1 accept_memory= [MM] 2 Format: { eager | lazy } 3 default: lazy 4 By default, unaccepted memory is accepted lazily to 5 avoid prolonged boot times. The lazy option will add 6 some runtime overhead until all memory is eventually 7 accepted. In most cases the overhead is negligible. 8 For some workloads or for debugging purposes 9 accept_memory=eager can be used to accept all memory 10 at once during boot. 11 12 acpi= [HW,ACPI,X86,ARM64,RISCV64,EARLY] 13 Advanced Configuration and Power Interface 14 Format: { force | on | off | strict | noirq | rsdt | 15 copy_dsdt | nospcr } 16 force -- enable ACPI if default was off 17 on -- enable ACPI but allow fallback to DT [arm64,riscv64] 18 off -- disable ACPI if default was on 19 noirq -- do not use ACPI for IRQ routing 20 strict -- Be less tolerant of platforms that are not 21 strictly ACPI specification compliant. 22 rsdt -- prefer RSDT over (default) XSDT 23 copy_dsdt -- copy DSDT to memory 24 nocmcff -- Disable firmware first mode for corrected 25 errors. This disables parsing the HEST CMC error 26 source to check if firmware has set the FF flag. This 27 may result in duplicate corrected error reports. 28 nospcr -- disable console in ACPI SPCR table as 29 default _serial_ console on ARM64 30 For ARM64, ONLY "acpi=off", "acpi=on", "acpi=force" or 31 "acpi=nospcr" are available 32 For RISCV64, ONLY "acpi=off", "acpi=on" or "acpi=force" 33 are available 34 35 See also Documentation/power/runtime_pm.rst, pci=noacpi 36 37 acpi_apic_instance= [ACPI,IOAPIC,EARLY] 38 Format: <int> 39 2: use 2nd APIC table, if available 40 1,0: use 1st APIC table 41 default: 0 42 43 acpi_backlight= [HW,ACPI] 44 { vendor | video | native | none } 45 If set to vendor, prefer vendor-specific driver 46 (e.g. thinkpad_acpi, sony_acpi, etc.) instead 47 of the ACPI video.ko driver. 48 If set to video, use the ACPI video.ko driver. 49 If set to native, use the device's native backlight mode. 50 If set to none, disable the ACPI backlight interface. 51 52 acpi_force_32bit_fadt_addr [ACPI,EARLY] 53 force FADT to use 32 bit addresses rather than the 54 64 bit X_* addresses. Some firmware have broken 64 55 bit addresses for force ACPI ignore these and use 56 the older legacy 32 bit addresses. 57 58 acpica_no_return_repair [HW, ACPI] 59 Disable AML predefined validation mechanism 60 This mechanism can repair the evaluation result to make 61 the return objects more ACPI specification compliant. 62 This option is useful for developers to identify the 63 root cause of an AML interpreter issue when the issue 64 has something to do with the repair mechanism. 65 66 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG] 67 acpi.debug_level= [HW,ACPI,ACPI_DEBUG] 68 Format: <int> 69 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI 70 debug output. Bits in debug_layer correspond to a 71 _COMPONENT in an ACPI source file, e.g., 72 #define _COMPONENT ACPI_EVENTS 73 Bits in debug_level correspond to a level in 74 ACPI_DEBUG_PRINT statements, e.g., 75 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ... 76 The debug_level mask defaults to "info". See 77 Documentation/firmware-guide/acpi/debug.rst for more information about 78 debug layers and levels. 79 80 Enable processor driver info messages: 81 acpi.debug_layer=0x20000000 82 Enable AML "Debug" output, i.e., stores to the Debug 83 object while interpreting AML: 84 acpi.debug_layer=0xffffffff acpi.debug_level=0x2 85 Enable all messages related to ACPI hardware: 86 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff 87 88 Some values produce so much output that the system is 89 unusable. The "log_buf_len" parameter may be useful 90 if you need to capture more output. 91 92 acpi_enforce_resources= [ACPI] 93 { strict | lax | no } 94 Check for resource conflicts between native drivers 95 and ACPI OperationRegions (SystemIO and SystemMemory 96 only). IO ports and memory declared in ACPI might be 97 used by the ACPI subsystem in arbitrary AML code and 98 can interfere with legacy drivers. 99 strict (default): access to resources claimed by ACPI 100 is denied; legacy drivers trying to access reserved 101 resources will fail to bind to device using them. 102 lax: access to resources claimed by ACPI is allowed; 103 legacy drivers trying to access reserved resources 104 will bind successfully but a warning message is logged. 105 no: ACPI OperationRegions are not marked as reserved, 106 no further checks are performed. 107 108 acpi_force_table_verification [HW,ACPI,EARLY] 109 Enable table checksum verification during early stage. 110 By default, this is disabled due to x86 early mapping 111 size limitation. 112 113 acpi_irq_balance [HW,ACPI] 114 ACPI will balance active IRQs 115 default in APIC mode 116 117 acpi_irq_nobalance [HW,ACPI] 118 ACPI will not move active IRQs (default) 119 default in PIC mode 120 121 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 122 Format: <irq>,<irq>... 123 124 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 125 use by PCI 126 Format: <irq>,<irq>... 127 128 acpi_mask_gpe= [HW,ACPI] 129 Due to the existence of _Lxx/_Exx, some GPEs triggered 130 by unsupported hardware/firmware features can result in 131 GPE floodings that cannot be automatically disabled by 132 the GPE dispatcher. 133 This facility can be used to prevent such uncontrolled 134 GPE floodings. 135 Format: <byte> or <bitmap-list> 136 137 acpi_no_auto_serialize [HW,ACPI] 138 Disable auto-serialization of AML methods 139 AML control methods that contain the opcodes to create 140 named objects will be marked as "Serialized" by the 141 auto-serialization feature. 142 This feature is enabled by default. 143 This option allows to turn off the feature. 144 145 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 146 kernels. 147 148 acpi_no_static_ssdt [HW,ACPI,EARLY] 149 Disable installation of static SSDTs at early boot time 150 By default, SSDTs contained in the RSDT/XSDT will be 151 installed automatically and they will appear under 152 /sys/firmware/acpi/tables. 153 This option turns off this feature. 154 Note that specifying this option does not affect 155 dynamic table installation which will install SSDT 156 tables to /sys/firmware/acpi/tables/dynamic. 157 158 acpi_no_watchdog [HW,ACPI,WDT] 159 Ignore the ACPI-based watchdog interface (WDAT) and let 160 a native driver control the watchdog device instead. 161 162 acpi_rsdp= [ACPI,EFI,KEXEC,EARLY] 163 Pass the RSDP address to the kernel, mostly used 164 on machines running EFI runtime service to boot the 165 second kernel for kdump. 166 167 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 168 Format: To spoof as Windows 98: ="Microsoft Windows" 169 170 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 171 of 2 which is mandated by ACPI 6) as the supported ACPI 172 specification revision (when using this switch, it may 173 be necessary to carry out a cold reboot _twice_ in a 174 row to make it take effect on the platform firmware). 175 176 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 177 acpi_osi="string1" # add string1 178 acpi_osi="!string2" # remove string2 179 acpi_osi=!* # remove all strings 180 acpi_osi=! # disable all built-in OS vendor 181 strings 182 acpi_osi=!! # enable all built-in OS vendor 183 strings 184 acpi_osi= # disable all strings 185 186 'acpi_osi=!' can be used in combination with single or 187 multiple 'acpi_osi="string1"' to support specific OS 188 vendor string(s). Note that such command can only 189 affect the default state of the OS vendor strings, thus 190 it cannot affect the default state of the feature group 191 strings and the current state of the OS vendor strings, 192 specifying it multiple times through kernel command line 193 is meaningless. This command is useful when one do not 194 care about the state of the feature group strings which 195 should be controlled by the OSPM. 196 Examples: 197 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 198 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 199 can make '_OSI("Windows 2000")' TRUE. 200 201 'acpi_osi=' cannot be used in combination with other 202 'acpi_osi=' command lines, the _OSI method will not 203 exist in the ACPI namespace. NOTE that such command can 204 only affect the _OSI support state, thus specifying it 205 multiple times through kernel command line is also 206 meaningless. 207 Examples: 208 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 209 FALSE. 210 211 'acpi_osi=!*' can be used in combination with single or 212 multiple 'acpi_osi="string1"' to support specific 213 string(s). Note that such command can affect the 214 current state of both the OS vendor strings and the 215 feature group strings, thus specifying it multiple times 216 through kernel command line is meaningful. But it may 217 still not able to affect the final state of a string if 218 there are quirks related to this string. This command 219 is useful when one want to control the state of the 220 feature group strings to debug BIOS issues related to 221 the OSPM features. 222 Examples: 223 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 224 '_OSI("Module Device")' FALSE. 225 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 226 '_OSI("Module Device")' TRUE. 227 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 228 equivalent to 229 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 230 and 231 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 232 they all will make '_OSI("Windows 2000")' TRUE. 233 234 acpi_pm_good [X86] 235 Override the pmtimer bug detection: force the kernel 236 to assume that this machine's pmtimer latches its value 237 and always returns good values. 238 239 acpi_sci= [HW,ACPI,EARLY] ACPI System Control Interrupt trigger mode 240 Format: { level | edge | high | low } 241 242 acpi_skip_timer_override [HW,ACPI,EARLY] 243 Recognize and ignore IRQ0/pin2 Interrupt Override. 244 For broken nForce2 BIOS resulting in XT-PIC timer. 245 246 acpi_sleep= [HW,ACPI] Sleep options 247 Format: { s3_bios, s3_mode, s3_beep, s4_hwsig, 248 s4_nohwsig, old_ordering, nonvs, 249 sci_force_enable, nobl } 250 See Documentation/power/video.rst for information on 251 s3_bios and s3_mode. 252 s3_beep is for debugging; it makes the PC's speaker beep 253 as soon as the kernel's real-mode entry point is called. 254 s4_hwsig causes the kernel to check the ACPI hardware 255 signature during resume from hibernation, and gracefully 256 refuse to resume if it has changed. This complies with 257 the ACPI specification but not with reality, since 258 Windows does not do this and many laptops do change it 259 on docking. So the default behaviour is to allow resume 260 and simply warn when the signature changes, unless the 261 s4_hwsig option is enabled. 262 s4_nohwsig prevents ACPI hardware signature from being 263 used (or even warned about) during resume. 264 old_ordering causes the ACPI 1.0 ordering of the _PTS 265 control method, with respect to putting devices into 266 low power states, to be enforced (the ACPI 2.0 ordering 267 of _PTS is used by default). 268 nonvs prevents the kernel from saving/restoring the 269 ACPI NVS memory during suspend/hibernation and resume. 270 sci_force_enable causes the kernel to set SCI_EN directly 271 on resume from S1/S3 (which is against the ACPI spec, 272 but some broken systems don't work without it). 273 nobl causes the internal blacklist of systems known to 274 behave incorrectly in some ways with respect to system 275 suspend and resume to be ignored (use wisely). 276 277 acpi_use_timer_override [HW,ACPI,EARLY] 278 Use timer override. For some broken Nvidia NF5 boards 279 that require a timer override, but don't have HPET 280 281 add_efi_memmap [EFI,X86,EARLY] Include EFI memory map in 282 kernel's map of available physical RAM. 283 284 agp= [AGP] 285 { off | try_unsupported } 286 off: disable AGP support 287 try_unsupported: try to drive unsupported chipsets 288 (may crash computer or cause data corruption) 289 290 ALSA [HW,ALSA] 291 See Documentation/sound/alsa-configuration.rst 292 293 alignment= [KNL,ARM] 294 Allow the default userspace alignment fault handler 295 behaviour to be specified. Bit 0 enables warnings, 296 bit 1 enables fixups, and bit 2 sends a segfault. 297 298 align_va_addr= [X86-64] 299 Align virtual addresses by clearing slice [14:12] when 300 allocating a VMA at process creation time. This option 301 gives you up to 3% performance improvement on AMD F15h 302 machines (where it is enabled by default) for a 303 CPU-intensive style benchmark, and it can vary highly in 304 a microbenchmark depending on workload and compiler. 305 306 32: only for 32-bit processes 307 64: only for 64-bit processes 308 on: enable for both 32- and 64-bit processes 309 off: disable for both 32- and 64-bit processes 310 311 alloc_snapshot [FTRACE] 312 Allocate the ftrace snapshot buffer on boot up when the 313 main buffer is allocated. This is handy if debugging 314 and you need to use tracing_snapshot() on boot up, and 315 do not want to use tracing_snapshot_alloc() as it needs 316 to be done where GFP_KERNEL allocations are allowed. 317 318 allow_mismatched_32bit_el0 [ARM64,EARLY] 319 Allow execve() of 32-bit applications and setting of the 320 PER_LINUX32 personality on systems where only a strict 321 subset of the CPUs support 32-bit EL0. When this 322 parameter is present, the set of CPUs supporting 32-bit 323 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0 324 and hot-unplug operations may be restricted. 325 326 See Documentation/arch/arm64/asymmetric-32bit.rst for more 327 information. 328 329 amd_iommu= [HW,X86-64] 330 Pass parameters to the AMD IOMMU driver in the system. 331 Possible values are: 332 fullflush - Deprecated, equivalent to iommu.strict=1 333 off - do not initialize any AMD IOMMU found in 334 the system 335 force_isolation - Force device isolation for all 336 devices. The IOMMU driver is not 337 allowed anymore to lift isolation 338 requirements as needed. This option 339 does not override iommu=pt 340 force_enable - Force enable the IOMMU on platforms known 341 to be buggy with IOMMU enabled. Use this 342 option with care. 343 pgtbl_v1 - Use v1 page table for DMA-API (Default). 344 pgtbl_v2 - Use v2 page table for DMA-API. 345 irtcachedis - Disable Interrupt Remapping Table (IRT) caching. 346 nohugepages - Limit page-sizes used for v1 page-tables 347 to 4 KiB. 348 v2_pgsizes_only - Limit page-sizes used for v1 page-tables 349 to 4KiB/2Mib/1GiB. 350 351 352 amd_iommu_dump= [HW,X86-64] 353 Enable AMD IOMMU driver option to dump the ACPI table 354 for AMD IOMMU. With this option enabled, AMD IOMMU 355 driver will print ACPI tables for AMD IOMMU during 356 IOMMU initialization. 357 358 amd_iommu_intr= [HW,X86-64] 359 Specifies one of the following AMD IOMMU interrupt 360 remapping modes: 361 legacy - Use legacy interrupt remapping mode. 362 vapic - Use virtual APIC mode, which allows IOMMU 363 to inject interrupts directly into guest. 364 This mode requires kvm-amd.avic=1. 365 (Default when IOMMU HW support is present.) 366 367 amd_pstate= [X86,EARLY] 368 disable 369 Do not enable amd_pstate as the default 370 scaling driver for the supported processors 371 passive 372 Use amd_pstate with passive mode as a scaling driver. 373 In this mode autonomous selection is disabled. 374 Driver requests a desired performance level and platform 375 tries to match the same performance level if it is 376 satisfied by guaranteed performance level. 377 active 378 Use amd_pstate_epp driver instance as the scaling driver, 379 driver provides a hint to the hardware if software wants 380 to bias toward performance (0x0) or energy efficiency (0xff) 381 to the CPPC firmware. then CPPC power algorithm will 382 calculate the runtime workload and adjust the realtime cores 383 frequency. 384 guided 385 Activate guided autonomous mode. Driver requests minimum and 386 maximum performance level and the platform autonomously 387 selects a performance level in this range and appropriate 388 to the current workload. 389 390 amd_prefcore= 391 [X86] 392 disable 393 Disable amd-pstate preferred core. 394 395 amijoy.map= [HW,JOY] Amiga joystick support 396 Map of devices attached to JOY0DAT and JOY1DAT 397 Format: <a>,<b> 398 See also Documentation/input/joydev/joystick.rst 399 400 analog.map= [HW,JOY] Analog joystick and gamepad support 401 Specifies type or capabilities of an analog joystick 402 connected to one of 16 gameports 403 Format: <type1>,<type2>,..<type16> 404 405 apc= [HW,SPARC] 406 Power management functions (SPARCstation-4/5 + deriv.) 407 Format: noidle 408 Disable APC CPU standby support. SPARCstation-Fox does 409 not play well with APC CPU idle - disable it if you have 410 APC and your system crashes randomly. 411 412 apic [APIC,X86-64] Use IO-APIC. Default. 413 414 apic= [APIC,X86,EARLY] Advanced Programmable Interrupt Controller 415 Change the output verbosity while booting 416 Format: { quiet (default) | verbose | debug } 417 Change the amount of debugging information output 418 when initialising the APIC and IO-APIC components. 419 420 apic_extnmi= [APIC,X86,EARLY] External NMI delivery setting 421 Format: { bsp (default) | all | none } 422 bsp: External NMI is delivered only to CPU 0 423 all: External NMIs are broadcast to all CPUs as a 424 backup of CPU 0 425 none: External NMI is masked for all CPUs. This is 426 useful so that a dump capture kernel won't be 427 shot down by NMI 428 429 apicpmtimer Do APIC timer calibration using the pmtimer. Implies 430 apicmaintimer. Useful when your PIT timer is totally 431 broken. 432 433 autoconf= [IPV6] 434 See Documentation/networking/ipv6.rst. 435 436 apm= [APM] Advanced Power Management 437 See header of arch/x86/kernel/apm_32.c. 438 439 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 440 Format: { "0" | "1" } 441 See security/apparmor/Kconfig help text 442 0 -- disable. 443 1 -- enable. 444 Default value is set via kernel config option. 445 446 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards 447 Format: <io>,<irq>,<nodeID> 448 449 arm64.no32bit_el0 [ARM64] Unconditionally disable the execution of 450 32 bit applications. 451 452 arm64.nobti [ARM64] Unconditionally disable Branch Target 453 Identification support 454 455 arm64.nogcs [ARM64] Unconditionally disable Guarded Control Stack 456 support 457 458 arm64.nomops [ARM64] Unconditionally disable Memory Copy and Memory 459 Set instructions support 460 461 arm64.nompam [ARM64] Unconditionally disable Memory Partitioning And 462 Monitoring support 463 464 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension 465 support 466 467 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 468 support 469 470 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix 471 Extension support 472 473 arm64.nosve [ARM64] Unconditionally disable Scalable Vector 474 Extension support 475 476 ataflop= [HW,M68k] 477 478 atarimouse= [HW,MOUSE] Atari Mouse 479 480 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 481 EzKey and similar keyboards 482 483 atkbd.reset= [HW] Reset keyboard during initialization 484 485 atkbd.set= [HW] Select keyboard code set 486 Format: <int> (2 = AT (default), 3 = PS/2) 487 488 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 489 keyboards 490 491 atkbd.softraw= [HW] Choose between synthetic and real raw mode 492 Format: <bool> (0 = real, 1 = synthetic (default)) 493 494 atkbd.softrepeat= [HW] 495 Use software keyboard repeat 496 497 audit= [KNL] Enable the audit sub-system 498 Format: { "0" | "1" | "off" | "on" } 499 0 | off - kernel audit is disabled and can not be 500 enabled until the next reboot 501 unset - kernel audit is initialized but disabled and 502 will be fully enabled by the userspace auditd. 503 1 | on - kernel audit is initialized and partially 504 enabled, storing at most audit_backlog_limit 505 messages in RAM until it is fully enabled by the 506 userspace auditd. 507 Default: unset 508 509 audit_backlog_limit= [KNL] Set the audit queue size limit. 510 Format: <int> (must be >=0) 511 Default: 64 512 513 bau= [X86_UV] Enable the BAU on SGI UV. The default 514 behavior is to disable the BAU (i.e. bau=0). 515 Format: { "0" | "1" } 516 0 - Disable the BAU. 517 1 - Enable the BAU. 518 unset - Disable the BAU. 519 520 baycom_epp= [HW,AX25] 521 Format: <io>,<mode> 522 523 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem 524 Format: <io>,<mode> 525 See header of drivers/net/hamradio/baycom_par.c. 526 527 baycom_ser_fdx= [HW,AX25] 528 BayCom Serial Port AX.25 Modem (Full Duplex Mode) 529 Format: <io>,<irq>,<mode>[,<baud>] 530 See header of drivers/net/hamradio/baycom_ser_fdx.c. 531 532 baycom_ser_hdx= [HW,AX25] 533 BayCom Serial Port AX.25 Modem (Half Duplex Mode) 534 Format: <io>,<irq>,<mode> 535 See header of drivers/net/hamradio/baycom_ser_hdx.c. 536 537 bdev_allow_write_mounted= 538 Format: <bool> 539 Control the ability to open a mounted block device 540 for writing, i.e., allow / disallow writes that bypass 541 the FS. This was implemented as a means to prevent 542 fuzzers from crashing the kernel by overwriting the 543 metadata underneath a mounted FS without its awareness. 544 This also prevents destructive formatting of mounted 545 filesystems by naive storage tooling that don't use 546 O_EXCL. Default is Y and can be changed through the 547 Kconfig option CONFIG_BLK_DEV_WRITE_MOUNTED. 548 549 bert_disable [ACPI] 550 Disable BERT OS support on buggy BIOSes. 551 552 bgrt_disable [ACPI,X86,EARLY] 553 Disable BGRT to avoid flickering OEM logo. 554 555 blkdevparts= Manual partition parsing of block device(s) for 556 embedded devices based on command line input. 557 See Documentation/block/cmdline-partition.rst 558 559 boot_delay= [KNL,EARLY] 560 Milliseconds to delay each printk during boot. 561 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled, 562 and you may also have to specify "lpj=". Boot_delay 563 values larger than 10 seconds (10000) are assumed 564 erroneous and ignored. 565 Format: integer 566 567 bootconfig [KNL,EARLY] 568 Extended command line options can be added to an initrd 569 and this will cause the kernel to look for it. 570 571 See Documentation/admin-guide/bootconfig.rst 572 573 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 574 bttv.radio= Most important insmod options are available as 575 kernel args too. 576 bttv.pll= See Documentation/admin-guide/media/bttv.rst 577 bttv.tuner= 578 579 bulk_remove=off [PPC] This parameter disables the use of the pSeries 580 firmware feature for flushing multiple hpte entries 581 at a time. 582 583 c101= [NET] Moxa C101 synchronous serial card 584 585 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 586 Sometimes CPU hardware bugs make them report the cache 587 size incorrectly. The kernel will attempt work arounds 588 to fix known problems, but for some CPUs it is not 589 possible to determine what the correct size should be. 590 This option provides an override for these situations. 