recover these locks, then set this 4155 parameter to '1'. 4156 The default parameter value of '0' causes the kernel 4157 not to attempt recovery of lost locks. 4158 4159 nfs.send_implementation_id= 4160 [NFSv4.1] Send client implementation identification 4161 information in exchange_id requests. 4162 If zero, no implementation identification information 4163 will be sent. 4164 The default is to send the implementation identification 4165 information. 4166 4167 nfs4.layoutstats_timer= 4168 [NFSv4.2] Change the rate at which the kernel sends 4169 layoutstats to the pNFS metadata server. 4170 4171 Setting this to value to 0 causes the kernel to use 4172 whatever value is the default set by the layout 4173 driver. A non-zero value sets the minimum interval 4174 in seconds between layoutstats transmissions. 4175 4176 nfsd.inter_copy_offload_enable= 4177 [NFSv4.2] When set to 1, the server will support 4178 server-to-server copies for which this server is 4179 the destination of the copy. 4180 4181 nfsd.nfs4_disable_idmapping= 4182 [NFSv4] When set to the default of '1', the NFSv4 4183 server will return only numeric uids and gids to 4184 clients using auth_sys, and will accept numeric uids 4185 and gids from such clients. This is intended to ease 4186 migration from NFSv2/v3. 4187 4188 nfsd.nfsd4_ssc_umount_timeout= 4189 [NFSv4.2] When used as the destination of a 4190 server-to-server copy, knfsd temporarily mounts 4191 the source server. It caches the mount in case 4192 it will be needed again, and discards it if not 4193 used for the number of milliseconds specified by 4194 this parameter. 4195 4196 nfsaddrs= [NFS] Deprecated. Use ip= instead. 4197 See Documentation/admin-guide/nfs/nfsroot.rst. 4198 4199 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 4200 See Documentation/admin-guide/nfs/nfsroot.rst. 4201 4202 nfsrootdebug [NFS] enable nfsroot debugging messages. 4203 See Documentation/admin-guide/nfs/nfsroot.rst. 4204 4205 nmi_backtrace.backtrace_idle [KNL] 4206 Dump stacks even of idle CPUs in response to an 4207 NMI stack-backtrace request. 4208 4209 nmi_debug= [KNL,SH] Specify one or more actions to take 4210 when a NMI is triggered. 4211 Format: [state][,regs][,debounce][,die] 4212 4213 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 4214 Format: [panic,][nopanic,][rNNN,][num] 4215 Valid num: 0 or 1 4216 0 - turn hardlockup detector in nmi_watchdog off 4217 1 - turn hardlockup detector in nmi_watchdog on 4218 rNNN - configure the watchdog with raw perf event 0xNNN 4219 4220 When panic is specified, panic when an NMI watchdog 4221 timeout occurs (or 'nopanic' to not panic on an NMI 4222 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 4223 To disable both hard and soft lockup detectors, 4224 please see 'nowatchdog'. 4225 This is useful when you use a panic=... timeout and 4226 need the box quickly up again. 4227 4228 These settings can be accessed at runtime via 4229 the nmi_watchdog and hardlockup_panic sysctls. 4230 4231 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 4232 emulation library even if a 387 maths coprocessor 4233 is present. 4234 4235 no4lvl [RISCV,EARLY] Disable 4-level and 5-level paging modes. 4236 Forces kernel to use 3-level paging instead. 4237 4238 no5lvl [X86-64,RISCV,EARLY] Disable 5-level paging mode. Forces 4239 kernel to use 4-level paging instead. 4240 4241 noalign [KNL,ARM] 4242 4243 noapic [SMP,APIC,EARLY] Tells the kernel to not make use of any 4244 IOAPICs that may be present in the system. 4245 4246 noapictimer [APIC,X86] Don't set up the APIC timer 4247 4248 noautogroup Disable scheduler automatic task group creation. 4249 4250 nocache [ARM,EARLY] 4251 4252 no_console_suspend 4253 [HW] Never suspend the console 4254 Disable suspending of consoles during suspend and 4255 hibernate operations. Once disabled, debugging 4256 messages can reach various consoles while the rest 4257 of the system is being put to sleep (ie, while 4258 debugging driver suspend/resume hooks). This may 4259 not work reliably with all consoles, but is known 4260 to work with serial and VGA consoles. 4261 To facilitate more flexible debugging, we also add 4262 console_suspend, a printk module parameter to control 4263 it. Users could use console_suspend (usually 4264 /sys/module/printk/parameters/console_suspend) to 4265 turn on/off it dynamically. 4266 4267 no_debug_objects 4268 [KNL,EARLY] Disable object debugging 4269 4270 nodsp [SH] Disable hardware DSP at boot time. 4271 4272 noefi [EFI,EARLY] Disable EFI runtime services support. 4273 4274 no_entry_flush [PPC,EARLY] Don't flush the L1-D cache when entering the kernel. 4275 4276 noexec32 [X86-64] 4277 This affects only 32-bit executables. 4278 noexec32=on: enable non-executable mappings (default) 4279 read doesn't imply executable mappings 4280 noexec32=off: disable non-executable mappings 4281 read implies executable mappings 4282 4283 no_file_caps Tells the kernel not to honor file capabilities. The 4284 only way then for a file to be executed with privilege 4285 is to be setuid root or executed by root. 4286 4287 nofpu [MIPS,SH] Disable hardware FPU at boot time. 4288 4289 nofsgsbase [X86] Disables FSGSBASE instructions. 4290 4291 nofxsr [BUGS=X86-32] Disables x86 floating point extended 4292 register save and restore. The kernel will only save 4293 legacy floating-point registers on task switch. 4294 4295 nogbpages [X86] Do not use GB pages for kernel direct mappings. 4296 4297 no_hash_pointers 4298 [KNL,EARLY] 4299 Alias for "hash_pointers=never". 4300 4301 nohibernate [HIBERNATION] Disable hibernation and resume. 4302 4303 nohlt [ARM,ARM64,MICROBLAZE,MIPS,PPC,RISCV,SH] Forces the kernel to 4304 busy wait in do_idle() and not use the arch_cpu_idle() 4305 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP 4306 to be effective. This is useful on platforms where the 4307 sleep(SH) or wfi(ARM,ARM64) instructions do not work 4308 correctly or when doing power measurements to evaluate 4309 the impact of the sleep instructions. This is also 4310 useful when using JTAG debugger. 4311 4312 nohpet [X86] Don't use the HPET timer. 4313 4314 nohugeiomap [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge I/O mappings. 4315 4316 nohugevmalloc [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge vmalloc mappings. 4317 4318 nohz= [KNL] Boottime enable/disable dynamic ticks 4319 Valid arguments: on, off 4320 Default: on 4321 4322 nohz_full= [KNL,BOOT,SMP,ISOL] 4323 The argument is a cpu list, as described above. 4324 In kernels built with CONFIG_NO_HZ_FULL=y, set 4325 the specified list of CPUs whose tick will be stopped 4326 whenever possible. The boot CPU will be forced outside 4327 the range to maintain the timekeeping. Any CPUs 4328 in this list will have their RCU callbacks offloaded, 4329 just as if they had also been called out in the 4330 rcu_nocbs= boot parameter. 4331 4332 Note that this argument takes precedence over 4333 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 4334 4335 noinitrd [RAM] Tells the kernel not to load any configured 4336 initial RAM disk. 4337 4338 nointremap [X86-64,Intel-IOMMU,EARLY] Do not enable interrupt 4339 remapping. 4340 [Deprecated - use intremap=off] 4341 4342 noinvpcid [X86,EARLY] Disable the INVPCID cpu feature. 4343 4344 noiotrap [SH] Disables trapped I/O port accesses. 4345 4346 noirqdebug [X86-32] Disables the code which attempts to detect and 4347 disable unhandled interrupt sources. 4348 4349 noisapnp [ISAPNP] Disables ISA PnP code. 4350 4351 nokaslr [KNL,EARLY] 4352 When CONFIG_RANDOMIZE_BASE is set, this disables 4353 kernel and module base offset ASLR (Address Space 4354 Layout Randomization). 4355 4356 no-kvmapf [X86,KVM,EARLY] Disable paravirtualized asynchronous page 4357 fault handling. 4358 4359 no-kvmclock [X86,KVM,EARLY] Disable paravirtualized KVM clock driver 4360 4361 nolapic [X86-32,APIC,EARLY] Do not enable or use the local APIC. 4362 4363 nolapic_timer [X86-32,APIC,EARLY] Do not use the local APIC timer. 4364 4365 nomce [X86-32] Disable Machine Check Exception 4366 4367 nomfgpt [X86-32] Disable Multi-Function General Purpose 4368 Timer usage (for AMD Geode machines). 4369 4370 nomodeset Disable kernel modesetting. Most systems' firmware 4371 sets up a display mode and provides framebuffer memory 4372 for output. With nomodeset, DRM and fbdev drivers will 4373 not load if they could possibly displace the pre- 4374 initialized output. Only the system framebuffer will 4375 be available for use. The respective drivers will not 4376 perform display-mode changes or accelerated rendering. 4377 4378 Useful as error fallback, or for testing and debugging. 4379 4380 nomodule Disable module load 4381 4382 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 4383 shutdown the other cpus. Instead use the REBOOT_VECTOR 4384 irq. 4385 4386 nopat [X86,EARLY] Disable PAT (page attribute table extension of 4387 pagetables) support. 4388 4389 nopcid [X86-64,EARLY] Disable the PCID cpu feature. 4390 4391 nopku [X86] Disable Memory Protection Keys CPU feature found 4392 in some Intel CPUs. 4393 4394 nopti [X86-64,EARLY] 4395 Equivalent to pti=off 4396 4397 nopv= [X86,XEN,KVM,HYPER_V,VMWARE,EARLY] 4398 Disables the PV optimizations forcing the guest to run 4399 as generic guest with no PV drivers. Currently support 4400 XEN HVM, KVM, HYPER_V and VMWARE guest. 4401 4402 nopvspin [X86,XEN,KVM,EARLY] 4403 Disables the qspinlock slow path using PV optimizations 4404 which allow the hypervisor to 'idle' the guest on lock 4405 contention. 4406 4407 norandmaps Don't use address space randomization. Equivalent to 4408 echo 0 > /proc/sys/kernel/randomize_va_space 4409 4410 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 4411 with UP alternatives 4412 4413 noresume [SWSUSP] Disables resume and restores original swap 4414 space. 4415 4416 no-scroll [VGA] Disables scrollback. 4417 This is required for the Braillex ib80-piezo Braille 4418 reader made by F.H. Papenmeier (Germany). 4419 4420 nosgx [X86-64,SGX,EARLY] Disables Intel SGX kernel support. 4421 4422 nosmap [PPC,EARLY] 4423 Disable SMAP (Supervisor Mode Access Prevention) 4424 even if it is supported by processor. 4425 4426 nosmep [PPC64s,EARLY] 4427 Disable SMEP (Supervisor Mode Execution Prevention) 4428 even if it is supported by processor. 4429 4430 nosmp [SMP,EARLY] Tells an SMP kernel to act as a UP kernel, 4431 and disable the IO APIC. legacy for "maxcpus=0". 4432 4433 nosmt [KNL,MIPS,PPC,EARLY] Disable symmetric multithreading (SMT). 4434 Equivalent to smt=1. 4435 4436 [KNL,X86,PPC,S390] Disable symmetric multithreading (SMT). 4437 nosmt=force: Force disable SMT, cannot be undone 4438 via the sysfs control file. 4439 4440 nosoftlockup [KNL] Disable the soft-lockup detector. 4441 4442 nospec_store_bypass_disable 4443 [HW,EARLY] Disable all mitigations for the Speculative 4444 Store Bypass vulnerability 4445 4446 nospectre_bhb [ARM64,EARLY] Disable all mitigations for Spectre-BHB (branch 4447 history injection) vulnerability. System may allow data leaks 4448 with this option. 4449 4450 nospectre_v1 [X86,PPC,EARLY] Disable mitigations for Spectre Variant 1 4451 (bounds check bypass). With this option data leaks are 4452 possible in the system. 4453 4454 nospectre_v2 [X86,PPC_E500,ARM64,EARLY] Disable all mitigations 4455 for the Spectre variant 2 (indirect branch 4456 prediction) vulnerability. System may allow data 4457 leaks with this option. 4458 4459 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES,RISCV,LOONGARCH,EARLY] 4460 Disable paravirtualized steal time accounting. steal time 4461 is computed, but won't influence scheduler behaviour 4462 4463 nosync [HW,M68K] Disables sync negotiation for all devices. 4464 4465 no_timer_check [X86,APIC] Disables the code which tests for broken 4466 timer IRQ sources, i.e., the IO-APIC timer. This can 4467 work around problems with incorrect timer 4468 initialization on some boards. 4469 4470 no_uaccess_flush 4471 [PPC,EARLY] Don't flush the L1-D cache after accessing user data. 4472 4473 novmcoredd [KNL,KDUMP] 4474 Disable device dump. Device dump allows drivers to 4475 append dump data to vmcore so you can collect driver 4476 specified debug info. Drivers can append the data 4477 without any limit and this data is stored in memory, 4478 so this may cause significant memory stress. Disabling 4479 device dump can help save memory but the driver debug 4480 data will be no longer available. This parameter 4481 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 4482 is set. 4483 4484 no-vmw-sched-clock 4485 [X86,PV_OPS,EARLY] Disable paravirtualized VMware 4486 scheduler clock and use the default one. 4487 4488 nowatchdog [KNL] Disable both lockup detectors, i.e. 4489 soft-lockup and NMI watchdog (hard-lockup). 4490 4491 nowb [ARM,EARLY] 4492 4493 nox2apic [X86-64,APIC,EARLY] Do not enable x2APIC mode. 4494 4495 NOTE: this parameter will be ignored on systems with the 4496 LEGACY_XAPIC_DISABLED bit set in the 4497 IA32_XAPIC_DISABLE_STATUS MSR. 4498 4499 noxsave [BUGS=X86] Disables x86 extended register state save 4500 and restore using xsave. The kernel will fallback to 4501 enabling legacy floating-point and sse state. 4502 4503 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 4504 register states. The kernel will fall back to use 4505 xsave to save the states. By using this parameter, 4506 performance of saving the states is degraded because 4507 xsave doesn't support modified optimization while 4508 xsaveopt supports it on xsaveopt enabled systems. 4509 4510 noxsaves [X86] Disables xsaves and xrstors used in saving and 4511 restoring x86 extended register state in compacted 4512 form of xsave area. The kernel will fall back to use 4513 xsaveopt and xrstor to save and restore the states 4514 in standard form of xsave area. By using this 4515 parameter, xsave area per process might occupy more 4516 memory on xsaves enabled systems. 4517 4518 nr_cpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 4519 could support. nr_cpus=n : n >= 1 limits the kernel to 4520 support 'n' processors. It could be larger than the 4521 number of already plugged CPU during bootup, later in 4522 runtime you can physically add extra cpu until it reaches 4523 n. So during boot up some boot time memory for per-cpu 4524 variables need be pre-allocated for later physical cpu 4525 hot plugging. 4526 4527 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 4528 4529 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86, EARLY] 4530 Disable NUMA, Only set up a single NUMA node 4531 spanning all memory. 4532 4533 numa=fake=<size>[MG] 4534 [KNL, ARM64, RISCV, X86, EARLY] 4535 If given as a memory unit, fills all system RAM with 4536 nodes of size interleaved over physical nodes. 4537 4538 numa=fake=<N> 4539 [KNL, ARM64, RISCV, X86, EARLY] 4540 If given as an integer, fills all system RAM with N 4541 fake nodes interleaved over physical nodes. 4542 4543 numa=fake=<N>U 4544 [KNL, ARM64, RISCV, X86, EARLY] 4545 If given as an integer followed by 'U', it will 4546 divide each physical node into N emulated nodes. 4547 4548 numa=noacpi [X86] Don't parse the SRAT table for NUMA setup 4549 4550 numa=nohmat [X86] Don't parse the HMAT table for NUMA setup, or 4551 soft-reserved memory partitioning. 4552 4553 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic 4554 NUMA balancing. 4555 Allowed values are enable and disable 4556 4557 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 4558 'node', 'default' can be specified 4559 This can be set from sysctl after boot. 4560 See Documentation/admin-guide/sysctl/vm.rst for details. 4561 4562 ohci1394_dma=early [HW,EARLY] enable debugging via the ohci1394 driver. 4563 See Documentation/core-api/debugging-via-ohci1394.rst for more 4564 info. 4565 4566 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 4567 Rather than timing out after 20 ms if an EC 4568 command is not properly ACKed, override the length 4569 of the timeout. We have interrupts disabled while 4570 waiting for the ACK, so if this is set too high 4571 interrupts *may* be lost! 4572 4573 omap_mux= [OMAP] Override bootloader pin multiplexing. 4574 Format: <mux_mode0.mode_name=value>... 4575 For example, to override I2C bus2: 4576 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 4577 4578 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 4579 4580 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 4581 4582 boundary - index of last SLC block on Flex-OneNAND. 4583 The remaining blocks are configured as MLC blocks. 4584 lock - Configure if Flex-OneNAND boundary should be locked. 4585 Once locked, the boundary cannot be changed. 4586 1 indicates lock status, 0 indicates unlock status. 4587 4588 oops=panic [KNL,EARLY] 4589 Always panic on oopses. Default is to just kill the 4590 process, but there is a small probability of 4591 deadlocking the machine. 4592 This will also cause panics on machine check exceptions. 4593 Useful together with panic=30 to trigger a reboot. 4594 4595 page_alloc.shuffle= 4596 [KNL] Boolean flag to control whether the page allocator 4597 should randomize its free lists. This parameter can be 4598 used to enable/disable page randomization. The state of 4599 the flag can be read from sysfs at: 4600 /sys/module/page_alloc/parameters/shuffle. 4601 This parameter is only available if CONFIG_SHUFFLE_PAGE_ALLOCATOR=y. 4602 4603 page_owner= [KNL,EARLY] Boot-time page_owner enabling option. 4604 Storage of the information about who allocated 4605 each page is disabled in default. With this switch, 4606 we can turn it on. 4607 on: enable the feature 4608 4609 page_poison= [KNL,EARLY] Boot-time parameter changing the state of 4610 poisoning on the buddy allocator, available with 4611 CONFIG_PAGE_POISONING=y. 4612 off: turn off poisoning (default) 4613 on: turn on poisoning 4614 4615 page_reporting.page_reporting_order= 4616 [KNL] Minimal page reporting order 4617 Format: <integer> 4618 Adjust the minimal page reporting order. The page 4619 reporting is disabled when it exceeds MAX_PAGE_ORDER. 4620 4621 panic= [KNL] Kernel behaviour on panic: delay <timeout> 4622 timeout > 0: seconds before rebooting 4623 timeout = 0: wait forever 4624 timeout < 0: reboot immediately 4625 Format: <timeout> 4626 4627 panic_on_taint= [KNL,EARLY] 4628 Bitmask for conditionally calling panic() in add_taint() 4629 Format: <hex>[,nousertaint] 4630 Hexadecimal bitmask representing the set of TAINT flags 4631 that will cause the kernel to panic when add_taint() is 4632 called with any of the flags in this set. 4633 The optional switch "nousertaint" can be utilized to 4634 prevent userspace forced crashes by writing to sysctl 4635 /proc/sys/kernel/tainted any flagset matching with the 4636 bitmask set on panic_on_taint. 4637 See Documentation/admin-guide/tainted-kernels.rst for 4638 extra details on the taint flags that users can pick 4639 to compose the bitmask to assign to panic_on_taint. 4640 4641 panic_on_warn=1 panic() instead of WARN(). Useful to cause kdump 4642 on a WARN(). 4643 4644 panic_print= Bitmask for printing system info when panic happens. 4645 User can chose combination of the following bits: 4646 bit 0: print all tasks info 4647 bit 1: print system memory info 4648 bit 2: print timer info 4649 bit 3: print locks info if CONFIG_LOCKDEP is on 4650 bit 4: print ftrace buffer 4651 bit 5: replay all kernel messages on consoles at the end of panic 4652 bit 6: print all CPUs backtrace (if available in the arch) 4653 bit 7: print only tasks in uninterruptible (blocked) state 4654 *Be aware* that this option may print a _lot_ of lines, 4655 so there are risks of losing older messages in the log. 4656 Use this option carefully, maybe worth to setup a 4657 bigger log buffer with "log_buf_len" along with this. 4658 4659 panic_sys_info= A comma separated list of extra information to be dumped 4660 on panic. 4661 Format: val[,val...] 4662 Where @val can be any of the following: 4663 4664 tasks: print all tasks info 4665 mem: print system memory info 4666 timers: print timers info 4667 locks: print locks info if CONFIG_LOCKDEP is on 4668 ftrace: print ftrace buffer 4669 all_bt: print all CPUs backtrace (if available in the arch) 4670 blocked_tasks: print only tasks in uninterruptible (blocked) state 4671 4672 This is a human readable alternative to the 'panic_print' option. 4673 4674 panic_console_replay 4675 When panic happens, replay all kernel messages on 4676 consoles at the end of panic. 4677 4678 parkbd.port= [HW] Parallel port number the keyboard adapter is 4679 connected to, default is 0. 