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