591 592 carrier_timeout= 593 [NET] Specifies amount of time (in seconds) that 594 the kernel should wait for a network carrier. By default 595 it waits 120 seconds. 596 597 ca_keys= [KEYS] This parameter identifies a specific key(s) on 598 the system trusted keyring to be used for certificate 599 trust validation. 600 format: { id:<keyid> | builtin } 601 602 cca= [MIPS,EARLY] Override the kernel pages' cache coherency 603 algorithm. Accepted values range from 0 to 7 604 inclusive. See arch/mips/include/asm/pgtable-bits.h 605 for platform specific values (SB1, Loongson3 and 606 others). 607 608 ccw_timeout_log [S390] 609 See Documentation/arch/s390/common_io.rst for details. 610 611 cfi= [X86-64] Set Control Flow Integrity checking features 612 when CONFIG_FINEIBT is enabled. 613 Format: feature[,feature...] 614 Default: auto 615 616 auto: Use FineIBT if IBT available, otherwise kCFI. 617 Under FineIBT, enable "paranoid" mode when 618 FRED is not available. 619 off: Turn off CFI checking. 620 kcfi: Use kCFI (disable FineIBT). 621 fineibt: Use FineIBT (even if IBT not available). 622 norand: Do not re-randomize CFI hashes. 623 paranoid: Add caller hash checking under FineIBT. 624 bhi: Enable register poisoning to stop speculation 625 across FineIBT. (Disabled by default.) 626 warn: Do not enforce CFI checking: warn only. 627 debug: Report CFI initialization details. 628 629 cgroup_disable= [KNL] Disable a particular controller or optional feature 630 Format: {name of the controller(s) or feature(s) to disable} 631 The effects of cgroup_disable=foo are: 632 - foo isn't auto-mounted if you mount all cgroups in 633 a single hierarchy 634 - foo isn't visible as an individually mountable 635 subsystem 636 - if foo is an optional feature then the feature is 637 disabled and corresponding cgroup files are not 638 created 639 {Currently only "memory" controller deal with this and 640 cut the overhead, others just disable the usage. So 641 only cgroup_disable=memory is actually worthy} 642 Specifying "pressure" disables per-cgroup pressure 643 stall information accounting feature 644 645 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 646 Format: { { controller | "all" | "named" } 647 [,{ controller | "all" | "named" }...] } 648 Like cgroup_disable, but only applies to cgroup v1; 649 the blacklisted controllers remain available in cgroup2. 650 "all" blacklists all controllers and "named" disables 651 named mounts. Specifying both "all" and "named" disables 652 all v1 hierarchies. 653 654 cgroup_v1_proc= [KNL] Show also missing controllers in /proc/cgroups 655 Format: { "true" | "false" } 656 /proc/cgroups lists only v1 controllers by default. 657 This compatibility option enables listing also v2 658 controllers (whose v1 code is not compiled!), so that 659 semi-legacy software can check this file to decide 660 about usage of v2 (sic) controllers. 661 662 cgroup_favordynmods= [KNL] Enable or Disable favordynmods. 663 Format: { "true" | "false" } 664 Defaults to the value of CONFIG_CGROUP_FAVOR_DYNMODS. 665 666 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 667 Format: <string> 668 nosocket -- Disable socket memory accounting. 669 nokmem -- Disable kernel memory accounting. 670 nobpf -- Disable BPF memory accounting. 671 672 checkreqprot= [SELINUX] Set initial checkreqprot flag value. 673 Format: { "0" | "1" } 674 See security/selinux/Kconfig help text. 675 0 -- check protection applied by kernel (includes 676 any implied execute protection). 677 1 -- check protection requested by application. 678 Default value is set via a kernel config option. 679 Value can be changed at runtime via 680 /sys/fs/selinux/checkreqprot. 681 Setting checkreqprot to 1 is deprecated. 682 683 cio_ignore= [S390] 684 See Documentation/arch/s390/common_io.rst for details. 685 686 clearcpuid=X[,X...] [X86] 687 Disable CPUID feature X for the kernel. See 688 arch/x86/include/asm/cpufeatures.h for the valid bit 689 numbers X. Note the Linux-specific bits are not necessarily 690 stable over kernel options, but the vendor-specific 691 ones should be. 692 X can also be a string as appearing in the flags: line 693 in /proc/cpuinfo which does not have the above 694 instability issue. However, not all features have names 695 in /proc/cpuinfo. 696 Note that using this option will taint your kernel. 697 Also note that user programs calling CPUID directly 698 or using the feature without checking anything 699 will still see it. This just prevents it from 700 being used by the kernel or shown in /proc/cpuinfo. 701 Also note the kernel might malfunction if you disable 702 some critical bits. 703 704 clk_ignore_unused 705 [CLK] 706 Prevents the clock framework from automatically gating 707 clocks that have not been explicitly enabled by a Linux 708 device driver but are enabled in hardware at reset or 709 by the bootloader/firmware. Note that this does not 710 force such clocks to be always-on nor does it reserve 711 those clocks in any way. This parameter is useful for 712 debug and development, but should not be needed on a 713 platform with proper driver support. For more 714 information, see Documentation/driver-api/clk.rst. 715 716 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 717 [Deprecated] 718 Forces specified clocksource (if available) to be used 719 when calculating gettimeofday(). If specified 720 clocksource is not available, it defaults to PIT. 721 Format: { pit | tsc | cyclone | pmtmr } 722 723 clocksource= Override the default clocksource 724 Format: <string> 725 Override the default clocksource and use the clocksource 726 with the name specified. 727 Some clocksource names to choose from, depending on 728 the platform: 729 [all] jiffies (this is the base, fallback clocksource) 730 [ACPI] acpi_pm 731 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 732 pxa_timer,timer3,32k_counter,timer0_1 733 [X86-32] pit,hpet,tsc; 734 scx200_hrt on Geode; cyclone on IBM x440 735 [MIPS] MIPS 736 [PARISC] cr16 737 [S390] tod 738 [SH] SuperH 739 [SPARC64] tick 740 [X86-64] hpet,tsc 741 742 clocksource.arm_arch_timer.evtstrm= 743 [ARM,ARM64,EARLY] 744 Format: <bool> 745 Enable/disable the eventstream feature of the ARM 746 architected timer so that code using WFE-based polling 747 loops can be debugged more effectively on production 748 systems. 749 750 clocksource.verify_n_cpus= [KNL] 751 Limit the number of CPUs checked for clocksources 752 marked with CLOCK_SOURCE_VERIFY_PERCPU that 753 are marked unstable due to excessive skew. 754 A negative value says to check all CPUs, while 755 zero says not to check any. Values larger than 756 nr_cpu_ids are silently truncated to nr_cpu_ids. 757 The actual CPUs are chosen randomly, with 758 no replacement if the same CPU is chosen twice. 759 760 clocksource-wdtest.holdoff= [KNL] 761 Set the time in seconds that the clocksource 762 watchdog test waits before commencing its tests. 763 Defaults to zero when built as a module and to 764 10 seconds when built into the kernel. 765 766 cma=nn[MG]@[start[MG][-end[MG]]] 767 [KNL,CMA,EARLY] 768 Sets the size of kernel global memory area for 769 contiguous memory allocations and optionally the 770 placement constraint by the physical address range of 771 memory allocations. A value of 0 disables CMA 772 altogether. For more information, see 773 kernel/dma/contiguous.c 774 775 cma_pernuma=nn[MG] 776 [KNL,CMA,EARLY] 777 Sets the size of kernel per-numa memory area for 778 contiguous memory allocations. A value of 0 disables 779 per-numa CMA altogether. And If this option is not 780 specified, the default value is 0. 781 With per-numa CMA enabled, DMA users on node nid will 782 first try to allocate buffer from the pernuma area 783 which is located in node nid, if the allocation fails, 784 they will fallback to the global default memory area. 785 786 numa_cma=<node>:nn[MG][,<node>:nn[MG]] 787 [KNL,CMA,EARLY] 788 Sets the size of kernel numa memory area for 789 contiguous memory allocations. It will reserve CMA 790 area for the specified node. 791 792 With numa CMA enabled, DMA users on node nid will 793 first try to allocate buffer from the numa area 794 which is located in node nid, if the allocation fails, 795 they will fallback to the global default memory area. 796 797 cmo_free_hint= [PPC] Format: { yes | no } 798 Specify whether pages are marked as being inactive 799 when they are freed. This is used in CMO environments 800 to determine OS memory pressure for page stealing by 801 a hypervisor. 802 Default: yes 803 804 coherent_pool=nn[KMG] [ARM,KNL,EARLY] 805 Sets the size of memory pool for coherent, atomic dma 806 allocations, by default set to 256K. 807 808 com20020= [HW,NET] ARCnet - COM20020 chipset 809 Format: 810 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 811 812 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 813 Format: <io>[,<irq>] 814 815 com90xx= [HW,NET] 816 ARCnet - COM90xx chipset (memory-mapped buffers) 817 Format: <io>[,<irq>[,<memstart>]] 818 819 condev= [HW,S390] console device 820 conmode= 821 822 con3215_drop= [S390,EARLY] 3215 console drop mode. 823 Format: y|n|Y|N|1|0 824 When set to true, drop data on the 3215 console when 825 the console buffer is full. In this case the 826 operator using a 3270 terminal emulator (for example 827 x3270) does not have to enter the clear key for the 828 console output to advance and the kernel to continue. 829 This leads to a much faster boot time when a 3270 830 terminal emulator is active. If no 3270 terminal 831 emulator is used, this parameter has no effect. 832 833 console= [KNL] Output console device and options. 834 835 tty<n> Use the virtual console device <n>. 836 837 ttyS<n>[,options] 838 ttyUSB0[,options] 839 Use the specified serial port. The options are of 840 the form "bbbbpnf", where "bbbb" is the baud rate, 841 "p" is parity ("n", "o", or "e"), "n" is number of 842 bits, and "f" is flow control ("r" for RTS or 843 omit it). Default is "9600n8". 844 845 See Documentation/admin-guide/serial-console.rst for more 846 information. See 847 Documentation/networking/netconsole.rst for an 848 alternative. 849 850 <DEVNAME>:<n>.<n>[,options] 851 Use the specified serial port on the serial core bus. 852 The addressing uses DEVNAME of the physical serial port 853 device, followed by the serial core controller instance, 854 and the serial port instance. The options are the same 855 as documented for the ttyS addressing above. 856 857 The mapping of the serial ports to the tty instances 858 can be viewed with: 859 860 $ ls -d /sys/bus/serial-base/devices/*:*.*/tty/* 861 /sys/bus/serial-base/devices/00:04:0.0/tty/ttyS0 862 863 In the above example, the console can be addressed with 864 console=00:04:0.0. Note that a console addressed this 865 way will only get added when the related device driver 866 is ready. The use of an earlycon parameter in addition to 867 the console may be desired for console output early on. 868 869 uart[8250],io,<addr>[,options] 870 uart[8250],mmio,<addr>[,options] 871 uart[8250],mmio16,<addr>[,options] 872 uart[8250],mmio32,<addr>[,options] 873 uart[8250],0x<addr>[,options] 874 Start an early, polled-mode console on the 8250/16550 875 UART at the specified I/O port or MMIO address, 876 switching to the matching ttyS device later. 877 MMIO inter-register address stride is either 8-bit 878 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 879 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 880 to be equivalent to 'mmio'. 'options' are specified in 881 the same format described for ttyS above; if unspecified, 882 the h/w is not re-initialized. 883 884 hvc<n> Use the hypervisor console device <n>. This is for 885 both Xen and PowerPC hypervisors. 886 887 { null | "" } 888 Use to disable console output, i.e., to have kernel 889 console messages discarded. 890 This must be the only console= parameter used on the 891 kernel command line. 892 893 If the device connected to the port is not a TTY but a braille 894 device, prepend "brl," before the device type, for instance 895 console=brl,ttyS0 896 For now, only VisioBraille is supported. 897 898 console_msg_format= 899 [KNL] Change console messages format 900 default 901 By default we print messages on consoles in 902 "[time stamp] text\n" format (time stamp may not be 903 printed, depending on CONFIG_PRINTK_TIME or 904 `printk_time' param). 905 syslog 906 Switch to syslog format: "<%u>[time stamp] text\n" 907 IOW, each message will have a facility and loglevel 908 prefix. The format is similar to one used by syslog() 909 syscall, or to executing "dmesg -S --raw" or to reading 910 from /proc/kmsg. 911 912 consoleblank= [KNL] The console blank (screen saver) timeout in 913 seconds. A value of 0 disables the blank timer. 914 Defaults to 0. 915 916 coredump_filter= 917 [KNL] Change the default value for 918 /proc/<pid>/coredump_filter. 919 See also Documentation/filesystems/proc.rst. 920 921 coresight_cpu_debug.enable 922 [ARM,ARM64] 923 Format: <bool> 924 Enable/disable the CPU sampling based debugging. 925 0: default value, disable debugging 926 1: enable debugging at boot time 927 928 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 929 Format: 930 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 931 932 cpuidle.off=1 [CPU_IDLE] 933 disable the cpuidle sub-system 934 935 cpuidle.governor= 936 [CPU_IDLE] Name of the cpuidle governor to use. 937 938 cpufreq.off=1 [CPU_FREQ] 939 disable the cpufreq sub-system 940 941 cpufreq.default_governor= 942 [CPU_FREQ] Name of the default cpufreq governor or 943 policy to use. This governor must be registered in the 944 kernel before the cpufreq driver probes. 945 946 cpu_init_udelay=N 947 [X86,EARLY] Delay for N microsec between assert and de-assert 948 of APIC INIT to start processors. This delay occurs 949 on every CPU online, such as boot, and resume from suspend. 950 Default: 10000 951 952 cpuhp.parallel= 953 [SMP] Enable/disable parallel bringup of secondary CPUs 954 Format: <bool> 955 Default is enabled if CONFIG_HOTPLUG_PARALLEL=y. Otherwise 956 the parameter has no effect. 957 958 crash_kexec_post_notifiers 959 Only jump to kdump kernel after running the panic 960 notifiers and dumping kmsg. This option increases 961 the risks of a kdump failure, since some panic 962 notifiers can make the crashed kernel more unstable. 963 In configurations where kdump may not be reliable, 964 running the panic notifiers could allow collecting 965 more data on dmesg, like stack traces from other CPUS 966 or extra data dumped by panic_print. Note that some 967 configurations enable this option unconditionally, 968 like Hyper-V, PowerPC (fadump) and AMD SEV-SNP. 969 970 crashkernel=size[KMG][@offset[KMG]] 971 [KNL,EARLY] Using kexec, Linux can switch to a 'crash kernel' 972 upon panic. This parameter reserves the physical 973 memory region [offset, offset + size] for that kernel 974 image. If '@offset' is omitted, then a suitable offset 975 is selected automatically. 976 [KNL, X86-64, ARM64, RISCV, LoongArch] Select a region 977 under 4G first, and fall back to reserve region above 978 4G when '@offset' hasn't been specified. 979 See Documentation/admin-guide/kdump/kdump.rst for further details. 980 981 crashkernel=range1:size1[,range2:size2,...][@offset] 982 [KNL] Same as above, but depends on the memory 983 in the running system. The syntax of range is 984 start-[end] where start and end are both 985 a memory unit (amount[KMG]). See also 986 Documentation/admin-guide/kdump/kdump.rst for an example. 987 988 crashkernel=size[KMG],high 989 [KNL, X86-64, ARM64, RISCV, LoongArch] range could be 990 above 4G. 991 Allow kernel to allocate physical memory region from top, 992 so could be above 4G if system have more than 4G ram 993 installed. Otherwise memory region will be allocated 994 below 4G, if available. 995 It will be ignored if crashkernel=X is specified. 996 crashkernel=size[KMG],low 997 [KNL, X86-64, ARM64, RISCV, LoongArch] range under 4G. 998 When crashkernel=X,high is passed, kernel could allocate 999 physical memory region above 4G, that cause second kernel 1000 crash on system that require some amount of low memory, 1001 e.g. swiotlb requires at least 64M+32K low memory, also 1002 enough extra low memory is needed to make sure DMA buffers 1003 for 32-bit devices won't run out. Kernel would try to allocate 1004 default size of memory below 4G automatically. The default 1005 size is platform dependent. 1006 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB) 1007 --> arm64: 128MiB 1008 --> riscv: 128MiB 1009 --> loongarch: 128MiB 1010 This one lets the user specify own low range under 4G 1011 for second kernel instead. 1012 0: to disable low allocation. 1013 It will be ignored when crashkernel=X,high is not used 1014 or memory reserved is below 4G. 1015 crashkernel=size[KMG],cma 1016 [KNL, X86] Reserve additional crash kernel memory from 1017 CMA. This reservation is usable by the first system's 1018 userspace memory and kernel movable allocations (memory 1019 balloon, zswap). Pages allocated from this memory range 1020 will not be included in the vmcore so this should not 1021 be used if dumping of userspace memory is intended and 1022 it has to be expected that some movable kernel pages 1023 may be missing from the dump. 1024 1025 A standard crashkernel reservation, as described above, 1026 is still needed to hold the crash kernel and initrd. 1027 1028 This option increases the risk of a kdump failure: DMA 1029 transfers configured by the first kernel may end up 1030 corrupting the second kernel's memory. 1031 1032 This reservation method is intended for systems that 1033 can't afford to sacrifice enough memory for standard 1034 crashkernel reservation and where less reliable and 1035 possibly incomplete kdump is preferable to no kdump at 1036 all. 1037 1038 cryptomgr.notests 1039 [KNL] Disable crypto self-tests 1040 1041 cs89x0_dma= [HW,NET] 1042 Format: <dma> 1043 1044 cs89x0_media= [HW,NET] 1045 Format: { rj45 | aui | bnc } 1046 1047 csdlock_debug= [KNL] Enable or disable debug add-ons of cross-CPU 1048 function call handling. When switched on, 1049 additional debug data is printed to the console 1050 in case a hanging CPU is detected, and that 1051 CPU is pinged again in order to try to resolve 1052 the hang situation. The default value of this 1053 option depends on the CSD_LOCK_WAIT_DEBUG_DEFAULT 1054 Kconfig option. 1055 1056 dasd= [HW,NET] 1057 See header of drivers/s390/block/dasd_devmap.c. 1058 1059 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 1060 (one device per port) 1061 Format: <port#>,<type> 1062 See also Documentation/input/devices/joystick-parport.rst 1063 1064 debug [KNL,EARLY] Enable kernel debugging (events log level). 1065 1066 debug_boot_weak_hash 1067 [KNL,EARLY] Enable printing [hashed] pointers early in the 1068 boot sequence. If enabled, we use a weak hash instead 1069 of siphash to hash pointers. Use this option if you are 1070 seeing instances of '(___ptrval___)') and need to see a 1071 value (hashed pointer) instead. Cryptographically 1072 insecure, please do not use on production kernels. 1073 1074 debug_locks_verbose= 1075 [KNL] verbose locking self-tests 1076 Format: <int> 1077 Print debugging info while doing the locking API 1078 self-tests. 1079 Bitmask for the various LOCKTYPE_ tests. Defaults to 0 1080 (no extra messages), setting it to -1 (all bits set) 1081 will print _a_lot_ more information - normally only 1082 useful to lockdep developers. 1083 1084 debug_objects [KNL,EARLY] Enable object debugging 1085 1086 debug_guardpage_minorder= 1087 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this 1088 parameter allows control of the order of pages that will 1089 be intentionally kept free (and hence protected) by the 1090 buddy allocator. Bigger value increase the probability 1091 of catching random memory corruption, but reduce the 1092 amount of memory for normal system use. The maximum 1093 possible value is MAX_PAGE_ORDER/2. Setting this 1094 parameter to 1 or 2 should be enough to identify most 1095 random memory corruption problems caused by bugs in 1096 kernel or driver code when a CPU writes to (or reads 1097 from) a random memory location. Note that there exists 1098 a class of memory corruptions problems caused by buggy 1099 H/W or F/W or by drivers badly programming DMA 1100 (basically when memory is written at bus level and the 1101 CPU MMU is bypassed) which are not detectable by 1102 CONFIG_DEBUG_PAGEALLOC, hence this option will not 1103 help tracking down these problems. 1104 1105 debug_pagealloc= 1106 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 1107 enables the feature at boot time. By default, it is 1108 disabled and the system will work mostly the same as a 1109 kernel built without CONFIG_DEBUG_PAGEALLOC. 1110 Note: to get most of debug_pagealloc error reports, it's 1111 useful to also enable the page_owner functionality. 1112 on: enable the feature 1113 1114 debugfs= [KNL,EARLY] This parameter enables what is exposed to 1115 userspace and debugfs internal clients. 1116 Format: { on, no-mount, off } 1117 on: All functions are enabled. 1118 no-mount: 1119 Filesystem is not registered but kernel clients can 1120 access APIs and a crashkernel can be used to read 1121 its content. There is nothing to mount. 1122 off: Filesystem is not registered and clients 1123 get a -EPERM as result when trying to register files 1124 or directories within debugfs. 1125 This is equivalent of the runtime functionality if 1126 debugfs was not enabled in the kernel at all. 1127 Default value is set in build-time with a kernel configuration. 1128 1129 debugpat [X86] Enable PAT debugging 1130 1131 default_hugepagesz= 1132 [HW] The size of the default HugeTLB page. This is 1133 the size represented by the legacy /proc/ hugepages 1134 APIs. In addition, this is the default hugetlb size 1135 used for shmget(), mmap() and mounting hugetlbfs 1136 filesystems. If not specified, defaults to the 1137 architecture's default huge page size. Huge page 1138 sizes are architecture dependent. See also 1139 Documentation/admin-guide/mm/hugetlbpage.rst. 1140 Format: size[KMG] 1141 1142 deferred_probe_timeout= 1143 [KNL] Debugging option to set a timeout in seconds for 1144 deferred probe to give up waiting on dependencies to 1145 probe. Only specific dependencies (subsystems or 1146 drivers) that have opted in will be ignored. A timeout 1147 of 0 will timeout at the end of initcalls. If the time 1148 out hasn't expired, it'll be restarted by each 1149 successful driver registration. This option will also 1150 dump out devices still on the deferred probe list after 1151 retrying. 1152 1153 delayacct [KNL] Enable per-task delay accounting 1154 1155 dell_smm_hwmon.ignore_dmi= 1156 [HW] Continue probing hardware even if DMI data 1157 indicates that the driver is running on unsupported 1158 hardware. 1159 1160 dell_smm_hwmon.force= 1161 [HW] Activate driver even if SMM BIOS signature does 1162 not match list of supported models and enable otherwise 1163 blacklisted features. 