4680 Format: <parport#> 4681 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 4682 0 for XT, 1 for AT (default is AT). 4683 Format: <mode> 4684 4685 parport= [HW,PPT] Specify parallel ports. 0 disables. 4686 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 4687 Use 'auto' to force the driver to use any 4688 IRQ/DMA settings detected (the default is to 4689 ignore detected IRQ/DMA settings because of 4690 possible conflicts). You can specify the base 4691 address, IRQ, and DMA settings; IRQ and DMA 4692 should be numbers, or 'auto' (for using detected 4693 settings on that particular port), or 'nofifo' 4694 (to avoid using a FIFO even if it is detected). 4695 Parallel ports are assigned in the order they 4696 are specified on the command line, starting 4697 with parport0. 4698 4699 parport_init_mode= [HW,PPT] 4700 Configure VIA parallel port to operate in 4701 a specific mode. This is necessary on Pegasos 4702 computer where firmware has no options for setting 4703 up parallel port mode and sets it to spp. 4704 Currently this function knows 686a and 8231 chips. 4705 Format: [spp|ps2|epp|ecp|ecpepp] 4706 4707 pata_legacy.all= [HW,LIBATA] 4708 Format: <int> 4709 Set to non-zero to probe primary and secondary ISA 4710 port ranges on PCI systems where no PCI PATA device 4711 has been found at either range. Disabled by default. 4712 4713 pata_legacy.autospeed= [HW,LIBATA] 4714 Format: <int> 4715 Set to non-zero if a chip is present that snoops speed 4716 changes. Disabled by default. 4717 4718 pata_legacy.ht6560a= [HW,LIBATA] 4719 Format: <int> 4720 Set to 1, 2, or 3 for HT 6560A on the primary channel, 4721 the secondary channel, or both channels respectively. 4722 Disabled by default. 4723 4724 pata_legacy.ht6560b= [HW,LIBATA] 4725 Format: <int> 4726 Set to 1, 2, or 3 for HT 6560B on the primary channel, 4727 the secondary channel, or both channels respectively. 4728 Disabled by default. 4729 4730 pata_legacy.iordy_mask= [HW,LIBATA] 4731 Format: <int> 4732 IORDY enable mask. Set individual bits to allow IORDY 4733 for the respective channel. Bit 0 is for the first 4734 legacy channel handled by this driver, bit 1 is for 4735 the second channel, and so on. The sequence will often 4736 correspond to the primary legacy channel, the secondary 4737 legacy channel, and so on, but the handling of a PCI 4738 bus and the use of other driver options may interfere 4739 with the sequence. By default IORDY is allowed across 4740 all channels. 4741 4742 pata_legacy.opti82c46x= [HW,LIBATA] 4743 Format: <int> 4744 Set to 1, 2, or 3 for Opti 82c611A on the primary 4745 channel, the secondary channel, or both channels 4746 respectively. Disabled by default. 4747 4748 pata_legacy.opti82c611a= [HW,LIBATA] 4749 Format: <int> 4750 Set to 1, 2, or 3 for Opti 82c465MV on the primary 4751 channel, the secondary channel, or both channels 4752 respectively. Disabled by default. 4753 4754 pata_legacy.pio_mask= [HW,LIBATA] 4755 Format: <int> 4756 PIO mode mask for autospeed devices. Set individual 4757 bits to allow the use of the respective PIO modes. 4758 Bit 0 is for mode 0, bit 1 is for mode 1, and so on. 4759 All modes allowed by default. 4760 4761 pata_legacy.probe_all= [HW,LIBATA] 4762 Format: <int> 4763 Set to non-zero to probe tertiary and further ISA 4764 port ranges on PCI systems. Disabled by default. 4765 4766 pata_legacy.probe_mask= [HW,LIBATA] 4767 Format: <int> 4768 Probe mask for legacy ISA PATA ports. Depending on 4769 platform configuration and the use of other driver 4770 options up to 6 legacy ports are supported: 0x1f0, 4771 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing 4772 of individual ports can be disabled by setting the 4773 corresponding bits in the mask to 1. Bit 0 is for 4774 the first port in the list above (0x1f0), and so on. 4775 By default all supported ports are probed. 4776 4777 pata_legacy.qdi= [HW,LIBATA] 4778 Format: <int> 4779 Set to non-zero to probe QDI controllers. By default 4780 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise. 4781 4782 pata_legacy.winbond= [HW,LIBATA] 4783 Format: <int> 4784 Set to non-zero to probe Winbond controllers. Use 4785 the standard I/O port (0x130) if 1, otherwise the 4786 value given is the I/O port to use (typically 0x1b0). 4787 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE, 4788 0 otherwise. 4789 4790 pata_platform.pio_mask= [HW,LIBATA] 4791 Format: <int> 4792 Supported PIO mode mask. Set individual bits to allow 4793 the use of the respective PIO modes. Bit 0 is for 4794 mode 0, bit 1 is for mode 1, and so on. Mode 0 only 4795 allowed by default. 4796 4797 pause_on_oops=<int> 4798 Halt all CPUs after the first oops has been printed for 4799 the specified number of seconds. This is to be used if 4800 your oopses keep scrolling off the screen. 4801 4802 pcbit= [HW,ISDN] 4803 4804 pci=option[,option...] [PCI,EARLY] various PCI subsystem options. 4805 4806 Some options herein operate on a specific device 4807 or a set of devices (<pci_dev>). These are 4808 specified in one of the following formats: 4809 4810 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 4811 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 4812 4813 Note: the first format specifies a PCI 4814 bus/device/function address which may change 4815 if new hardware is inserted, if motherboard 4816 firmware changes, or due to changes caused 4817 by other kernel parameters. If the 4818 domain is left unspecified, it is 4819 taken to be zero. Optionally, a path 4820 to a device through multiple device/function 4821 addresses can be specified after the base 4822 address (this is more robust against 4823 renumbering issues). The second format 4824 selects devices using IDs from the 4825 configuration space which may match multiple 4826 devices in the system. 4827 4828 earlydump dump PCI config space before the kernel 4829 changes anything 4830 off [X86] don't probe for the PCI bus 4831 bios [X86-32] force use of PCI BIOS, don't access 4832 the hardware directly. Use this if your machine 4833 has a non-standard PCI host bridge. 4834 nobios [X86-32] disallow use of PCI BIOS, only direct 4835 hardware access methods are allowed. Use this 4836 if you experience crashes upon bootup and you 4837 suspect they are caused by the BIOS. 4838 conf1 [X86] Force use of PCI Configuration Access 4839 Mechanism 1 (config address in IO port 0xCF8, 4840 data in IO port 0xCFC, both 32-bit). 4841 conf2 [X86] Force use of PCI Configuration Access 4842 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 4843 the function, IO port 0xCFA, also 8-bit, sets 4844 bus number. The config space is then accessed 4845 through ports 0xC000-0xCFFF). 4846 See http://wiki.osdev.org/PCI for more info 4847 on the configuration access mechanisms. 4848 noaer [PCIE] If the PCIEAER kernel config parameter is 4849 enabled, this kernel boot option can be used to 4850 disable the use of PCIE advanced error reporting. 4851 nodomains [PCI] Disable support for multiple PCI 4852 root domains (aka PCI segments, in ACPI-speak). 4853 nommconf [X86] Disable use of MMCONFIG for PCI 4854 Configuration 4855 check_enable_amd_mmconf [X86] check for and enable 4856 properly configured MMIO access to PCI 4857 config space on AMD family 10h CPU 4858 nomsi [MSI] If the PCI_MSI kernel config parameter is 4859 enabled, this kernel boot option can be used to 4860 disable the use of MSI interrupts system-wide. 4861 noioapicquirk [APIC] Disable all boot interrupt quirks. 4862 Safety option to keep boot IRQs enabled. This 4863 should never be necessary. 4864 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 4865 primary IO-APIC for bridges that cannot disable 4866 boot IRQs. This fixes a source of spurious IRQs 4867 when the system masks IRQs. 4868 noioapicreroute [APIC] Disable workaround that uses the 4869 boot IRQ equivalent of an IRQ that connects to 4870 a chipset where boot IRQs cannot be disabled. 4871 The opposite of ioapicreroute. 4872 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 4873 routing table. These calls are known to be buggy 4874 on several machines and they hang the machine 4875 when used, but on other computers it's the only 4876 way to get the interrupt routing table. Try 4877 this option if the kernel is unable to allocate 4878 IRQs or discover secondary PCI buses on your 4879 motherboard. 4880 rom [X86] Assign address space to expansion ROMs. 4881 Use with caution as certain devices share 4882 address decoders between ROMs and other 4883 resources. 4884 norom [X86] Do not assign address space to 4885 expansion ROMs that do not already have 4886 BIOS assigned address ranges. 4887 nobar [X86] Do not assign address space to the 4888 BARs that weren't assigned by the BIOS. 4889 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 4890 assigned automatically to PCI devices. You can 4891 make the kernel exclude IRQs of your ISA cards 4892 this way. 4893 pirqaddr=0xAAAAA [X86] Specify the physical address 4894 of the PIRQ table (normally generated 4895 by the BIOS) if it is outside the 4896 F0000h-100000h range. 4897 lastbus=N [X86] Scan all buses thru bus #N. Can be 4898 useful if the kernel is unable to find your 4899 secondary buses and you want to tell it 4900 explicitly which ones they are. 4901 assign-busses [X86] Always assign all PCI bus 4902 numbers ourselves, overriding 4903 whatever the firmware may have done. 4904 usepirqmask [X86] Honor the possible IRQ mask stored 4905 in the BIOS $PIR table. This is needed on 4906 some systems with broken BIOSes, notably 4907 some HP Pavilion N5400 and Omnibook XE3 4908 notebooks. This will have no effect if ACPI 4909 IRQ routing is enabled. 4910 noacpi [X86] Do not use ACPI for IRQ routing 4911 or for PCI scanning. 4912 use_crs [X86] Use PCI host bridge window information 4913 from ACPI. On BIOSes from 2008 or later, this 4914 is enabled by default. If you need to use this, 4915 please report a bug. 4916 nocrs [X86] Ignore PCI host bridge windows from ACPI. 4917 If you need to use this, please report a bug. 4918 use_e820 [X86] Use E820 reservations to exclude parts of 4919 PCI host bridge windows. This is a workaround 4920 for BIOS defects in host bridge _CRS methods. 4921 If you need to use this, please report a bug to 4922 <linux-pci@vger.kernel.org>. 4923 no_e820 [X86] Ignore E820 reservations for PCI host 4924 bridge windows. This is the default on modern 4925 hardware. If you need to use this, please report 4926 a bug to <linux-pci@vger.kernel.org>. 4927 routeirq Do IRQ routing for all PCI devices. 4928 This is normally done in pci_enable_device(), 4929 so this option is a temporary workaround 4930 for broken drivers that don't call it. 4931 skip_isa_align [X86] do not align io start addr, so can 4932 handle more pci cards 4933 noearly [X86] Don't do any early type 1 scanning. 4934 This might help on some broken boards which 4935 machine check when some devices' config space 4936 is read. But various workarounds are disabled 4937 and some IOMMU drivers will not work. 4938 bfsort Sort PCI devices into breadth-first order. 4939 This sorting is done to get a device 4940 order compatible with older (<= 2.4) kernels. 4941 nobfsort Don't sort PCI devices into breadth-first order. 4942 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 4943 tuning and use the BIOS-configured MPS defaults. 4944 pcie_bus_safe Set every device's MPS to the largest value 4945 supported by all devices below the root complex. 4946 pcie_bus_perf Set device MPS to the largest allowable MPS 4947 based on its parent bus. Also set MRRS (Max 4948 Read Request Size) to the largest supported 4949 value (no larger than the MPS that the device 4950 or bus can support) for best performance. 4951 pcie_bus_peer2peer Set every device's MPS to 128B, which 4952 every device is guaranteed to support. This 4953 configuration allows peer-to-peer DMA between 4954 any pair of devices, possibly at the cost of 4955 reduced performance. This also guarantees 4956 that hot-added devices will work. 4957 cbiosize=nn[KMG] The fixed amount of bus space which is 4958 reserved for the CardBus bridge's IO window. 4959 The default value is 256 bytes. 4960 cbmemsize=nn[KMG] The fixed amount of bus space which is 4961 reserved for the CardBus bridge's memory 4962 window. The default value is 64 megabytes. 4963 resource_alignment= 4964 Format: 4965 [<order of align>@]<pci_dev>[; ...] 4966 Specifies alignment and device to reassign 4967 aligned memory resources. How to 4968 specify the device is described above. 4969 If <order of align> is not specified, 4970 PAGE_SIZE is used as alignment. 4971 A PCI-PCI bridge can be specified if resource 4972 windows need to be expanded. 4973 To specify the alignment for several 4974 instances of a device, the PCI vendor, 4975 device, subvendor, and subdevice may be 4976 specified, e.g., 12@pci:8086:9c22:103c:198f 4977 for 4096-byte alignment. 4978 ecrc= Enable/disable PCIe ECRC (transaction layer 4979 end-to-end CRC checking). Only effective if 4980 OS has native AER control (either granted by 4981 ACPI _OSC or forced via "pcie_ports=native") 4982 bios: Use BIOS/firmware settings. This is the 4983 the default. 4984 off: Turn ECRC off 4985 on: Turn ECRC on. 4986 hpiosize=nn[KMG] The fixed amount of bus space which is 4987 reserved for hotplug bridge's IO window. 4988 Default size is 256 bytes. 4989 hpmmiosize=nn[KMG] The fixed amount of bus space which is 4990 reserved for hotplug bridge's MMIO window. 4991 Default size is 2 megabytes. 4992 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 4993 reserved for hotplug bridge's MMIO_PREF window. 4994 Default size is 2 megabytes. 4995 hpmemsize=nn[KMG] The fixed amount of bus space which is 4996 reserved for hotplug bridge's MMIO and 4997 MMIO_PREF window. 4998 Default size is 2 megabytes. 4999 hpbussize=nn The minimum amount of additional bus numbers 5000 reserved for buses below a hotplug bridge. 5001 Default is 1. 5002 realloc= Enable/disable reallocating PCI bridge resources 5003 if allocations done by BIOS are too small to 5004 accommodate resources required by all child 5005 devices. 5006 off: Turn realloc off 5007 on: Turn realloc on 5008 realloc same as realloc=on 5009 noari do not use PCIe ARI. 5010 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 5011 do not use PCIe ATS (and IOMMU device IOTLB). 5012 pcie_scan_all Scan all possible PCIe devices. Otherwise we 5013 only look for one device below a PCIe downstream 5014 port. 5015 big_root_window Try to add a big 64bit memory window to the PCIe 5016 root complex on AMD CPUs. Some GFX hardware 5017 can resize a BAR to allow access to all VRAM. 5018 Adding the window is slightly risky (it may 5019 conflict with unreported devices), so this 5020 taints the kernel. 5021 disable_acs_redir=<pci_dev>[; ...] 5022 Specify one or more PCI devices (in the format 5023 specified above) separated by semicolons. 5024 Each device specified will have the PCI ACS 5025 redirect capabilities forced off which will 5026 allow P2P traffic between devices through 5027 bridges without forcing it upstream. Note: 5028 this removes isolation between devices and 5029 may put more devices in an IOMMU group. 5030 config_acs= 5031 Format: 5032 <ACS flags>@<pci_dev>[; ...] 5033 Specify one or more PCI devices (in the format 5034 specified above) optionally prepended with flags 5035 and separated by semicolons. The respective 5036 capabilities will be enabled, disabled or 5037 unchanged based on what is specified in 5038 flags. 5039 5040 ACS Flags is defined as follows: 5041 bit-0 : ACS Source Validation 5042 bit-1 : ACS Translation Blocking 5043 bit-2 : ACS P2P Request Redirect 5044 bit-3 : ACS P2P Completion Redirect 5045 bit-4 : ACS Upstream Forwarding 5046 bit-5 : ACS P2P Egress Control 5047 bit-6 : ACS Direct Translated P2P 5048 Each bit can be marked as: 5049 '0' – force disabled 5050 '1' – force enabled 5051 'x' – unchanged 5052 For example, 5053 pci=config_acs=10x@pci:0:0 5054 would configure all devices that support 5055 ACS to enable P2P Request Redirect, disable 5056 Translation Blocking, and leave Source 5057 Validation unchanged from whatever power-up 5058 or firmware set it to. 5059 5060 Note: this may remove isolation between devices 5061 and may put more devices in an IOMMU group. 5062 force_floating [S390] Force usage of floating interrupts. 5063 nomio [S390] Do not use MIO instructions. 5064 norid [S390] ignore the RID field and force use of 5065 one PCI domain per PCI function 5066 notph [PCIE] If the PCIE_TPH kernel config parameter 5067 is enabled, this kernel boot option can be used 5068 to disable PCIe TLP Processing Hints support 5069 system-wide. 5070 5071 pcie_aspm= [PCIE] Forcibly enable or ignore PCIe Active State Power 5072 Management. 5073 off Don't touch ASPM configuration at all. Leave any 5074 configuration done by firmware unchanged. 5075 force Enable ASPM even on devices that claim not to support it. 5076 WARNING: Forcing ASPM on may cause system lockups. 5077 5078 pcie_ports= [PCIE] PCIe port services handling: 5079 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 5080 even if the platform doesn't give the OS permission to 5081 use them. This may cause conflicts if the platform 5082 also tries to use these services. 5083 dpc-native Use native PCIe service for DPC only. May 5084 cause conflicts if firmware uses AER or DPC. 5085 compat Disable native PCIe services (PME, AER, DPC, PCIe 5086 hotplug). 5087 5088 pcie_port_pm= [PCIE] PCIe port power management handling: 5089 off Disable power management of all PCIe ports 5090 force Forcibly enable power management of all PCIe ports 5091 5092 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 5093 nomsi Do not use MSI for native PCIe PME signaling (this makes 5094 all PCIe root ports use INTx for all services). 5095 5096 pcmv= [HW,PCMCIA] BadgePAD 4 5097 5098 pd_ignore_unused 5099 [PM] 5100 Keep all power-domains already enabled by bootloader on, 5101 even if no driver has claimed them. This is useful 5102 for debug and development, but should not be 5103 needed on a platform with proper driver support. 5104 5105 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 5106 boot time. 5107 Format: { 0 | 1 } 5108 See arch/parisc/kernel/pdc_chassis.c 5109 5110 percpu_alloc= [MM,EARLY] 5111 Select which percpu first chunk allocator to use. 5112 Currently supported values are "embed" and "page". 5113 Archs may support subset or none of the selections. 5114 See comments in mm/percpu.c for details on each 5115 allocator. This parameter is primarily for debugging 5116 and performance comparison. 5117 5118 pirq= [SMP,APIC] Manual mp-table setup 5119 See Documentation/arch/x86/i386/IO-APIC.rst. 5120 5121 plip= [PPT,NET] Parallel port network link 5122 Format: { parport<nr> | timid | 0 } 5123 See also Documentation/admin-guide/parport.rst. 5124 5125 pmtmr= [X86] Manual setup of pmtmr I/O Port. 5126 Override pmtimer IOPort with a hex value. 5127 e.g. pmtmr=0x508 5128 5129 pmu_override= [PPC] Override the PMU. 5130 This option takes over the PMU facility, so it is no 5131 longer usable by perf. Setting this option starts the 5132 PMU counters by setting MMCR0 to 0 (the FC bit is 5133 cleared). If a number is given, then MMCR1 is set to 5134 that number, otherwise (e.g., 'pmu_override=on'), MMCR1 5135 remains 0. 5136 5137 pm_async= [PM] 5138 Format: off 5139 This parameter sets the initial value of the 5140 /sys/power/pm_async sysfs knob at boot time. 5141 If set to "off", disables asynchronous suspend and 5142 resume of devices during system-wide power transitions. 5143 This can be useful on platforms where device 5144 dependencies are not well-defined, or for debugging 5145 power management issues. Asynchronous operations are 5146 enabled by default. 5147 5148 5149 pm_debug_messages [SUSPEND,KNL] 5150 Enable suspend/resume debug messages during boot up. 5151 5152 pnp.debug=1 [PNP] 5153 Enable PNP debug messages (depends on the 5154 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 5155 via /sys/module/pnp/parameters/debug. We always show 5156 current resource usage; turning this on also shows 5157 possible settings and some assignment information. 5158 5159 pnpacpi= [ACPI] 5160 { off } 5161 5162 pnpbios= [ISAPNP] 5163 { on | off | curr | res | no-curr | no-res } 5164 5165 pnp_reserve_irq= 5166 [ISAPNP] Exclude IRQs for the autoconfiguration 5167 5168 pnp_reserve_dma= 5169 [ISAPNP] Exclude DMAs for the autoconfiguration 5170 5171 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 5172 Ranges are in pairs (I/O port base and size). 5173 5174 pnp_reserve_mem= 5175 [ISAPNP] Exclude memory regions for the 5176 autoconfiguration. 