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