1164 1165 dell_smm_hwmon.power_status= 1166 [HW] Report power status in /proc/i8k 1167 (disabled by default). 1168 1169 dell_smm_hwmon.restricted= 1170 [HW] Allow controlling fans only if SYS_ADMIN 1171 capability is set. 1172 1173 dell_smm_hwmon.fan_mult= 1174 [HW] Factor to multiply fan speed with. 1175 1176 dell_smm_hwmon.fan_max= 1177 [HW] Maximum configurable fan speed. 1178 1179 dfltcc= [HW,S390] 1180 Format: { on | off | def_only | inf_only | always } 1181 on: s390 zlib hardware support for compression on 1182 level 1 and decompression (default) 1183 off: No s390 zlib hardware support 1184 def_only: s390 zlib hardware support for deflate 1185 only (compression on level 1) 1186 inf_only: s390 zlib hardware support for inflate 1187 only (decompression) 1188 always: Same as 'on' but ignores the selected compression 1189 level always using hardware support (used for debugging) 1190 1191 dhash_entries= [KNL] 1192 Set number of hash buckets for dentry cache. 1193 1194 disable_1tb_segments [PPC,EARLY] 1195 Disables the use of 1TB hash page table segments. This 1196 causes the kernel to fall back to 256MB segments which 1197 can be useful when debugging issues that require an SLB 1198 miss to occur. 1199 1200 disable= [IPV6] 1201 See Documentation/networking/ipv6.rst. 1202 1203 disable_radix [PPC,EARLY] 1204 Disable RADIX MMU mode on POWER9 1205 1206 disable_tlbie [PPC] 1207 Disable TLBIE instruction. Currently does not work 1208 with KVM, with HASH MMU, or with coherent accelerators. 1209 1210 disable_ddw [PPC/PSERIES,EARLY] 1211 Disable Dynamic DMA Window support. Use this 1212 to workaround buggy firmware. 1213 1214 disable_ipv6= [IPV6] 1215 See Documentation/networking/ipv6.rst. 1216 1217 disable_mtrr_cleanup [X86,EARLY] 1218 The kernel tries to adjust MTRR layout from continuous 1219 to discrete, to make X server driver able to add WB 1220 entry later. This parameter disables that. 1221 1222 disable_mtrr_trim [X86, Intel and AMD only,EARLY] 1223 By default the kernel will trim any uncacheable 1224 memory out of your available memory pool based on 1225 MTRR settings. This parameter disables that behavior, 1226 possibly causing your machine to run very slowly. 1227 1228 disable_timer_pin_1 [X86,EARLY] 1229 Disable PIN 1 of APIC timer 1230 Can be useful to work around chipset bugs. 1231 1232 dis_ucode_ldr [X86] Disable the microcode loader. 1233 1234 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 1235 this option disables the debugging code at boot. 1236 1237 dma_debug_entries=<number> 1238 This option allows to tune the number of preallocated 1239 entries for DMA-API debugging code. One entry is 1240 required per DMA-API allocation. Use this if the 1241 DMA-API debugging code disables itself because the 1242 architectural default is too low. 1243 1244 dma_debug_driver=<driver_name> 1245 With this option the DMA-API debugging driver 1246 filter feature can be enabled at boot time. Just 1247 pass the driver to filter for as the parameter. 1248 The filter can be disabled or changed to another 1249 driver later using sysfs. 1250 1251 reg_file_data_sampling= 1252 [X86] Controls mitigation for Register File Data 1253 Sampling (RFDS) vulnerability. RFDS is a CPU 1254 vulnerability which may allow userspace to infer 1255 kernel data values previously stored in floating point 1256 registers, vector registers, or integer registers. 1257 RFDS only affects Intel Atom processors. 1258 1259 on: Turns ON the mitigation. 1260 off: Turns OFF the mitigation. 1261 1262 This parameter overrides the compile time default set 1263 by CONFIG_MITIGATION_RFDS. Mitigation cannot be 1264 disabled when other VERW based mitigations (like MDS) 1265 are enabled. In order to disable RFDS mitigation all 1266 VERW based mitigations need to be disabled. 1267 1268 For details see: 1269 Documentation/admin-guide/hw-vuln/reg-file-data-sampling.rst 1270 1271 driver_async_probe= [KNL] 1272 List of driver names to be probed asynchronously. * 1273 matches with all driver names. If * is specified, the 1274 rest of the listed driver names are those that will NOT 1275 match the *. 1276 Format: <driver_name1>,<driver_name2>... 1277 1278 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1279 Broken monitors, graphic adapters, KVMs and EDIDless 1280 panels may send no or incorrect EDID data sets. 1281 This parameter allows to specify an EDID data sets 1282 in the /lib/firmware directory that are used instead. 1283 An EDID data set will only be used for a particular 1284 connector, if its name and a colon are prepended to 1285 the EDID name. Each connector may use a unique EDID 1286 data set by separating the files with a comma. An EDID 1287 data set with no connector name will be used for 1288 any connectors not explicitly specified. 1289 1290 dscc4.setup= [NET] 1291 1292 dt_cpu_ftrs= [PPC,EARLY] 1293 Format: {"off" | "known"} 1294 Control how the dt_cpu_ftrs device-tree binding is 1295 used for CPU feature discovery and setup (if it 1296 exists). 1297 off: Do not use it, fall back to legacy cpu table. 1298 known: Do not pass through unknown features to guests 1299 or userspace, only those that the kernel is aware of. 1300 1301 dump_apple_properties [X86] 1302 Dump name and content of EFI device properties on 1303 x86 Macs. Useful for driver authors to determine 1304 what data is available or for reverse-engineering. 1305 1306 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1307 <module>.dyndbg[="val"] 1308 Enable debug messages at boot time. See 1309 Documentation/admin-guide/dynamic-debug-howto.rst 1310 for details. 1311 1312 early_ioremap_debug [KNL,EARLY] 1313 Enable debug messages in early_ioremap support. This 1314 is useful for tracking down temporary early mappings 1315 which are not unmapped. 1316 1317 earlycon= [KNL,EARLY] Output early console device and options. 1318 1319 When used with no options, the early console is 1320 determined by stdout-path property in device tree's 1321 chosen node or the ACPI SPCR table if supported by 1322 the platform. 1323 1324 cdns,<addr>[,options] 1325 Start an early, polled-mode console on a Cadence 1326 (xuartps) serial port at the specified address. Only 1327 supported option is baud rate. If baud rate is not 1328 specified, the serial port must already be setup and 1329 configured. 1330 1331 uart[8250],io,<addr>[,options[,uartclk]] 1332 uart[8250],mmio,<addr>[,options[,uartclk]] 1333 uart[8250],mmio32,<addr>[,options[,uartclk]] 1334 uart[8250],mmio32be,<addr>[,options[,uartclk]] 1335 uart[8250],0x<addr>[,options] 1336 Start an early, polled-mode console on the 8250/16550 1337 UART at the specified I/O port or MMIO address. 1338 MMIO inter-register address stride is either 8-bit 1339 (mmio) or 32-bit (mmio32 or mmio32be). 1340 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1341 to be equivalent to 'mmio'. 'options' are specified 1342 in the same format described for "console=ttyS<n>"; if 1343 unspecified, the h/w is not initialized. 'uartclk' is 1344 the uart clock frequency; if unspecified, it is set 1345 to 'BASE_BAUD' * 16. 1346 1347 pl011,<addr> 1348 pl011,mmio32,<addr> 1349 Start an early, polled-mode console on a pl011 serial 1350 port at the specified address. The pl011 serial port 1351 must already be setup and configured. Options are not 1352 yet supported. If 'mmio32' is specified, then only 1353 the driver will use only 32-bit accessors to read/write 1354 the device registers. 1355 1356 liteuart,<addr> 1357 Start an early console on a litex serial port at the 1358 specified address. The serial port must already be 1359 setup and configured. Options are not yet supported. 1360 1361 meson,<addr> 1362 Start an early, polled-mode console on a meson serial 1363 port at the specified address. The serial port must 1364 already be setup and configured. Options are not yet 1365 supported. 1366 1367 msm_serial,<addr> 1368 Start an early, polled-mode console on an msm serial 1369 port at the specified address. The serial port 1370 must already be setup and configured. Options are not 1371 yet supported. 1372 1373 msm_serial_dm,<addr> 1374 Start an early, polled-mode console on an msm serial 1375 dm port at the specified address. The serial port 1376 must already be setup and configured. Options are not 1377 yet supported. 1378 1379 owl,<addr> 1380 Start an early, polled-mode console on a serial port 1381 of an Actions Semi SoC, such as S500 or S900, at the 1382 specified address. The serial port must already be 1383 setup and configured. Options are not yet supported. 1384 1385 rda,<addr> 1386 Start an early, polled-mode console on a serial port 1387 of an RDA Micro SoC, such as RDA8810PL, at the 1388 specified address. The serial port must already be 1389 setup and configured. Options are not yet supported. 1390 1391 sbi 1392 Use RISC-V SBI (Supervisor Binary Interface) for early 1393 console. 1394 1395 smh Use ARM semihosting calls for early console. 1396 1397 s3c2410,<addr> 1398 s3c2412,<addr> 1399 s3c2440,<addr> 1400 s3c6400,<addr> 1401 s5pv210,<addr> 1402 exynos4210,<addr> 1403 Use early console provided by serial driver available 1404 on Samsung SoCs, requires selecting proper type and 1405 a correct base address of the selected UART port. The 1406 serial port must already be setup and configured. 1407 Options are not yet supported. 1408 1409 lantiq,<addr> 1410 Start an early, polled-mode console on a lantiq serial 1411 (lqasc) port at the specified address. The serial port 1412 must already be setup and configured. Options are not 1413 yet supported. 1414 1415 lpuart,<addr> 1416 lpuart32,<addr> 1417 Use early console provided by Freescale LP UART driver 1418 found on Freescale Vybrid and QorIQ LS1021A processors. 1419 A valid base address must be provided, and the serial 1420 port must already be setup and configured. 1421 1422 ec_imx21,<addr> 1423 ec_imx6q,<addr> 1424 Start an early, polled-mode, output-only console on the 1425 Freescale i.MX UART at the specified address. The UART 1426 must already be setup and configured. 1427 1428 ar3700_uart,<addr> 1429 Start an early, polled-mode console on the 1430 Armada 3700 serial port at the specified 1431 address. The serial port must already be setup 1432 and configured. Options are not yet supported. 1433 1434 qcom_geni,<addr> 1435 Start an early, polled-mode console on a Qualcomm 1436 Generic Interface (GENI) based serial port at the 1437 specified address. The serial port must already be 1438 setup and configured. Options are not yet supported. 1439 1440 efifb,[options] 1441 Start an early, unaccelerated console on the EFI 1442 memory mapped framebuffer (if available). On cache 1443 coherent non-x86 systems that use system memory for 1444 the framebuffer, pass the 'ram' option so that it is 1445 mapped with the correct attributes. 1446 1447 linflex,<addr> 1448 Use early console provided by Freescale LINFlexD UART 1449 serial driver for NXP S32V234 SoCs. A valid base 1450 address must be provided, and the serial port must 1451 already be setup and configured. 1452 1453 earlyprintk= [X86,SH,ARM,M68k,S390,UM,EARLY] 1454 earlyprintk=vga 1455 earlyprintk=sclp 1456 earlyprintk=xen 1457 earlyprintk=serial[,ttySn[,baudrate]] 1458 earlyprintk=serial[,0x...[,baudrate]] 1459 earlyprintk=ttySn[,baudrate] 1460 earlyprintk=dbgp[debugController#] 1461 earlyprintk=mmio32,membase[,{nocfg|baudrate}] 1462 earlyprintk=pciserial[,force],bus:device.function[,{nocfg|baudrate}] 1463 earlyprintk=xdbc[xhciController#] 1464 earlyprintk=bios 1465 1466 earlyprintk is useful when the kernel crashes before 1467 the normal console is initialized. It is not enabled by 1468 default because it has some cosmetic problems. 1469 1470 Use "nocfg" to skip UART configuration, assume 1471 BIOS/firmware has configured UART correctly. 1472 1473 Append ",keep" to not disable it when the real console 1474 takes over. 1475 1476 Only one of vga, serial, or usb debug port can 1477 be used at a time. 1478 1479 Currently only ttyS0 and ttyS1 may be specified by 1480 name. Other I/O ports may be explicitly specified 1481 on some architectures (x86 and arm at least) by 1482 replacing ttySn with an I/O port address, like this: 1483 earlyprintk=serial,0x1008,115200 1484 You can find the port for a given device in 1485 /proc/tty/driver/serial: 1486 2: uart:ST16650V2 port:00001008 irq:18 ... 1487 1488 Interaction with the standard serial driver is not 1489 very good. 1490 1491 The VGA output is eventually overwritten by 1492 the real console. 1493 1494 The xen option can only be used in Xen domains. 1495 1496 The sclp output can only be used on s390. 1497 1498 The bios output can only be used on SuperH. 1499 1500 The optional "force" to "pciserial" enables use of a 1501 PCI device even when its classcode is not of the 1502 UART class. 1503 1504 edac_report= [HW,EDAC] Control how to report EDAC event 1505 Format: {"on" | "off" | "force"} 1506 on: enable EDAC to report H/W event. May be overridden 1507 by other higher priority error reporting module. 1508 off: disable H/W event reporting through EDAC. 1509 force: enforce the use of EDAC to report H/W event. 1510 default: on. 1511 1512 edd= [EDD] 1513 Format: {"off" | "on" | "skip[mbr]"} 1514 1515 efi= [EFI,EARLY] 1516 Format: { "debug", "disable_early_pci_dma", 1517 "nochunk", "noruntime", "nosoftreserve", 1518 "novamap", "no_disable_early_pci_dma" } 1519 debug: enable misc debug output. 1520 disable_early_pci_dma: disable the busmaster bit on all 1521 PCI bridges while in the EFI boot stub. 1522 nochunk: disable reading files in "chunks" in the EFI 1523 boot stub, as chunking can cause problems with some 1524 firmware implementations. 1525 noruntime : disable EFI runtime services support 1526 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1527 attribute may cause the kernel to reserve the 1528 memory range for a memory mapping driver to 1529 claim. Specify efi=nosoftreserve to disable this 1530 reservation and treat the memory by its base type 1531 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1532 novamap: do not call SetVirtualAddressMap(). 1533 no_disable_early_pci_dma: Leave the busmaster bit set 1534 on all PCI bridges while in the EFI boot stub 1535 1536 efi_no_storage_paranoia [EFI,X86,EARLY] 1537 Using this parameter you can use more than 50% of 1538 your efi variable storage. Use this parameter only if 1539 you are really sure that your UEFI does sane gc and 1540 fulfills the spec otherwise your board may brick. 1541 1542 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1543 that is to be dynamically loaded by Linux. If there are 1544 multiple variables with the same name but with different 1545 vendor GUIDs, all of them will be loaded. See 1546 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1547 1548 1549 eisa_irq_edge= [PARISC,HW] 1550 See header of drivers/parisc/eisa.c. 1551 1552 ekgdboc= [X86,KGDB,EARLY] Allow early kernel console debugging 1553 Format: ekgdboc=kbd 1554 1555 This is designed to be used in conjunction with 1556 the boot argument: earlyprintk=vga 1557 1558 This parameter works in place of the kgdboc parameter 1559 but can only be used if the backing tty is available 1560 very early in the boot process. For early debugging 1561 via a serial port see kgdboc_earlycon instead. 1562 1563 elanfreq= [X86-32] 1564 See comment before function elanfreq_setup() in 1565 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1566 1567 elfcorehdr=[size[KMG]@]offset[KMG] [PPC,SH,X86,S390,EARLY] 1568 Specifies physical address of start of kernel core 1569 image elf header and optionally the size. Generally 1570 kexec loader will pass this option to capture kernel. 1571 See Documentation/admin-guide/kdump/kdump.rst for details. 1572 1573 enable_mtrr_cleanup [X86,EARLY] 1574 The kernel tries to adjust MTRR layout from continuous 1575 to discrete, to make X server driver able to add WB 1576 entry later. This parameter enables that. 1577 1578 enable_timer_pin_1 [X86] 1579 Enable PIN 1 of APIC timer 1580 Can be useful to work around chipset bugs 1581 (in particular on some ATI chipsets). 1582 The kernel tries to set a reasonable default. 1583 1584 enforcing= [SELINUX] Set initial enforcing status. 1585 Format: {"0" | "1"} 1586 See security/selinux/Kconfig help text. 1587 0 -- permissive (log only, no denials). 1588 1 -- enforcing (deny and log). 1589 Default value is 0. 1590 Value can be changed at runtime via 1591 /sys/fs/selinux/enforce. 1592 1593 erst_disable [ACPI] 1594 Disable Error Record Serialization Table (ERST) 1595 support. 1596 1597 ether= [HW,NET] Ethernet cards parameters 1598 This option is obsoleted by the "netdev=" option, which 1599 has equivalent usage. See its documentation for details. 1600 1601 evm= [EVM] 1602 Format: { "fix" } 1603 Permit 'security.evm' to be updated regardless of 1604 current integrity status. 1605 1606 early_page_ext [KNL,EARLY] Enforces page_ext initialization to earlier 1607 stages so cover more early boot allocations. 1608 Please note that as side effect some optimizations 1609 might be disabled to achieve that (e.g. parallelized 1610 memory initialization is disabled) so the boot process 1611 might take longer, especially on systems with a lot of 1612 memory. Available with CONFIG_PAGE_EXTENSION=y. 1613 1614 failslab= 1615 fail_usercopy= 1616 fail_page_alloc= 1617 fail_skb_realloc= 1618 fail_make_request=[KNL] 1619 General fault injection mechanism. 1620 Format: <interval>,<probability>,<space>,<times> 1621 See also Documentation/fault-injection/. 1622 1623 fb_tunnels= [NET] 1624 Format: { initns | none } 1625 See Documentation/admin-guide/sysctl/net.rst for 1626 fb_tunnels_only_for_init_ns 1627 1628 floppy= [HW] 1629 See Documentation/admin-guide/blockdev/floppy.rst. 1630 1631 forcepae [X86-32] 1632 Forcefully enable Physical Address Extension (PAE). 1633 Many Pentium M systems disable PAE but may have a 1634 functionally usable PAE implementation. 1635 Warning: use of this parameter will taint the kernel 1636 and may cause unknown problems. 1637 1638 fred= [X86-64] 1639 Enable/disable Flexible Return and Event Delivery. 1640 Format: { on | off } 1641 on: enable FRED when it's present. 1642 off: disable FRED, the default setting. 1643 1644 ftrace=[tracer] 1645 [FTRACE] will set and start the specified tracer 1646 as early as possible in order to facilitate early 1647 boot debugging. 1648 1649 ftrace_boot_snapshot 1650 [FTRACE] On boot up, a snapshot will be taken of the 1651 ftrace ring buffer that can be read at: 1652 /sys/kernel/tracing/snapshot. 1653 This is useful if you need tracing information from kernel 1654 boot up that is likely to be overridden by user space 1655 start up functionality. 1656 1657 Optionally, the snapshot can also be defined for a tracing 1658 instance that was created by the trace_instance= command 1659 line parameter. 1660 1661 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo 1662 1663 The above will cause the "foo" tracing instance to trigger 1664 a snapshot at the end of boot up. 1665 1666 ftrace_dump_on_oops[=2(orig_cpu) | =<instance>][,<instance> | 1667 ,<instance>=2(orig_cpu)] 1668 [FTRACE] will dump the trace buffers on oops. 1669 If no parameter is passed, ftrace will dump global 1670 buffers of all CPUs, if you pass 2 or orig_cpu, it 1671 will dump only the buffer of the CPU that triggered 1672 the oops, or the specific instance will be dumped if 1673 its name is passed. Multiple instance dump is also 1674 supported, and instances are separated by commas. Each 1675 instance supports only dump on CPU that triggered the 1676 oops by passing 2 or orig_cpu to it. 1677 1678 ftrace_dump_on_oops=foo=orig_cpu 1679 1680 The above will dump only the buffer of "foo" instance 1681 on CPU that triggered the oops. 1682 1683 ftrace_dump_on_oops,foo,bar=orig_cpu 1684 1685 The above will dump global buffer on all CPUs, the 1686 buffer of "foo" instance on all CPUs and the buffer 1687 of "bar" instance on CPU that triggered the oops. 1688 1689 ftrace_filter=[function-list] 1690 [FTRACE] Limit the functions traced by the function 1691 tracer at boot up. function-list is a comma-separated 1692 list of functions. This list can be changed at run 1693 time by the set_ftrace_filter file in the debugfs 1694 tracing directory. 1695 1696 ftrace_notrace=[function-list] 1697 [FTRACE] Do not trace the functions specified in 1698 function-list. This list can be changed at run time 1699 by the set_ftrace_notrace file in the debugfs 1700 tracing directory. 1701 1702 ftrace_graph_filter=[function-list] 1703 [FTRACE] Limit the top level callers functions traced 1704 by the function graph tracer at boot up. 1705 function-list is a comma-separated list of functions 1706 that can be changed at run time by the 1707 set_graph_function file in the debugfs tracing directory. 1708 1709 ftrace_graph_notrace=[function-list] 1710 [FTRACE] Do not trace from the functions specified in 1711 function-list. This list is a comma-separated list of 1712 functions that can be changed at run time by the 1713 set_graph_notrace file in the debugfs tracing directory. 1714 1715 ftrace_graph_max_depth=<uint> 1716 [FTRACE] Used with the function graph tracer. This is 1717 the max depth it will trace into a function. This value 1718 can be changed at run time by the max_graph_depth file 1719 in the tracefs tracing directory. default: 0 (no limit) 1720 1721 fw_devlink= [KNL,EARLY] Create device links between consumer and supplier 1722 devices by scanning the firmware to infer the 1723 consumer/supplier relationships. This feature is 1724 especially useful when drivers are loaded as modules as 1725 it ensures proper ordering of tasks like device probing 1726 (suppliers first, then consumers), supplier boot state 1727 clean up (only after all consumers have probed), 1728 suspend/resume & runtime PM (consumers first, then 1729 suppliers). 1730 Format: { off | permissive | on | rpm } 1731 off -- Don't create device links from firmware info. 1732 permissive -- Create device links from firmware info 1733 but use it only for ordering boot state clean 1734 up (sync_state() calls). 1735 on -- Create device links from firmware info and use it 1736 to enforce probe and suspend/resume ordering. 1737 rpm -- Like "on", but also use to order runtime PM. 1738 1739 fw_devlink.strict=<bool> 1740 [KNL,EARLY] Treat all inferred dependencies as mandatory 1741 dependencies. This only applies for fw_devlink=on|rpm. 1742 Format: <bool> 1743 1744 fw_devlink.sync_state = 1745 [KNL,EARLY] When all devices that could probe have finished 1746 probing, this parameter controls what to do with 1747 devices that haven't yet received their sync_state() 1748 calls. 1749 Format: { strict | timeout } 1750 strict -- Default. Continue waiting on consumers to 1751 probe successfully. 1752 timeout -- Give up waiting on consumers and call 1753 sync_state() on any devices that haven't yet 1754 received their sync_state() calls after 1755 deferred_probe_timeout has expired or by 1756 late_initcall() if !CONFIG_MODULES. 