5177 Ranges are in pairs (memory base and size). 5178 5179 ports= [IP_VS_FTP] IPVS ftp helper module 5180 Default is 21. 5181 Up to 8 (IP_VS_APP_MAX_PORTS) ports 5182 may be specified. 5183 Format: <port>,<port>.... 5184 5185 possible_cpus= [SMP,S390,X86] 5186 Format: <unsigned int> 5187 Set the number of possible CPUs, overriding the 5188 regular discovery mechanisms (such as ACPI/FW, etc). 5189 5190 powersave=off [PPC] This option disables power saving features. 5191 It specifically disables cpuidle and sets the 5192 platform machine description specific power_save 5193 function to NULL. On Idle the CPU just reduces 5194 execution priority. 5195 5196 ppc_strict_facility_enable 5197 [PPC,ENABLE] This option catches any kernel floating point, 5198 Altivec, VSX and SPE outside of regions specifically 5199 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 5200 There is some performance impact when enabling this. 5201 5202 ppc_tm= [PPC,EARLY] 5203 Format: {"off"} 5204 Disable Hardware Transactional Memory 5205 5206 preempt= [KNL] 5207 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC 5208 none - Limited to cond_resched() calls 5209 voluntary - Limited to cond_resched() and might_sleep() calls 5210 full - Any section that isn't explicitly preempt disabled 5211 can be preempted anytime. Tasks will also yield 5212 contended spinlocks (if the critical section isn't 5213 explicitly preempt disabled beyond the lock itself). 5214 lazy - Scheduler controlled. Similar to full but instead 5215 of preempting the task immediately, the task gets 5216 one HZ tick time to yield itself before the 5217 preemption will be forced. One preemption is when the 5218 task returns to user space. 5219 5220 print-fatal-signals= 5221 [KNL] debug: print fatal signals 5222 5223 If enabled, warn about various signal handling 5224 related application anomalies: too many signals, 5225 too many POSIX.1 timers, fatal signals causing a 5226 coredump - etc. 5227 5228 If you hit the warning due to signal overflow, 5229 you might want to try "ulimit -i unlimited". 5230 5231 default: off. 5232 5233 printk.always_kmsg_dump= 5234 Trigger kmsg_dump for cases other than kernel oops or 5235 panics 5236 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5237 default: disabled 5238 5239 printk.console_no_auto_verbose= 5240 Disable console loglevel raise on oops, panic 5241 or lockdep-detected issues (only if lock debug is on). 5242 With an exception to setups with low baudrate on 5243 serial console, keeping this 0 is a good choice 5244 in order to provide more debug information. 5245 Format: <bool> 5246 default: 0 (auto_verbose is enabled) 5247 5248 printk.debug_non_panic_cpus= 5249 Allows storing messages from non-panic CPUs into 5250 the printk log buffer during panic(). They are 5251 flushed to consoles by the panic-CPU on 5252 a best-effort basis. 5253 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5254 Default: disabled 5255 5256 printk.devkmsg={on,off,ratelimit} 5257 Control writing to /dev/kmsg. 5258 on - unlimited logging to /dev/kmsg from userspace 5259 off - logging to /dev/kmsg disabled 5260 ratelimit - ratelimit the logging 5261 Default: ratelimit 5262 5263 printk.time= Show timing data prefixed to each printk message line 5264 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5265 5266 proc_mem.force_override= [KNL] 5267 Format: {always | ptrace | never} 5268 Traditionally /proc/pid/mem allows memory permissions to be 5269 overridden without restrictions. This option may be set to 5270 restrict that. Can be one of: 5271 - 'always': traditional behavior always allows mem overrides. 5272 - 'ptrace': only allow mem overrides for active ptracers. 5273 - 'never': never allow mem overrides. 5274 If not specified, default is the CONFIG_PROC_MEM_* choice. 5275 5276 processor.max_cstate= [HW,ACPI] 5277 Limit processor to maximum C-state 5278 max_cstate=9 overrides any DMI blacklist limit. 5279 5280 processor.nocst [HW,ACPI] 5281 Ignore the _CST method to determine C-states, 5282 instead using the legacy FADT method 5283 5284 profile= [KNL] Enable kernel profiling via /proc/profile 5285 Format: [<profiletype>,]<number> 5286 Param: <profiletype>: "schedule" or "kvm" 5287 [defaults to kernel profiling] 5288 Param: "schedule" - profile schedule points. 5289 Param: "kvm" - profile VM exits. 5290 Param: <number> - step/bucket size as a power of 2 for 5291 statistical time based profiling. 5292 5293 prompt_ramdisk= [RAM] [Deprecated] 5294 5295 prot_virt= [S390] enable hosting protected virtual machines 5296 isolated from the hypervisor (if hardware supports 5297 that). If enabled, the default kernel base address 5298 might be overridden even when Kernel Address Space 5299 Layout Randomization is disabled. 5300 Format: <bool> 5301 5302 psi= [KNL] Enable or disable pressure stall information 5303 tracking. 5304 Format: <bool> 5305 5306 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 5307 probe for; one of (bare|imps|exps|lifebook|any). 5308 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 5309 per second. 5310 psmouse.resetafter= [HW,MOUSE] 5311 Try to reset the device after so many bad packets 5312 (0 = never). 5313 psmouse.resolution= 5314 [HW,MOUSE] Set desired mouse resolution, in dpi. 5315 psmouse.smartscroll= 5316 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 5317 0 = disabled, 1 = enabled (default). 5318 5319 pstore.backend= Specify the name of the pstore backend to use 5320 5321 pti= [X86-64] Control Page Table Isolation of user and 5322 kernel address spaces. Disabling this feature 5323 removes hardening, but improves performance of 5324 system calls and interrupts. 5325 5326 on - unconditionally enable 5327 off - unconditionally disable 5328 auto - kernel detects whether your CPU model is 5329 vulnerable to issues that PTI mitigates 5330 5331 Not specifying this option is equivalent to pti=auto. 5332 5333 pty.legacy_count= 5334 [KNL] Number of legacy pty's. Overwrites compiled-in 5335 default number. 5336 5337 quiet [KNL,EARLY] Disable most log messages 5338 5339 r128= [HW,DRM] 5340 5341 radix_hcall_invalidate=on [PPC/PSERIES] 5342 Disable RADIX GTSE feature and use hcall for TLB 5343 invalidate. 5344 5345 raid= [HW,RAID] 5346 See Documentation/admin-guide/md.rst. 5347 5348 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 5349 See Documentation/admin-guide/blockdev/ramdisk.rst. 5350 5351 ramdisk_start= [RAM] RAM disk image start address 5352 5353 random.trust_cpu=off 5354 [KNL,EARLY] Disable trusting the use of the CPU's 5355 random number generator (if available) to 5356 initialize the kernel's RNG. 5357 5358 random.trust_bootloader=off 5359 [KNL,EARLY] Disable trusting the use of the a seed 5360 passed by the bootloader (if available) to 5361 initialize the kernel's RNG. 5362 5363 randomize_kstack_offset= 5364 [KNL,EARLY] Enable or disable kernel stack offset 5365 randomization, which provides roughly 5 bits of 5366 entropy, frustrating memory corruption attacks 5367 that depend on stack address determinism or 5368 cross-syscall address exposures. This is only 5369 available on architectures that have defined 5370 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET. 5371 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5372 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT. 5373 5374 ras=option[,option,...] [KNL] RAS-specific options 5375 5376 cec_disable [X86] 5377 Disable the Correctable Errors Collector, 5378 see CONFIG_RAS_CEC help text. 5379 5380 rcu_nocbs[=cpu-list] 5381 [KNL] The optional argument is a cpu list, 5382 as described above. 5383 5384 In kernels built with CONFIG_RCU_NOCB_CPU=y, 5385 enable the no-callback CPU mode, which prevents 5386 such CPUs' callbacks from being invoked in 5387 softirq context. Invocation of such CPUs' RCU 5388 callbacks will instead be offloaded to "rcuox/N" 5389 kthreads created for that purpose, where "x" is 5390 "p" for RCU-preempt, "s" for RCU-sched, and "g" 5391 for the kthreads that mediate grace periods; and 5392 "N" is the CPU number. This reduces OS jitter on 5393 the offloaded CPUs, which can be useful for HPC 5394 and real-time workloads. It can also improve 5395 energy efficiency for asymmetric multiprocessors. 5396 5397 If a cpulist is passed as an argument, the specified 5398 list of CPUs is set to no-callback mode from boot. 5399 5400 Otherwise, if the '=' sign and the cpulist 5401 arguments are omitted, no CPU will be set to 5402 no-callback mode from boot but the mode may be 5403 toggled at runtime via cpusets. 5404 5405 Note that this argument takes precedence over 5406 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 5407 5408 rcu_nocb_poll [KNL] 5409 Rather than requiring that offloaded CPUs 5410 (specified by rcu_nocbs= above) explicitly 5411 awaken the corresponding "rcuoN" kthreads, 5412 make these kthreads poll for callbacks. 5413 This improves the real-time response for the 5414 offloaded CPUs by relieving them of the need to 5415 wake up the corresponding kthread, but degrades 5416 energy efficiency by requiring that the kthreads 5417 periodically wake up to do the polling. 5418 5419 rcutree.blimit= [KNL] 5420 Set maximum number of finished RCU callbacks to 5421 process in one batch. 5422 5423 rcutree.csd_lock_suppress_rcu_stall= [KNL] 5424 Do only a one-line RCU CPU stall warning when 5425 there is an ongoing too-long CSD-lock wait. 5426 5427 rcutree.do_rcu_barrier= [KNL] 5428 Request a call to rcu_barrier(). This is 5429 throttled so that userspace tests can safely 5430 hammer on the sysfs variable if they so choose. 5431 If triggered before the RCU grace-period machinery 5432 is fully active, this will error out with EAGAIN. 5433 5434 rcutree.dump_tree= [KNL] 5435 Dump the structure of the rcu_node combining tree 5436 out at early boot. This is used for diagnostic 5437 purposes, to verify correct tree setup. 5438 5439 rcutree.gp_cleanup_delay= [KNL] 5440 Set the number of jiffies to delay each step of 5441 RCU grace-period cleanup. 5442 5443 rcutree.gp_init_delay= [KNL] 5444 Set the number of jiffies to delay each step of 5445 RCU grace-period initialization. 5446 5447 rcutree.gp_preinit_delay= [KNL] 5448 Set the number of jiffies to delay each step of 5449 RCU grace-period pre-initialization, that is, 5450 the propagation of recent CPU-hotplug changes up 5451 the rcu_node combining tree. 5452 5453 rcutree.jiffies_till_first_fqs= [KNL] 5454 Set delay from grace-period initialization to 5455 first attempt to force quiescent states. 5456 Units are jiffies, minimum value is zero, 5457 and maximum value is HZ. 5458 5459 rcutree.jiffies_till_next_fqs= [KNL] 5460 Set delay between subsequent attempts to force 5461 quiescent states. Units are jiffies, minimum 5462 value is one, and maximum value is HZ. 5463 5464 rcutree.jiffies_till_sched_qs= [KNL] 5465 Set required age in jiffies for a 5466 given grace period before RCU starts 5467 soliciting quiescent-state help from 5468 rcu_note_context_switch() and cond_resched(). 5469 If not specified, the kernel will calculate 5470 a value based on the most recent settings 5471 of rcutree.jiffies_till_first_fqs 5472 and rcutree.jiffies_till_next_fqs. 5473 This calculated value may be viewed in 5474 rcutree.jiffies_to_sched_qs. Any attempt to set 5475 rcutree.jiffies_to_sched_qs will be cheerfully 5476 overwritten. 5477 5478 rcutree.kthread_prio= [KNL,BOOT] 5479 Set the SCHED_FIFO priority of the RCU per-CPU 5480 kthreads (rcuc/N). This value is also used for 5481 the priority of the RCU boost threads (rcub/N) 5482 and for the RCU grace-period kthreads (rcu_bh, 5483 rcu_preempt, and rcu_sched). If RCU_BOOST is 5484 set, valid values are 1-99 and the default is 1 5485 (the least-favored priority). Otherwise, when 5486 RCU_BOOST is not set, valid values are 0-99 and 5487 the default is zero (non-realtime operation). 5488 When RCU_NOCB_CPU is set, also adjust the 5489 priority of NOCB callback kthreads. 5490 5491 rcutree.nocb_nobypass_lim_per_jiffy= [KNL] 5492 On callback-offloaded (rcu_nocbs) CPUs, 5493 RCU reduces the lock contention that would 5494 otherwise be caused by callback floods through 5495 use of the ->nocb_bypass list. However, in the 5496 common non-flooded case, RCU queues directly to 5497 the main ->cblist in order to avoid the extra 5498 overhead of the ->nocb_bypass list and its lock. 5499 But if there are too many callbacks queued during 5500 a single jiffy, RCU pre-queues the callbacks into 5501 the ->nocb_bypass queue. The definition of "too 5502 many" is supplied by this kernel boot parameter. 5503 5504 rcutree.nohz_full_patience_delay= [KNL] 5505 On callback-offloaded (rcu_nocbs) CPUs, avoid 5506 disturbing RCU unless the grace period has 5507 reached the specified age in milliseconds. 5508 Defaults to zero. Large values will be capped 5509 at five seconds. All values will be rounded down 5510 to the nearest value representable by jiffies. 5511 5512 rcutree.qhimark= [KNL] 5513 Set threshold of queued RCU callbacks beyond which 5514 batch limiting is disabled. 5515 5516 rcutree.qlowmark= [KNL] 5517 Set threshold of queued RCU callbacks below which 5518 batch limiting is re-enabled. 5519 5520 rcutree.qovld= [KNL] 5521 Set threshold of queued RCU callbacks beyond which 5522 RCU's force-quiescent-state scan will aggressively 5523 enlist help from cond_resched() and sched IPIs to 5524 help CPUs more quickly reach quiescent states. 5525 Set to less than zero to make this be set based 5526 on rcutree.qhimark at boot time and to zero to 5527 disable more aggressive help enlistment. 5528 5529 rcutree.rcu_delay_page_cache_fill_msec= [KNL] 5530 Set the page-cache refill delay (in milliseconds) 5531 in response to low-memory conditions. The range 5532 of permitted values is in the range 0:100000. 5533 5534 rcutree.rcu_divisor= [KNL] 5535 Set the shift-right count to use to compute 5536 the callback-invocation batch limit bl from 5537 the number of callbacks queued on this CPU. 5538 The result will be bounded below by the value of 5539 the rcutree.blimit kernel parameter. Every bl 5540 callbacks, the softirq handler will exit in 5541 order to allow the CPU to do other work. 5542 5543 Please note that this callback-invocation batch 5544 limit applies only to non-offloaded callback 5545 invocation. Offloaded callbacks are instead 5546 invoked in the context of an rcuoc kthread, which 5547 scheduler will preempt as it does any other task. 5548 5549 rcutree.rcu_fanout_exact= [KNL] 5550 Disable autobalancing of the rcu_node combining 5551 tree. This is used by rcutorture, and might 5552 possibly be useful for architectures having high 5553 cache-to-cache transfer latencies. 5554 5555 rcutree.rcu_fanout_leaf= [KNL] 5556 Change the number of CPUs assigned to each 5557 leaf rcu_node structure. Useful for very 5558 large systems, which will choose the value 64, 5559 and for NUMA systems with large remote-access 5560 latencies, which will choose a value aligned 5561 with the appropriate hardware boundaries. 5562 5563 rcutree.rcu_min_cached_objs= [KNL] 5564 Minimum number of objects which are cached and 5565 maintained per one CPU. Object size is equal 5566 to PAGE_SIZE. The cache allows to reduce the 5567 pressure to page allocator, also it makes the 5568 whole algorithm to behave better in low memory 5569 condition. 5570 5571 rcutree.rcu_nocb_gp_stride= [KNL] 5572 Set the number of NOCB callback kthreads in 5573 each group, which defaults to the square root 5574 of the number of CPUs. Larger numbers reduce 5575 the wakeup overhead on the global grace-period 5576 kthread, but increases that same overhead on 5577 each group's NOCB grace-period kthread. 5578 5579 rcutree.rcu_kick_kthreads= [KNL] 5580 Cause the grace-period kthread to get an extra 5581 wake_up() if it sleeps three times longer than 5582 it should at force-quiescent-state time. 5583 This wake_up() will be accompanied by a 5584 WARN_ONCE() splat and an ftrace_dump(). 5585 5586 rcutree.rcu_resched_ns= [KNL] 5587 Limit the time spend invoking a batch of RCU 5588 callbacks to the specified number of nanoseconds. 5589 By default, this limit is checked only once 5590 every 32 callbacks in order to limit the pain 5591 inflicted by local_clock() overhead. 5592 5593 rcutree.rcu_unlock_delay= [KNL] 5594 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 5595 this specifies an rcu_read_unlock()-time delay 5596 in microseconds. This defaults to zero. 5597 Larger delays increase the probability of 5598 catching RCU pointer leaks, that is, buggy use 5599 of RCU-protected pointers after the relevant 5600 rcu_read_unlock() has completed. 5601 5602 rcutree.sysrq_rcu= [KNL] 5603 Commandeer a sysrq key to dump out Tree RCU's 5604 rcu_node tree with an eye towards determining 5605 why a new grace period has not yet started. 5606 5607 rcutree.use_softirq= [KNL] 5608 If set to zero, move all RCU_SOFTIRQ processing to 5609 per-CPU rcuc kthreads. Defaults to a non-zero 5610 value, meaning that RCU_SOFTIRQ is used by default. 5611 Specify rcutree.use_softirq=0 to use rcuc kthreads. 5612 5613 But note that CONFIG_PREEMPT_RT=y kernels disable 5614 this kernel boot parameter, forcibly setting it 5615 to zero. 5616 5617 rcutree.enable_rcu_lazy= [KNL] 5618 To save power, batch RCU callbacks and flush after 5619 delay, memory pressure or callback list growing too 5620 big. 5621 5622 rcutree.rcu_normal_wake_from_gp= [KNL] 5623 Reduces a latency of synchronize_rcu() call. This approach 5624 maintains its own track of synchronize_rcu() callers, so it 5625 does not interact with regular callbacks because it does not 5626 use a call_rcu[_hurry]() path. Please note, this is for a 5627 normal grace period. 5628 5629 How to enable it: 5630 5631 echo 1 > /sys/module/rcutree/parameters/rcu_normal_wake_from_gp 5632 or pass a boot parameter "rcutree.rcu_normal_wake_from_gp=1" 5633 5634 Default is 1 if num_possible_cpus() <= 16 and it is not explicitly 5635 disabled by the boot parameter passing 0. 5636 5637 rcuscale.gp_async= [KNL] 5638 Measure performance of asynchronous 5639 grace-period primitives such as call_rcu(). 5640 5641 rcuscale.gp_async_max= [KNL] 5642 Specify the maximum number of outstanding 5643 callbacks per writer thread. When a writer 5644 thread exceeds this limit, it invokes the 5645 corresponding flavor of rcu_barrier() to allow 5646 previously posted callbacks to drain. 5647 5648 rcuscale.gp_exp= [KNL] 5649 Measure performance of expedited synchronous 5650 grace-period primitives. 5651 5652 rcuscale.holdoff= [KNL] 5653 Set test-start holdoff period. The purpose of 5654 this parameter is to delay the start of the 5655 test until boot completes in order to avoid 5656 interference. 5657 5658 rcuscale.kfree_by_call_rcu= [KNL] 5659 In kernels built with CONFIG_RCU_LAZY=y, test 5660 call_rcu() instead of kfree_rcu(). 5661 5662 rcuscale.kfree_mult= [KNL] 5663 Instead of allocating an object of size kfree_obj, 5664 allocate one of kfree_mult * sizeof(kfree_obj). 5665 Defaults to 1. 5666 5667 rcuscale.kfree_rcu_test= [KNL] 5668 Set to measure performance of kfree_rcu() flooding. 5669 5670 rcuscale.kfree_rcu_test_double= [KNL] 5671 Test the double-argument variant of kfree_rcu(). 5672 If this parameter has the same value as 5673 rcuscale.kfree_rcu_test_single, both the single- 5674 and double-argument variants are tested. 5675 5676 rcuscale.kfree_rcu_test_single= [KNL] 5677 Test the single-argument variant of kfree_rcu(). 5678 If this parameter has the same value as 5679 rcuscale.kfree_rcu_test_double, both the single- 5680 and double-argument variants are tested. 5681 5682 rcuscale.kfree_nthreads= [KNL] 5683 The number of threads running loops of kfree_rcu(). 5684 5685 rcuscale.kfree_alloc_num= [KNL] 5686 Number of allocations and frees done in an iteration. 5687 5688 rcuscale.kfree_loops= [KNL] 5689 Number of loops doing rcuscale.kfree_alloc_num number 5690 of allocations and frees. 5691 5692 rcuscale.minruntime= [KNL] 5693 Set the minimum test run time in seconds. This 5694 does not affect the data-collection interval, 5695 but instead allows better measurement of things 5696 like CPU consumption. 5697 5698 rcuscale.nreaders= [KNL] 5699 Set number of RCU readers. The value -1 selects 5700 N, where N is the number of CPUs. A value 5701 "n" less than -1 selects N-n+1, where N is again 5702 the number of CPUs. For example, -2 selects N 5703 (the number of CPUs), -3 selects N+1, and so on. 5704 A value of "n" less than or equal to -N selects 5705 a single reader. 5706 5707 rcuscale.nwriters= [KNL] 5708 Set number of RCU writers. The values operate 5709 the same as for rcuscale.nreaders. 