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