1757 1758 gamecon.map[2|3]= 1759 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1760 support via parallel port (up to 5 devices per port) 1761 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1762 See also Documentation/input/devices/joystick-parport.rst 1763 1764 gamma= [HW,DRM] 1765 1766 gart_fix_e820= [X86-64,EARLY] disable the fix e820 for K8 GART 1767 Format: off | on 1768 default: on 1769 1770 gather_data_sampling= 1771 [X86,INTEL,EARLY] Control the Gather Data Sampling (GDS) 1772 mitigation. 1773 1774 Gather Data Sampling is a hardware vulnerability which 1775 allows unprivileged speculative access to data which was 1776 previously stored in vector registers. 1777 1778 This issue is mitigated by default in updated microcode. 1779 The mitigation may have a performance impact but can be 1780 disabled. On systems without the microcode mitigation 1781 disabling AVX serves as a mitigation. 1782 1783 force: Disable AVX to mitigate systems without 1784 microcode mitigation. No effect if the microcode 1785 mitigation is present. Known to cause crashes in 1786 userspace with buggy AVX enumeration. 1787 1788 off: Disable GDS mitigation. 1789 1790 gbpages [X86] Use GB pages for kernel direct mappings. 1791 1792 gcov_persist= [GCOV] When non-zero (default), profiling data for 1793 kernel modules is saved and remains accessible via 1794 debugfs, even when the module is unloaded/reloaded. 1795 When zero, profiling data is discarded and associated 1796 debugfs files are removed at module unload time. 1797 1798 goldfish [X86] Enable the goldfish android emulator platform. 1799 Don't use this when you are not running on the 1800 android emulator 1801 1802 gpio-mockup.gpio_mockup_ranges 1803 [HW] Sets the ranges of gpiochip of for this device. 1804 Format: <start1>,<end1>,<start2>,<end2>... 1805 gpio-mockup.gpio_mockup_named_lines 1806 [HW] Let the driver know GPIO lines should be named. 1807 1808 gpt [EFI] Forces disk with valid GPT signature but 1809 invalid Protective MBR to be treated as GPT. If the 1810 primary GPT is corrupted, it enables the backup/alternate 1811 GPT to be used instead. 1812 1813 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1814 the "Enable 0" bit of the configuration register. 1815 Format: 0 | 1 1816 Default: 0 1817 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1818 the "Enable 0" bit of the configuration register. 1819 Format: 0 | 1 1820 Default: 0 1821 grcan.select= [HW] Select which physical interface to use. 1822 Format: 0 | 1 1823 Default: 0 1824 grcan.txsize= [HW] Sets the size of the tx buffer. 1825 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1826 Default: 1024 1827 grcan.rxsize= [HW] Sets the size of the rx buffer. 1828 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1829 Default: 1024 1830 1831 hardened_usercopy= 1832 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 1833 hardening is enabled for this boot. Hardened 1834 usercopy checking is used to protect the kernel 1835 from reading or writing beyond known memory 1836 allocation boundaries as a proactive defense 1837 against bounds-checking flaws in the kernel's 1838 copy_to_user()/copy_from_user() interface. 1839 The default is determined by 1840 CONFIG_HARDENED_USERCOPY_DEFAULT_ON. 1841 on Perform hardened usercopy checks. 1842 off Disable hardened usercopy checks. 1843 1844 hardlockup_all_cpu_backtrace= 1845 [KNL] Should the hard-lockup detector generate 1846 backtraces on all cpus. 1847 Format: 0 | 1 1848 1849 hash_pointers= 1850 [KNL,EARLY] 1851 By default, when pointers are printed to the console 1852 or buffers via the %p format string, that pointer is 1853 "hashed", i.e. obscured by hashing the pointer value. 1854 This is a security feature that hides actual kernel 1855 addresses from unprivileged users, but it also makes 1856 debugging the kernel more difficult since unequal 1857 pointers can no longer be compared. The choices are: 1858 Format: { auto | always | never } 1859 Default: auto 1860 1861 auto - Hash pointers unless slab_debug is enabled. 1862 always - Always hash pointers (even if slab_debug is 1863 enabled). 1864 never - Never hash pointers. This option should only 1865 be specified when debugging the kernel. Do 1866 not use on production kernels. The boot 1867 param "no_hash_pointers" is an alias for 1868 this mode. 1869 1870 hashdist= [KNL,NUMA] Large hashes allocated during boot 1871 are distributed across NUMA nodes. Defaults on 1872 for 64-bit NUMA, off otherwise. 1873 Format: 0 | 1 (for off | on) 1874 1875 hd= [EIDE] (E)IDE hard drive subsystem geometry 1876 Format: <cyl>,<head>,<sect> 1877 1878 hest_disable [ACPI] 1879 Disable Hardware Error Source Table (HEST) support; 1880 corresponding firmware-first mode error processing 1881 logic will be disabled. 1882 1883 hibernate= [HIBERNATION] 1884 noresume Don't check if there's a hibernation image 1885 present during boot. 1886 nocompress Don't compress/decompress hibernation images. 1887 no Disable hibernation and resume. 1888 protect_image Turn on image protection during restoration 1889 (that will set all pages holding image data 1890 during restoration read-only). 1891 1892 hibernate.compressor= [HIBERNATION] Compression algorithm to be 1893 used with hibernation. 1894 Format: { lzo | lz4 } 1895 Default: lzo 1896 1897 lzo: Select LZO compression algorithm to 1898 compress/decompress hibernation image. 1899 1900 lz4: Select LZ4 compression algorithm to 1901 compress/decompress hibernation image. 1902 1903 hibernate.pm_test_delay= 1904 [HIBERNATION] 1905 Sets the number of seconds to remain in a hibernation test 1906 mode before resuming the system (see 1907 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 1908 is set. Default value is 5. 1909 1910 hibernate_compression_threads= 1911 [HIBERNATION] 1912 Set the number of threads used for compressing or decompressing 1913 hibernation images. 1914 1915 Format: <integer> 1916 Default: 3 1917 Minimum: 1 1918 Example: hibernate_compression_threads=4 1919 1920 highmem=nn[KMG] [KNL,BOOT,EARLY] forces the highmem zone to have an exact 1921 size of <nn>. This works even on boxes that have no 1922 highmem otherwise. This also works to reduce highmem 1923 size on bigger boxes. 1924 1925 highres= [KNL] Enable/disable high resolution timer mode. 1926 Valid parameters: "on", "off" 1927 Default: "on" 1928 1929 hlt [BUGS=ARM,SH] 1930 1931 hostname= [KNL,EARLY] Set the hostname (aka UTS nodename). 1932 Format: <string> 1933 This allows setting the system's hostname during early 1934 startup. This sets the name returned by gethostname. 1935 Using this parameter to set the hostname makes it 1936 possible to ensure the hostname is correctly set before 1937 any userspace processes run, avoiding the possibility 1938 that a process may call gethostname before the hostname 1939 has been explicitly set, resulting in the calling 1940 process getting an incorrect result. The string must 1941 not exceed the maximum allowed hostname length (usually 1942 64 characters) and will be truncated otherwise. 1943 1944 hpet= [X86-32,HPET] option to control HPET usage 1945 Format: { enable (default) | disable | force | 1946 verbose } 1947 disable: disable HPET and use PIT instead 1948 force: allow force enabled of undocumented chips (ICH4, 1949 VIA, nVidia) 1950 verbose: show contents of HPET registers during setup 1951 1952 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 1953 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 1954 1955 hugepages= [HW,EARLY] Number of HugeTLB pages to allocate at boot. 1956 If this follows hugepagesz (below), it specifies 1957 the number of pages of hugepagesz to be allocated. 1958 If this is the first HugeTLB parameter on the command 1959 line, it specifies the number of pages to allocate for 1960 the default huge page size. If using node format, the 1961 number of pages to allocate per-node can be specified. 1962 See also Documentation/admin-guide/mm/hugetlbpage.rst. 1963 Format: <integer> or (node format) 1964 <node>:<integer>[,<node>:<integer>] 1965 1966 hugepagesz= 1967 [HW,EARLY] The size of the HugeTLB pages. This is 1968 used in conjunction with hugepages (above) to 1969 allocate huge pages of a specific size at boot. The 1970 pair hugepagesz=X hugepages=Y can be specified once 1971 for each supported huge page size. Huge page sizes 1972 are architecture dependent. See also 1973 Documentation/admin-guide/mm/hugetlbpage.rst. 1974 Format: size[KMG] 1975 1976 hugepage_alloc_threads= 1977 [HW] The number of threads that should be used to 1978 allocate hugepages during boot. This option can be 1979 used to improve system bootup time when allocating 1980 a large amount of huge pages. 1981 The default value is 25% of the available hardware threads. 1982 1983 Note that this parameter only applies to non-gigantic huge pages. 1984 1985 hugetlb_cma= [HW,CMA,EARLY] The size of a CMA area used for allocation 1986 of gigantic hugepages. Or using node format, the size 1987 of a CMA area per node can be specified. 1988 Format: nn[KMGTPE] or (node format) 1989 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]] 1990 1991 Reserve a CMA area of given size and allocate gigantic 1992 hugepages using the CMA allocator. If enabled, the 1993 boot-time allocation of gigantic hugepages is skipped. 1994 1995 hugetlb_cma_only= 1996 [HW,CMA,EARLY] When allocating new HugeTLB pages, only 1997 try to allocate from the CMA areas. 1998 1999 This option does nothing if hugetlb_cma= is not also 2000 specified. 2001 2002 hugetlb_free_vmemmap= 2003 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP 2004 enabled. 2005 Control if HugeTLB Vmemmap Optimization (HVO) is enabled. 2006 Allows heavy hugetlb users to free up some more 2007 memory (7 * PAGE_SIZE for each 2MB hugetlb page). 2008 Format: { on | off (default) } 2009 2010 on: enable HVO 2011 off: disable HVO 2012 2013 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y, 2014 the default is on. 2015 2016 Note that the vmemmap pages may be allocated from the added 2017 memory block itself when memory_hotplug.memmap_on_memory is 2018 enabled, those vmemmap pages cannot be optimized even if this 2019 feature is enabled. Other vmemmap pages not allocated from 2020 the added memory block itself do not be affected. 2021 2022 hung_task_panic= 2023 [KNL] Should the hung task detector generate panics. 2024 Format: 0 | 1 2025 2026 A value of 1 instructs the kernel to panic when a 2027 hung task is detected. The default value is controlled 2028 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time 2029 option. The value selected by this boot parameter can 2030 be changed later by the kernel.hung_task_panic sysctl. 2031 2032 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 2033 terminal devices. Valid values: 0..8 2034 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 2035 If specified, z/VM IUCV HVC accepts connections 2036 from listed z/VM user IDs only. 2037 2038 hv_nopvspin [X86,HYPER_V,EARLY] 2039 Disables the paravirt spinlock optimizations 2040 which allow the hypervisor to 'idle' the guest 2041 on lock contention. 2042 2043 hw_protection= [HW] 2044 Format: reboot | shutdown 2045 2046 Hardware protection action taken on critical events like 2047 overtemperature or imminent voltage loss. 2048 2049 i2c_bus= [HW] Override the default board specific I2C bus speed 2050 or register an additional I2C bus that is not 2051 registered from board initialization code. 2052 Format: 2053 <bus_id>,<clkrate> 2054 2055 i2c_touchscreen_props= [HW,ACPI,X86] 2056 Set device-properties for ACPI-enumerated I2C-attached 2057 touchscreen, to e.g. fix coordinates of upside-down 2058 mounted touchscreens. If you need this option please 2059 submit a drivers/platform/x86/touchscreen_dmi.c patch 2060 adding a DMI quirk for this. 2061 2062 Format: 2063 <ACPI_HW_ID>:<prop_name>=<val>[:prop_name=val][:...] 2064 Where <val> is one of: 2065 Omit "=<val>" entirely Set a boolean device-property 2066 Unsigned number Set a u32 device-property 2067 Anything else Set a string device-property 2068 2069 Examples (split over multiple lines): 2070 i2c_touchscreen_props=GDIX1001:touchscreen-inverted-x: 2071 touchscreen-inverted-y 2072 2073 i2c_touchscreen_props=MSSL1680:touchscreen-size-x=1920: 2074 touchscreen-size-y=1080:touchscreen-inverted-y: 2075 firmware-name=gsl1680-vendor-model.fw:silead,home-button 2076 2077 i8042.debug [HW] Toggle i8042 debug mode 2078 i8042.unmask_kbd_data 2079 [HW] Enable printing of interrupt data from the KBD port 2080 (disabled by default, and as a pre-condition 2081 requires that i8042.debug=1 be enabled) 2082 i8042.direct [HW] Put keyboard port into non-translated mode 2083 i8042.dumbkbd [HW] Pretend that controller can only read data from 2084 keyboard and cannot control its state 2085 (Don't attempt to blink the leds) 2086 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 2087 i8042.nokbd [HW] Don't check/create keyboard port 2088 i8042.noloop [HW] Disable the AUX Loopback command while probing 2089 for the AUX port 2090 i8042.nomux [HW] Don't check presence of an active multiplexing 2091 controller 2092 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 2093 controllers 2094 i8042.notimeout [HW] Ignore timeout condition signalled by controller 2095 i8042.reset [HW] Reset the controller during init, cleanup and 2096 suspend-to-ram transitions, only during s2r 2097 transitions, or never reset 2098 Format: { 1 | Y | y | 0 | N | n } 2099 1, Y, y: always reset controller 2100 0, N, n: don't ever reset controller 2101 Default: only on s2r transitions on x86; most other 2102 architectures force reset to be always executed 2103 i8042.unlock [HW] Unlock (ignore) the keylock 2104 i8042.kbdreset [HW] Reset device connected to KBD port 2105 i8042.probe_defer 2106 [HW] Allow deferred probing upon i8042 probe errors 2107 2108 i810= [HW,DRM] 2109 2110 i915.invert_brightness= 2111 [DRM] Invert the sense of the variable that is used to 2112 set the brightness of the panel backlight. Normally a 2113 brightness value of 0 indicates backlight switched off, 2114 and the maximum of the brightness value sets the backlight 2115 to maximum brightness. If this parameter is set to 0 2116 (default) and the machine requires it, or this parameter 2117 is set to 1, a brightness value of 0 sets the backlight 2118 to maximum brightness, and the maximum of the brightness 2119 value switches the backlight off. 2120 -1 -- never invert brightness 2121 0 -- machine default 2122 1 -- force brightness inversion 2123 2124 ia32_emulation= [X86-64] 2125 Format: <bool> 2126 When true, allows loading 32-bit programs and executing 32-bit 2127 syscalls, essentially overriding IA32_EMULATION_DEFAULT_DISABLED at 2128 boot time. When false, unconditionally disables IA32 emulation. 2129 2130 icn= [HW,ISDN] 2131 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 2132 2133 2134 idle= [X86,EARLY] 2135 Format: idle=poll, idle=halt, idle=nomwait 2136 2137 idle=poll: Don't do power saving in the idle loop 2138 using HLT, but poll for rescheduling event. This will 2139 make the CPUs eat a lot more power, but may be useful 2140 to get slightly better performance in multiprocessor 2141 benchmarks. It also makes some profiling using 2142 performance counters more accurate. Please note that 2143 on systems with MONITOR/MWAIT support (like Intel 2144 EM64T CPUs) this option has no performance advantage 2145 over the normal idle loop. It may also interact badly 2146 with hyperthreading. 2147 2148 idle=halt: Halt is forced to be used for CPU idle. 2149 In such case C2/C3 won't be used again. 2150 2151 idle=nomwait: Disable mwait for CPU C-states 2152 2153 idxd.sva= [HW] 2154 Format: <bool> 2155 Allow force disabling of Shared Virtual Memory (SVA) 2156 support for the idxd driver. By default it is set to 2157 true (1). 2158 2159 idxd.tc_override= [HW] 2160 Format: <bool> 2161 Allow override of default traffic class configuration 2162 for the device. By default it is set to false (0). 2163 2164 ieee754= [MIPS] Select IEEE Std 754 conformance mode 2165 Format: { strict | legacy | 2008 | relaxed | emulated } 2166 Default: strict 2167 2168 Choose which programs will be accepted for execution 2169 based on the IEEE 754 NaN encoding(s) supported by 2170 the FPU and the NaN encoding requested with the value 2171 of an ELF file header flag individually set by each 2172 binary. Hardware implementations are permitted to 2173 support either or both of the legacy and the 2008 NaN 2174 encoding mode. 2175 2176 Available settings are as follows: 2177 strict accept binaries that request a NaN encoding 2178 supported by the FPU 2179 legacy only accept legacy-NaN binaries, if supported 2180 by the FPU 2181 2008 only accept 2008-NaN binaries, if supported 2182 by the FPU 2183 relaxed accept any binaries regardless of whether 2184 supported by the FPU 2185 emulated accept any binaries but enable FPU emulator 2186 if binary mode is unsupported by the FPU. 2187 2188 The FPU emulator is always able to support both NaN 2189 encodings, so if no FPU hardware is present or it has 2190 been disabled with 'nofpu', then the settings of 2191 'legacy' and '2008' strap the emulator accordingly, 2192 'relaxed' straps the emulator for both legacy-NaN and 2193 2008-NaN, whereas 'strict' enables legacy-NaN only on 2194 legacy processors and both NaN encodings on MIPS32 or 2195 MIPS64 CPUs. 2196 2197 The setting for ABS.fmt/NEG.fmt instruction execution 2198 mode generally follows that for the NaN encoding, 2199 except where unsupported by hardware. 2200 2201 ignore_loglevel [KNL,EARLY] 2202 Ignore loglevel setting - this will print /all/ 2203 kernel messages to the console. Useful for debugging. 2204 We also add it as printk module parameter, so users 2205 could change it dynamically, usually by 2206 /sys/module/printk/parameters/ignore_loglevel. 2207 2208 ignore_rlimit_data 2209 Ignore RLIMIT_DATA setting for data mappings, 2210 print warning at first misuse. Can be changed via 2211 /sys/module/kernel/parameters/ignore_rlimit_data. 2212 2213 ihash_entries= [KNL] 2214 Set number of hash buckets for inode cache. 2215 2216 ima_appraise= [IMA] appraise integrity measurements 2217 Format: { "off" | "enforce" | "fix" | "log" } 2218 default: "enforce" 2219 2220 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 2221 The builtin appraise policy appraises all files 2222 owned by uid=0. 2223 2224 ima_canonical_fmt [IMA] 2225 Use the canonical format for the binary runtime 2226 measurements, instead of host native format. 2227 2228 ima_hash= [IMA] 2229 Format: { md5 | sha1 | rmd160 | sha256 | sha384 2230 | sha512 | ... } 2231 default: "sha1" 2232 2233 The list of supported hash algorithms is defined 2234 in crypto/hash_info.h. 2235 2236 ima_policy= [IMA] 2237 The builtin policies to load during IMA setup. 2238 Format: "tcb | appraise_tcb | secure_boot | 2239 fail_securely | critical_data" 2240 2241 The "tcb" policy measures all programs exec'd, files 2242 mmap'd for exec, and all files opened with the read 2243 mode bit set by either the effective uid (euid=0) or 2244 uid=0. 2245 2246 The "appraise_tcb" policy appraises the integrity of 2247 all files owned by root. 2248 2249 The "secure_boot" policy appraises the integrity 2250 of files (eg. kexec kernel image, kernel modules, 2251 firmware, policy, etc) based on file signatures. 2252 2253 The "fail_securely" policy forces file signature 2254 verification failure also on privileged mounted 2255 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 2256 flag. 2257 2258 The "critical_data" policy measures kernel integrity 2259 critical data. 2260 2261 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 2262 Load a policy which meets the needs of the Trusted 2263 Computing Base. This means IMA will measure all 2264 programs exec'd, files mmap'd for exec, and all files 2265 opened for read by uid=0. 2266 2267 ima_template= [IMA] 2268 Select one of defined IMA measurements template formats. 2269 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" | 2270 "ima-sigv2" } 2271 Default: "ima-ng" 2272 2273 ima_template_fmt= 2274 [IMA] Define a custom template format. 2275 Format: { "field1|...|fieldN" } 2276 2277 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 2278 Format: <min_file_size> 2279 Set the minimal file size for using asynchronous hash. 2280 If left unspecified, ahash usage is disabled. 2281 2282 ahash performance varies for different data sizes on 2283 different crypto accelerators. This option can be used 2284 to achieve the best performance for a particular HW. 2285 2286 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size 2287 Format: <bufsize> 2288 Set hashing buffer size. Default: 4k. 2289 2290 ahash performance varies for different chunk sizes on 2291 different crypto accelerators. This option can be used 2292 to achieve best performance for particular HW. 2293 2294 ima= [IMA] Enable or disable IMA 2295 Format: { "off" | "on" } 2296 Default: "on" 2297 Note that disabling IMA is limited to kdump kernel. 2298 2299 indirect_target_selection= [X86,Intel] Mitigation control for Indirect 2300 Target Selection(ITS) bug in Intel CPUs. Updated 2301 microcode is also required for a fix in IBPB. 2302 2303 on: Enable mitigation (default). 2304 off: Disable mitigation. 2305 force: Force the ITS bug and deploy default 2306 mitigation. 2307 vmexit: Only deploy mitigation if CPU is affected by 2308 guest/host isolation part of ITS. 2309 stuff: Deploy RSB-fill mitigation when retpoline is 2310 also deployed. Otherwise, deploy the default 2311 mitigation. 2312 2313 For details see: 2314 Documentation/admin-guide/hw-vuln/indirect-target-selection.rst 2315 2316 init= [KNL] 2317 Format: <full_path> 2318 Run specified binary instead of /sbin/init as init 2319 process. 2320 2321 initcall_debug [KNL] Trace initcalls as they are executed. Useful 2322 for working out where the kernel is dying during 2323 startup. 2324 2325 initcall_blacklist= [KNL] Do not execute a comma-separated list of 2326 initcall functions. Useful for debugging built-in 2327 modules and initcalls. 2328 2329 initramfs_async= [KNL] 2330 Format: <bool> 2331 Default: 1 2332 This parameter controls whether the initramfs 2333 image is unpacked asynchronously, concurrently 2334 with devices being probed and 2335 initialized. This should normally just work, 2336 but as a debugging aid, one can get the 2337 historical behaviour of the initramfs 2338 unpacking being completed before device_ and 2339 late_ initcalls. 2340 2341 initrd= [BOOT,EARLY] Specify the location of the initial ramdisk 2342 2343 initrdmem= [KNL,EARLY] Specify a physical address and size from which to 2344 load the initrd. If an initrd is compiled in or 2345 specified in the bootparams, it takes priority over this 2346 setting. 2347 Format: ss[KMG],nn[KMG] 2348 Default is 0, 0 2349 2350 init_on_alloc= [MM,EARLY] Fill newly allocated pages and heap objects with 2351 zeroes. 2352 Format: 0 | 1 2353 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 2354 2355 init_on_free= [MM,EARLY] Fill freed pages and heap objects with zeroes. 