5710 N, where N is the number of CPUs 5711 5712 rcuscale.scale_type= [KNL] 5713 Specify the RCU implementation to test. 5714 5715 rcuscale.shutdown= [KNL] 5716 Shut the system down after performance tests 5717 complete. This is useful for hands-off automated 5718 testing. 5719 5720 rcuscale.verbose= [KNL] 5721 Enable additional printk() statements. 5722 5723 rcuscale.writer_holdoff= [KNL] 5724 Write-side holdoff between grace periods, 5725 in microseconds. The default of zero says 5726 no holdoff. 5727 5728 rcuscale.writer_holdoff_jiffies= [KNL] 5729 Additional write-side holdoff between grace 5730 periods, but in jiffies. The default of zero 5731 says no holdoff. 5732 5733 rcutorture.fqs_duration= [KNL] 5734 Set duration of force_quiescent_state bursts 5735 in microseconds. 5736 5737 rcutorture.fqs_holdoff= [KNL] 5738 Set holdoff time within force_quiescent_state bursts 5739 in microseconds. 5740 5741 rcutorture.fqs_stutter= [KNL] 5742 Set wait time between force_quiescent_state bursts 5743 in seconds. 5744 5745 rcutorture.fwd_progress= [KNL] 5746 Specifies the number of kthreads to be used 5747 for RCU grace-period forward-progress testing 5748 for the types of RCU supporting this notion. 5749 Defaults to 1 kthread, values less than zero or 5750 greater than the number of CPUs cause the number 5751 of CPUs to be used. 5752 5753 rcutorture.fwd_progress_div= [KNL] 5754 Specify the fraction of a CPU-stall-warning 5755 period to do tight-loop forward-progress testing. 5756 5757 rcutorture.fwd_progress_holdoff= [KNL] 5758 Number of seconds to wait between successive 5759 forward-progress tests. 5760 5761 rcutorture.fwd_progress_need_resched= [KNL] 5762 Enclose cond_resched() calls within checks for 5763 need_resched() during tight-loop forward-progress 5764 testing. 5765 5766 rcutorture.gp_cond= [KNL] 5767 Use conditional/asynchronous update-side 5768 normal-grace-period primitives, if available. 5769 5770 rcutorture.gp_cond_exp= [KNL] 5771 Use conditional/asynchronous update-side 5772 expedited-grace-period primitives, if available. 5773 5774 rcutorture.gp_cond_full= [KNL] 5775 Use conditional/asynchronous update-side 5776 normal-grace-period primitives that also take 5777 concurrent expedited grace periods into account, 5778 if available. 5779 5780 rcutorture.gp_cond_exp_full= [KNL] 5781 Use conditional/asynchronous update-side 5782 expedited-grace-period primitives that also take 5783 concurrent normal grace periods into account, 5784 if available. 5785 5786 rcutorture.gp_cond_wi= [KNL] 5787 Nominal wait interval for normal conditional 5788 grace periods (specified by rcutorture's 5789 gp_cond and gp_cond_full module parameters), 5790 in microseconds. The actual wait interval will 5791 be randomly selected to nanosecond granularity up 5792 to this wait interval. Defaults to 16 jiffies, 5793 for example, 16,000 microseconds on a system 5794 with HZ=1000. 5795 5796 rcutorture.gp_cond_wi_exp= [KNL] 5797 Nominal wait interval for expedited conditional 5798 grace periods (specified by rcutorture's 5799 gp_cond_exp and gp_cond_exp_full module 5800 parameters), in microseconds. The actual wait 5801 interval will be randomly selected to nanosecond 5802 granularity up to this wait interval. Defaults to 5803 128 microseconds. 5804 5805 rcutorture.gp_exp= [KNL] 5806 Use expedited update-side primitives, if available. 5807 5808 rcutorture.gp_normal= [KNL] 5809 Use normal (non-expedited) asynchronous 5810 update-side primitives, if available. 5811 5812 rcutorture.gp_poll= [KNL] 5813 Use polled update-side normal-grace-period 5814 primitives, if available. 5815 5816 rcutorture.gp_poll_exp= [KNL] 5817 Use polled update-side expedited-grace-period 5818 primitives, if available. 5819 5820 rcutorture.gp_poll_full= [KNL] 5821 Use polled update-side normal-grace-period 5822 primitives that also take concurrent expedited 5823 grace periods into account, if available. 5824 5825 rcutorture.gp_poll_exp_full= [KNL] 5826 Use polled update-side expedited-grace-period 5827 primitives that also take concurrent normal 5828 grace periods into account, if available. 5829 5830 rcutorture.gp_poll_wi= [KNL] 5831 Nominal wait interval for normal conditional 5832 grace periods (specified by rcutorture's 5833 gp_poll and gp_poll_full module parameters), 5834 in microseconds. The actual wait interval will 5835 be randomly selected to nanosecond granularity up 5836 to this wait interval. Defaults to 16 jiffies, 5837 for example, 16,000 microseconds on a system 5838 with HZ=1000. 5839 5840 rcutorture.gp_poll_wi_exp= [KNL] 5841 Nominal wait interval for expedited conditional 5842 grace periods (specified by rcutorture's 5843 gp_poll_exp and gp_poll_exp_full module 5844 parameters), in microseconds. The actual wait 5845 interval will be randomly selected to nanosecond 5846 granularity up to this wait interval. Defaults to 5847 128 microseconds. 5848 5849 rcutorture.gp_sync= [KNL] 5850 Use normal (non-expedited) synchronous 5851 update-side primitives, if available. If all 5852 of rcutorture.gp_cond=, rcutorture.gp_exp=, 5853 rcutorture.gp_normal=, and rcutorture.gp_sync= 5854 are zero, rcutorture acts as if is interpreted 5855 they are all non-zero. 5856 5857 rcutorture.gpwrap_lag= [KNL] 5858 Enable grace-period wrap lag testing. Setting 5859 to false prevents the gpwrap lag test from 5860 running. Default is true. 5861 5862 rcutorture.gpwrap_lag_gps= [KNL] 5863 Set the value for grace-period wrap lag during 5864 active lag testing periods. This controls how many 5865 grace periods differences we tolerate between 5866 rdp and rnp's gp_seq before setting overflow flag. 5867 The default is always set to 8. 5868 5869 rcutorture.gpwrap_lag_cycle_mins= [KNL] 5870 Set the total cycle duration for gpwrap lag 5871 testing in minutes. This is the total time for 5872 one complete cycle of active and inactive 5873 testing periods. Default is 30 minutes. 5874 5875 rcutorture.gpwrap_lag_active_mins= [KNL] 5876 Set the duration for which gpwrap lag is active 5877 within each cycle, in minutes. During this time, 5878 the grace-period wrap lag will be set to the 5879 value specified by gpwrap_lag_gps. Default is 5880 5 minutes. 5881 5882 rcutorture.irqreader= [KNL] 5883 Run RCU readers from irq handlers, or, more 5884 accurately, from a timer handler. Not all RCU 5885 flavors take kindly to this sort of thing. 5886 5887 rcutorture.leakpointer= [KNL] 5888 Leak an RCU-protected pointer out of the reader. 5889 This can of course result in splats, and is 5890 intended to test the ability of things like 5891 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 5892 such leaks. 5893 5894 rcutorture.n_barrier_cbs= [KNL] 5895 Set callbacks/threads for rcu_barrier() testing. 5896 5897 rcutorture.nfakewriters= [KNL] 5898 Set number of concurrent RCU writers. These just 5899 stress RCU, they don't participate in the actual 5900 test, hence the "fake". 5901 5902 rcutorture.nocbs_nthreads= [KNL] 5903 Set number of RCU callback-offload togglers. 5904 Zero (the default) disables toggling. 5905 5906 rcutorture.nocbs_toggle= [KNL] 5907 Set the delay in milliseconds between successive 5908 callback-offload toggling attempts. 5909 5910 rcutorture.nreaders= [KNL] 5911 Set number of RCU readers. The value -1 selects 5912 N-1, where N is the number of CPUs. A value 5913 "n" less than -1 selects N-n-2, where N is again 5914 the number of CPUs. For example, -2 selects N 5915 (the number of CPUs), -3 selects N+1, and so on. 5916 5917 rcutorture.object_debug= [KNL] 5918 Enable debug-object double-call_rcu() testing. 5919 5920 rcutorture.onoff_holdoff= [KNL] 5921 Set time (s) after boot for CPU-hotplug testing. 5922 5923 rcutorture.onoff_interval= [KNL] 5924 Set time (jiffies) between CPU-hotplug operations, 5925 or zero to disable CPU-hotplug testing. 5926 5927 rcutorture.preempt_duration= [KNL] 5928 Set duration (in milliseconds) of preemptions 5929 by a high-priority FIFO real-time task. Set to 5930 zero (the default) to disable. The CPUs to 5931 preempt are selected randomly from the set that 5932 are online at a given point in time. Races with 5933 CPUs going offline are ignored, with that attempt 5934 at preemption skipped. 5935 5936 rcutorture.preempt_interval= [KNL] 5937 Set interval (in milliseconds, defaulting to one 5938 second) between preemptions by a high-priority 5939 FIFO real-time task. This delay is mediated 5940 by an hrtimer and is further fuzzed to avoid 5941 inadvertent synchronizations. 5942 5943 rcutorture.read_exit_burst= [KNL] 5944 The number of times in a given read-then-exit 5945 episode that a set of read-then-exit kthreads 5946 is spawned. 5947 5948 rcutorture.read_exit_delay= [KNL] 5949 The delay, in seconds, between successive 5950 read-then-exit testing episodes. 5951 5952 rcutorture.reader_flavor= [KNL] 5953 A bit mask indicating which readers to use. 5954 If there is more than one bit set, the readers 5955 are entered from low-order bit up, and are 5956 exited in the opposite order. For SRCU, the 5957 0x1 bit is normal readers, 0x2 NMI-safe readers, 5958 and 0x4 light-weight readers. 5959 5960 rcutorture.shuffle_interval= [KNL] 5961 Set task-shuffle interval (s). Shuffling tasks 5962 allows some CPUs to go into dyntick-idle mode 5963 during the rcutorture test. 5964 5965 rcutorture.shutdown_secs= [KNL] 5966 Set time (s) after boot system shutdown. This 5967 is useful for hands-off automated testing. 5968 5969 rcutorture.stall_cpu= [KNL] 5970 Duration of CPU stall (s) to test RCU CPU stall 5971 warnings, zero to disable. 5972 5973 rcutorture.stall_cpu_block= [KNL] 5974 Sleep while stalling if set. This will result 5975 in warnings from preemptible RCU in addition to 5976 any other stall-related activity. Note that 5977 in kernels built with CONFIG_PREEMPTION=n and 5978 CONFIG_PREEMPT_COUNT=y, this parameter will 5979 cause the CPU to pass through a quiescent state. 5980 Given CONFIG_PREEMPTION=n, this will suppress 5981 RCU CPU stall warnings, but will instead result 5982 in scheduling-while-atomic splats. 5983 5984 Use of this module parameter results in splats. 5985 5986 5987 rcutorture.stall_cpu_holdoff= [KNL] 5988 Time to wait (s) after boot before inducing stall. 5989 5990 rcutorture.stall_cpu_irqsoff= [KNL] 5991 Disable interrupts while stalling if set, but only 5992 on the first stall in the set. 5993 5994 rcutorture.stall_cpu_repeat= [KNL] 5995 Number of times to repeat the stall sequence, 5996 so that rcutorture.stall_cpu_repeat=3 will result 5997 in four stall sequences. 5998 5999 rcutorture.stall_gp_kthread= [KNL] 6000 Duration (s) of forced sleep within RCU 6001 grace-period kthread to test RCU CPU stall 6002 warnings, zero to disable. If both stall_cpu 6003 and stall_gp_kthread are specified, the 6004 kthread is starved first, then the CPU. 6005 6006 rcutorture.stat_interval= [KNL] 6007 Time (s) between statistics printk()s. 6008 6009 rcutorture.stutter= [KNL] 6010 Time (s) to stutter testing, for example, specifying 6011 five seconds causes the test to run for five seconds, 6012 wait for five seconds, and so on. This tests RCU's 6013 ability to transition abruptly to and from idle. 6014 6015 rcutorture.test_boost= [KNL] 6016 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 6017 "Maybe" means test if the RCU implementation 6018 under test support RCU priority boosting. 6019 6020 rcutorture.test_boost_duration= [KNL] 6021 Duration (s) of each individual boost test. 6022 6023 rcutorture.test_boost_holdoff= [KNL] 6024 Holdoff time (s) from start of test to the start 6025 of RCU priority-boost testing. Defaults to zero, 6026 that is, no holdoff. 6027 6028 rcutorture.test_boost_interval= [KNL] 6029 Interval (s) between each boost test. 6030 6031 rcutorture.test_no_idle_hz= [KNL] 6032 Test RCU's dyntick-idle handling. See also the 6033 rcutorture.shuffle_interval parameter. 6034 6035 rcutorture.torture_type= [KNL] 6036 Specify the RCU implementation to test. 6037 6038 rcutorture.verbose= [KNL] 6039 Enable additional printk() statements. 6040 6041 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 6042 Dump ftrace buffer after reporting RCU CPU 6043 stall warning. 6044 6045 rcupdate.rcu_cpu_stall_notifiers= [KNL] 6046 Provide RCU CPU stall notifiers, but see the 6047 warnings in the RCU_CPU_STALL_NOTIFIER Kconfig 6048 option's help text. TL;DR: You almost certainly 6049 do not want rcupdate.rcu_cpu_stall_notifiers. 6050 6051 rcupdate.rcu_cpu_stall_suppress= [KNL] 6052 Suppress RCU CPU stall warning messages. 6053 6054 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 6055 Suppress RCU CPU stall warning messages and 6056 rcutorture writer stall warnings that occur 6057 during early boot, that is, during the time 6058 before the init task is spawned. 6059 6060 rcupdate.rcu_cpu_stall_timeout= [KNL] 6061 Set timeout for RCU CPU stall warning messages. 6062 The value is in seconds and the maximum allowed 6063 value is 300 seconds. 6064 6065 rcupdate.rcu_exp_cpu_stall_timeout= [KNL] 6066 Set timeout for expedited RCU CPU stall warning 6067 messages. The value is in milliseconds 6068 and the maximum allowed value is 21000 6069 milliseconds. Please note that this value is 6070 adjusted to an arch timer tick resolution. 6071 Setting this to zero causes the value from 6072 rcupdate.rcu_cpu_stall_timeout to be used (after 6073 conversion from seconds to milliseconds). 6074 6075 rcupdate.rcu_cpu_stall_cputime= [KNL] 6076 Provide statistics on the cputime and count of 6077 interrupts and tasks during the sampling period. For 6078 multiple continuous RCU stalls, all sampling periods 6079 begin at half of the first RCU stall timeout. 6080 6081 rcupdate.rcu_exp_stall_task_details= [KNL] 6082 Print stack dumps of any tasks blocking the 6083 current expedited RCU grace period during an 6084 expedited RCU CPU stall warning. 6085 6086 rcupdate.rcu_expedited= [KNL] 6087 Use expedited grace-period primitives, for 6088 example, synchronize_rcu_expedited() instead 6089 of synchronize_rcu(). This reduces latency, 6090 but can increase CPU utilization, degrade 6091 real-time latency, and degrade energy efficiency. 6092 No effect on CONFIG_TINY_RCU kernels. 6093 6094 rcupdate.rcu_normal= [KNL] 6095 Use only normal grace-period primitives, 6096 for example, synchronize_rcu() instead of 6097 synchronize_rcu_expedited(). This improves 6098 real-time latency, CPU utilization, and 6099 energy efficiency, but can expose users to 6100 increased grace-period latency. This parameter 6101 overrides rcupdate.rcu_expedited. No effect on 6102 CONFIG_TINY_RCU kernels. 6103 6104 rcupdate.rcu_normal_after_boot= [KNL] 6105 Once boot has completed (that is, after 6106 rcu_end_inkernel_boot() has been invoked), use 6107 only normal grace-period primitives. No effect 6108 on CONFIG_TINY_RCU kernels. 6109 6110 But note that CONFIG_PREEMPT_RT=y kernels enables 6111 this kernel boot parameter, forcibly setting 6112 it to the value one, that is, converting any 6113 post-boot attempt at an expedited RCU grace 6114 period to instead use normal non-expedited 6115 grace-period processing. 6116 6117 rcupdate.rcu_task_collapse_lim= [KNL] 6118 Set the maximum number of callbacks present 6119 at the beginning of a grace period that allows 6120 the RCU Tasks flavors to collapse back to using 6121 a single callback queue. This switching only 6122 occurs when rcupdate.rcu_task_enqueue_lim is 6123 set to the default value of -1. 6124 6125 rcupdate.rcu_task_contend_lim= [KNL] 6126 Set the minimum number of callback-queuing-time 6127 lock-contention events per jiffy required to 6128 cause the RCU Tasks flavors to switch to per-CPU 6129 callback queuing. This switching only occurs 6130 when rcupdate.rcu_task_enqueue_lim is set to 6131 the default value of -1. 6132 6133 rcupdate.rcu_task_enqueue_lim= [KNL] 6134 Set the number of callback queues to use for the 6135 RCU Tasks family of RCU flavors. The default 6136 of -1 allows this to be automatically (and 6137 dynamically) adjusted. This parameter is intended 6138 for use in testing. 6139 6140 rcupdate.rcu_task_ipi_delay= [KNL] 6141 Set time in jiffies during which RCU tasks will 6142 avoid sending IPIs, starting with the beginning 6143 of a given grace period. Setting a large 6144 number avoids disturbing real-time workloads, 6145 but lengthens grace periods. 6146 6147 rcupdate.rcu_task_lazy_lim= [KNL] 6148 Number of callbacks on a given CPU that will 6149 cancel laziness on that CPU. Use -1 to disable 6150 cancellation of laziness, but be advised that 6151 doing so increases the danger of OOM due to 6152 callback flooding. 6153 6154 rcupdate.rcu_task_stall_info= [KNL] 6155 Set initial timeout in jiffies for RCU task stall 6156 informational messages, which give some indication 6157 of the problem for those not patient enough to 6158 wait for ten minutes. Informational messages are 6159 only printed prior to the stall-warning message 6160 for a given grace period. Disable with a value 6161 less than or equal to zero. Defaults to ten 6162 seconds. A change in value does not take effect 6163 until the beginning of the next grace period. 6164 6165 rcupdate.rcu_task_stall_info_mult= [KNL] 6166 Multiplier for time interval between successive 6167 RCU task stall informational messages for a given 6168 RCU tasks grace period. This value is clamped 6169 to one through ten, inclusive. It defaults to 6170 the value three, so that the first informational 6171 message is printed 10 seconds into the grace 6172 period, the second at 40 seconds, the third at 6173 160 seconds, and then the stall warning at 600 6174 seconds would prevent a fourth at 640 seconds. 6175 6176 rcupdate.rcu_task_stall_timeout= [KNL] 6177 Set timeout in jiffies for RCU task stall 6178 warning messages. Disable with a value less 6179 than or equal to zero. Defaults to ten minutes. 6180 A change in value does not take effect until 6181 the beginning of the next grace period. 6182 6183 rcupdate.rcu_tasks_lazy_ms= [KNL] 6184 Set timeout in milliseconds RCU Tasks asynchronous 6185 callback batching for call_rcu_tasks(). 6186 A negative value will take the default. A value 6187 of zero will disable batching. Batching is 6188 always disabled for synchronize_rcu_tasks(). 6189 6190 rcupdate.rcu_tasks_trace_lazy_ms= [KNL] 6191 Set timeout in milliseconds RCU Tasks 6192 Trace asynchronous callback batching for 6193 call_rcu_tasks_trace(). A negative value 6194 will take the default. A value of zero will 6195 disable batching. Batching is always disabled 6196 for synchronize_rcu_tasks_trace(). 6197 6198 rcupdate.rcu_self_test= [KNL] 6199 Run the RCU early boot self tests 6200 6201 rdinit= [KNL] 6202 Format: <full_path> 6203 Run specified binary instead of /init from the ramdisk, 6204 used for early userspace startup. See initrd. 6205 6206 rdrand= [X86,EARLY] 6207 force - Override the decision by the kernel to hide the 6208 advertisement of RDRAND support (this affects 6209 certain AMD processors because of buggy BIOS 6210 support, specifically around the suspend/resume 6211 path). 6212 6213 rdt= [HW,X86,RDT] 6214 Turn on/off individual RDT features. List is: 6215 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 6216 mba, smba, bmec, abmc. 6217 E.g. to turn on cmt and turn off mba use: 6218 rdt=cmt,!mba 6219 6220 reboot= [KNL] 6221 Format (x86 or x86_64): 6222 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \ 6223 [[,]s[mp]#### \ 6224 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 6225 [[,]f[orce] 6226 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 6227 (prefix with 'panic_' to set mode for panic 6228 reboot only), 6229 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 6230 reboot_force is either force or not specified, 6231 reboot_cpu is s[mp]#### with #### being the processor 6232 to be used for rebooting. 6233 6234 acpi 6235 Use the ACPI RESET_REG in the FADT. If ACPI is not 6236 configured or the ACPI reset does not work, the reboot 6237 path attempts the reset using the keyboard controller. 6238 6239 bios 6240 Use the CPU reboot vector for warm reset 6241 6242 cold 6243 Set the cold reboot flag 6244 6245 default 6246 There are some built-in platform specific "quirks" 6247 - you may see: "reboot: <name> series board detected. 6248 Selecting <type> for reboots." In the case where you 6249 think the quirk is in error (e.g. you have newer BIOS, 6250 or newer board) using this option will ignore the 6251 built-in quirk table, and use the generic default 6252 reboot actions. 6253 6254 efi 6255 Use efi reset_system runtime service. If EFI is not 6256 configured or the EFI reset does not work, the reboot 6257 path attempts the reset using the keyboard controller. 6258 6259 force 6260 Don't stop other CPUs on reboot. This can make reboot 6261 more reliable in some cases. 6262 6263 kbd 6264 Use the keyboard controller. cold reset (default) 6265 6266 pci 6267 Use a write to the PCI config space register 0xcf9 to 6268 trigger reboot. 6269 6270 triple 6271 Force a triple fault (init) 6272 6273 warm 6274 Don't set the cold reboot flag 6275 6276 Using warm reset will be much faster especially on big 6277 memory systems because the BIOS will not go through 6278 the memory check. Disadvantage is that not all 6279 hardware will be completely reinitialized on reboot so 6280 there may be boot problems on some systems. 6281 6282 6283 refscale.holdoff= [KNL] 6284 Set test-start holdoff period. The purpose of 6285 this parameter is to delay the start of the 6286 test until boot completes in order to avoid 6287 interference. 6288 6289 refscale.lookup_instances= [KNL] 6290 Number of data elements to use for the forms of 6291 SLAB_TYPESAFE_BY_RCU testing. A negative number 6292 is negated and multiplied by nr_cpu_ids, while 6293 zero specifies nr_cpu_ids. 6294 6295 refscale.loops= [KNL] 6296 Set the number of loops over the synchronization 6297 primitive under test. Increasing this number 6298 reduces noise due to loop start/end overhead, 6299 but the default has already reduced the per-pass 6300 noise to a handful of picoseconds on ca. 2020 6301 x86 laptops. 6302 6303 refscale.nreaders= [KNL] 6304 Set number of readers. The default value of -1 6305 selects N, where N is roughly 75% of the number 6306 of CPUs. A value of zero is an interesting choice. 6307 6308 refscale.nruns= [KNL] 6309 Set number of runs, each of which is dumped onto 6310 the console log. 6311 6312 refscale.readdelay= [KNL] 6313 Set the read-side critical-section duration, 6314 measured in microseconds. 6315 6316 refscale.scale_type= [KNL] 6317 Specify the read-protection implementation to test. 6318 6319 refscale.shutdown= [KNL] 6320 Shut down the system at the end of the performance 6321 test. This defaults to 1 (shut it down) when 6322 refscale is built into the kernel and to 0 (leave 6323 it running) when refscale is built as a module. 6324 6325 refscale.verbose= [KNL] 6326 Enable additional printk() statements. 6327 6328 refscale.verbose_batched= [KNL] 6329 Batch the additional printk() statements. If zero 6330 (the default) or negative, print everything. Otherwise, 6331 print every Nth verbose statement, where N is the value 6332 specified. 6333 6334 regulator_ignore_unused 6335 [REGULATOR] 6336 Prevents regulator framework from disabling regulators 6337 that are unused, due no driver claiming them. This may 6338 be useful for debug and development, but should not be 6339 needed on a platform with proper driver support. 6340 6341 relax_domain_level= 6342 [KNL, SMP] Set scheduler's default relax_domain_level. 6343 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 6344 6345 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 6346 Format: <base1>,<size1>[,<base2>,<size2>,...] 6347 Reserve I/O ports or memory so the kernel won't use 6348 them. If <base> is less than 0x10000, the region 6349 is assumed to be I/O ports; otherwise it is memory. 6350 6351 reserve_mem= [RAM] 6352 Format: nn[KMG]:<align>:<label> 6353 Reserve physical memory and label it with a name that 6354 other subsystems can use to access it. This is typically 6355 used for systems that do not wipe the RAM, and this command 6356 line will try to reserve the same physical memory on 6357 soft reboots. Note, it is not guaranteed to be the same 6358 location. For example, if anything about the system changes 6359 or if booting a different kernel. It can also fail if KASLR 6360 places the kernel at the location of where the RAM reservation 6361 was from a previous boot, the new reservation will be at a 6362 different location. 6363 Any subsystem using this feature must add a way to verify 6364 that the contents of the physical memory is from a previous 6365 boot, as there may be cases where the memory will not be 6366 located at the same location. 6367 6368 The format is size:align:label for example, to request 6369 12 megabytes of 4096 alignment for ramoops: 6370 6371 reserve_mem=12M:4096:oops ramoops.mem_name=oops 6372 6373 reservetop= [X86-32,EARLY] 6374 Format: nn[KMG] 6375 Reserves a hole at the top of the kernel virtual 6376 address space. 6377 6378 reset_devices [KNL] Force drivers to reset the underlying device 6379 during initialization. 6380 6381 resume= [SWSUSP] 6382 Specify the partition device for software suspend 6383 Format: 6384 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 6385 6386 resume_offset= [SWSUSP] 6387 Specify the offset from the beginning of the partition 6388 given by "resume=" at which the swap header is located, 6389 in <PAGE_SIZE> units (needed only for swap files). 6390 See Documentation/power/swsusp-and-swap-files.rst 6391 6392 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 6393 read the resume files 6394 6395 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 6396 Useful for devices that are detected asynchronously 6397 (e.g. USB and MMC devices). 6398 6399 retain_initrd [RAM] Keep initrd memory after extraction. After boot, it will 6400 be accessible via /sys/firmware/initrd. 6401 6402 retbleed= [X86] Control mitigation of RETBleed (Arbitrary 6403 Speculative Code Execution with Return Instructions) 6404 vulnerability. 6405 6406 AMD-based UNRET and IBPB mitigations alone do not stop 6407 sibling threads from influencing the predictions of other 6408 sibling threads. For that reason, STIBP is used on pro- 6409 cessors that support it, and mitigate SMT on processors 6410 that don't. 6411 6412 off - no mitigation 6413 auto - automatically select a migitation 6414 auto,nosmt - automatically select a mitigation, 6415 disabling SMT if necessary for 6416 the full mitigation (only on Zen1 6417 and older without STIBP). 6418 ibpb - On AMD, mitigate short speculation 6419 windows on basic block boundaries too. 6420 Safe, highest perf impact. It also 6421 enables STIBP if present. Not suitable 6422 on Intel. 6423 ibpb,nosmt - Like "ibpb" above but will disable SMT 6424 when STIBP is not available. This is 6425 the alternative for systems which do not 6426 have STIBP. 6427 unret - Force enable untrained return thunks, 6428 only effective on AMD f15h-f17h based 6429 systems. 6430 unret,nosmt - Like unret, but will disable SMT when STIBP 6431 is not available. This is the alternative for 6432 systems which do not have STIBP. 6433 6434 Selecting 'auto' will choose a mitigation method at run 6435 time according to the CPU. 6436 6437 Not specifying this option is equivalent to retbleed=auto. 6438 6439 rfkill.default_state= 6440 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 6441 etc. communication is blocked by default. 6442 1 Unblocked. 6443 6444 rfkill.master_switch_mode= 6445 0 The "airplane mode" button does nothing. 6446 1 The "airplane mode" button toggles between everything 6447 blocked and the previous configuration. 6448 2 The "airplane mode" button toggles between everything 6449 blocked and everything unblocked. 6450 6451 ring3mwait=disable 6452 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 6453 CPUs. 6454 6455 riscv_isa_fallback [RISCV,EARLY] 6456 When CONFIG_RISCV_ISA_FALLBACK is not enabled, permit 6457 falling back to detecting extension support by parsing 6458 "riscv,isa" property on devicetree systems when the 6459 replacement properties are not found. See the Kconfig 6460 entry for RISCV_ISA_FALLBACK. 6461 6462 ro [KNL] Mount root device read-only on boot 6463 6464 rodata= [KNL,EARLY] 6465 on Mark read-only kernel memory as read-only (default). 6466 off Leave read-only kernel memory writable for debugging. 6467 noalias Mark read-only kernel memory as read-only but retain 6468 writable aliases in the direct map for regions outside 6469 of the kernel image. [arm64] 6470 6471 rockchip.usb_uart 6472 [EARLY] 6473 Enable the uart passthrough on the designated usb port 6474 on Rockchip SoCs. When active, the signals of the 6475 debug-uart get routed to the D+ and D- pins of the usb 6476 port and the regular usb controller gets disabled. 6477 6478 root= [KNL] Root filesystem 6479 Usually this is a block device specifier of some kind, 6480 see the early_lookup_bdev comment in 6481 block/early-lookup.c for details. 6482 Alternatively this can be "ram" for the legacy initial 6483 ramdisk, "nfs" and "cifs" for root on a network file 6484 system, or "mtd" and "ubi" for mounting from raw flash. 6485 6486 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 6487 mount the root filesystem 6488 6489 rootflags= [KNL] Set root filesystem mount option string 6490 6491 initramfs_options= [KNL] 6492 Specify mount options for for the initramfs mount. 6493 6494 rootfstype= [KNL] Set root filesystem type 6495 6496 rootwait [KNL] Wait (indefinitely) for root device to show up. 6497 Useful for devices that are detected asynchronously 6498 (e.g. USB and MMC devices). 6499 6500 rootwait= [KNL] Maximum time (in seconds) to wait for root device 6501 to show up before attempting to mount the root 6502 filesystem. 6503 6504 rproc_mem=nn[KMG][@address] 6505 [KNL,ARM,CMA] Remoteproc physical memory block. 6506 Memory area to be used by remote processor image, 6507 managed by CMA. 6508 6509 rt_group_sched= [KNL] Enable or disable SCHED_RR/FIFO group scheduling 6510 when CONFIG_RT_GROUP_SCHED=y. Defaults to 6511 !CONFIG_RT_GROUP_SCHED_DEFAULT_DISABLED. 6512 Format: <bool> 6513 6514 rw [KNL] Mount root device read-write on boot 6515 6516 S [KNL] Run init in single mode 6517 6518 s390_iommu= [HW,S390] 6519 Set s390 IOTLB flushing mode 6520 strict 6521 With strict flushing every unmap operation will result 6522 in an IOTLB flush. Default is lazy flushing before 6523 reuse, which is faster. Deprecated, equivalent to 6524 iommu.strict=1. 6525 6526 s390_iommu_aperture= [KNL,S390] 6527 Specifies the size of the per device DMA address space 6528 accessible through the DMA and IOMMU APIs as a decimal 6529 factor of the size of main memory. 6530 The default is 1 meaning that one can concurrently use 6531 as many DMA addresses as physical memory is installed, 6532 if supported by hardware, and thus map all of memory 6533 once. With a value of 2 one can map all of memory twice 6534 and so on. As a special case a factor of 0 imposes no 6535 restrictions other than those given by hardware at the 6536 cost of significant additional memory use for tables. 6537 6538 sa1100ir [NET] 6539 See drivers/net/irda/sa1100_ir.c. 6540 6541 sched_proxy_exec= [KNL] 6542 Enables or disables "proxy execution" style 6543 solution to mutex-based priority inversion. 6544 Format: <bool> 6545 6546 sched_verbose [KNL,EARLY] Enables verbose scheduler debug messages. 6547 6548 schedstats= [KNL,X86] Enable or disable scheduled statistics. 6549 Allowed values are enable and disable. This feature 6550 incurs a small amount of overhead in the scheduler 6551 but is useful for debugging and performance tuning. 6552 6553 sched_thermal_decay_shift= 6554 [Deprecated] 6555 [KNL, SMP] Set a decay shift for scheduler thermal 6556 pressure signal. Thermal pressure signal follows the 6557 default decay period of other scheduler pelt 6558 signals(usually 32 ms but configurable). Setting 6559 sched_thermal_decay_shift will left shift the decay 6560 period for the thermal pressure signal by the shift 6561 value. 6562 i.e. with the default pelt decay period of 32 ms 6563 sched_thermal_decay_shift thermal pressure decay pr 6564 1 64 ms 6565 2 128 ms 6566 and so on. 6567 Format: integer between 0 and 10 6568 Default is 0. 6569 6570 scftorture.holdoff= [KNL] 6571 Number of seconds to hold off before starting 6572 test. Defaults to zero for module insertion and 6573 to 10 seconds for built-in smp_call_function() 6574 tests. 6575 6576 scftorture.longwait= [KNL] 6577 Request ridiculously long waits randomly selected 6578 up to the chosen limit in seconds. Zero (the 6579 default) disables this feature. Please note 6580 that requesting even small non-zero numbers of 6581 seconds can result in RCU CPU stall warnings, 6582 softlockup complaints, and so on. 6583 6584 scftorture.nthreads= [KNL] 6585 Number of kthreads to spawn to invoke the 6586 smp_call_function() family of functions. 6587 The default of -1 specifies a number of kthreads 6588 equal to the number of CPUs. 6589 6590 scftorture.onoff_holdoff= [KNL] 6591 Number seconds to wait after the start of the 6592 test before initiating CPU-hotplug operations. 6593 6594 scftorture.onoff_interval= [KNL] 6595 Number seconds to wait between successive 6596 CPU-hotplug operations. Specifying zero (which 6597 is the default) disables CPU-hotplug operations. 6598 6599 scftorture.shutdown_secs= [KNL] 6600 The number of seconds following the start of the 6601 test after which to shut down the system. The 6602 default of zero avoids shutting down the system. 6603 Non-zero values are useful for automated tests. 6604 6605 scftorture.stat_interval= [KNL] 6606 The number of seconds between outputting the 6607 current test statistics to the console. A value 6608 of zero disables statistics output. 6609 6610 scftorture.stutter_cpus= [KNL] 6611 The number of jiffies to wait between each change 6612 to the set of CPUs under test. 6613 6614 scftorture.use_cpus_read_lock= [KNL] 6615 Use use_cpus_read_lock() instead of the default 6616 preempt_disable() to disable CPU hotplug 6617 while invoking one of the smp_call_function*() 6618 functions. 6619 6620 scftorture.verbose= [KNL] 6621 Enable additional printk() statements. 6622 6623 scftorture.weight_single= [KNL] 6624 The probability weighting to use for the 6625 smp_call_function_single() function with a zero 6626 "wait" parameter. A value of -1 selects the 6627 default if all other weights are -1. However, 6628 if at least one weight has some other value, a 6629 value of -1 will instead select a weight of zero. 6630 6631 scftorture.weight_single_wait= [KNL] 6632 The probability weighting to use for the 6633 smp_call_function_single() function with a 6634 non-zero "wait" parameter. See weight_single. 6635 6636 scftorture.weight_many= [KNL] 6637 The probability weighting to use for the 6638 smp_call_function_many() function with a zero 6639 "wait" parameter. See weight_single. 6640 Note well that setting a high probability for 6641 this weighting can place serious IPI load 6642 on the system. 6643 6644 scftorture.weight_many_wait= [KNL] 6645 The probability weighting to use for the 6646 smp_call_function_many() function with a 6647 non-zero "wait" parameter. See weight_single 6648 and weight_many. 6649 6650 scftorture.weight_all= [KNL] 6651 The probability weighting to use for the 6652 smp_call_function_all() function with a zero 6653 "wait" parameter. See weight_single and 6654 weight_many. 6655 6656 scftorture.weight_all_wait= [KNL] 6657 The probability weighting to use for the 6658 smp_call_function_all() function with a 6659 non-zero "wait" parameter. See weight_single 6660 and weight_many. 6661 6662 sdw_mclk_divider=[SDW] 6663 Specify the MCLK divider for Intel SoundWire buses in 6664 case the BIOS does not provide the clock rate properly. 6665 6666 skew_tick= [KNL,EARLY] Offset the periodic timer tick per cpu to mitigate 6667 xtime_lock contention on larger systems, and/or RCU lock 6668 contention on all systems with CONFIG_MAXSMP set. 6669 Format: { "0" | "1" } 6670 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 6671 1 -- enable. 6672 Note: increases power consumption, thus should only be 6673 enabled if running jitter sensitive (HPC/RT) workloads. 6674 6675 security= [SECURITY] Choose a legacy "major" security module to 6676 enable at boot. This has been deprecated by the 6677 "lsm=" parameter. 6678 6679 selinux= [SELINUX] Disable or enable SELinux at boot time. 6680 Format: { "0" | "1" } 6681 See security/selinux/Kconfig help text. 6682 0 -- disable. 6683 1 -- enable. 6684 Default value is 1. 6685 6686 serialnumber [BUGS=X86-32] 6687 6688 sev=option[,option...] [X86-64] 6689 6690 debug 6691 Enable debug messages. 6692 6693 nosnp 6694 Do not enable SEV-SNP (applies to host/hypervisor 6695 only). Setting 'nosnp' avoids the RMP check overhead 6696 in memory accesses when users do not want to run 6697 SEV-SNP guests. 6698 6699 shapers= [NET] 6700 Maximal number of shapers. 6701 6702 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 6703 Limit apic dumping. The parameter defines the maximal 6704 number of local apics being dumped. Also it is possible 6705 to set it to "all" by meaning -- no limit here. 6706 Format: { 1 (default) | 2 | ... | all }. 6707 The parameter valid if only apic=debug or 6708 apic=verbose is specified. 6709 Example: apic=debug show_lapic=all 6710 6711 slab_debug[=options[,slabs][;[options[,slabs]]...] [MM] 6712 Enabling slab_debug allows one to determine the 6713 culprit if slab objects become corrupted. Enabling 6714 slab_debug can create guard zones around objects and 6715 may poison objects when not in use. Also tracks the 6716 last alloc / free. For more information see 6717 Documentation/admin-guide/mm/slab.rst. 6718 (slub_debug legacy name also accepted for now) 6719 6720 Using this option implies the "no_hash_pointers" 6721 option which can be undone by adding the 6722 "hash_pointers=always" option. 6723 6724 slab_max_order= [MM] 6725 Determines the maximum allowed order for slabs. 6726 A high setting may cause OOMs due to memory 6727 fragmentation. For more information see 6728 Documentation/admin-guide/mm/slab.rst. 6729 (slub_max_order legacy name also accepted for now) 6730 6731 slab_merge [MM] 6732 Enable merging of slabs with similar size when the 6733 kernel is built without CONFIG_SLAB_MERGE_DEFAULT. 6734 (slub_merge legacy name also accepted for now) 6735 6736 slab_min_objects= [MM] 6737 The minimum number of objects per slab. SLUB will 6738 increase the slab order up to slab_max_order to 6739 generate a sufficiently large slab able to contain 6740 the number of objects indicated. The higher the number 6741 of objects the smaller the overhead of tracking slabs 6742 and the less frequently locks need to be acquired. 6743 For more information see 6744 Documentation/admin-guide/mm/slab.rst. 6745 (slub_min_objects legacy name also accepted for now) 6746 6747 slab_min_order= [MM] 6748 Determines the minimum page order for slabs. Must be 6749 lower or equal to slab_max_order. For more information see 6750 Documentation/admin-guide/mm/slab.rst. 6751 (slub_min_order legacy name also accepted for now) 6752 6753 slab_nomerge [MM] 6754 Disable merging of slabs with similar size. May be 6755 necessary if there is some reason to distinguish 6756 allocs to different slabs, especially in hardened 6757 environments where the risk of heap overflows and 6758 layout control by attackers can usually be 6759 frustrated by disabling merging. This will reduce 6760 most of the exposure of a heap attack to a single 6761 cache (risks via metadata attacks are mostly 6762 unchanged). Debug options disable merging on their 6763 own. 6764 For more information see 6765 Documentation/admin-guide/mm/slab.rst. 6766 (slub_nomerge legacy name also accepted for now) 6767 6768 slab_strict_numa [MM] 6769 Support memory policies on a per object level 6770 in the slab allocator. The default is for memory 6771 policies to be applied at the folio level when 6772 a new folio is needed or a partial folio is 6773 retrieved from the lists. Increases overhead 6774 in the slab fastpaths but gains more accurate 6775 NUMA kernel object placement which helps with slow 6776 interconnects in NUMA systems. 6777 6778 slram= [HW,MTD] 6779 6780 smart2= [HW] 6781 Format: <io1>[,<io2>[,...,<io8>]] 6782 6783 smp.csd_lock_timeout= [KNL] 6784 Specify the period of time in milliseconds 6785 that smp_call_function() and friends will wait 6786 for a CPU to release the CSD lock. This is 6787 useful when diagnosing bugs involving CPUs 6788 disabling interrupts for extended periods 6789 of time. Defaults to 5,000 milliseconds, and 6790 setting a value of zero disables this feature. 6791 This feature may be more efficiently disabled 6792 using the csdlock_debug- kernel parameter. 6793 6794 smp.panic_on_ipistall= [KNL] 6795 If a csd_lock_timeout extends for more than 6796 the specified number of milliseconds, panic the 6797 system. By default, let CSD-lock acquisition 6798 take as long as they take. Specifying 300,000 6799 for this value provides a 5-minute timeout. 