2356 Format: 0 | 1 2357 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 2358 2359 init_pkru= [X86] Specify the default memory protection keys rights 2360 register contents for all processes. 0x55555554 by 2361 default (disallow access to all but pkey 0). Can 2362 override in debugfs after boot. 2363 2364 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 2365 Format: <irq> 2366 2367 int_pln_enable [X86] Enable power limit notification interrupt 2368 2369 integrity_audit=[IMA] 2370 Format: { "0" | "1" } 2371 0 -- basic integrity auditing messages. (Default) 2372 1 -- additional integrity auditing messages. 2373 2374 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 2375 on 2376 Enable intel iommu driver. 2377 off 2378 Disable intel iommu driver. 2379 igfx_off [Default Off] 2380 By default, gfx is mapped as normal device. If a gfx 2381 device has a dedicated DMAR unit, the DMAR unit is 2382 bypassed by not enabling DMAR with this option. In 2383 this case, gfx device will use physical address for 2384 DMA. 2385 strict [Default Off] 2386 Deprecated, equivalent to iommu.strict=1. 2387 sp_off [Default Off] 2388 By default, super page will be supported if Intel IOMMU 2389 has the capability. With this option, super page will 2390 not be supported. 2391 sm_on 2392 Enable the Intel IOMMU scalable mode if the hardware 2393 advertises that it has support for the scalable mode 2394 translation. 2395 sm_off 2396 Disallow use of the Intel IOMMU scalable mode. 2397 tboot_noforce [Default Off] 2398 Do not force the Intel IOMMU enabled under tboot. 2399 By default, tboot will force Intel IOMMU on, which 2400 could harm performance of some high-throughput 2401 devices like 40GBit network cards, even if identity 2402 mapping is enabled. 2403 Note that using this option lowers the security 2404 provided by tboot because it makes the system 2405 vulnerable to DMA attacks. 2406 2407 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 2408 0 disables intel_idle and fall back on acpi_idle. 2409 1 to 9 specify maximum depth of C-state. 2410 2411 intel_pstate= [X86,EARLY] 2412 disable 2413 Do not enable intel_pstate as the default 2414 scaling driver for the supported processors 2415 active 2416 Use intel_pstate driver to bypass the scaling 2417 governors layer of cpufreq and provides it own 2418 algorithms for p-state selection. There are two 2419 P-state selection algorithms provided by 2420 intel_pstate in the active mode: powersave and 2421 performance. The way they both operate depends 2422 on whether or not the hardware managed P-states 2423 (HWP) feature has been enabled in the processor 2424 and possibly on the processor model. 2425 passive 2426 Use intel_pstate as a scaling driver, but configure it 2427 to work with generic cpufreq governors (instead of 2428 enabling its internal governor). This mode cannot be 2429 used along with the hardware-managed P-states (HWP) 2430 feature. 2431 force 2432 Enable intel_pstate on systems that prohibit it by default 2433 in favor of acpi-cpufreq. Forcing the intel_pstate driver 2434 instead of acpi-cpufreq may disable platform features, such 2435 as thermal controls and power capping, that rely on ACPI 2436 P-States information being indicated to OSPM and therefore 2437 should be used with caution. This option does not work with 2438 processors that aren't supported by the intel_pstate driver 2439 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 2440 no_hwp 2441 Do not enable hardware P state control (HWP) 2442 if available. 2443 hwp_only 2444 Only load intel_pstate on systems which support 2445 hardware P state control (HWP) if available. 2446 support_acpi_ppc 2447 Enforce ACPI _PPC performance limits. If the Fixed ACPI 2448 Description Table, specifies preferred power management 2449 profile as "Enterprise Server" or "Performance Server", 2450 then this feature is turned on by default. 2451 per_cpu_perf_limits 2452 Allow per-logical-CPU P-State performance control limits using 2453 cpufreq sysfs interface 2454 no_cas 2455 Do not enable capacity-aware scheduling (CAS) on 2456 hybrid systems 2457 2458 intremap= [X86-64,Intel-IOMMU,EARLY] 2459 on enable Interrupt Remapping (default) 2460 off disable Interrupt Remapping 2461 nosid disable Source ID checking 2462 no_x2apic_optout 2463 BIOS x2APIC opt-out request will be ignored 2464 nopost disable Interrupt Posting 2465 posted_msi 2466 enable MSIs delivered as posted interrupts 2467 2468 iomem= Disable strict checking of access to MMIO memory 2469 strict regions from userspace. 2470 relaxed 2471 2472 iommu= [X86,EARLY] 2473 2474 off 2475 Don't initialize and use any kind of IOMMU. 2476 2477 force 2478 Force the use of the hardware IOMMU even when 2479 it is not actually needed (e.g. because < 3 GB 2480 memory). 2481 2482 noforce 2483 Don't force hardware IOMMU usage when it is not 2484 needed. (default). 2485 2486 biomerge 2487 panic 2488 nopanic 2489 merge 2490 nomerge 2491 2492 soft 2493 Use software bounce buffering (SWIOTLB) (default for 2494 Intel machines). This can be used to prevent the usage 2495 of an available hardware IOMMU. 2496 2497 [X86] 2498 pt 2499 [X86] 2500 nopt 2501 [PPC/POWERNV] 2502 nobypass 2503 Disable IOMMU bypass, using IOMMU for PCI devices. 2504 2505 [X86] 2506 AMD Gart HW IOMMU-specific options: 2507 2508 <size> 2509 Set the size of the remapping area in bytes. 2510 2511 allowed 2512 Overwrite iommu off workarounds for specific chipsets 2513 2514 fullflush 2515 Flush IOMMU on each allocation (default). 2516 2517 nofullflush 2518 Don't use IOMMU fullflush. 2519 2520 memaper[=<order>] 2521 Allocate an own aperture over RAM with size 2522 32MB<<order. (default: order=1, i.e. 64MB) 2523 2524 merge 2525 Do scatter-gather (SG) merging. Implies "force" 2526 (experimental). 2527 2528 nomerge 2529 Don't do scatter-gather (SG) merging. 2530 2531 noaperture 2532 Ask the IOMMU not to touch the aperture for AGP. 2533 2534 noagp 2535 Don't initialize the AGP driver and use full aperture. 2536 2537 panic 2538 Always panic when IOMMU overflows. 2539 2540 iommu.forcedac= [ARM64,X86,EARLY] Control IOVA allocation for PCI devices. 2541 Format: { "0" | "1" } 2542 0 - Try to allocate a 32-bit DMA address first, before 2543 falling back to the full range if needed. 2544 1 - Allocate directly from the full usable range, 2545 forcing Dual Address Cycle for PCI cards supporting 2546 greater than 32-bit addressing. 2547 2548 iommu.strict= [ARM64,X86,S390,EARLY] Configure TLB invalidation behaviour 2549 Format: { "0" | "1" } 2550 0 - Lazy mode. 2551 Request that DMA unmap operations use deferred 2552 invalidation of hardware TLBs, for increased 2553 throughput at the cost of reduced device isolation. 2554 Will fall back to strict mode if not supported by 2555 the relevant IOMMU driver. 2556 1 - Strict mode. 2557 DMA unmap operations invalidate IOMMU hardware TLBs 2558 synchronously. 2559 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}. 2560 Note: on x86, strict mode specified via one of the 2561 legacy driver-specific options takes precedence. 2562 2563 iommu.passthrough= 2564 [ARM64,X86,EARLY] Configure DMA to bypass the IOMMU by default. 2565 Format: { "0" | "1" } 2566 0 - Use IOMMU translation for DMA. 2567 1 - Bypass the IOMMU for DMA. 2568 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2569 2570 io7= [HW] IO7 for Marvel-based Alpha systems 2571 See comment before marvel_specify_io7 in 2572 arch/alpha/kernel/core_marvel.c. 2573 2574 io_delay= [X86,EARLY] I/O delay method 2575 0x80 2576 Standard port 0x80 based delay 2577 0xed 2578 Alternate port 0xed based delay (needed on some systems) 2579 udelay 2580 Simple two microseconds delay 2581 none 2582 No delay 2583 2584 ip= [IP_PNP] 2585 See Documentation/admin-guide/nfs/nfsroot.rst. 2586 2587 ipcmni_extend [KNL,EARLY] Extend the maximum number of unique System V 2588 IPC identifiers from 32,768 to 16,777,216. 2589 2590 ipe.enforce= [IPE] 2591 Format: <bool> 2592 Determine whether IPE starts in permissive (0) or 2593 enforce (1) mode. The default is enforce. 2594 2595 ipe.success_audit= 2596 [IPE] 2597 Format: <bool> 2598 Start IPE with success auditing enabled, emitting 2599 an audit event when a binary is allowed. The default 2600 is 0. 2601 2602 irqaffinity= [SMP] Set the default irq affinity mask 2603 The argument is a cpu list, as described above. 2604 2605 irqchip.gicv2_force_probe= 2606 [ARM,ARM64,EARLY] 2607 Format: <bool> 2608 Force the kernel to look for the second 4kB page 2609 of a GICv2 controller even if the memory range 2610 exposed by the device tree is too small. 2611 2612 irqchip.gicv3_nolpi= 2613 [ARM,ARM64,EARLY] 2614 Force the kernel to ignore the availability of 2615 LPIs (and by consequence ITSs). Intended for system 2616 that use the kernel as a bootloader, and thus want 2617 to let secondary kernels in charge of setting up 2618 LPIs. 2619 2620 irqchip.gicv3_pseudo_nmi= [ARM64,EARLY] 2621 Enables support for pseudo-NMIs in the kernel. This 2622 requires the kernel to be built with 2623 CONFIG_ARM64_PSEUDO_NMI. 2624 2625 irqchip.riscv_imsic_noipi 2626 [RISC-V,EARLY] 2627 Force the kernel to not use IMSIC software injected MSIs 2628 as IPIs. Intended for system where IMSIC is trap-n-emulated, 2629 and thus want to reduce MMIO traps when triggering IPIs 2630 to multiple harts. 2631 2632 irqfixup [HW] 2633 When an interrupt is not handled search all handlers 2634 for it. Intended to get systems with badly broken 2635 firmware running. 2636 2637 irqhandler.duration_warn_us= [KNL] 2638 Warn if an IRQ handler exceeds the specified duration 2639 threshold in microseconds. Useful for identifying 2640 long-running IRQs in the system. 2641 2642 irqpoll [HW] 2643 When an interrupt is not handled search all handlers 2644 for it. Also check all handlers each timer 2645 interrupt. Intended to get systems with badly broken 2646 firmware running. 2647 2648 isapnp= [ISAPNP] 2649 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2650 2651 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2652 [Deprecated - use cpusets instead] 2653 Format: [flag-list,]<cpu-list> 2654 2655 Specify one or more CPUs to isolate from disturbances 2656 specified in the flag list (default: domain): 2657 2658 nohz 2659 Disable the tick when a single task runs as well as 2660 disabling other kernel noises like having RCU callbacks 2661 offloaded. This is equivalent to the nohz_full parameter. 2662 2663 A residual 1Hz tick is offloaded to workqueues, which you 2664 need to affine to housekeeping through the global 2665 workqueue's affinity configured via the 2666 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2667 by using the 'domain' flag described below. 2668 2669 NOTE: by default the global workqueue runs on all CPUs, 2670 so to protect individual CPUs the 'cpumask' file has to 2671 be configured manually after bootup. 2672 2673 domain 2674 Isolate from the general SMP balancing and scheduling 2675 algorithms. Note that performing domain isolation this way 2676 is irreversible: it's not possible to bring back a CPU to 2677 the domains once isolated through isolcpus. It's strongly 2678 advised to use cpusets instead to disable scheduler load 2679 balancing through the "cpuset.sched_load_balance" file. 2680 It offers a much more flexible interface where CPUs can 2681 move in and out of an isolated set anytime. 2682 2683 You can move a process onto or off an "isolated" CPU via 2684 the CPU affinity syscalls or cpuset. 2685 <cpu number> begins at 0 and the maximum value is 2686 "number of CPUs in system - 1". 2687 2688 managed_irq 2689 2690 Isolate from being targeted by managed interrupts 2691 which have an interrupt mask containing isolated 2692 CPUs. The affinity of managed interrupts is 2693 handled by the kernel and cannot be changed via 2694 the /proc/irq/* interfaces. 2695 2696 This isolation is best effort and only effective 2697 if the automatically assigned interrupt mask of a 2698 device queue contains isolated and housekeeping 2699 CPUs. If housekeeping CPUs are online then such 2700 interrupts are directed to the housekeeping CPU 2701 so that IO submitted on the housekeeping CPU 2702 cannot disturb the isolated CPU. 2703 2704 If a queue's affinity mask contains only isolated 2705 CPUs then this parameter has no effect on the 2706 interrupt routing decision, though interrupts are 2707 only delivered when tasks running on those 2708 isolated CPUs submit IO. IO submitted on 2709 housekeeping CPUs has no influence on those 2710 queues. 2711 2712 The format of <cpu-list> is described above. 2713 2714 iucv= [HW,NET] 2715 2716 ivrs_ioapic [HW,X86-64] 2717 Provide an override to the IOAPIC-ID<->DEVICE-ID 2718 mapping provided in the IVRS ACPI table. 2719 By default, PCI segment is 0, and can be omitted. 2720 2721 For example, to map IOAPIC-ID decimal 10 to 2722 PCI segment 0x1 and PCI device 00:14.0, 2723 write the parameter as: 2724 ivrs_ioapic=10@0001:00:14.0 2725 2726 Deprecated formats: 2727 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0 2728 write the parameter as: 2729 ivrs_ioapic[10]=00:14.0 2730 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and 2731 PCI device 00:14.0 write the parameter as: 2732 ivrs_ioapic[10]=0001:00:14.0 2733 2734 ivrs_hpet [HW,X86-64] 2735 Provide an override to the HPET-ID<->DEVICE-ID 2736 mapping provided in the IVRS ACPI table. 2737 By default, PCI segment is 0, and can be omitted. 2738 2739 For example, to map HPET-ID decimal 10 to 2740 PCI segment 0x1 and PCI device 00:14.0, 2741 write the parameter as: 2742 ivrs_hpet=10@0001:00:14.0 2743 2744 Deprecated formats: 2745 * To map HPET-ID decimal 0 to PCI device 00:14.0 2746 write the parameter as: 2747 ivrs_hpet[0]=00:14.0 2748 * To map HPET-ID decimal 10 to PCI segment 0x1 and 2749 PCI device 00:14.0 write the parameter as: 2750 ivrs_ioapic[10]=0001:00:14.0 2751 2752 ivrs_acpihid [HW,X86-64] 2753 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2754 mapping provided in the IVRS ACPI table. 2755 By default, PCI segment is 0, and can be omitted. 2756 2757 For example, to map UART-HID:UID AMD0020:0 to 2758 PCI segment 0x1 and PCI device ID 00:14.5, 2759 write the parameter as: 2760 ivrs_acpihid=AMD0020:0@0001:00:14.5 2761 2762 Deprecated formats: 2763 * To map UART-HID:UID AMD0020:0 to PCI segment is 0, 2764 PCI device ID 00:14.5, write the parameter as: 2765 ivrs_acpihid[00:14.5]=AMD0020:0 2766 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and 2767 PCI device ID 00:14.5, write the parameter as: 2768 ivrs_acpihid[0001:00:14.5]=AMD0020:0 2769 2770 js= [HW,JOY] Analog joystick 2771 See Documentation/input/joydev/joystick.rst. 2772 2773 kasan_multi_shot 2774 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2775 report on every invalid memory access. Without this 2776 parameter KASAN will print report only for the first 2777 invalid access. 2778 2779 keep_bootcon [KNL,EARLY] 2780 Do not unregister boot console at start. This is only 2781 useful for debugging when something happens in the window 2782 between unregistering the boot console and initializing 2783 the real console. 2784 2785 keepinitrd [HW,ARM] See retain_initrd. 2786 2787 kernelcore= [KNL,X86,PPC,EARLY] 2788 Format: nn[KMGTPE] | nn% | "mirror" 2789 This parameter specifies the amount of memory usable by 2790 the kernel for non-movable allocations. The requested 2791 amount is spread evenly throughout all nodes in the 2792 system as ZONE_NORMAL. The remaining memory is used for 2793 movable memory in its own zone, ZONE_MOVABLE. In the 2794 event, a node is too small to have both ZONE_NORMAL and 2795 ZONE_MOVABLE, kernelcore memory will take priority and 2796 other nodes will have a larger ZONE_MOVABLE. 2797 2798 ZONE_MOVABLE is used for the allocation of pages that 2799 may be reclaimed or moved by the page migration 2800 subsystem. Note that allocations like PTEs-from-HighMem 2801 still use the HighMem zone if it exists, and the Normal 2802 zone if it does not. 2803 2804 It is possible to specify the exact amount of memory in 2805 the form of "nn[KMGTPE]", a percentage of total system 2806 memory in the form of "nn%", or "mirror". If "mirror" 2807 option is specified, mirrored (reliable) memory is used 2808 for non-movable allocations and remaining memory is used 2809 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2810 are exclusive, so you cannot specify multiple forms. 2811 2812 kgdbdbgp= [KGDB,HW,EARLY] kgdb over EHCI usb debug port. 2813 Format: <Controller#>[,poll interval] 2814 The controller # is the number of the ehci usb debug 2815 port as it is probed via PCI. The poll interval is 2816 optional and is the number seconds in between 2817 each poll cycle to the debug port in case you need 2818 the functionality for interrupting the kernel with 2819 gdb or control-c on the dbgp connection. When 2820 not using this parameter you use sysrq-g to break into 2821 the kernel debugger. 2822 2823 kgdboc= [KGDB,HW] kgdb over consoles. 2824 Requires a tty driver that supports console polling, 2825 or a supported polling keyboard driver (non-usb). 2826 Serial only format: <serial_device>[,baud] 2827 keyboard only format: kbd 2828 keyboard and serial format: kbd,<serial_device>[,baud] 2829 Optional Kernel mode setting: 2830 kms, kbd format: kms,kbd 2831 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2832 2833 kgdboc_earlycon= [KGDB,HW,EARLY] 2834 If the boot console provides the ability to read 2835 characters and can work in polling mode, you can use 2836 this parameter to tell kgdb to use it as a backend 2837 until the normal console is registered. Intended to 2838 be used together with the kgdboc parameter which 2839 specifies the normal console to transition to. 2840 2841 The name of the early console should be specified 2842 as the value of this parameter. Note that the name of 2843 the early console might be different than the tty 2844 name passed to kgdboc. It's OK to leave the value 2845 blank and the first boot console that implements 2846 read() will be picked. 2847 2848 kgdbwait [KGDB,EARLY] Stop kernel execution and enter the 2849 kernel debugger at the earliest opportunity. 2850 2851 kho= [KEXEC,EARLY] 2852 Format: { "0" | "1" | "off" | "on" | "y" | "n" } 2853 Enables or disables Kexec HandOver. 2854 "0" | "off" | "n" - kexec handover is disabled 2855 "1" | "on" | "y" - kexec handover is enabled 2856 2857 kho_scratch= [KEXEC,EARLY] 2858 Format: ll[KMG],mm[KMG],nn[KMG] | nn% 2859 Defines the size of the KHO scratch region. The KHO 2860 scratch regions are physically contiguous memory 2861 ranges that can only be used for non-kernel 2862 allocations. That way, even when memory is heavily 2863 fragmented with handed over memory, the kexeced 2864 kernel will always have enough contiguous ranges to 2865 bootstrap itself. 2866 2867 It is possible to specify the exact amount of 2868 memory in the form of "ll[KMG],mm[KMG],nn[KMG]" 2869 where the first parameter defines the size of a low 2870 memory scratch area, the second parameter defines 2871 the size of a global scratch area and the third 2872 parameter defines the size of additional per-node 2873 scratch areas. The form "nn%" defines scale factor 2874 (in percents) of memory that was used during boot. 2875 2876 kmac= [MIPS] Korina ethernet MAC address. 2877 Configure the RouterBoard 532 series on-chip 2878 Ethernet adapter MAC address. 2879 2880 kmemleak= [KNL,EARLY] Boot-time kmemleak enable/disable 2881 Valid arguments: on, off 2882 Default: on 2883 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 2884 the default is off. 2885 2886 kprobe_event=[probe-list] 2887 [FTRACE] Add kprobe events and enable at boot time. 2888 The probe-list is a semicolon delimited list of probe 2889 definitions. Each definition is same as kprobe_events 2890 interface, but the parameters are comma delimited. 2891 For example, to add a kprobe event on vfs_read with 2892 arg1 and arg2, add to the command line; 2893 2894 kprobe_event=p,vfs_read,$arg1,$arg2 2895 2896 See also Documentation/trace/kprobetrace.rst "Kernel 2897 Boot Parameter" section. 2898 2899 kpti= [ARM64,EARLY] Control page table isolation of 2900 user and kernel address spaces. 2901 Default: enabled on cores which need mitigation. 2902 0: force disabled 2903 1: force enabled 2904 2905 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires 2906 CONFIG_KUNIT to be set to be fully enabled. The 2907 default value can be overridden via 2908 KUNIT_DEFAULT_ENABLED. 2909 Default is 1 (enabled) 2910 2911 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 2912 Default is 0 (don't ignore, but inject #GP) 2913 2914 kvm.eager_page_split= 2915 [KVM,X86] Controls whether or not KVM will try to 2916 proactively split all huge pages during dirty logging. 2917 Eager page splitting reduces interruptions to vCPU 2918 execution by eliminating the write-protection faults 2919 and MMU lock contention that would otherwise be 2920 required to split huge pages lazily. 2921 2922 VM workloads that rarely perform writes or that write 2923 only to a small region of VM memory may benefit from 2924 disabling eager page splitting to allow huge pages to 2925 still be used for reads. 2926 2927 The behavior of eager page splitting depends on whether 2928 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If 2929 disabled, all huge pages in a memslot will be eagerly 2930 split when dirty logging is enabled on that memslot. If 2931 enabled, eager page splitting will be performed during 2932 the KVM_CLEAR_DIRTY ioctl, and only for the pages being 2933 cleared. 2934 2935 Eager page splitting is only supported when kvm.tdp_mmu=Y. 2936 2937 Default is Y (on). 2938 2939 kvm.enable_virt_at_load=[KVM,ARM64,LOONGARCH,MIPS,RISCV,X86] 2940 If enabled, KVM will enable virtualization in hardware 2941 when KVM is loaded, and disable virtualization when KVM 2942 is unloaded (if KVM is built as a module). 2943 2944 If disabled, KVM will dynamically enable and disable 2945 virtualization on-demand when creating and destroying 2946 VMs, i.e. on the 0=>1 and 1=>0 transitions of the 2947 number of VMs. 2948 2949 Enabling virtualization at module load avoids potential 2950 latency for creation of the 0=>1 VM, as KVM serializes 2951 virtualization enabling across all online CPUs. The 2952 "cost" of enabling virtualization when KVM is loaded, 2953 is that doing so may interfere with using out-of-tree 2954 hypervisors that want to "own" virtualization hardware. 2955 2956 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 2957 Default is false (don't support). 2958 2959 kvm.nx_huge_pages= 2960 [KVM] Controls the software workaround for the 2961 X86_BUG_ITLB_MULTIHIT bug. 2962 force : Always deploy workaround. 2963 off : Never deploy workaround. 2964 auto : Deploy workaround based on the presence of 2965 X86_BUG_ITLB_MULTIHIT. 2966 2967 Default is 'auto'. 2968 2969 If the software workaround is enabled for the host, 2970 guests do need not to enable it for nested guests. 