6800 6801 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 6802 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 6803 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 6804 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 6805 smsc-ircc2.ircc_irq= [HW] IRQ line 6806 smsc-ircc2.ircc_dma= [HW] DMA channel 6807 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 6808 0: Toshiba Satellite 1800 (GP data pin select) 6809 1: Fast pin select (default) 6810 2: ATC IRMode 6811 6812 smt= [KNL,MIPS,S390,EARLY] Set the maximum number of threads 6813 (logical CPUs) to use per physical CPU on systems 6814 capable of symmetric multithreading (SMT). Will 6815 be capped to the actual hardware limit. 6816 Format: <integer> 6817 Default: -1 (no limit) 6818 6819 softlockup_panic= 6820 [KNL] Should the soft-lockup detector generate panics. 6821 Format: 0 | 1 6822 6823 A value of 1 instructs the soft-lockup detector 6824 to panic the machine when a soft-lockup occurs. It is 6825 also controlled by the kernel.softlockup_panic sysctl 6826 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 6827 respective build-time switch to that functionality. 6828 6829 softlockup_all_cpu_backtrace= 6830 [KNL] Should the soft-lockup detector generate 6831 backtraces on all cpus. 6832 Format: 0 | 1 6833 6834 sonypi.*= [HW] Sony Programmable I/O Control Device driver 6835 See Documentation/admin-guide/laptops/sonypi.rst 6836 6837 spectre_bhi= [X86] Control mitigation of Branch History Injection 6838 (BHI) vulnerability. This setting affects the 6839 deployment of the HW BHI control and the SW BHB 6840 clearing sequence. 6841 6842 on - (default) Enable the HW or SW mitigation as 6843 needed. This protects the kernel from 6844 both syscalls and VMs. 6845 vmexit - On systems which don't have the HW mitigation 6846 available, enable the SW mitigation on vmexit 6847 ONLY. On such systems, the host kernel is 6848 protected from VM-originated BHI attacks, but 6849 may still be vulnerable to syscall attacks. 6850 off - Disable the mitigation. 6851 6852 spectre_v2= [X86,EARLY] Control mitigation of Spectre variant 2 6853 (indirect branch speculation) vulnerability. 6854 The default operation protects the kernel from 6855 user space attacks. 6856 6857 on - unconditionally enable, implies 6858 spectre_v2_user=on 6859 off - unconditionally disable, implies 6860 spectre_v2_user=off 6861 auto - kernel detects whether your CPU model is 6862 vulnerable 6863 6864 Selecting 'on' will, and 'auto' may, choose a 6865 mitigation method at run time according to the 6866 CPU, the available microcode, the setting of the 6867 CONFIG_MITIGATION_RETPOLINE configuration option, 6868 and the compiler with which the kernel was built. 6869 6870 Selecting 'on' will also enable the mitigation 6871 against user space to user space task attacks. 6872 Selecting specific mitigation does not force enable 6873 user mitigations. 6874 6875 Selecting 'off' will disable both the kernel and 6876 the user space protections. 6877 6878 Specific mitigations can also be selected manually: 6879 6880 retpoline - replace indirect branches 6881 retpoline,generic - Retpolines 6882 retpoline,lfence - LFENCE; indirect branch 6883 retpoline,amd - alias for retpoline,lfence 6884 eibrs - Enhanced/Auto IBRS 6885 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines 6886 eibrs,lfence - Enhanced/Auto IBRS + LFENCE 6887 ibrs - use IBRS to protect kernel 6888 6889 Not specifying this option is equivalent to 6890 spectre_v2=auto. 6891 6892 spectre_v2_user= 6893 [X86] Control mitigation of Spectre variant 2 6894 (indirect branch speculation) vulnerability between 6895 user space tasks 6896 6897 on - Unconditionally enable mitigations. Is 6898 enforced by spectre_v2=on 6899 6900 off - Unconditionally disable mitigations. Is 6901 enforced by spectre_v2=off 6902 6903 prctl - Indirect branch speculation is enabled, 6904 but mitigation can be enabled via prctl 6905 per thread. The mitigation control state 6906 is inherited on fork. 6907 6908 prctl,ibpb 6909 - Like "prctl" above, but only STIBP is 6910 controlled per thread. IBPB is issued 6911 always when switching between different user 6912 space processes. 6913 6914 seccomp 6915 - Same as "prctl" above, but all seccomp 6916 threads will enable the mitigation unless 6917 they explicitly opt out. 6918 6919 seccomp,ibpb 6920 - Like "seccomp" above, but only STIBP is 6921 controlled per thread. IBPB is issued 6922 always when switching between different 6923 user space processes. 6924 6925 auto - Kernel selects the mitigation depending on 6926 the available CPU features and vulnerability. 6927 6928 Default mitigation: "prctl" 6929 6930 Not specifying this option is equivalent to 6931 spectre_v2_user=auto. 6932 6933 spec_rstack_overflow= 6934 [X86,EARLY] Control RAS overflow mitigation on AMD Zen CPUs 6935 6936 off - Disable mitigation 6937 microcode - Enable microcode mitigation only 6938 safe-ret - Enable sw-only safe RET mitigation (default) 6939 ibpb - Enable mitigation by issuing IBPB on 6940 kernel entry 6941 ibpb-vmexit - Issue IBPB only on VMEXIT 6942 (cloud-specific mitigation) 6943 6944 spec_store_bypass_disable= 6945 [HW,EARLY] Control Speculative Store Bypass (SSB) Disable mitigation 6946 (Speculative Store Bypass vulnerability) 6947 6948 Certain CPUs are vulnerable to an exploit against a 6949 a common industry wide performance optimization known 6950 as "Speculative Store Bypass" in which recent stores 6951 to the same memory location may not be observed by 6952 later loads during speculative execution. The idea 6953 is that such stores are unlikely and that they can 6954 be detected prior to instruction retirement at the 6955 end of a particular speculation execution window. 6956 6957 In vulnerable processors, the speculatively forwarded 6958 store can be used in a cache side channel attack, for 6959 example to read memory to which the attacker does not 6960 directly have access (e.g. inside sandboxed code). 6961 6962 This parameter controls whether the Speculative Store 6963 Bypass optimization is used. 6964 6965 On x86 the options are: 6966 6967 on - Unconditionally disable Speculative Store Bypass 6968 off - Unconditionally enable Speculative Store Bypass 6969 auto - Kernel detects whether the CPU model contains an 6970 implementation of Speculative Store Bypass and 6971 picks the most appropriate mitigation. If the 6972 CPU is not vulnerable, "off" is selected. If the 6973 CPU is vulnerable the default mitigation is 6974 architecture and Kconfig dependent. See below. 6975 prctl - Control Speculative Store Bypass per thread 6976 via prctl. Speculative Store Bypass is enabled 6977 for a process by default. The state of the control 6978 is inherited on fork. 6979 seccomp - Same as "prctl" above, but all seccomp threads 6980 will disable SSB unless they explicitly opt out. 6981 6982 Default mitigations: 6983 X86: "prctl" 6984 6985 On powerpc the options are: 6986 6987 on,auto - On Power8 and Power9 insert a store-forwarding 6988 barrier on kernel entry and exit. On Power7 6989 perform a software flush on kernel entry and 6990 exit. 6991 off - No action. 6992 6993 Not specifying this option is equivalent to 6994 spec_store_bypass_disable=auto. 6995 6996 split_lock_detect= 6997 [X86] Enable split lock detection or bus lock detection 6998 6999 When enabled (and if hardware support is present), atomic 7000 instructions that access data across cache line 7001 boundaries will result in an alignment check exception 7002 for split lock detection or a debug exception for 7003 bus lock detection. 7004 7005 off - not enabled 7006 7007 warn - the kernel will emit rate-limited warnings 7008 about applications triggering the #AC 7009 exception or the #DB exception. This mode is 7010 the default on CPUs that support split lock 7011 detection or bus lock detection. Default 7012 behavior is by #AC if both features are 7013 enabled in hardware. 7014 7015 fatal - the kernel will send SIGBUS to applications 7016 that trigger the #AC exception or the #DB 7017 exception. Default behavior is by #AC if 7018 both features are enabled in hardware. 7019 7020 ratelimit:N - 7021 Set system wide rate limit to N bus locks 7022 per second for bus lock detection. 7023 0 < N <= 1000. 7024 7025 N/A for split lock detection. 7026 7027 7028 If an #AC exception is hit in the kernel or in 7029 firmware (i.e. not while executing in user mode) 7030 the kernel will oops in either "warn" or "fatal" 7031 mode. 7032 7033 #DB exception for bus lock is triggered only when 7034 CPL > 0. 7035 7036 srbds= [X86,INTEL,EARLY] 7037 Control the Special Register Buffer Data Sampling 7038 (SRBDS) mitigation. 7039 7040 Certain CPUs are vulnerable to an MDS-like 7041 exploit which can leak bits from the random 7042 number generator. 7043 7044 By default, this issue is mitigated by 7045 microcode. However, the microcode fix can cause 7046 the RDRAND and RDSEED instructions to become 7047 much slower. Among other effects, this will 7048 result in reduced throughput from /dev/urandom. 7049 7050 The microcode mitigation can be disabled with 7051 the following option: 7052 7053 off: Disable mitigation and remove 7054 performance impact to RDRAND and RDSEED 7055 7056 srcutree.big_cpu_lim [KNL] 7057 Specifies the number of CPUs constituting a 7058 large system, such that srcu_struct structures 7059 should immediately allocate an srcu_node array. 7060 This kernel-boot parameter defaults to 128, 7061 but takes effect only when the low-order four 7062 bits of srcutree.convert_to_big is equal to 3 7063 (decide at boot). 7064 7065 srcutree.convert_to_big [KNL] 7066 Specifies under what conditions an SRCU tree 7067 srcu_struct structure will be converted to big 7068 form, that is, with an rcu_node tree: 7069 7070 0: Never. 7071 1: At init_srcu_struct() time. 7072 2: When rcutorture decides to. 7073 3: Decide at boot time (default). 7074 0x1X: Above plus if high contention. 7075 7076 Either way, the srcu_node tree will be sized based 7077 on the actual runtime number of CPUs (nr_cpu_ids) 7078 instead of the compile-time CONFIG_NR_CPUS. 7079 7080 srcutree.counter_wrap_check [KNL] 7081 Specifies how frequently to check for 7082 grace-period sequence counter wrap for the 7083 srcu_data structure's ->srcu_gp_seq_needed field. 7084 The greater the number of bits set in this kernel 7085 parameter, the less frequently counter wrap will 7086 be checked for. Note that the bottom two bits 7087 are ignored. 7088 7089 srcutree.exp_holdoff [KNL] 7090 Specifies how many nanoseconds must elapse 7091 since the end of the last SRCU grace period for 7092 a given srcu_struct until the next normal SRCU 7093 grace period will be considered for automatic 7094 expediting. Set to zero to disable automatic 7095 expediting. 7096 7097 srcutree.srcu_max_nodelay [KNL] 7098 Specifies the number of no-delay instances 7099 per jiffy for which the SRCU grace period 7100 worker thread will be rescheduled with zero 7101 delay. Beyond this limit, worker thread will 7102 be rescheduled with a sleep delay of one jiffy. 7103 7104 srcutree.srcu_max_nodelay_phase [KNL] 7105 Specifies the per-grace-period phase, number of 7106 non-sleeping polls of readers. Beyond this limit, 7107 grace period worker thread will be rescheduled 7108 with a sleep delay of one jiffy, between each 7109 rescan of the readers, for a grace period phase. 7110 7111 srcutree.srcu_retry_check_delay [KNL] 7112 Specifies number of microseconds of non-sleeping 7113 delay between each non-sleeping poll of readers. 7114 7115 srcutree.small_contention_lim [KNL] 7116 Specifies the number of update-side contention 7117 events per jiffy will be tolerated before 7118 initiating a conversion of an srcu_struct 7119 structure to big form. Note that the value of 7120 srcutree.convert_to_big must have the 0x10 bit 7121 set for contention-based conversions to occur. 7122 7123 ssbd= [ARM64,HW,EARLY] 7124 Speculative Store Bypass Disable control 7125 7126 On CPUs that are vulnerable to the Speculative 7127 Store Bypass vulnerability and offer a 7128 firmware based mitigation, this parameter 7129 indicates how the mitigation should be used: 7130 7131 force-on: Unconditionally enable mitigation for 7132 for both kernel and userspace 7133 force-off: Unconditionally disable mitigation for 7134 for both kernel and userspace 7135 kernel: Always enable mitigation in the 7136 kernel, and offer a prctl interface 7137 to allow userspace to register its 7138 interest in being mitigated too. 7139 7140 stack_guard_gap= [MM] 7141 override the default stack gap protection. The value 7142 is in page units and it defines how many pages prior 7143 to (for stacks growing down) resp. after (for stacks 7144 growing up) the main stack are reserved for no other 7145 mapping. Default value is 256 pages. 7146 7147 stack_depot_disable= [KNL,EARLY] 7148 Setting this to true through kernel command line will 7149 disable the stack depot thereby saving the static memory 7150 consumed by the stack hash table. By default this is set 7151 to false. 7152 7153 stack_depot_max_pools= [KNL,EARLY] 7154 Specify the maximum number of pools to use for storing 7155 stack traces. Pools are allocated on-demand up to this 7156 limit. Default value is 8191 pools. 7157 7158 stacktrace [FTRACE] 7159 Enabled the stack tracer on boot up. 7160 7161 stacktrace_filter=[function-list] 7162 [FTRACE] Limit the functions that the stack tracer 7163 will trace at boot up. function-list is a comma-separated 7164 list of functions. This list can be changed at run 7165 time by the stack_trace_filter file in the debugfs 7166 tracing directory. Note, this enables stack tracing 7167 and the stacktrace above is not needed. 7168 7169 sti= [PARISC,HW] 7170 Format: <num> 7171 Set the STI (builtin display/keyboard on the HP-PARISC 7172 machines) console (graphic card) which should be used 7173 as the initial boot-console. 7174 See also comment in drivers/video/console/sticore.c. 7175 7176 sti_font= [HW] 7177 See comment in drivers/video/console/sticore.c. 7178 7179 stifb= [HW] 7180 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 7181 7182 strict_sas_size= 7183 [X86] 7184 Format: <bool> 7185 Enable or disable strict sigaltstack size checks 7186 against the required signal frame size which 7187 depends on the supported FPU features. This can 7188 be used to filter out binaries which have 7189 not yet been made aware of AT_MINSIGSTKSZ. 7190 7191 stress_hpt [PPC,EARLY] 7192 Limits the number of kernel HPT entries in the hash 7193 page table to increase the rate of hash page table 7194 faults on kernel addresses. 7195 7196 stress_slb [PPC,EARLY] 7197 Limits the number of kernel SLB entries, and flushes 7198 them frequently to increase the rate of SLB faults 7199 on kernel addresses. 7200 7201 sunrpc.min_resvport= 7202 sunrpc.max_resvport= 7203 [NFS,SUNRPC] 7204 SunRPC servers often require that client requests 7205 originate from a privileged port (i.e. a port in the 7206 range 0 < portnr < 1024). 7207 An administrator who wishes to reserve some of these 7208 ports for other uses may adjust the range that the 7209 kernel's sunrpc client considers to be privileged 7210 using these two parameters to set the minimum and 7211 maximum port values. 7212 7213 sunrpc.svc_rpc_per_connection_limit= 7214 [NFS,SUNRPC] 7215 Limit the number of requests that the server will 7216 process in parallel from a single connection. 7217 The default value is 0 (no limit). 7218 7219 sunrpc.pool_mode= 7220 [NFS] 7221 Control how the NFS server code allocates CPUs to 7222 service thread pools. Depending on how many NICs 7223 you have and where their interrupts are bound, this 7224 option will affect which CPUs will do NFS serving. 7225 Note: this parameter cannot be changed while the 7226 NFS server is running. 7227 7228 auto the server chooses an appropriate mode 7229 automatically using heuristics 7230 global a single global pool contains all CPUs 7231 percpu one pool for each CPU 7232 pernode one pool for each NUMA node (equivalent 7233 to global on non-NUMA machines) 7234 7235 sunrpc.tcp_slot_table_entries= 7236 sunrpc.udp_slot_table_entries= 7237 [NFS,SUNRPC] 7238 Sets the upper limit on the number of simultaneous 7239 RPC calls that can be sent from the client to a 7240 server. Increasing these values may allow you to 7241 improve throughput, but will also increase the 7242 amount of memory reserved for use by the client. 7243 7244 suspend.pm_test_delay= 7245 [SUSPEND] 7246 Sets the number of seconds to remain in a suspend test 7247 mode before resuming the system (see 7248 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 7249 is set. Default value is 5. 7250 7251 svm= [PPC] 7252 Format: { on | off | y | n | 1 | 0 } 7253 This parameter controls use of the Protected 7254 Execution Facility on pSeries. 7255 7256 swiotlb= [ARM,PPC,MIPS,X86,S390,EARLY] 7257 Format: { <int> [,<int>] | force | noforce } 7258 <int> -- Number of I/O TLB slabs 7259 <int> -- Second integer after comma. Number of swiotlb 7260 areas with their own lock. Will be rounded up 7261 to a power of 2. 7262 force -- force using of bounce buffers even if they 7263 wouldn't be automatically used by the kernel 7264 noforce -- Never use bounce buffers (for debugging) 7265 7266 switches= [HW,M68k,EARLY] 7267 7268 sysctl.*= [KNL] 7269 Set a sysctl parameter, right before loading the init 7270 process, as if the value was written to the respective 7271 /proc/sys/... file. Both '.' and '/' are recognized as 7272 separators. Unrecognized parameters and invalid values 7273 are reported in the kernel log. Sysctls registered 7274 later by a loaded module cannot be set this way. 7275 Example: sysctl.vm.swappiness=40 7276 7277 sysrq_always_enabled 7278 [KNL] 7279 Ignore sysrq setting - this boot parameter will 7280 neutralize any effect of /proc/sys/kernel/sysrq. 7281 Useful for debugging. 7282 7283 tcpmhash_entries= [KNL,NET] 7284 Set the number of tcp_metrics_hash slots. 7285 Default value is 8192 or 16384 depending on total 7286 ram pages. This is used to specify the TCP metrics 7287 cache size. See Documentation/networking/ip-sysctl.rst 7288 "tcp_no_metrics_save" section for more details. 7289 7290 tdfx= [HW,DRM] 7291 7292 test_suspend= [SUSPEND] 7293 Format: { "mem" | "standby" | "freeze" }[,N] 7294 Specify "mem" (for Suspend-to-RAM) or "standby" (for 7295 standby suspend) or "freeze" (for suspend type freeze) 7296 as the system sleep state during system startup with 7297 the optional capability to repeat N number of times. 7298 The system is woken from this state using a 7299 wakeup-capable RTC alarm. 7300 7301 thash_entries= [KNL,NET] 7302 Set number of hash buckets for TCP connection 7303 7304 thermal.act= [HW,ACPI] 7305 -1: disable all active trip points in all thermal zones 7306 <degrees C>: override all lowest active trip points 7307 7308 thermal.crt= [HW,ACPI] 7309 -1: disable all critical trip points in all thermal zones 7310 <degrees C>: override all critical trip points 7311 7312 thermal.off= [HW,ACPI] 7313 1: disable ACPI thermal control 7314 7315 thermal.psv= [HW,ACPI] 7316 -1: disable all passive trip points 7317 <degrees C>: override all passive trip points to this 7318 value 7319 7320 thermal.tzp= [HW,ACPI] 7321 Specify global default ACPI thermal zone polling rate 7322 <deci-seconds>: poll all this frequency 7323 0: no polling (default) 7324 7325 thp_anon= [KNL] 7326 Format: <size>[KMG],<size>[KMG]:<state>;<size>[KMG]-<size>[KMG]:<state> 7327 state is one of "always", "madvise", "never" or "inherit". 7328 Control the default behavior of the system with respect 7329 to anonymous transparent hugepages. 7330 Can be used multiple times for multiple anon THP sizes. 7331 See Documentation/admin-guide/mm/transhuge.rst for more 7332 details. 7333 7334 threadirqs [KNL,EARLY] 7335 Force threading of all interrupt handlers except those 7336 marked explicitly IRQF_NO_THREAD. 7337 7338 thp_shmem= [KNL] 7339 Format: <size>[KMG],<size>[KMG]:<policy>;<size>[KMG]-<size>[KMG]:<policy> 7340 Control the default policy of each hugepage size for the 7341 internal shmem mount. <policy> is one of policies available 7342 for the shmem mount ("always", "inherit", "never", "within_size", 7343 and "advise"). 7344 It can be used multiple times for multiple shmem THP sizes. 7345 See Documentation/admin-guide/mm/transhuge.rst for more 7346 details. 7347 7348 topology= [S390,EARLY] 7349 Format: {off | on} 7350 Specify if the kernel should make use of the cpu 7351 topology information if the hardware supports this. 7352 The scheduler will make use of this information and 7353 e.g. base its process migration decisions on it. 7354 Default is on. 7355 7356 torture.disable_onoff_at_boot= [KNL] 7357 Prevent the CPU-hotplug component of torturing 7358 until after init has spawned. 7359 7360 torture.ftrace_dump_at_shutdown= [KNL] 7361 Dump the ftrace buffer at torture-test shutdown, 7362 even if there were no errors. This can be a 7363 very costly operation when many torture tests 7364 are running concurrently, especially on systems 7365 with rotating-rust storage. 7366 7367 torture.verbose_sleep_frequency= [KNL] 7368 Specifies how many verbose printk()s should be 7369 emitted between each sleep. The default of zero 7370 disables verbose-printk() sleeping. 7371 7372 torture.verbose_sleep_duration= [KNL] 7373 Duration of each verbose-printk() sleep in jiffies. 7374 7375 tpm.disable_pcr_integrity= [HW,TPM] 7376 Do not protect PCR registers from unintended physical 7377 access, or interposers in the bus by the means of 7378 having an integrity protected session wrapped around 7379 TPM2_PCR_Extend command. Consider this in a situation 7380 where TPM is heavily utilized by IMA, thus protection 7381 causing a major performance hit, and the space where 7382 machines are deployed is by other means guarded. 7383 7384 tpm_crb_ffa.busy_timeout_ms= [ARM64,TPM] 7385 Maximum time in milliseconds to retry sending a message 7386 to the TPM service before giving up. This parameter controls 7387 how long the system will continue retrying when the TPM 7388 service is busy. 7389 Format: <unsigned int> 7390 Default: 2000 (2 seconds) 7391 7392 tpm_suspend_pcr=[HW,TPM] 7393 Format: integer pcr id 7394 Specify that at suspend time, the tpm driver 7395 should extend the specified pcr with zeros, 7396 as a workaround for some chips which fail to 7397 flush the last written pcr on TPM_SaveState. 7398 This will guarantee that all the other pcrs 7399 are saved. 7400 7401 tpm_tis.interrupts= [HW,TPM] 7402 Enable interrupts for the MMIO based physical layer 7403 for the FIFO interface. By default it is set to false 7404 (0). For more information about TPM hardware interfaces 7405 defined by Trusted Computing Group (TCG) see 7406 https://trustedcomputinggroup.org/resource/pc-client-platform-tpm-profile-ptp-specification/ 7407 7408 tp_printk [FTRACE] 7409 Have the tracepoints sent to printk as well as the 7410 tracing ring buffer. This is useful for early boot up 7411 where the system hangs or reboots and does not give the 7412 option for reading the tracing buffer or performing a 7413 ftrace_dump_on_oops. 7414 7415 To turn off having tracepoints sent to printk, 7416 echo 0 > /proc/sys/kernel/tracepoint_printk 7417 Note, echoing 1 into this file without the 7418 tp_printk kernel cmdline option has no effect. 7419 7420 The tp_printk_stop_on_boot (see below) can also be used 7421 to stop the printing of events to console at 7422 late_initcall_sync. 7423 7424 ** CAUTION ** 7425 7426 Having tracepoints sent to printk() and activating high 7427 frequency tracepoints such as irq or sched, can cause 7428 the system to live lock. 7429 7430 tp_printk_stop_on_boot [FTRACE] 7431 When tp_printk (above) is set, it can cause a lot of noise 7432 on the console. It may be useful to only include the 7433 printing of events during boot up, as user space may 7434 make the system inoperable. 7435 7436 This command line option will stop the printing of events 7437 to console at the late_initcall_sync() time frame. 7438 7439 trace_buf_size=nn[KMG] 7440 [FTRACE] will set tracing buffer size on each cpu. 7441 7442 trace_clock= [FTRACE] Set the clock used for tracing events 7443 at boot up. 7444 local - Use the per CPU time stamp counter 7445 (converted into nanoseconds). Fast, but 7446 depending on the architecture, may not be 7447 in sync between CPUs. 7448 global - Event time stamps are synchronized across 7449 CPUs. May be slower than the local clock, 7450 but better for some race conditions. 7451 counter - Simple counting of events (1, 2, ..) 7452 note, some counts may be skipped due to the 7453 infrastructure grabbing the clock more than 7454 once per event. 7455 uptime - Use jiffies as the time stamp. 7456 perf - Use the same clock that perf uses. 7457 mono - Use ktime_get_mono_fast_ns() for time stamps. 7458 mono_raw - Use ktime_get_raw_fast_ns() for time 7459 stamps. 7460 boot - Use ktime_get_boot_fast_ns() for time stamps. 7461 Architectures may add more clocks. See 7462 Documentation/trace/ftrace.rst for more details. 7463 7464 trace_event=[event-list] 7465 [FTRACE] Set and start specified trace events in order 7466 to facilitate early boot debugging. The event-list is a 7467 comma-separated list of trace events to enable. See 7468 also Documentation/trace/events.rst 7469 7470 To enable modules, use :mod: keyword: 7471 7472 trace_event=:mod:<module> 7473 7474 The value before :mod: will only enable specific events 7475 that are part of the module. See the above mentioned 7476 document for more information. 7477 7478 trace_instance=[instance-info] 7479 [FTRACE] Create a ring buffer instance early in boot up. 7480 This will be listed in: 7481 7482 /sys/kernel/tracing/instances 7483 7484 Events can be enabled at the time the instance is created 7485 via: 7486 7487 trace_instance=<name>,<system1>:<event1>,<system2>:<event2> 7488 7489 Note, the "<system*>:" portion is optional if the event is 7490 unique. 7491 7492 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall 7493 7494 will enable the "sched_switch" event (note, the "sched:" is optional, and 7495 the same thing would happen if it was left off). The irq_handler_entry 7496 event, and all events under the "initcall" system. 7497 7498 Flags can be added to the instance to modify its behavior when it is 7499 created. The flags are separated by '^'. 7500 7501 The available flags are: 7502 7503 traceoff - Have the tracing instance tracing disabled after it is created. 7504 traceprintk - Have trace_printk() write into this trace instance 7505 (note, "printk" and "trace_printk" can also be used) 7506 7507 trace_instance=foo^traceoff^traceprintk,sched,irq 7508 7509 The flags must come before the defined events. 7510 7511 If memory has been reserved (see memmap for x86), the instance 7512 can use that memory: 7513 7514 memmap=12M$0x284500000 trace_instance=boot_map@0x284500000:12M 7515 7516 The above will create a "boot_map" instance that uses the physical 7517 memory at 0x284500000 that is 12Megs. The per CPU buffers of that 7518 instance will be split up accordingly. 7519 7520 Alternatively, the memory can be reserved by the reserve_mem option: 7521 7522 reserve_mem=12M:4096:trace trace_instance=boot_map@trace 7523 7524 This will reserve 12 megabytes at boot up with a 4096 byte alignment 7525 and place the ring buffer in this memory. Note that due to KASLR, the 7526 memory may not be the same location each time, which will not preserve 7527 the buffer content. 7528 7529 Also note that the layout of the ring buffer data may change between 7530 kernel versions where the validator will fail and reset the ring buffer 7531 if the layout is not the same as the previous kernel. 7532 7533 If the ring buffer is used for persistent bootups and has events enabled, 7534 it is recommend to disable tracing so that events from a previous boot do not 7535 mix with events of the current boot (unless you are debugging a random crash 7536 at boot up). 7537 7538 reserve_mem=12M:4096:trace trace_instance=boot_map^traceoff^traceprintk@trace,sched,irq 7539 7540 Note, saving the trace buffer across reboots does require that the system 7541 is set up to not wipe memory. For instance, CONFIG_RESET_ATTACK_MITIGATION 7542 can force a memory reset on boot which will clear any trace that was stored. 7543 This is just one of many ways that can clear memory. Make sure your system 7544 keeps the content of memory across reboots before relying on this option. 7545 7546 NB: Both the mapped address and size must be page aligned for the architecture. 7547 7548 See also Documentation/trace/debugging.rst 7549 7550 7551 trace_options=[option-list] 7552 [FTRACE] Enable or disable tracer options at boot. 7553 The option-list is a comma delimited list of options 7554 that can be enabled or disabled just as if you were 7555 to echo the option name into 7556 7557 /sys/kernel/tracing/trace_options 7558 7559 For example, to enable stacktrace option (to dump the 7560 stack trace of each event), add to the command line: 7561 7562 trace_options=stacktrace 7563 7564 See also Documentation/trace/ftrace.rst "trace options" 7565 section. 7566 7567 trace_trigger=[trigger-list] 7568 [FTRACE] Add an event trigger on specific events. 7569 Set a trigger on top of a specific event, with an optional 7570 filter. 7571 7572 The format is "trace_trigger=<event>.<trigger>[ if <filter>],..." 7573 Where more than one trigger may be specified that are comma delimited. 7574 7575 For example: 7576 7577 trace_trigger="sched_switch.stacktrace if prev_state == 2" 7578 7579 The above will enable the "stacktrace" trigger on the "sched_switch" 7580 event but only trigger it if the "prev_state" of the "sched_switch" 7581 event is "2" (TASK_UNINTERRUPTIBLE). 7582 7583 See also "Event triggers" in Documentation/trace/events.rst 7584 7585 7586 traceoff_after_boot 7587 [FTRACE] Sometimes tracing is used to debug issues 7588 during the boot process. Since the trace buffer has a 7589 limited amount of storage, it may be prudent to 7590 disable tracing after the boot is finished, otherwise 7591 the critical information may be overwritten. With this 7592 option, the main tracing buffer will be turned off at 7593 the end of the boot process. 7594 7595 traceoff_on_warning 7596 [FTRACE] enable this option to disable tracing when a 7597 warning is hit. This turns off "tracing_on". Tracing can 7598 be enabled again by echoing '1' into the "tracing_on" 7599 file located in /sys/kernel/tracing/ 7600 7601 This option is useful, as it disables the trace before 7602 the WARNING dump is called, which prevents the trace to 7603 be filled with content caused by the warning output. 7604 7605 This option can also be set at run time via the sysctl 7606 option: kernel/traceoff_on_warning 7607 7608 transparent_hugepage= 7609 [KNL] 7610 Format: [always|madvise|never] 7611 Can be used to control the default behavior of the system 7612 with respect to transparent hugepages. 7613 See Documentation/admin-guide/mm/transhuge.rst 7614 for more details. 7615 7616 transparent_hugepage_shmem= [KNL] 7617 Format: [always|within_size|advise|never|deny|force] 7618 Can be used to control the hugepage allocation policy for 7619 the internal shmem mount. 7620 See Documentation/admin-guide/mm/transhuge.rst 7621 for more details. 7622 7623 transparent_hugepage_tmpfs= [KNL] 7624 Format: [always|within_size|advise|never] 7625 Can be used to control the default hugepage allocation policy 7626 for the tmpfs mount. 7627 See Documentation/admin-guide/mm/transhuge.rst 7628 for more details. 7629 7630 trusted.source= [KEYS] 7631 Format: <string> 7632 This parameter identifies the trust source as a backend 7633 for trusted keys implementation. Supported trust 7634 sources: 7635 - "tpm" 7636 - "tee" 7637 - "caam" 7638 - "dcp" 7639 If not specified then it defaults to iterating through 7640 the trust source list starting with TPM and assigns the 7641 first trust source as a backend which is initialized 7642 successfully during iteration. 7643 7644 trusted.rng= [KEYS] 7645 Format: <string> 7646 The RNG used to generate key material for trusted keys. 7647 Can be one of: 7648 - "kernel" 7649 - the same value as trusted.source: "tpm" or "tee" 7650 - "default" 7651 If not specified, "default" is used. In this case, 7652 the RNG's choice is left to each individual trust source. 7653 7654 trusted.dcp_use_otp_key 7655 This is intended to be used in combination with 7656 trusted.source=dcp and will select the DCP OTP key 7657 instead of the DCP UNIQUE key blob encryption. 7658 7659 trusted.dcp_skip_zk_test 7660 This is intended to be used in combination with 7661 trusted.source=dcp and will disable the check if the 7662 blob key is all zeros. This is helpful for situations where 7663 having this key zero'ed is acceptable. E.g. in testing 7664 scenarios. 7665 7666 tsa= [X86] Control mitigation for Transient Scheduler 7667 Attacks on AMD CPUs. Search the following in your 7668 favourite search engine for more details: 7669 7670 "Technical guidance for mitigating transient scheduler 7671 attacks". 7672 7673 off - disable the mitigation 7674 on - enable the mitigation (default) 7675 user - mitigate only user/kernel transitions 7676 vm - mitigate only guest/host transitions 7677 7678 7679 tsc= Disable clocksource stability checks for TSC. 7680 Format: <string> 7681 [x86] reliable: mark tsc clocksource as reliable, this 7682 disables clocksource verification at runtime, as well 7683 as the stability checks done at bootup. Used to enable 7684 high-resolution timer mode on older hardware, and in 7685 virtualized environment. 7686 [x86] noirqtime: Do not use TSC to do irq accounting. 7687 Used to run time disable IRQ_TIME_ACCOUNTING on any 7688 platforms where RDTSC is slow and this accounting 7689 can add overhead. 7690 [x86] unstable: mark the TSC clocksource as unstable, this 7691 marks the TSC unconditionally unstable at bootup and 7692 avoids any further wobbles once the TSC watchdog notices. 7693 [x86] nowatchdog: disable clocksource watchdog. Used 7694 in situations with strict latency requirements (where 7695 interruptions from clocksource watchdog are not 7696 acceptable). 7697 [x86] recalibrate: force recalibration against a HW timer 7698 (HPET or PM timer) on systems whose TSC frequency was 7699 obtained from HW or FW using either an MSR or CPUID(0x15). 7700 Warn if the difference is more than 500 ppm. 7701 [x86] watchdog: Use TSC as the watchdog clocksource with 7702 which to check other HW timers (HPET or PM timer), but 7703 only on systems where TSC has been deemed trustworthy. 7704 This will be suppressed by an earlier tsc=nowatchdog and 7705 can be overridden by a later tsc=nowatchdog. A console 7706 message will flag any such suppression or overriding. 7707 7708 tsc_early_khz= [X86,EARLY] Skip early TSC calibration and use the given 7709 value instead. Useful when the early TSC frequency discovery 7710 procedure is not reliable, such as on overclocked systems 7711 with CPUID.16h support and partial CPUID.15h support. 7712 Format: <unsigned int> 7713 7714 tsx= [X86] Control Transactional Synchronization 7715 Extensions (TSX) feature in Intel processors that 7716 support TSX control. 7717 7718 This parameter controls the TSX feature. The options are: 7719 7720 on - Enable TSX on the system. Although there are 7721 mitigations for all known security vulnerabilities, 7722 TSX has been known to be an accelerator for 7723 several previous speculation-related CVEs, and 7724 so there may be unknown security risks associated 7725 with leaving it enabled. 7726 7727 off - Disable TSX on the system. (Note that this 7728 option takes effect only on newer CPUs which are 7729 not vulnerable to MDS, i.e., have 7730 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 7731 the new IA32_TSX_CTRL MSR through a microcode 7732 update. This new MSR allows for the reliable 7733 deactivation of the TSX functionality.) 7734 7735 auto - Disable TSX if X86_BUG_TAA is present, 7736 otherwise enable TSX on the system. 7737 7738 Not specifying this option is equivalent to tsx=off. 7739 7740 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 7741 for more details. 7742 7743 tsx_async_abort= [X86,INTEL,EARLY] Control mitigation for the TSX Async 7744 Abort (TAA) vulnerability. 7745 7746 Similar to Micro-architectural Data Sampling (MDS) 7747 certain CPUs that support Transactional 7748 Synchronization Extensions (TSX) are vulnerable to an 7749 exploit against CPU internal buffers which can forward 7750 information to a disclosure gadget under certain 7751 conditions. 7752 7753 In vulnerable processors, the speculatively forwarded 7754 data can be used in a cache side channel attack, to 7755 access data to which the attacker does not have direct 7756 access. 7757 7758 This parameter controls the TAA mitigation. The 7759 options are: 7760 7761 full - Enable TAA mitigation on vulnerable CPUs 7762 if TSX is enabled. 7763 7764 full,nosmt - Enable TAA mitigation and disable SMT on 7765 vulnerable CPUs. If TSX is disabled, SMT 7766 is not disabled because CPU is not 7767 vulnerable to cross-thread TAA attacks. 7768 off - Unconditionally disable TAA mitigation 7769 7770 On MDS-affected machines, tsx_async_abort=off can be 7771 prevented by an active MDS mitigation as both vulnerabilities 7772 are mitigated with the same mechanism so in order to disable 7773 this mitigation, you need to specify mds=off too. 7774 7775 Not specifying this option is equivalent to 7776 tsx_async_abort=full. On CPUs which are MDS affected 7777 and deploy MDS mitigation, TAA mitigation is not 7778 required and doesn't provide any additional 7779 mitigation. 7780 7781 For details see: 7782 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 7783 7784 turbografx.map[2|3]= [HW,JOY] 7785 TurboGraFX parallel port interface 7786 Format: 7787 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 7788 See also Documentation/input/devices/joystick-parport.rst 7789 7790 udbg-immortal [PPC] When debugging early kernel crashes that 7791 happen after console_init() and before a proper 7792 console driver takes over, this boot options might 7793 help "seeing" what's going on. 7794 7795 uhash_entries= [KNL,NET] 7796 Set number of hash buckets for UDP/UDP-Lite connections 7797 7798 uhci-hcd.ignore_oc= 7799 [USB] Ignore overcurrent events (default N). 7800 Some badly-designed motherboards generate lots of 7801 bogus events, for ports that aren't wired to 7802 anything. Set this parameter to avoid log spamming. 7803 Note that genuine overcurrent events won't be 7804 reported either. 7805 7806 unaligned_scalar_speed= 7807 [RISCV] 7808 Format: {slow | fast | unsupported} 7809 Allow skipping scalar unaligned access speed tests. This 7810 is useful for testing alternative code paths and to skip 7811 the tests in environments where they run too slowly. All 7812 CPUs must have the same scalar unaligned access speed. 7813 7814 unaligned_vector_speed= 7815 [RISCV] 7816 Format: {slow | fast | unsupported} 7817 Allow skipping vector unaligned access speed tests. This 7818 is useful for testing alternative code paths and to skip 7819 the tests in environments where they run too slowly. All 7820 CPUs must have the same vector unaligned access speed. 7821 7822 unknown_nmi_panic 7823 [X86] Cause panic on unknown NMI. 7824 7825 unwind_debug [X86-64,EARLY] 7826 Enable unwinder debug output. This can be 7827 useful for debugging certain unwinder error 7828 conditions, including corrupt stacks and 7829 bad/missing unwinder metadata. 7830 7831 usbcore.authorized_default= 7832 [USB] Default USB device authorization: 7833 (default -1 = authorized (same as 1), 7834 0 = not authorized, 1 = authorized, 2 = authorized 7835 if device connected to internal port) 7836 7837 usbcore.autosuspend= 7838 [USB] The autosuspend time delay (in seconds) used 7839 for newly-detected USB devices (default 2). This 7840 is the time required before an idle device will be 7841 autosuspended. Devices for which the delay is set 7842 to a negative value won't be autosuspended at all. 7843 7844 usbcore.usbfs_snoop= 7845 [USB] Set to log all usbfs traffic (default 0 = off). 7846 7847 usbcore.usbfs_snoop_max= 7848 [USB] Maximum number of bytes to snoop in each URB 7849 (default = 65536). 7850 7851 usbcore.blinkenlights= 7852 [USB] Set to cycle leds on hubs (default 0 = off). 7853 7854 usbcore.old_scheme_first= 7855 [USB] Start with the old device initialization 7856 scheme (default 0 = off). 7857 7858 usbcore.usbfs_memory_mb= 7859 [USB] Memory limit (in MB) for buffers allocated by 7860 usbfs (default = 16, 0 = max = 2047). 7861 7862 usbcore.use_both_schemes= 7863 [USB] Try the other device initialization scheme 7864 if the first one fails (default 1 = enabled). 7865 7866 usbcore.initial_descriptor_timeout= 7867 [USB] Specifies timeout for the initial 64-byte 7868 USB_REQ_GET_DESCRIPTOR request in milliseconds 7869 (default 5000 = 5.0 seconds). 