2971 2972 kvm.nx_huge_pages_recovery_ratio= 2973 [KVM] Controls how many 4KiB pages are periodically zapped 2974 back to huge pages. 0 disables the recovery, otherwise if 2975 the value is N KVM will zap 1/Nth of the 4KiB pages every 2976 period (see below). The default is 60. 2977 2978 kvm.nx_huge_pages_recovery_period_ms= 2979 [KVM] Controls the time period at which KVM zaps 4KiB pages 2980 back to huge pages. If the value is a non-zero N, KVM will 2981 zap a portion (see ratio above) of the pages every N msecs. 2982 If the value is 0 (the default), KVM will pick a period based 2983 on the ratio, such that a page is zapped after 1 hour on average. 2984 2985 kvm-amd.nested= [KVM,AMD] Control nested virtualization feature in 2986 KVM/SVM. Default is 1 (enabled). 2987 2988 kvm-amd.npt= [KVM,AMD] Control KVM's use of Nested Page Tables, 2989 a.k.a. Two-Dimensional Page Tables. Default is 1 2990 (enabled). Disable by KVM if hardware lacks support 2991 for NPT. 2992 2993 kvm-amd.ciphertext_hiding_asids= 2994 [KVM,AMD] Ciphertext hiding prevents disallowed accesses 2995 to SNP private memory from reading ciphertext. Instead, 2996 reads will see constant default values (0xff). 2997 2998 If ciphertext hiding is enabled, the joint SEV-ES and 2999 SEV-SNP ASID space is partitioned into separate SEV-ES 3000 and SEV-SNP ASID ranges, with the SEV-SNP range being 3001 [1..max_snp_asid] and the SEV-ES range being 3002 (max_snp_asid..min_sev_asid), where min_sev_asid is 3003 enumerated by CPUID.0x.8000_001F[EDX]. 3004 3005 A non-zero value enables SEV-SNP ciphertext hiding and 3006 adjusts the ASID ranges for SEV-ES and SEV-SNP guests. 3007 KVM caps the number of SEV-SNP ASIDs at the maximum 3008 possible value, e.g. specifying -1u will assign all 3009 joint SEV-ES and SEV-SNP ASIDs to SEV-SNP. Note, 3010 assigning all joint ASIDs to SEV-SNP, i.e. configuring 3011 max_snp_asid == min_sev_asid-1, will effectively make 3012 SEV-ES unusable. 3013 3014 kvm-arm.mode= 3015 [KVM,ARM,EARLY] Select one of KVM/arm64's modes of 3016 operation. 3017 3018 none: Forcefully disable KVM. 3019 3020 nvhe: Standard nVHE-based mode, without support for 3021 protected guests. 3022 3023 protected: Mode with support for guests whose state is 3024 kept private from the host, using VHE or 3025 nVHE depending on HW support. 3026 3027 nested: VHE-based mode with support for nested 3028 virtualization. Requires at least ARMv8.4 3029 hardware (with FEAT_NV2). 3030 3031 Defaults to VHE/nVHE based on hardware support. Setting 3032 mode to "protected" will disable kexec and hibernation 3033 for the host. To force nVHE on VHE hardware, add 3034 "arm64_sw.hvhe=0 id_aa64mmfr1.vh=0" to the 3035 command-line. 3036 "nested" is experimental and should be used with 3037 extreme caution. 3038 3039 kvm-arm.vgic_v3_group0_trap= 3040 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-0 3041 system registers 3042 3043 kvm-arm.vgic_v3_group1_trap= 3044 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-1 3045 system registers 3046 3047 kvm-arm.vgic_v3_common_trap= 3048 [KVM,ARM,EARLY] Trap guest accesses to GICv3 common 3049 system registers 3050 3051 kvm-arm.vgic_v4_enable= 3052 [KVM,ARM,EARLY] Allow use of GICv4 for direct 3053 injection of LPIs. 3054 3055 kvm-arm.wfe_trap_policy= 3056 [KVM,ARM] Control when to set WFE instruction trap for 3057 KVM VMs. Traps are allowed but not guaranteed by the 3058 CPU architecture. 3059 3060 trap: set WFE instruction trap 3061 3062 notrap: clear WFE instruction trap 3063 3064 kvm-arm.wfi_trap_policy= 3065 [KVM,ARM] Control when to set WFI instruction trap for 3066 KVM VMs. Traps are allowed but not guaranteed by the 3067 CPU architecture. 3068 3069 trap: set WFI instruction trap 3070 3071 notrap: clear WFI instruction trap 3072 3073 kvm_cma_resv_ratio=n [PPC,EARLY] 3074 Reserves given percentage from system memory area for 3075 contiguous memory allocation for KVM hash pagetable 3076 allocation. 3077 By default it reserves 5% of total system memory. 3078 Format: <integer> 3079 Default: 5 3080 3081 kvm-intel.ept= [KVM,Intel] Control KVM's use of Extended Page Tables, 3082 a.k.a. Two-Dimensional Page Tables. Default is 1 3083 (enabled). Disable by KVM if hardware lacks support 3084 for EPT. 3085 3086 kvm-intel.emulate_invalid_guest_state= 3087 [KVM,Intel] Control whether to emulate invalid guest 3088 state. Ignored if kvm-intel.enable_unrestricted_guest=1, 3089 as guest state is never invalid for unrestricted 3090 guests. This param doesn't apply to nested guests (L2), 3091 as KVM never emulates invalid L2 guest state. 3092 Default is 1 (enabled). 3093 3094 kvm-intel.flexpriority= 3095 [KVM,Intel] Control KVM's use of FlexPriority feature 3096 (TPR shadow). Default is 1 (enabled). Disable by KVM if 3097 hardware lacks support for it. 3098 3099 kvm-intel.nested= 3100 [KVM,Intel] Control nested virtualization feature in 3101 KVM/VMX. Default is 1 (enabled). 3102 3103 kvm-intel.unrestricted_guest= 3104 [KVM,Intel] Control KVM's use of unrestricted guest 3105 feature (virtualized real and unpaged mode). Default 3106 is 1 (enabled). Disable by KVM if EPT is disabled or 3107 hardware lacks support for it. 3108 3109 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 3110 CVE-2018-3620. 3111 3112 Valid arguments: never, cond, always 3113 3114 always: L1D cache flush on every VMENTER. 3115 cond: Flush L1D on VMENTER only when the code between 3116 VMEXIT and VMENTER can leak host memory. 3117 never: Disables the mitigation 3118 3119 Default is cond (do L1 cache flush in specific instances) 3120 3121 kvm-intel.vpid= [KVM,Intel] Control KVM's use of Virtual Processor 3122 Identification feature (tagged TLBs). Default is 1 3123 (enabled). Disable by KVM if hardware lacks support 3124 for it. 3125 3126 l1d_flush= [X86,INTEL,EARLY] 3127 Control mitigation for L1D based snooping vulnerability. 3128 3129 Certain CPUs are vulnerable to an exploit against CPU 3130 internal buffers which can forward information to a 3131 disclosure gadget under certain conditions. 3132 3133 In vulnerable processors, the speculatively 3134 forwarded data can be used in a cache side channel 3135 attack, to access data to which the attacker does 3136 not have direct access. 3137 3138 This parameter controls the mitigation. The 3139 options are: 3140 3141 on - enable the interface for the mitigation 3142 3143 l1tf= [X86,EARLY] Control mitigation of the L1TF vulnerability on 3144 affected CPUs 3145 3146 The kernel PTE inversion protection is unconditionally 3147 enabled and cannot be disabled. 3148 3149 full 3150 Provides all available mitigations for the 3151 L1TF vulnerability. Disables SMT and 3152 enables all mitigations in the 3153 hypervisors, i.e. unconditional L1D flush. 3154 3155 SMT control and L1D flush control via the 3156 sysfs interface is still possible after 3157 boot. Hypervisors will issue a warning 3158 when the first VM is started in a 3159 potentially insecure configuration, 3160 i.e. SMT enabled or L1D flush disabled. 3161 3162 full,force 3163 Same as 'full', but disables SMT and L1D 3164 flush runtime control. Implies the 3165 'nosmt=force' command line option. 3166 (i.e. sysfs control of SMT is disabled.) 3167 3168 flush 3169 Leaves SMT enabled and enables the default 3170 hypervisor mitigation, i.e. conditional 3171 L1D flush. 3172 3173 SMT control and L1D flush control via the 3174 sysfs interface is still possible after 3175 boot. Hypervisors will issue a warning 3176 when the first VM is started in a 3177 potentially insecure configuration, 3178 i.e. SMT enabled or L1D flush disabled. 3179 3180 flush,nosmt 3181 3182 Disables SMT and enables the default 3183 hypervisor mitigation. 3184 3185 SMT control and L1D flush control via the 3186 sysfs interface is still possible after 3187 boot. Hypervisors will issue a warning 3188 when the first VM is started in a 3189 potentially insecure configuration, 3190 i.e. SMT enabled or L1D flush disabled. 3191 3192 flush,nowarn 3193 Same as 'flush', but hypervisors will not 3194 warn when a VM is started in a potentially 3195 insecure configuration. 3196 3197 off 3198 Disables hypervisor mitigations and doesn't 3199 emit any warnings. 3200 It also drops the swap size and available 3201 RAM limit restriction on both hypervisor and 3202 bare metal. 3203 3204 Default is 'flush'. 3205 3206 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 3207 3208 l2cr= [PPC] 3209 3210 l3cr= [PPC] 3211 3212 lapic [X86-32,APIC,EARLY] Enable the local APIC even if BIOS 3213 disabled it. 3214 3215 lapic= [X86,APIC] Do not use TSC deadline 3216 value for LAPIC timer one-shot implementation. Default 3217 back to the programmable timer unit in the LAPIC. 3218 Format: notscdeadline 3219 3220 lapic_timer_c2_ok [X86,APIC,EARLY] trust the local apic timer 3221 in C2 power state. 3222 3223 libata.dma= [LIBATA] DMA control 3224 libata.dma=0 Disable all PATA and SATA DMA 3225 libata.dma=1 PATA and SATA Disk DMA only 3226 libata.dma=2 ATAPI (CDROM) DMA only 3227 libata.dma=4 Compact Flash DMA only 3228 Combinations also work, so libata.dma=3 enables DMA 3229 for disks and CDROMs, but not CFs. 3230 3231 libata.ignore_hpa= [LIBATA] Ignore HPA limit 3232 libata.ignore_hpa=0 keep BIOS limits (default) 3233 libata.ignore_hpa=1 ignore limits, using full disk 3234 3235 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 3236 when set. 3237 Format: <int> 3238 3239 libata.force= [LIBATA] Force configurations. The format is a comma- 3240 separated list of "[ID:]VAL" where ID is PORT[.DEVICE]. 3241 PORT and DEVICE are decimal numbers matching port, link 3242 or device. Basically, it matches the ATA ID string 3243 printed on console by libata. If the whole ID part is 3244 omitted, the last PORT and DEVICE values are used. If 3245 ID hasn't been specified yet, the configuration applies 3246 to all ports, links and devices. 3247 3248 If only DEVICE is omitted, the parameter applies to 3249 the port and all links and devices behind it. DEVICE 3250 number of 0 either selects the first device or the 3251 first fan-out link behind PMP device. It does not 3252 select the host link. DEVICE number of 15 selects the 3253 host link and device attached to it. 3254 3255 The VAL specifies the configuration to force. As long 3256 as there is no ambiguity, shortcut notation is allowed. 3257 For example, both 1.5 and 1.5G would work for 1.5Gbps. 3258 The following configurations can be forced. 3259 3260 * Cable type: 40c, 80c, short40c, unk, ign or sata. 3261 Any ID with matching PORT is used. 3262 3263 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 3264 3265 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 3266 udma[/][16,25,33,44,66,100,133] notation is also 3267 allowed. 3268 3269 * nohrst, nosrst, norst: suppress hard, soft and both 3270 resets. 3271 3272 * rstonce: only attempt one reset during hot-unplug 3273 link recovery. 3274 3275 * [no]dbdelay: Enable or disable the extra 200ms delay 3276 before debouncing a link PHY and device presence 3277 detection. 3278 3279 * [no]ncq: Turn on or off NCQ. 3280 3281 * [no]ncqtrim: Enable or disable queued DSM TRIM. 3282 3283 * [no]ncqati: Enable or disable NCQ trim on ATI chipset. 3284 3285 * [no]trim: Enable or disable (unqueued) TRIM. 3286 3287 * trim_zero: Indicate that TRIM command zeroes data. 3288 3289 * max_trim_128m: Set 128M maximum trim size limit. 3290 3291 * [no]dma: Turn on or off DMA transfers. 3292 3293 * atapi_dmadir: Enable ATAPI DMADIR bridge support. 3294 3295 * atapi_mod16_dma: Enable the use of ATAPI DMA for 3296 commands that are not a multiple of 16 bytes. 3297 3298 * [no]dmalog: Enable or disable the use of the 3299 READ LOG DMA EXT command to access logs. 3300 3301 * [no]iddevlog: Enable or disable access to the 3302 identify device data log. 3303 3304 * [no]logdir: Enable or disable access to the general 3305 purpose log directory. 3306 3307 * max_sec_128: Set transfer size limit to 128 sectors. 3308 3309 * max_sec_1024: Set or clear transfer size limit to 3310 1024 sectors. 3311 3312 * max_sec_lba48: Set or clear transfer size limit to 3313 65535 sectors. 3314 3315 * external: Mark port as external (hotplug-capable). 3316 3317 * [no]lpm: Enable or disable link power management. 3318 3319 * [no]setxfer: Indicate if transfer speed mode setting 3320 should be skipped. 3321 3322 * [no]fua: Disable or enable FUA (Force Unit Access) 3323 support for devices supporting this feature. 3324 3325 * dump_id: Dump IDENTIFY data. 3326 3327 * disable: Disable this device. 3328 3329 If there are multiple matching configurations changing 3330 the same attribute, the last one is used. 3331 3332 load_ramdisk= [RAM] [Deprecated] 3333 3334 lockd.nlm_grace_period=P [NFS] Assign grace period. 3335 Format: <integer> 3336 3337 lockd.nlm_tcpport=N [NFS] Assign TCP port. 3338 Format: <integer> 3339 3340 lockd.nlm_timeout=T [NFS] Assign timeout value. 3341 Format: <integer> 3342 3343 lockd.nlm_udpport=M [NFS] Assign UDP port. 3344 Format: <integer> 3345 3346 lockdown= [SECURITY,EARLY] 3347 { integrity | confidentiality } 3348 Enable the kernel lockdown feature. If set to 3349 integrity, kernel features that allow userland to 3350 modify the running kernel are disabled. If set to 3351 confidentiality, kernel features that allow userland 3352 to extract confidential information from the kernel 3353 are also disabled. 3354 3355 locktorture.acq_writer_lim= [KNL] 3356 Set the time limit in jiffies for a lock 3357 acquisition. Acquisitions exceeding this limit 3358 will result in a splat once they do complete. 3359 3360 locktorture.bind_readers= [KNL] 3361 Specify the list of CPUs to which the readers are 3362 to be bound. 3363 3364 locktorture.bind_writers= [KNL] 3365 Specify the list of CPUs to which the writers are 3366 to be bound. 3367 3368 locktorture.call_rcu_chains= [KNL] 3369 Specify the number of self-propagating call_rcu() 3370 chains to set up. These are used to ensure that 3371 there is a high probability of an RCU grace period 3372 in progress at any given time. Defaults to 0, 3373 which disables these call_rcu() chains. 3374 3375 locktorture.long_hold= [KNL] 3376 Specify the duration in milliseconds for the 3377 occasional long-duration lock hold time. Defaults 3378 to 100 milliseconds. Select 0 to disable. 3379 3380 locktorture.nested_locks= [KNL] 3381 Specify the maximum lock nesting depth that 3382 locktorture is to exercise, up to a limit of 8 3383 (MAX_NESTED_LOCKS). Specify zero to disable. 3384 Note that this parameter is ineffective on types 3385 of locks that do not support nested acquisition. 3386 3387 locktorture.nreaders_stress= [KNL] 3388 Set the number of locking read-acquisition kthreads. 3389 Defaults to being automatically set based on the 3390 number of online CPUs. 3391 3392 locktorture.nwriters_stress= [KNL] 3393 Set the number of locking write-acquisition kthreads. 3394 3395 locktorture.onoff_holdoff= [KNL] 3396 Set time (s) after boot for CPU-hotplug testing. 3397 3398 locktorture.onoff_interval= [KNL] 3399 Set time (s) between CPU-hotplug operations, or 3400 zero to disable CPU-hotplug testing. 3401 3402 locktorture.rt_boost= [KNL] 3403 Do periodic testing of real-time lock priority 3404 boosting. Select 0 to disable, 1 to boost 3405 only rt_mutex, and 2 to boost unconditionally. 3406 Defaults to 2, which might seem to be an 3407 odd choice, but which should be harmless for 3408 non-real-time spinlocks, due to their disabling 3409 of preemption. Note that non-realtime mutexes 3410 disable boosting. 3411 3412 locktorture.rt_boost_factor= [KNL] 3413 Number that determines how often and for how 3414 long priority boosting is exercised. This is 3415 scaled down by the number of writers, so that the 3416 number of boosts per unit time remains roughly 3417 constant as the number of writers increases. 3418 On the other hand, the duration of each boost 3419 increases with the number of writers. 3420 3421 locktorture.shuffle_interval= [KNL] 3422 Set task-shuffle interval (jiffies). Shuffling 3423 tasks allows some CPUs to go into dyntick-idle 3424 mode during the locktorture test. 3425 3426 locktorture.shutdown_secs= [KNL] 3427 Set time (s) after boot system shutdown. This 3428 is useful for hands-off automated testing. 3429 3430 locktorture.stat_interval= [KNL] 3431 Time (s) between statistics printk()s. 3432 3433 locktorture.stutter= [KNL] 3434 Time (s) to stutter testing, for example, 3435 specifying five seconds causes the test to run for 3436 five seconds, wait for five seconds, and so on. 3437 This tests the locking primitive's ability to 3438 transition abruptly to and from idle. 3439 3440 locktorture.torture_type= [KNL] 3441 Specify the locking implementation to test. 3442 3443 locktorture.verbose= [KNL] 3444 Enable additional printk() statements. 3445 3446 locktorture.writer_fifo= [KNL] 3447 Run the write-side locktorture kthreads at 3448 sched_set_fifo() real-time priority. 3449 3450 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 3451 Format: <irq> 3452 3453 loglevel= [KNL,EARLY] 3454 All Kernel Messages with a loglevel smaller than the 3455 console loglevel will be printed to the console. It can 3456 also be changed with klogd or other programs. The 3457 loglevels are defined as follows: 3458 3459 0 (KERN_EMERG) system is unusable 3460 1 (KERN_ALERT) action must be taken immediately 3461 2 (KERN_CRIT) critical conditions 3462 3 (KERN_ERR) error conditions 3463 4 (KERN_WARNING) warning conditions 3464 5 (KERN_NOTICE) normal but significant condition 3465 6 (KERN_INFO) informational 3466 7 (KERN_DEBUG) debug-level messages 3467 3468 log_buf_len=n[KMG] [KNL,EARLY] 3469 Sets the size of the printk ring buffer, in bytes. 3470 n must be a power of two and greater than the 3471 minimal size. The minimal size is defined by 3472 LOG_BUF_SHIFT kernel config parameter. There 3473 is also CONFIG_LOG_CPU_MAX_BUF_SHIFT config 3474 parameter that allows to increase the default size 3475 depending on the number of CPUs. See init/Kconfig 3476 for more details. 3477 3478 logo.nologo [FB] Disables display of the built-in Linux logo. 3479 This may be used to provide more screen space for 3480 kernel log messages and is useful when debugging 3481 kernel boot problems. 3482 3483 lp=0 [LP] Specify parallel ports to use, e.g, 3484 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 3485 lp=reset first parallel port). 'lp=0' disables the 3486 lp=auto printer driver. 'lp=reset' (which can be 3487 specified in addition to the ports) causes 3488 attached printers to be reset. Using 3489 lp=port1,port2,... specifies the parallel ports 3490 to associate lp devices with, starting with 3491 lp0. A port specification may be 'none' to skip 3492 that lp device, or a parport name such as 3493 'parport0'. Specifying 'lp=auto' instead of a 3494 port specification list means that device IDs 3495 from each port should be examined, to see if 3496 an IEEE 1284-compliant printer is attached; if 3497 so, the driver will manage that printer. 3498 See also header of drivers/char/lp.c. 3499 3500 lpj=n [KNL] 3501 Sets loops_per_jiffy to given constant, thus avoiding 3502 time-consuming boot-time autodetection (up to 250 ms per 3503 CPU). 0 enables autodetection (default). To determine 3504 the correct value for your kernel, boot with normal 3505 autodetection and see what value is printed. Note that 3506 on SMP systems the preset will be applied to all CPUs, 3507 which is likely to cause problems if your CPUs need 3508 significantly divergent settings. An incorrect value 3509 will cause delays in the kernel to be wrong, leading to 3510 unpredictable I/O errors and other breakage. Although 3511 unlikely, in the extreme case this might damage your 3512 hardware. 3513 3514 lsm.debug [SECURITY] Enable LSM initialization debugging output. 3515 3516 lsm=lsm1,...,lsmN 3517 [SECURITY] Choose order of LSM initialization. This 3518 overrides CONFIG_LSM, and the "security=" parameter. 3519 3520 machtype= [Loongson] Share the same kernel image file between 3521 different yeeloong laptops. 3522 Example: machtype=lemote-yeeloong-2f-7inch 3523 3524 maxcpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 3525 will bring up during bootup. maxcpus=n : n >= 0 limits 3526 the kernel to bring up 'n' processors. Surely after 3527 bootup you can bring up the other plugged cpu by executing 3528 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 3529 only takes effect during system bootup. 3530 While n=0 is a special case, it is equivalent to "nosmp", 3531 which also disables the IO APIC. 3532 3533 max_loop= [LOOP] The number of loop block devices that get 3534 (loop.max_loop) unconditionally pre-created at init time. The default 3535 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 3536 of statically allocating a predefined number, loop 3537 devices can be requested on-demand with the 3538 /dev/loop-control interface. 3539 3540 mce= [X86-{32,64}] 3541 3542 Please see Documentation/arch/x86/x86_64/machinecheck.rst for sysfs runtime tunables. 3543 3544 off 3545 disable machine check 3546 3547 no_cmci 3548 disable CMCI(Corrected Machine Check Interrupt) that 3549 Intel processor supports. Usually this disablement is 3550 not recommended, but it might be handy if your 3551 hardware is misbehaving. 3552 3553 Note that you'll get more problems without CMCI than 3554 with due to the shared banks, i.e. you might get 3555 duplicated error logs. 3556 3557 dont_log_ce 3558 don't make logs for corrected errors. All events 3559 reported as corrected are silently cleared by OS. This 3560 option will be useful if you have no interest in any 3561 of corrected errors. 3562 3563 ignore_ce 3564 disable features for corrected errors, e.g. 3565 polling timer and CMCI. All events reported as 3566 corrected are not cleared by OS and remained in its 3567 error banks. 3568 3569 Usually this disablement is not recommended, however 3570 if there is an agent checking/clearing corrected 3571 errors (e.g. BIOS or hardware monitoring 3572 applications), conflicting with OS's error handling, 3573 and you cannot deactivate the agent, then this option 3574 will be a help. 3575 3576 no_lmce 3577 do not opt-in to Local MCE delivery. Use legacy method 3578 to broadcast MCEs. 3579 3580 bootlog 3581 enable logging of machine checks left over from 3582 booting. Disabled by default on AMD Fam10h and older 3583 because some BIOS leave bogus ones. 3584 3585 If your BIOS doesn't do that it's a good idea to 3586 enable though to make sure you log even machine check 3587 events that result in a reboot. On Intel systems it is 3588 enabled by default. 3589 3590 nobootlog 3591 disable boot machine check logging. 3592 3593 monarchtimeout (number) 3594 sets the time in us to wait for other CPUs on machine 3595 checks. 0 to disable. 3596 3597 bios_cmci_threshold 3598 don't overwrite the bios-set CMCI threshold. This boot 3599 option prevents Linux from overwriting the CMCI 3600 threshold set by the bios. Without this option, Linux 3601 always sets the CMCI threshold to 1. Enabling this may 3602 make memory predictive failure analysis less effective 3603 if the bios sets thresholds for memory errors since we 3604 will not see details for all errors. 3605 3606 recovery 3607 force-enable recoverable machine check code paths 3608 3609 Everything else is in sysfs now. 3610 3611 3612 md= [HW] RAID subsystems devices and level 3613 See Documentation/admin-guide/md.rst. 3614 3615 mdacon= [MDA] 3616 Format: <first>,<last> 3617 Specifies range of consoles to be captured by the MDA. 3618 3619 mds= [X86,INTEL,EARLY] 3620 Control mitigation for the Micro-architectural Data 3621 Sampling (MDS) vulnerability. 3622 3623 Certain CPUs are vulnerable to an exploit against CPU 3624 internal buffers which can forward information to a 3625 disclosure gadget under certain conditions. 3626 3627 In vulnerable processors, the speculatively 3628 forwarded data can be used in a cache side channel 3629 attack, to access data to which the attacker does 3630 not have direct access. 3631 3632 This parameter controls the MDS mitigation. The 3633 options are: 3634 3635 full - Enable MDS mitigation on vulnerable CPUs 3636 full,nosmt - Enable MDS mitigation and disable 3637 SMT on vulnerable CPUs 3638 off - Unconditionally disable MDS mitigation 3639 3640 On TAA-affected machines, mds=off can be prevented by 3641 an active TAA mitigation as both vulnerabilities are 3642 mitigated with the same mechanism so in order to disable 3643 this mitigation, you need to specify tsx_async_abort=off 3644 too. 3645 3646 Not specifying this option is equivalent to 3647 mds=full. 3648 3649 For details see: Documentation/admin-guide/hw-vuln/mds.rst 3650 3651 mem=nn[KMG] [HEXAGON,EARLY] Set the memory size. 3652 Must be specified, otherwise memory size will be 0. 3653 3654 mem=nn[KMG] [KNL,BOOT,EARLY] Force usage of a specific amount 3655 of memory Amount of memory to be used in cases 3656 as follows: 3657 3658 1 for test; 3659 2 when the kernel is not able to see the whole system memory; 3660 3 memory that lies after 'mem=' boundary is excluded from 3661 the hypervisor, then assigned to KVM guests. 3662 4 to limit the memory available for kdump kernel. 3663 3664 [ARC,MICROBLAZE] - the limit applies only to low memory, 3665 high memory is not affected. 3666 3667 [ARM64] - only limits memory covered by the linear 3668 mapping. The NOMAP regions are not affected. 3669 3670 [X86] Work as limiting max address. Use together 3671 with memmap= to avoid physical address space collisions. 3672 Without memmap= PCI devices could be placed at addresses 3673 belonging to unused RAM. 3674 3675 Note that this only takes effects during boot time since 3676 in above case 3, memory may need be hot added after boot 3677 if system memory of hypervisor is not sufficient. 3678 3679 mem=nn[KMG]@ss[KMG] 3680 [ARM,MIPS,EARLY] - override the memory layout 3681 reported by firmware. 3682 Define a memory region of size nn[KMG] starting at 3683 ss[KMG]. 3684 Multiple different regions can be specified with 3685 multiple mem= parameters on the command line. 3686 3687 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 3688 memory. 3689 3690 memblock=debug [KNL,EARLY] Enable memblock debug messages. 3691 3692 memchunk=nn[KMG] 3693 [KNL,SH] Allow user to override the default size for 3694 per-device physically contiguous DMA buffers. 3695 3696 memhp_default_state=online/offline/online_kernel/online_movable 3697 [KNL] Set the initial state for the memory hotplug 3698 onlining policy. If not specified, the default value is 3699 set according to the 3700 CONFIG_MHP_DEFAULT_ONLINE_TYPE kernel config 3701 options. 3702 See Documentation/admin-guide/mm/memory-hotplug.rst. 3703 3704 memmap=exactmap [KNL,X86,EARLY] Enable setting of an exact 3705 E820 memory map, as specified by the user. 3706 Such memmap=exactmap lines can be constructed based on 3707 BIOS output or other requirements. See the memmap=nn@ss 3708 option description. 3709 3710 memmap=nn[KMG]@ss[KMG] 3711 [KNL, X86,MIPS,XTENSA,EARLY] Force usage of a specific region of memory. 3712 Region of memory to be used is from ss to ss+nn. 3713 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 3714 which limits max address to nn[KMG]. 3715 Multiple different regions can be specified, 3716 comma delimited. 3717 Example: 3718 memmap=100M@2G,100M#3G,1G!1024G 3719 3720 memmap=nn[KMG]#ss[KMG] 3721 [KNL,ACPI,EARLY] Mark specific memory as ACPI data. 3722 Region of memory to be marked is from ss to ss+nn. 3723 3724 memmap=nn[KMG]$ss[KMG] 3725 [KNL,ACPI,EARLY] Mark specific memory as reserved. 3726 Region of memory to be reserved is from ss to ss+nn. 3727 Example: Exclude memory from 0x18690000-0x1869ffff 3728 memmap=64K$0x18690000 3729 or 3730 memmap=0x10000$0x18690000 3731 Some bootloaders may need an escape character before '$', 3732 like Grub2, otherwise '$' and the following number 3733 will be eaten. 3734 3735 memmap=nn[KMG]!ss[KMG,EARLY] 3736 [KNL,X86] Mark specific memory as protected. 3737 Region of memory to be used, from ss to ss+nn. 3738 The memory region may be marked as e820 type 12 (0xc) 3739 and is NVDIMM or ADR memory. 3740 3741 memmap=<size>%<offset>-<oldtype>+<newtype> 3742 [KNL,ACPI,EARLY] Convert memory within the specified region 3743 from <oldtype> to <newtype>. If "-<oldtype>" is left 3744 out, the whole region will be marked as <newtype>, 3745 even if previously unavailable. If "+<newtype>" is left 3746 out, matching memory will be removed. Types are 3747 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 3748 3 = ACPI, 12 = PRAM. 3749 3750 memory_corruption_check=0/1 [X86,EARLY] 3751 Some BIOSes seem to corrupt the first 64k of 3752 memory when doing things like suspend/resume. 3753 Setting this option will scan the memory 3754 looking for corruption. Enabling this will 3755 both detect corruption and prevent the kernel 3756 from using the memory being corrupted. 3757 However, it's intended as a diagnostic tool; if 3758 repeatable BIOS-originated corruption always 3759 affects the same memory, you can use memmap= 3760 to prevent the kernel from using that memory. 3761 3762 memory_corruption_check_size=size [X86,EARLY] 3763 By default it checks for corruption in the low 3764 64k, making this memory unavailable for normal 3765 use. Use this parameter to scan for 3766 corruption in more or less memory. 3767 3768 memory_corruption_check_period=seconds [X86,EARLY] 3769 By default it checks for corruption every 60 3770 seconds. Use this parameter to check at some 3771 other rate. 0 disables periodic checking. 3772 3773 memory_hotplug.memmap_on_memory 3774 [KNL,X86,ARM] Boolean flag to enable this feature. 3775 Format: {on | off (default)} 3776 When enabled, runtime hotplugged memory will 3777 allocate its internal metadata (struct pages, 3778 those vmemmap pages cannot be optimized even 3779 if hugetlb_free_vmemmap is enabled) from the 3780 hotadded memory which will allow to hotadd a 3781 lot of memory without requiring additional 3782 memory to do so. 3783 This feature is disabled by default because it 3784 has some implication on large (e.g. GB) 3785 allocations in some configurations (e.g. small 3786 memory blocks). 3787 The state of the flag can be read in 3788 /sys/module/memory_hotplug/parameters/memmap_on_memory. 3789 Note that even when enabled, there are a few cases where 3790 the feature is not effective. 3791 3792 memtest= [KNL,X86,ARM,M68K,PPC,RISCV,EARLY] Enable memtest 3793 Format: <integer> 3794 default : 0 <disable> 3795 Specifies the number of memtest passes to be 3796 performed. Each pass selects another test 3797 pattern from a given set of patterns. Memtest 3798 fills the memory with this pattern, validates 3799 memory contents and reserves bad memory 3800 regions that are detected. 3801 3802 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 3803 Valid arguments: on, off 3804 Default: off 3805 mem_encrypt=on: Activate SME 3806 mem_encrypt=off: Do not activate SME 3807 3808 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst 3809 for details on when memory encryption can be activated. 3810 3811 mem_sleep_default= [SUSPEND] Default system suspend mode: 3812 s2idle - Suspend-To-Idle 3813 shallow - Power-On Suspend or equivalent (if supported) 3814 deep - Suspend-To-RAM or equivalent (if supported) 3815 See Documentation/admin-guide/pm/sleep-states.rst. 3816 3817 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 3818 the BIOS has incorrectly applied a workaround. TinyBIOS 3819 version 0.98 is known to be affected, 0.99 fixes the 3820 problem by letting the user disable the workaround. 3821 3822 mga= [HW,DRM] 3823 3824 microcode= [X86] Control the behavior of the microcode loader. 3825 Available options, comma separated: 3826 3827 base_rev=X - with <X> with format: <u32> 3828 Set the base microcode revision of each thread when in 3829 debug mode. 3830 3831 dis_ucode_ldr: disable the microcode loader 3832 3833 force_minrev: 3834 Enable or disable the microcode minimal revision 3835 enforcement for the runtime microcode loader. 3836 3837 mini2440= [ARM,HW,KNL] 3838 Format:[0..2][b][c][t] 3839 Default: "0tb" 3840 MINI2440 configuration specification: 3841 0 - The attached screen is the 3.5" TFT 3842 1 - The attached screen is the 7" TFT 3843 2 - The VGA Shield is attached (1024x768) 3844 Leaving out the screen size parameter will not load 3845 the TFT driver, and the framebuffer will be left 3846 unconfigured. 3847 b - Enable backlight. The TFT backlight pin will be 3848 linked to the kernel VESA blanking code and a GPIO 3849 LED. This parameter is not necessary when using the 3850 VGA shield. 3851 c - Enable the s3c camera interface. 3852 t - Reserved for enabling touchscreen support. The 3853 touchscreen support is not enabled in the mainstream 3854 kernel as of 2.6.30, a preliminary port can be found 3855 in the "bleeding edge" mini2440 support kernel at 3856 https://repo.or.cz/w/linux-2.6/mini2440.git 3857 3858 mitigations= 3859 [X86,PPC,S390,ARM64,EARLY] Control optional mitigations for 3860 CPU vulnerabilities. This is a set of curated, 3861 arch-independent options, each of which is an 3862 aggregation of existing arch-specific options. 3863 3864 Note, "mitigations" is supported if and only if the 3865 kernel was built with CPU_MITIGATIONS=y. 3866 3867 off 3868 Disable all optional CPU mitigations. This 3869 improves system performance, but it may also 3870 expose users to several CPU vulnerabilities. 3871 Equivalent to: if nokaslr then kpti=0 [ARM64] 3872 gather_data_sampling=off [X86] 3873 indirect_target_selection=off [X86] 3874 kvm.nx_huge_pages=off [X86] 3875 l1tf=off [X86] 3876 mds=off [X86] 3877 mmio_stale_data=off [X86] 3878 no_entry_flush [PPC] 3879 no_uaccess_flush [PPC] 3880 nobp=0 [S390] 3881 nopti [X86,PPC] 3882 nospectre_bhb [ARM64] 3883 nospectre_v1 [X86,PPC] 3884 nospectre_v2 [X86,PPC,S390,ARM64] 3885 reg_file_data_sampling=off [X86] 3886 retbleed=off [X86] 3887 spec_rstack_overflow=off [X86] 3888 spec_store_bypass_disable=off [X86,PPC] 3889 spectre_bhi=off [X86] 3890 spectre_v2_user=off [X86] 3891 srbds=off [X86,INTEL] 3892 ssbd=force-off [ARM64] 3893 tsx_async_abort=off [X86] 3894 vmscape=off [X86] 3895 3896 Exceptions: 3897 This does not have any effect on 3898 kvm.nx_huge_pages when 3899 kvm.nx_huge_pages=force. 3900 3901 auto (default) 3902 Mitigate all CPU vulnerabilities, but leave SMT 3903 enabled, even if it's vulnerable. This is for 3904 users who don't want to be surprised by SMT 3905 getting disabled across kernel upgrades, or who 3906 have other ways of avoiding SMT-based attacks. 3907 Equivalent to: (default behavior) 3908 3909 auto,nosmt 3910 Mitigate all CPU vulnerabilities, disabling SMT 3911 if needed. This is for users who always want to 3912 be fully mitigated, even if it means losing SMT. 3913 Equivalent to: l1tf=flush,nosmt [X86] 3914 mds=full,nosmt [X86] 3915 tsx_async_abort=full,nosmt [X86] 3916 mmio_stale_data=full,nosmt [X86] 3917 retbleed=auto,nosmt [X86] 3918 3919 [X86] After one of the above options, additionally 3920 supports attack-vector based controls as documented in 3921 Documentation/admin-guide/hw-vuln/attack_vector_controls.rst 3922 3923 mminit_loglevel= 3924 [KNL,EARLY] When CONFIG_DEBUG_MEMORY_INIT is set, this 3925 parameter allows control of the logging verbosity for 3926 the additional memory initialisation checks. A value 3927 of 0 disables mminit logging and a level of 4 will 3928 log everything. Information is printed at KERN_DEBUG 3929 so loglevel=8 may also need to be specified. 3930 3931 mmio_stale_data= 3932 [X86,INTEL,EARLY] Control mitigation for the Processor 3933 MMIO Stale Data vulnerabilities. 3934 3935 Processor MMIO Stale Data is a class of 3936 vulnerabilities that may expose data after an MMIO 3937 operation. Exposed data could originate or end in 3938 the same CPU buffers as affected by MDS and TAA. 3939 Therefore, similar to MDS and TAA, the mitigation 3940 is to clear the affected CPU buffers. 3941 3942 This parameter controls the mitigation. The 3943 options are: 3944 3945 full - Enable mitigation on vulnerable CPUs 3946 3947 full,nosmt - Enable mitigation and disable SMT on 3948 vulnerable CPUs. 3949 3950 off - Unconditionally disable mitigation 3951 3952 On MDS or TAA affected machines, 3953 mmio_stale_data=off can be prevented by an active 3954 MDS or TAA mitigation as these vulnerabilities are 3955 mitigated with the same mechanism so in order to 3956 disable this mitigation, you need to specify 3957 mds=off and tsx_async_abort=off too. 3958 3959 Not specifying this option is equivalent to 3960 mmio_stale_data=full. 3961 3962 For details see: 3963 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst 3964 3965 <module>.async_probe[=<bool>] [KNL] 3966 If no <bool> value is specified or if the value 3967 specified is not a valid <bool>, enable asynchronous 3968 probe on this module. Otherwise, enable/disable 3969 asynchronous probe on this module as indicated by the 3970 <bool> value. See also: module.async_probe 3971 3972 module.async_probe=<bool> 3973 [KNL] When set to true, modules will use async probing 3974 by default. To enable/disable async probing for a 3975 specific module, use the module specific control that 3976 is documented under <module>.async_probe. When both 3977 module.async_probe and <module>.async_probe are 3978 specified, <module>.async_probe takes precedence for 3979 the specific module. 3980 3981 module.enable_dups_trace 3982 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set, 3983 this means that duplicate request_module() calls will 3984 trigger a WARN_ON() instead of a pr_warn(). Note that 3985 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s 3986 will always be issued and this option does nothing. 3987 module.sig_enforce 3988 [KNL] When CONFIG_MODULE_SIG is set, this means that 3989 modules without (valid) signatures will fail to load. 3990 Note that if CONFIG_MODULE_SIG_FORCE is set, that 3991 is always true, so this option does nothing. 3992 3993 module_blacklist= [KNL] Do not load a comma-separated list of 3994 modules. Useful for debugging problem modules. 3995 3996 mousedev.tap_time= 3997 [MOUSE] Maximum time between finger touching and 3998 leaving touchpad surface for touch to be considered 3999 a tap and be reported as a left button click (for 4000 touchpads working in absolute mode only). 4001 Format: <msecs> 4002 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 4003 reporting absolute coordinates, such as tablets 4004 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 4005 reporting absolute coordinates, such as tablets 4006 4007 movablecore= [KNL,X86,PPC,EARLY] 4008 Format: nn[KMGTPE] | nn% 4009 This parameter is the complement to kernelcore=, it 4010 specifies the amount of memory used for migratable 4011 allocations. If both kernelcore and movablecore is 4012 specified, then kernelcore will be at *least* the 4013 specified value but may be more. If movablecore on its 4014 own is specified, the administrator must be careful 4015 that the amount of memory usable for all allocations 4016 is not too small. 4017 4018 movable_node [KNL,EARLY] Boot-time switch to make hotplugable memory 4019 NUMA nodes to be movable. This means that the memory 4020 of such nodes will be usable only for movable 4021 allocations which rules out almost all kernel 4022 allocations. Use with caution! 4023 4024 MTD_Partition= [MTD] 4025 Format: <name>,<region-number>,<size>,<offset> 4026 4027 MTD_Region= [MTD] Format: 4028 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 4029 4030 mtdparts= [MTD] 4031 See drivers/mtd/parsers/cmdlinepart.c 4032 4033 mtouchusb.raw_coordinates= 4034 [HW] Make the MicroTouch USB driver use raw coordinates 4035 ('y', default) or cooked coordinates ('n') 4036 4037 mtrr=debug [X86,EARLY] 4038 Enable printing debug information related to MTRR 4039 registers at boot time. 4040 4041 mtrr_chunk_size=nn[KMG,X86,EARLY] 4042 used for mtrr cleanup. It is largest continuous chunk 4043 that could hold holes aka. UC entries. 4044 4045 mtrr_gran_size=nn[KMG,X86,EARLY] 4046 Used for mtrr cleanup. It is granularity of mtrr block. 4047 Default is 1. 4048 Large value could prevent small alignment from 4049 using up MTRRs. 4050 4051 mtrr_spare_reg_nr=n [X86,EARLY] 4052 Format: <integer> 4053 Range: 0,7 : spare reg number 4054 Default : 1 4055 Used for mtrr cleanup. It is spare mtrr entries number. 4056 Set to 2 or more if your graphical card needs more. 4057 4058 multitce=off [PPC] This parameter disables the use of the pSeries 4059 firmware feature for updating multiple TCE entries 4060 at a time. 4061 4062 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 4063 4064 netdev= [NET] Network devices parameters 4065 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 4066 Note that mem_start is often overloaded to mean 4067 something different and driver-specific. 4068 This usage is only documented in each driver source 4069 file if at all. 4070 4071 netpoll.carrier_timeout= 4072 [NET] Specifies amount of time (in seconds) that 4073 netpoll should wait for a carrier. By default netpoll 4074 waits 4 seconds. 4075 4076 nf_conntrack.acct= 4077 [NETFILTER] Enable connection tracking flow accounting 4078 0 to disable accounting 4079 1 to enable accounting 4080 Default value is 0. 4081 4082 nfs.cache_getent= 4083 [NFS] sets the pathname to the program which is used 4084 to update the NFS client cache entries. 4085 4086 nfs.cache_getent_timeout= 4087 [NFS] sets the timeout after which an attempt to 4088 update a cache entry is deemed to have failed. 4089 4090 nfs.callback_nr_threads= 4091 [NFSv4] set the total number of threads that the 4092 NFS client will assign to service NFSv4 callback 4093 requests. 4094 4095 nfs.callback_tcpport= 4096 [NFS] set the TCP port on which the NFSv4 callback 4097 channel should listen. 4098 4099 nfs.delay_retrans= 4100 [NFS] specifies the number of times the NFSv4 client 4101 retries the request before returning an EAGAIN error, 4102 after a reply of NFS4ERR_DELAY from the server. 4103 Only applies if the softerr mount option is enabled, 4104 and the specified value is >= 0. 4105 4106 nfs.enable_ino64= 4107 [NFS] enable 64-bit inode numbers. 4108 If zero, the NFS client will fake up a 32-bit inode 4109 number for the readdir() and stat() syscalls instead 4110 of returning the full 64-bit number. 4111 The default is to return 64-bit inode numbers. 4112 4113 nfs.idmap_cache_timeout= 4114 [NFS] set the maximum lifetime for idmapper cache 4115 entries. 4116 4117 nfs.max_session_cb_slots= 4118 [NFSv4.1] Sets the maximum number of session 4119 slots the client will assign to the callback 4120 channel. This determines the maximum number of 4121 callbacks the client will process in parallel for 4122 a particular server. 4123 4124 nfs.max_session_slots= 4125 [NFSv4.1] Sets the maximum number of session slots 4126 the client will attempt to negotiate with the server. 4127 This limits the number of simultaneous RPC requests 4128 that the client can send to the NFSv4.1 server. 4129 Note that there is little point in setting this 4130 value higher than the max_tcp_slot_table_limit. 4131 4132 nfs.nfs4_disable_idmapping= 4133 [NFSv4] When set to the default of '1', this option 4134 ensures that both the RPC level authentication 4135 scheme and the NFS level operations agree to use 4136 numeric uids/gids if the mount is using the 4137 'sec=sys' security flavour. In effect it is 4138 disabling idmapping, which can make migration from 4139 legacy NFSv2/v3 systems to NFSv4 easier. 4140 Servers that do not support this mode of operation 4141 will be autodetected by the client, and it will fall 4142 back to using the idmapper. 4143 To turn off this behaviour, set the value to '0'. 4144 4145 nfs.nfs4_unique_id= 4146 [NFS4] Specify an additional fixed unique ident- 4147 ification string that NFSv4 clients can insert into 4148 their nfs_client_id4 string. This is typically a 4149 UUID that is generated at system install time. 4150 4151 nfs.recover_lost_locks= 4152 [NFSv4] Attempt to recover locks that were lost due 4153 to a lease timeout on the server. Please note that 4154 doing this risks data corruption, since there are 4155 no guarantees that the file will remain unchanged 4156 after the locks are lost. 4157 If you want to enable the kernel legacy behaviour of 4158 attempting to recover these locks, then set this 4159 parameter to '1'. 4160 The default parameter value of '0' causes the kernel 4161 not to attempt recovery of lost locks. 4162 4163 nfs.send_implementation_id= 4164 [NFSv4.1] Send client implementation identification 4165 information in exchange_id requests. 4166 If zero, no implementation identification information 4167 will be sent. 4168 The default is to send the implementation identification 4169 information. 4170 4171 nfs4.layoutstats_timer= 4172 [NFSv4.2] Change the rate at which the kernel sends 4173 layoutstats to the pNFS metadata server. 4174 4175 Setting this to value to 0 causes the kernel to use 4176 whatever value is the default set by the layout 4177 driver. A non-zero value sets the minimum interval 4178 in seconds between layoutstats transmissions. 4179 4180 nfsd.inter_copy_offload_enable= 4181 [NFSv4.2] When set to 1, the server will support 4182 server-to-server copies for which this server is 4183 the destination of the copy. 4184 4185 nfsd.nfs4_disable_idmapping= 4186 [NFSv4] When set to the default of '1', the NFSv4 4187 server will return only numeric uids and gids to 4188 clients using auth_sys, and will accept numeric uids 4189 and gids from such clients. This is intended to ease 4190 migration from NFSv2/v3. 4191 4192 nfsd.nfsd4_ssc_umount_timeout= 4193 [NFSv4.2] When used as the destination of a 4194 server-to-server copy, knfsd temporarily mounts 4195 the source server. It caches the mount in case 4196 it will be needed again, and discards it if not 4197 used for the number of milliseconds specified by 4198 this parameter. 4199 4200 nfsaddrs= [NFS] Deprecated. Use ip= instead. 