7870 7871 usbcore.nousb [USB] Disable the USB subsystem 7872 7873 usbcore.quirks= 7874 [USB] A list of quirk entries to augment the built-in 7875 usb core quirk list. List entries are separated by 7876 commas. Each entry has the form 7877 VendorID:ProductID:Flags. The IDs are 4-digit hex 7878 numbers and Flags is a set of letters. Each letter 7879 will change the built-in quirk; setting it if it is 7880 clear and clearing it if it is set. The letters have 7881 the following meanings: 7882 a = USB_QUIRK_STRING_FETCH_255 (string 7883 descriptors must not be fetched using 7884 a 255-byte read); 7885 b = USB_QUIRK_RESET_RESUME (device can't resume 7886 correctly so reset it instead); 7887 c = USB_QUIRK_NO_SET_INTF (device can't handle 7888 Set-Interface requests); 7889 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 7890 handle its Configuration or Interface 7891 strings); 7892 e = USB_QUIRK_RESET (device can't be reset 7893 (e.g morph devices), don't use reset); 7894 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 7895 more interface descriptions than the 7896 bNumInterfaces count, and can't handle 7897 talking to these interfaces); 7898 g = USB_QUIRK_DELAY_INIT (device needs a pause 7899 during initialization, after we read 7900 the device descriptor); 7901 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 7902 high speed and super speed interrupt 7903 endpoints, the USB 2.0 and USB 3.0 spec 7904 require the interval in microframes (1 7905 microframe = 125 microseconds) to be 7906 calculated as interval = 2 ^ 7907 (bInterval-1). 7908 Devices with this quirk report their 7909 bInterval as the result of this 7910 calculation instead of the exponent 7911 variable used in the calculation); 7912 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 7913 handle device_qualifier descriptor 7914 requests); 7915 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 7916 generates spurious wakeup, ignore 7917 remote wakeup capability); 7918 k = USB_QUIRK_NO_LPM (device can't handle Link 7919 Power Management); 7920 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 7921 (Device reports its bInterval as linear 7922 frames instead of the USB 2.0 7923 calculation); 7924 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 7925 to be disconnected before suspend to 7926 prevent spurious wakeup); 7927 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 7928 pause after every control message); 7929 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 7930 delay after resetting its port); 7931 p = USB_QUIRK_SHORT_SET_ADDRESS_REQ_TIMEOUT 7932 (Reduce timeout of the SET_ADDRESS 7933 request from 5000 ms to 500 ms); 7934 Example: quirks=0781:5580:bk,0a5c:5834:gij 7935 7936 usbhid.mousepoll= 7937 [USBHID] The interval which mice are to be polled at. 7938 7939 usbhid.jspoll= 7940 [USBHID] The interval which joysticks are to be polled at. 7941 7942 usbhid.kbpoll= 7943 [USBHID] The interval which keyboards are to be polled at. 7944 7945 usb-storage.delay_use= 7946 [UMS] The delay in seconds before a new device is 7947 scanned for Logical Units (default 1). 7948 Optionally the delay in milliseconds if the value has 7949 suffix with "ms". 7950 Example: delay_use=2567ms 7951 7952 usb-storage.quirks= 7953 [UMS] A list of quirks entries to supplement or 7954 override the built-in unusual_devs list. List 7955 entries are separated by commas. Each entry has 7956 the form VID:PID:Flags where VID and PID are Vendor 7957 and Product ID values (4-digit hex numbers) and 7958 Flags is a set of characters, each corresponding 7959 to a common usb-storage quirk flag as follows: 7960 a = SANE_SENSE (collect more than 18 bytes 7961 of sense data, not on uas); 7962 b = BAD_SENSE (don't collect more than 18 7963 bytes of sense data, not on uas); 7964 c = FIX_CAPACITY (decrease the reported 7965 device capacity by one sector); 7966 d = NO_READ_DISC_INFO (don't use 7967 READ_DISC_INFO command, not on uas); 7968 e = NO_READ_CAPACITY_16 (don't use 7969 READ_CAPACITY_16 command); 7970 f = NO_REPORT_OPCODES (don't use report opcodes 7971 command, uas only); 7972 g = MAX_SECTORS_240 (don't transfer more than 7973 240 sectors at a time, uas only); 7974 h = CAPACITY_HEURISTICS (decrease the 7975 reported device capacity by one 7976 sector if the number is odd); 7977 i = IGNORE_DEVICE (don't bind to this 7978 device); 7979 j = NO_REPORT_LUNS (don't use report luns 7980 command, uas only); 7981 k = NO_SAME (do not use WRITE_SAME, uas only) 7982 l = NOT_LOCKABLE (don't try to lock and 7983 unlock ejectable media, not on uas); 7984 m = MAX_SECTORS_64 (don't transfer more 7985 than 64 sectors = 32 KB at a time, 7986 not on uas); 7987 n = INITIAL_READ10 (force a retry of the 7988 initial READ(10) command, not on uas); 7989 o = CAPACITY_OK (accept the capacity 7990 reported by the device, not on uas); 7991 p = WRITE_CACHE (the device cache is ON 7992 by default, not on uas); 7993 r = IGNORE_RESIDUE (the device reports 7994 bogus residue values, not on uas); 7995 s = SINGLE_LUN (the device has only one 7996 Logical Unit); 7997 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 7998 commands, uas only); 7999 u = IGNORE_UAS (don't bind to the uas driver); 8000 w = NO_WP_DETECT (don't test whether the 8001 medium is write-protected). 8002 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 8003 even if the device claims no cache, 8004 not on uas) 8005 Example: quirks=0419:aaf5:rl,0421:0433:rc 8006 8007 user_debug= [KNL,ARM] 8008 Format: <int> 8009 See arch/arm/Kconfig.debug help text. 8010 1 - undefined instruction events 8011 2 - system calls 8012 4 - invalid data aborts 8013 8 - SIGSEGV faults 8014 16 - SIGBUS faults 8015 Example: user_debug=31 8016 8017 vdso= [X86,SH,SPARC] 8018 On X86_32, this is an alias for vdso32=. Otherwise: 8019 8020 vdso=1: enable VDSO (the default) 8021 vdso=0: disable VDSO mapping 8022 8023 vdso32= [X86] Control the 32-bit vDSO 8024 vdso32=1: enable 32-bit VDSO 8025 vdso32=0 or vdso32=2: disable 32-bit VDSO 8026 8027 See the help text for CONFIG_COMPAT_VDSO for more 8028 details. If CONFIG_COMPAT_VDSO is set, the default is 8029 vdso32=0; otherwise, the default is vdso32=1. 8030 8031 For compatibility with older kernels, vdso32=2 is an 8032 alias for vdso32=0. 8033 8034 Try vdso32=0 if you encounter an error that says: 8035 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 8036 8037 video= [FB,EARLY] Frame buffer configuration 8038 See Documentation/fb/modedb.rst. 8039 8040 video.brightness_switch_enabled= [ACPI] 8041 Format: [0|1] 8042 If set to 1, on receiving an ACPI notify event 8043 generated by hotkey, video driver will adjust brightness 8044 level and then send out the event to user space through 8045 the allocated input device. If set to 0, video driver 8046 will only send out the event without touching backlight 8047 brightness level. 8048 default: 1 8049 8050 virtio_mmio.device= 8051 [VMMIO] Memory mapped virtio (platform) device. 8052 8053 <size>@<baseaddr>:<irq>[:<id>] 8054 where: 8055 <size> := size (can use standard suffixes 8056 like K, M and G) 8057 <baseaddr> := physical base address 8058 <irq> := interrupt number (as passed to 8059 request_irq()) 8060 <id> := (optional) platform device id 8061 example: 8062 virtio_mmio.device=1K@0x100b0000:48:7 8063 8064 Can be used multiple times for multiple devices. 8065 8066 vga= [BOOT,X86-32] Select a particular video mode 8067 See Documentation/arch/x86/boot.rst and 8068 Documentation/admin-guide/svga.rst. 8069 Use vga=ask for menu. 8070 This is actually a boot loader parameter; the value is 8071 passed to the kernel using a special protocol. 8072 8073 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 8074 May slow down system boot speed, especially when 8075 enabled on systems with a large amount of memory. 8076 All options are enabled by default, and this 8077 interface is meant to allow for selectively 8078 enabling or disabling specific virtual memory 8079 debugging features. 8080 8081 Available options are: 8082 P Enable page structure init time poisoning 8083 - Disable all of the above options 8084 8085 vmalloc=nn[KMG] [KNL,BOOT,EARLY] Forces the vmalloc area to have an 8086 exact size of <nn>. This can be used to increase 8087 the minimum size (128MB on x86, arm32 platforms). 8088 It can also be used to decrease the size and leave more room 8089 for directly mapped kernel RAM. Note that this parameter does 8090 not exist on many other platforms (including arm64, alpha, 8091 loongarch, arc, csky, hexagon, microblaze, mips, nios2, openrisc, 8092 parisc, m64k, powerpc, riscv, sh, um, xtensa, s390, sparc). 8093 8094 vmcp_cma=nn[MG] [KNL,S390,EARLY] 8095 Sets the memory size reserved for contiguous memory 8096 allocations for the vmcp device driver. 8097 8098 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 8099 Format: <command> 8100 8101 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 8102 Format: <command> 8103 8104 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 8105 Format: <command> 8106 8107 vmscape= [X86] Controls mitigation for VMscape attacks. 8108 VMscape attacks can leak information from a userspace 8109 hypervisor to a guest via speculative side-channels. 8110 8111 off - disable the mitigation 8112 ibpb - use Indirect Branch Prediction Barrier 8113 (IBPB) mitigation (default) 8114 force - force vulnerability detection even on 8115 unaffected processors 8116 8117 vsyscall= [X86-64,EARLY] 8118 Controls the behavior of vsyscalls (i.e. calls to 8119 fixed addresses of 0xffffffffff600x00 from legacy 8120 code). Most statically-linked binaries and older 8121 versions of glibc use these calls. Because these 8122 functions are at fixed addresses, they make nice 8123 targets for exploits that can control RIP. 8124 8125 emulate Vsyscalls turn into traps and are emulated 8126 reasonably safely. The vsyscall page is 8127 readable. 8128 8129 xonly [default] Vsyscalls turn into traps and are 8130 emulated reasonably safely. The vsyscall 8131 page is not readable. 8132 8133 none Vsyscalls don't work at all. This makes 8134 them quite hard to use for exploits but 8135 might break your system. 8136 8137 vt.color= [VT] Default text color. 8138 Format: 0xYX, X = foreground, Y = background. 8139 Default: 0x07 = light gray on black. 8140 8141 vt.cur_default= [VT] Default cursor shape. 8142 Format: 0xCCBBAA, where AA, BB, and CC are the same as 8143 the parameters of the <Esc>[?A;B;Cc escape sequence; 8144 see vga-softcursor.rst. Default: 2 = underline. 8145 8146 vt.default_blu= [VT] 8147 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 8148 Change the default blue palette of the console. 8149 This is a 16-member array composed of values 8150 ranging from 0-255. 8151 8152 vt.default_grn= [VT] 8153 Format: <green0>,<green1>,<green2>,...,<green15> 8154 Change the default green palette of the console. 8155 This is a 16-member array composed of values 8156 ranging from 0-255. 8157 8158 vt.default_red= [VT] 8159 Format: <red0>,<red1>,<red2>,...,<red15> 8160 Change the default red palette of the console. 8161 This is a 16-member array composed of values 8162 ranging from 0-255. 8163 8164 vt.default_utf8= 8165 [VT] 8166 Format=<0|1> 8167 Set system-wide default UTF-8 mode for all tty's. 8168 Default is 1, i.e. UTF-8 mode is enabled for all 8169 newly opened terminals. 8170 8171 vt.global_cursor_default= 8172 [VT] 8173 Format=<-1|0|1> 8174 Set system-wide default for whether a cursor 8175 is shown on new VTs. Default is -1, 8176 i.e. cursors will be created by default unless 8177 overridden by individual drivers. 0 will hide 8178 cursors, 1 will display them. 8179 8180 vt.italic= [VT] Default color for italic text; 0-15. 8181 Default: 2 = green. 8182 8183 vt.underline= [VT] Default color for underlined text; 0-15. 8184 Default: 3 = cyan. 8185 8186 watchdog timers [HW,WDT] For information on watchdog timers, 8187 see Documentation/watchdog/watchdog-parameters.rst 8188 or other driver-specific files in the 8189 Documentation/watchdog/ directory. 8190 8191 watchdog_thresh= 8192 [KNL] 8193 Set the hard lockup detector stall duration 8194 threshold in seconds. The soft lockup detector 8195 threshold is set to twice the value. A value of 0 8196 disables both lockup detectors. Default is 10 8197 seconds. 8198 8199 workqueue.unbound_cpus= 8200 [KNL,SMP] Specify to constrain one or some CPUs 8201 to use in unbound workqueues. 8202 Format: <cpu-list> 8203 By default, all online CPUs are available for 8204 unbound workqueues. 8205 8206 workqueue.watchdog_thresh= 8207 If CONFIG_WQ_WATCHDOG is configured, workqueue can 8208 warn stall conditions and dump internal state to 8209 help debugging. 0 disables workqueue stall 8210 detection; otherwise, it's the stall threshold 8211 duration in seconds. The default value is 30 and 8212 it can be updated at runtime by writing to the 8213 corresponding sysfs file. 8214 8215 workqueue.panic_on_stall=<uint> 8216 Panic when workqueue stall is detected by 8217 CONFIG_WQ_WATCHDOG. It sets the number times of the 8218 stall to trigger panic. 8219 8220 The default is 0, which disables the panic on stall. 8221 8222 workqueue.cpu_intensive_thresh_us= 8223 Per-cpu work items which run for longer than this 8224 threshold are automatically considered CPU intensive 8225 and excluded from concurrency management to prevent 8226 them from noticeably delaying other per-cpu work 8227 items. Default is 10000 (10ms). 8228 8229 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8230 will report the work functions which violate this 8231 threshold repeatedly. They are likely good 8232 candidates for using WQ_UNBOUND workqueues instead. 8233 8234 workqueue.cpu_intensive_warning_thresh=<uint> 8235 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8236 will report the work functions which violate the 8237 intensive_threshold_us repeatedly. In order to prevent 8238 spurious warnings, start printing only after a work 8239 function has violated this threshold number of times. 8240 8241 The default is 4 times. 0 disables the warning. 8242 8243 workqueue.power_efficient 8244 Per-cpu workqueues are generally preferred because 8245 they show better performance thanks to cache 8246 locality; unfortunately, per-cpu workqueues tend to 8247 be more power hungry than unbound workqueues. 8248 8249 Enabling this makes the per-cpu workqueues which 8250 were observed to contribute significantly to power 8251 consumption unbound, leading to measurably lower 8252 power usage at the cost of small performance 8253 overhead. 8254 8255 The default value of this parameter is determined by 8256 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 8257 8258 workqueue.default_affinity_scope= 8259 Select the default affinity scope to use for unbound 8260 workqueues. Can be one of "cpu", "smt", "cache", 8261 "numa" and "system". Default is "cache". For more 8262 information, see the Affinity Scopes section in 8263 Documentation/core-api/workqueue.rst. 8264 8265 This can be changed after boot by writing to the 8266 matching /sys/module/workqueue/parameters file. All 8267 workqueues with the "default" affinity scope will be 8268 updated accordingly. 8269 8270 workqueue.debug_force_rr_cpu 8271 Workqueue used to implicitly guarantee that work 8272 items queued without explicit CPU specified are put 8273 on the local CPU. This guarantee is no longer true 8274 and while local CPU is still preferred work items 8275 may be put on foreign CPUs. This debug option 8276 forces round-robin CPU selection to flush out 8277 usages which depend on the now broken guarantee. 8278 When enabled, memory and cache locality will be 8279 impacted. 8280 8281 writecombine= [LOONGARCH,EARLY] Control the MAT (Memory Access 8282 Type) of ioremap_wc(). 8283 8284 on - Enable writecombine, use WUC for ioremap_wc() 8285 off - Disable writecombine, use SUC for ioremap_wc() 8286 8287 x2apic_phys [X86-64,APIC,EARLY] Use x2apic physical mode instead of 8288 default x2apic cluster mode on platforms 8289 supporting x2apic. 8290 8291 xen_512gb_limit [KNL,X86-64,XEN] 8292 Restricts the kernel running paravirtualized under Xen 8293 to use only up to 512 GB of RAM. The reason to do so is 8294 crash analysis tools and Xen tools for doing domain 8295 save/restore/migration must be enabled to handle larger 8296 domains. 8297 8298 xen_emul_unplug= [HW,X86,XEN,EARLY] 8299 Unplug Xen emulated devices 8300 Format: [unplug0,][unplug1] 8301 ide-disks -- unplug primary master IDE devices 8302 aux-ide-disks -- unplug non-primary-master IDE devices 8303 nics -- unplug network devices 8304 all -- unplug all emulated devices (NICs and IDE disks) 8305 unnecessary -- unplugging emulated devices is 8306 unnecessary even if the host did not respond to 8307 the unplug protocol 8308 never -- do not unplug even if version check succeeds 8309 8310 xen_legacy_crash [X86,XEN,EARLY] 8311 Crash from Xen panic notifier, without executing late 8312 panic() code such as dumping handler. 8313 8314 xen_mc_debug [X86,XEN,EARLY] 8315 Enable multicall debugging when running as a Xen PV guest. 8316 Enabling this feature will reduce performance a little 8317 bit, so it should only be enabled for obtaining extended 8318 debug data in case of multicall errors. 8319 8320 xen_msr_safe= [X86,XEN,EARLY] 8321 Format: <bool> 8322 Select whether to always use non-faulting (safe) MSR 8323 access functions when running as Xen PV guest. The 8324 default value is controlled by CONFIG_XEN_PV_MSR_SAFE. 8325 8326 xen_nopv [X86] 8327 Disables the PV optimizations forcing the HVM guest to 8328 run as generic HVM guest with no PV drivers. 8329 This option is obsoleted by the "nopv" option, which 8330 has equivalent effect for XEN platform. 8331 8332 xen_no_vector_callback 8333 [KNL,X86,XEN,EARLY] Disable the vector callback for Xen 8334 event channel interrupts. 8335 8336 xen_scrub_pages= [XEN] 8337 Boolean option to control scrubbing pages before giving them back 8338 to Xen, for use by other domains. Can be also changed at runtime 8339 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 8340 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 8341 8342 xen_timer_slop= [X86-64,XEN,EARLY] 8343 Set the timer slop (in nanoseconds) for the virtual Xen 8344 timers (default is 100000). This adjusts the minimum 8345 delta of virtualized Xen timers, where lower values 8346 improve timer resolution at the expense of processing 8347 more timer interrupts. 8348 8349 xen.balloon_boot_timeout= [XEN] 8350 The time (in seconds) to wait before giving up to boot 8351 in case initial ballooning fails to free enough memory. 8352 Applies only when running as HVM or PVH guest and 8353 started with less memory configured than allowed at 8354 max. Default is 180. 8355 8356 xen.event_eoi_delay= [XEN] 8357 How long to delay EOI handling in case of event 8358 storms (jiffies). Default is 10. 8359 8360 xen.event_loop_timeout= [XEN] 8361 After which time (jiffies) the event handling loop 8362 should start to delay EOI handling. Default is 2. 8363 8364 xen.fifo_events= [XEN] 8365 Boolean parameter to disable using fifo event handling 8366 even if available. Normally fifo event handling is 8367 preferred over the 2-level event handling, as it is 8368 fairer and the number of possible event channels is 8369 much higher. Default is on (use fifo events). 8370 8371 xirc2ps_cs= [NET,PCMCIA] 8372 Format: 8373 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 8374 8375 xive= [PPC] 8376 By default on POWER9 and above, the kernel will 8377 natively use the XIVE interrupt controller. This option 8378 allows the fallback firmware mode to be used: 8379 8380 off Fallback to firmware control of XIVE interrupt 8381 controller on both pseries and powernv 8382 platforms. Only useful on POWER9 and above. 8383 8384 xive.store-eoi=off [PPC] 8385 By default on POWER10 and above, the kernel will use 8386 stores for EOI handling when the XIVE interrupt mode 8387 is active. This option allows the XIVE driver to use 8388 loads instead, as on POWER9. 8389 8390 xhci-hcd.quirks [USB,KNL] 8391 A hex value specifying bitmask with supplemental xhci 8392 host controller quirks. Meaning of each bit can be 8393 consulted in header drivers/usb/host/xhci.h. 8394 8395 xmon [PPC,EARLY] 8396 Format: { early | on | rw | ro | off } 8397 Controls if xmon debugger is enabled. Default is off. 8398 Passing only "xmon" is equivalent to "xmon=early". 8399 early Call xmon as early as possible on boot; xmon 8400 debugger is called from setup_arch(). 8401 on xmon debugger hooks will be installed so xmon 8402 is only called on a kernel crash. Default mode, 8403 i.e. either "ro" or "rw" mode, is controlled 8404 with CONFIG_XMON_DEFAULT_RO_MODE. 8405 rw xmon debugger hooks will be installed so xmon 8406 is called only on a kernel crash, mode is write, 8407 meaning SPR registers, memory and, other data 8408 can be written using xmon commands. 8409 ro same as "rw" option above but SPR registers, 8410 memory, and other data can't be written using 8411 xmon commands. 8412 off xmon is disabled. 8413