4201 See Documentation/admin-guide/nfs/nfsroot.rst. 4202 4203 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 4204 See Documentation/admin-guide/nfs/nfsroot.rst. 4205 4206 nfsrootdebug [NFS] enable nfsroot debugging messages. 4207 See Documentation/admin-guide/nfs/nfsroot.rst. 4208 4209 nmi_backtrace.backtrace_idle [KNL] 4210 Dump stacks even of idle CPUs in response to an 4211 NMI stack-backtrace request. 4212 4213 nmi_debug= [KNL,SH] Specify one or more actions to take 4214 when a NMI is triggered. 4215 Format: [state][,regs][,debounce][,die] 4216 4217 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 4218 Format: [panic,][nopanic,][rNNN,][num] 4219 Valid num: 0 or 1 4220 0 - turn hardlockup detector in nmi_watchdog off 4221 1 - turn hardlockup detector in nmi_watchdog on 4222 rNNN - configure the watchdog with raw perf event 0xNNN 4223 4224 When panic is specified, panic when an NMI watchdog 4225 timeout occurs (or 'nopanic' to not panic on an NMI 4226 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 4227 To disable both hard and soft lockup detectors, 4228 please see 'nowatchdog'. 4229 This is useful when you use a panic=... timeout and 4230 need the box quickly up again. 4231 4232 These settings can be accessed at runtime via 4233 the nmi_watchdog and hardlockup_panic sysctls. 4234 4235 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 4236 emulation library even if a 387 maths coprocessor 4237 is present. 4238 4239 no4lvl [RISCV,EARLY] Disable 4-level and 5-level paging modes. 4240 Forces kernel to use 3-level paging instead. 4241 4242 no5lvl [X86-64,RISCV,EARLY] Disable 5-level paging mode. Forces 4243 kernel to use 4-level paging instead. 4244 4245 noalign [KNL,ARM] 4246 4247 noapic [SMP,APIC,EARLY] Tells the kernel to not make use of any 4248 IOAPICs that may be present in the system. 4249 4250 noapictimer [APIC,X86] Don't set up the APIC timer 4251 4252 noautogroup Disable scheduler automatic task group creation. 4253 4254 nocache [ARM,EARLY] 4255 4256 no_console_suspend 4257 [HW] Never suspend the console 4258 Disable suspending of consoles during suspend and 4259 hibernate operations. Once disabled, debugging 4260 messages can reach various consoles while the rest 4261 of the system is being put to sleep (ie, while 4262 debugging driver suspend/resume hooks). This may 4263 not work reliably with all consoles, but is known 4264 to work with serial and VGA consoles. 4265 To facilitate more flexible debugging, we also add 4266 console_suspend, a printk module parameter to control 4267 it. Users could use console_suspend (usually 4268 /sys/module/printk/parameters/console_suspend) to 4269 turn on/off it dynamically. 4270 4271 no_debug_objects 4272 [KNL,EARLY] Disable object debugging 4273 4274 nodsp [SH] Disable hardware DSP at boot time. 4275 4276 noefi [EFI,EARLY] Disable EFI runtime services support. 4277 4278 no_entry_flush [PPC,EARLY] Don't flush the L1-D cache when entering the kernel. 4279 4280 noexec32 [X86-64] 4281 This affects only 32-bit executables. 4282 noexec32=on: enable non-executable mappings (default) 4283 read doesn't imply executable mappings 4284 noexec32=off: disable non-executable mappings 4285 read implies executable mappings 4286 4287 no_file_caps Tells the kernel not to honor file capabilities. The 4288 only way then for a file to be executed with privilege 4289 is to be setuid root or executed by root. 4290 4291 nofpu [MIPS,SH] Disable hardware FPU at boot time. 4292 4293 nofsgsbase [X86] Disables FSGSBASE instructions. 4294 4295 nofxsr [BUGS=X86-32] Disables x86 floating point extended 4296 register save and restore. The kernel will only save 4297 legacy floating-point registers on task switch. 4298 4299 nogbpages [X86] Do not use GB pages for kernel direct mappings. 4300 4301 no_hash_pointers 4302 [KNL,EARLY] 4303 Alias for "hash_pointers=never". 4304 4305 nohibernate [HIBERNATION] Disable hibernation and resume. 4306 4307 nohlt [ARM,ARM64,MICROBLAZE,MIPS,PPC,RISCV,SH] Forces the kernel to 4308 busy wait in do_idle() and not use the arch_cpu_idle() 4309 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP 4310 to be effective. This is useful on platforms where the 4311 sleep(SH) or wfi(ARM,ARM64) instructions do not work 4312 correctly or when doing power measurements to evaluate 4313 the impact of the sleep instructions. This is also 4314 useful when using JTAG debugger. 4315 4316 nohpet [X86] Don't use the HPET timer. 4317 4318 nohugeiomap [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge I/O mappings. 4319 4320 nohugevmalloc [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge vmalloc mappings. 4321 4322 nohz= [KNL] Boottime enable/disable dynamic ticks 4323 Valid arguments: on, off 4324 Default: on 4325 4326 nohz_full= [KNL,BOOT,SMP,ISOL] 4327 The argument is a cpu list, as described above. 4328 In kernels built with CONFIG_NO_HZ_FULL=y, set 4329 the specified list of CPUs whose tick will be stopped 4330 whenever possible. The boot CPU will be forced outside 4331 the range to maintain the timekeeping. Any CPUs 4332 in this list will have their RCU callbacks offloaded, 4333 just as if they had also been called out in the 4334 rcu_nocbs= boot parameter. 4335 4336 Note that this argument takes precedence over 4337 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 4338 4339 noinitrd [RAM] Tells the kernel not to load any configured 4340 initial RAM disk. 4341 4342 nointremap [X86-64,Intel-IOMMU,EARLY] Do not enable interrupt 4343 remapping. 4344 [Deprecated - use intremap=off] 4345 4346 noinvpcid [X86,EARLY] Disable the INVPCID cpu feature. 4347 4348 noiotrap [SH] Disables trapped I/O port accesses. 4349 4350 noirqdebug [X86-32] Disables the code which attempts to detect and 4351 disable unhandled interrupt sources. 4352 4353 noisapnp [ISAPNP] Disables ISA PnP code. 4354 4355 nokaslr [KNL,EARLY] 4356 When CONFIG_RANDOMIZE_BASE is set, this disables 4357 kernel and module base offset ASLR (Address Space 4358 Layout Randomization). 4359 4360 no-kvmapf [X86,KVM,EARLY] Disable paravirtualized asynchronous page 4361 fault handling. 4362 4363 no-kvmclock [X86,KVM,EARLY] Disable paravirtualized KVM clock driver 4364 4365 nolapic [X86-32,APIC,EARLY] Do not enable or use the local APIC. 4366 4367 nolapic_timer [X86-32,APIC,EARLY] Do not use the local APIC timer. 4368 4369 nomce [X86-32] Disable Machine Check Exception 4370 4371 nomfgpt [X86-32] Disable Multi-Function General Purpose 4372 Timer usage (for AMD Geode machines). 4373 4374 nomodeset Disable kernel modesetting. Most systems' firmware 4375 sets up a display mode and provides framebuffer memory 4376 for output. With nomodeset, DRM and fbdev drivers will 4377 not load if they could possibly displace the pre- 4378 initialized output. Only the system framebuffer will 4379 be available for use. The respective drivers will not 4380 perform display-mode changes or accelerated rendering. 4381 4382 Useful as error fallback, or for testing and debugging. 4383 4384 nomodule Disable module load 4385 4386 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 4387 shutdown the other cpus. Instead use the REBOOT_VECTOR 4388 irq. 4389 4390 nopat [X86,EARLY] Disable PAT (page attribute table extension of 4391 pagetables) support. 4392 4393 nopcid [X86-64,EARLY] Disable the PCID cpu feature. 4394 4395 nopku [X86] Disable Memory Protection Keys CPU feature found 4396 in some Intel CPUs. 4397 4398 nopti [X86-64,EARLY] 4399 Equivalent to pti=off 4400 4401 nopv= [X86,XEN,KVM,HYPER_V,VMWARE,EARLY] 4402 Disables the PV optimizations forcing the guest to run 4403 as generic guest with no PV drivers. Currently support 4404 XEN HVM, KVM, HYPER_V and VMWARE guest. 4405 4406 nopvspin [X86,XEN,KVM,EARLY] 4407 Disables the qspinlock slow path using PV optimizations 4408 which allow the hypervisor to 'idle' the guest on lock 4409 contention. 4410 4411 norandmaps Don't use address space randomization. Equivalent to 4412 echo 0 > /proc/sys/kernel/randomize_va_space 4413 4414 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 4415 with UP alternatives 4416 4417 noresume [SWSUSP] Disables resume and restores original swap 4418 space. 4419 4420 no-scroll [VGA] Disables scrollback. 4421 This is required for the Braillex ib80-piezo Braille 4422 reader made by F.H. Papenmeier (Germany). 4423 4424 nosgx [X86-64,SGX,EARLY] Disables Intel SGX kernel support. 4425 4426 nosmap [PPC,EARLY] 4427 Disable SMAP (Supervisor Mode Access Prevention) 4428 even if it is supported by processor. 4429 4430 nosmep [PPC64s,EARLY] 4431 Disable SMEP (Supervisor Mode Execution Prevention) 4432 even if it is supported by processor. 4433 4434 nosmp [SMP,EARLY] Tells an SMP kernel to act as a UP kernel, 4435 and disable the IO APIC. legacy for "maxcpus=0". 4436 4437 nosmt [KNL,MIPS,PPC,EARLY] Disable symmetric multithreading (SMT). 4438 Equivalent to smt=1. 4439 4440 [KNL,X86,PPC,S390] Disable symmetric multithreading (SMT). 4441 nosmt=force: Force disable SMT, cannot be undone 4442 via the sysfs control file. 4443 4444 nosoftlockup [KNL] Disable the soft-lockup detector. 4445 4446 nospec_store_bypass_disable 4447 [HW,EARLY] Disable all mitigations for the Speculative 4448 Store Bypass vulnerability 4449 4450 nospectre_bhb [ARM64,EARLY] Disable all mitigations for Spectre-BHB (branch 4451 history injection) vulnerability. System may allow data leaks 4452 with this option. 4453 4454 nospectre_v1 [X86,PPC,EARLY] Disable mitigations for Spectre Variant 1 4455 (bounds check bypass). With this option data leaks are 4456 possible in the system. 4457 4458 nospectre_v2 [X86,PPC_E500,ARM64,EARLY] Disable all mitigations 4459 for the Spectre variant 2 (indirect branch 4460 prediction) vulnerability. System may allow data 4461 leaks with this option. 4462 4463 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES,RISCV,LOONGARCH,EARLY] 4464 Disable paravirtualized steal time accounting. steal time 4465 is computed, but won't influence scheduler behaviour 4466 4467 nosync [HW,M68K] Disables sync negotiation for all devices. 4468 4469 no_timer_check [X86,APIC] Disables the code which tests for broken 4470 timer IRQ sources, i.e., the IO-APIC timer. This can 4471 work around problems with incorrect timer 4472 initialization on some boards. 4473 4474 no_uaccess_flush 4475 [PPC,EARLY] Don't flush the L1-D cache after accessing user data. 4476 4477 novmcoredd [KNL,KDUMP] 4478 Disable device dump. Device dump allows drivers to 4479 append dump data to vmcore so you can collect driver 4480 specified debug info. Drivers can append the data 4481 without any limit and this data is stored in memory, 4482 so this may cause significant memory stress. Disabling 4483 device dump can help save memory but the driver debug 4484 data will be no longer available. This parameter 4485 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 4486 is set. 4487 4488 no-vmw-sched-clock 4489 [X86,PV_OPS,EARLY] Disable paravirtualized VMware 4490 scheduler clock and use the default one. 4491 4492 nowatchdog [KNL] Disable both lockup detectors, i.e. 4493 soft-lockup and NMI watchdog (hard-lockup). 4494 4495 nowb [ARM,EARLY] 4496 4497 nox2apic [X86-64,APIC,EARLY] Do not enable x2APIC mode. 4498 4499 NOTE: this parameter will be ignored on systems with the 4500 LEGACY_XAPIC_DISABLED bit set in the 4501 IA32_XAPIC_DISABLE_STATUS MSR. 4502 4503 noxsave [BUGS=X86] Disables x86 extended register state save 4504 and restore using xsave. The kernel will fallback to 4505 enabling legacy floating-point and sse state. 4506 4507 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 4508 register states. The kernel will fall back to use 4509 xsave to save the states. By using this parameter, 4510 performance of saving the states is degraded because 4511 xsave doesn't support modified optimization while 4512 xsaveopt supports it on xsaveopt enabled systems. 4513 4514 noxsaves [X86] Disables xsaves and xrstors used in saving and 4515 restoring x86 extended register state in compacted 4516 form of xsave area. The kernel will fall back to use 4517 xsaveopt and xrstor to save and restore the states 4518 in standard form of xsave area. By using this 4519 parameter, xsave area per process might occupy more 4520 memory on xsaves enabled systems. 4521 4522 nr_cpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 4523 could support. nr_cpus=n : n >= 1 limits the kernel to 4524 support 'n' processors. It could be larger than the 4525 number of already plugged CPU during bootup, later in 4526 runtime you can physically add extra cpu until it reaches 4527 n. So during boot up some boot time memory for per-cpu 4528 variables need be pre-allocated for later physical cpu 4529 hot plugging. 4530 4531 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 4532 4533 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86, EARLY] 4534 Disable NUMA, Only set up a single NUMA node 4535 spanning all memory. 4536 4537 numa=fake=<size>[MG] 4538 [KNL, ARM64, RISCV, X86, EARLY] 4539 If given as a memory unit, fills all system RAM with 4540 nodes of size interleaved over physical nodes. 4541 4542 numa=fake=<N> 4543 [KNL, ARM64, RISCV, X86, EARLY] 4544 If given as an integer, fills all system RAM with N 4545 fake nodes interleaved over physical nodes. 4546 4547 numa=fake=<N>U 4548 [KNL, ARM64, RISCV, X86, EARLY] 4549 If given as an integer followed by 'U', it will 4550 divide each physical node into N emulated nodes. 4551 4552 numa=noacpi [X86] Don't parse the SRAT table for NUMA setup 4553 4554 numa=nohmat [X86] Don't parse the HMAT table for NUMA setup, or 4555 soft-reserved memory partitioning. 4556 4557 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic 4558 NUMA balancing. 4559 Allowed values are enable and disable 4560 4561 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 4562 'node', 'default' can be specified 4563 This can be set from sysctl after boot. 4564 See Documentation/admin-guide/sysctl/vm.rst for details. 4565 4566 ohci1394_dma=early [HW,EARLY] enable debugging via the ohci1394 driver. 4567 See Documentation/core-api/debugging-via-ohci1394.rst for more 4568 info. 4569 4570 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 4571 Rather than timing out after 20 ms if an EC 4572 command is not properly ACKed, override the length 4573 of the timeout. We have interrupts disabled while 4574 waiting for the ACK, so if this is set too high 4575 interrupts *may* be lost! 4576 4577 omap_mux= [OMAP] Override bootloader pin multiplexing. 4578 Format: <mux_mode0.mode_name=value>... 4579 For example, to override I2C bus2: 4580 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 4581 4582 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 4583 4584 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 4585 4586 boundary - index of last SLC block on Flex-OneNAND. 4587 The remaining blocks are configured as MLC blocks. 4588 lock - Configure if Flex-OneNAND boundary should be locked. 4589 Once locked, the boundary cannot be changed. 4590 1 indicates lock status, 0 indicates unlock status. 4591 4592 oops=panic [KNL,EARLY] 4593 Always panic on oopses. Default is to just kill the 4594 process, but there is a small probability of 4595 deadlocking the machine. 4596 This will also cause panics on machine check exceptions. 4597 Useful together with panic=30 to trigger a reboot. 4598 4599 page_alloc.shuffle= 4600 [KNL] Boolean flag to control whether the page allocator 4601 should randomize its free lists. This parameter can be 4602 used to enable/disable page randomization. The state of 4603 the flag can be read from sysfs at: 4604 /sys/module/page_alloc/parameters/shuffle. 4605 This parameter is only available if CONFIG_SHUFFLE_PAGE_ALLOCATOR=y. 4606 4607 page_owner= [KNL,EARLY] Boot-time page_owner enabling option. 4608 Storage of the information about who allocated 4609 each page is disabled in default. With this switch, 4610 we can turn it on. 4611 on: enable the feature 4612 4613 page_poison= [KNL,EARLY] Boot-time parameter changing the state of 4614 poisoning on the buddy allocator, available with 4615 CONFIG_PAGE_POISONING=y. 4616 off: turn off poisoning (default) 4617 on: turn on poisoning 4618 4619 page_reporting.page_reporting_order= 4620 [KNL] Minimal page reporting order 4621 Format: <integer> 4622 Adjust the minimal page reporting order. The page 4623 reporting is disabled when it exceeds MAX_PAGE_ORDER. 4624 4625 panic= [KNL] Kernel behaviour on panic: delay <timeout> 4626 timeout > 0: seconds before rebooting 4627 timeout = 0: wait forever 4628 timeout < 0: reboot immediately 4629 Format: <timeout> 4630 4631 panic_on_taint= [KNL,EARLY] 4632 Bitmask for conditionally calling panic() in add_taint() 4633 Format: <hex>[,nousertaint] 4634 Hexadecimal bitmask representing the set of TAINT flags 4635 that will cause the kernel to panic when add_taint() is 4636 called with any of the flags in this set. 4637 The optional switch "nousertaint" can be utilized to 4638 prevent userspace forced crashes by writing to sysctl 4639 /proc/sys/kernel/tainted any flagset matching with the 4640 bitmask set on panic_on_taint. 4641 See Documentation/admin-guide/tainted-kernels.rst for 4642 extra details on the taint flags that users can pick 4643 to compose the bitmask to assign to panic_on_taint. 4644 4645 panic_on_warn=1 panic() instead of WARN(). Useful to cause kdump 4646 on a WARN(). 4647 4648 panic_print= Bitmask for printing system info when panic happens. 4649 User can chose combination of the following bits: 4650 bit 0: print all tasks info 4651 bit 1: print system memory info 4652 bit 2: print timer info 4653 bit 3: print locks info if CONFIG_LOCKDEP is on 4654 bit 4: print ftrace buffer 4655 bit 5: replay all kernel messages on consoles at the end of panic 4656 bit 6: print all CPUs backtrace (if available in the arch) 4657 bit 7: print only tasks in uninterruptible (blocked) state 4658 *Be aware* that this option may print a _lot_ of lines, 4659 so there are risks of losing older messages in the log. 4660 Use this option carefully, maybe worth to setup a 4661 bigger log buffer with "log_buf_len" along with this. 4662 4663 panic_sys_info= A comma separated list of extra information to be dumped 4664 on panic. 4665 Format: val[,val...] 4666 Where @val can be any of the following: 4667 4668 tasks: print all tasks info 4669 mem: print system memory info 4670 timers: print timers info 4671 locks: print locks info if CONFIG_LOCKDEP is on 4672 ftrace: print ftrace buffer 4673 all_bt: print all CPUs backtrace (if available in the arch) 4674 blocked_tasks: print only tasks in uninterruptible (blocked) state 4675 4676 This is a human readable alternative to the 'panic_print' option. 4677 4678 panic_console_replay 4679 When panic happens, replay all kernel messages on 4680 consoles at the end of panic. 4681 4682 parkbd.port= [HW] Parallel port number the keyboard adapter is 4683 connected to, default is 0. 4684 Format: <parport#> 4685 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 4686 0 for XT, 1 for AT (default is AT). 4687 Format: <mode> 4688 4689 parport= [HW,PPT] Specify parallel ports. 0 disables. 4690 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 4691 Use 'auto' to force the driver to use any 4692 IRQ/DMA settings detected (the default is to 4693 ignore detected IRQ/DMA settings because of 4694 possible conflicts). You can specify the base 4695 address, IRQ, and DMA settings; IRQ and DMA 4696 should be numbers, or 'auto' (for using detected 4697 settings on that particular port), or 'nofifo' 4698 (to avoid using a FIFO even if it is detected). 4699 Parallel ports are assigned in the order they 4700 are specified on the command line, starting 4701 with parport0. 4702 4703 parport_init_mode= [HW,PPT] 4704 Configure VIA parallel port to operate in 4705 a specific mode. This is necessary on Pegasos 4706 computer where firmware has no options for setting 4707 up parallel port mode and sets it to spp. 4708 Currently this function knows 686a and 8231 chips. 4709 Format: [spp|ps2|epp|ecp|ecpepp] 4710 4711 pata_legacy.all= [HW,LIBATA] 4712 Format: <int> 4713 Set to non-zero to probe primary and secondary ISA 4714 port ranges on PCI systems where no PCI PATA device 4715 has been found at either range. Disabled by default. 4716 4717 pata_legacy.autospeed= [HW,LIBATA] 4718 Format: <int> 4719 Set to non-zero if a chip is present that snoops speed 4720 changes. Disabled by default. 4721 4722 pata_legacy.ht6560a= [HW,LIBATA] 4723 Format: <int> 4724 Set to 1, 2, or 3 for HT 6560A on the primary channel, 4725 the secondary channel, or both channels respectively. 4726 Disabled by default. 4727 4728 pata_legacy.ht6560b= [HW,LIBATA] 4729 Format: <int> 4730 Set to 1, 2, or 3 for HT 6560B on the primary channel, 4731 the secondary channel, or both channels respectively. 4732 Disabled by default. 4733 4734 pata_legacy.iordy_mask= [HW,LIBATA] 4735 Format: <int> 4736 IORDY enable mask. Set individual bits to allow IORDY 4737 for the respective channel. Bit 0 is for the first 4738 legacy channel handled by this driver, bit 1 is for 4739 the second channel, and so on. The sequence will often 4740 correspond to the primary legacy channel, the secondary 4741 legacy channel, and so on, but the handling of a PCI 4742 bus and the use of other driver options may interfere 4743 with the sequence. By default IORDY is allowed across 4744 all channels. 4745 4746 pata_legacy.opti82c46x= [HW,LIBATA] 4747 Format: <int> 4748 Set to 1, 2, or 3 for Opti 82c611A on the primary 4749 channel, the secondary channel, or both channels 4750 respectively. Disabled by default. 4751 4752 pata_legacy.opti82c611a= [HW,LIBATA] 4753 Format: <int> 4754 Set to 1, 2, or 3 for Opti 82c465MV on the primary 4755 channel, the secondary channel, or both channels 4756 respectively. Disabled by default. 4757 4758 pata_legacy.pio_mask= [HW,LIBATA] 4759 Format: <int> 4760 PIO mode mask for autospeed devices. Set individual 4761 bits to allow the use of the respective PIO modes. 4762 Bit 0 is for mode 0, bit 1 is for mode 1, and so on. 4763 All modes allowed by default. 4764 4765 pata_legacy.probe_all= [HW,LIBATA] 4766 Format: <int> 4767 Set to non-zero to probe tertiary and further ISA 4768 port ranges on PCI systems. Disabled by default. 4769 4770 pata_legacy.probe_mask= [HW,LIBATA] 4771 Format: <int> 4772 Probe mask for legacy ISA PATA ports. Depending on 4773 platform configuration and the use of other driver 4774 options up to 6 legacy ports are supported: 0x1f0, 4775 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing 4776 of individual ports can be disabled by setting the 4777 corresponding bits in the mask to 1. Bit 0 is for 4778 the first port in the list above (0x1f0), and so on. 4779 By default all supported ports are probed. 4780 4781 pata_legacy.qdi= [HW,LIBATA] 4782 Format: <int> 4783 Set to non-zero to probe QDI controllers. By default 4784 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise. 4785 4786 pata_legacy.winbond= [HW,LIBATA] 4787 Format: <int> 4788 Set to non-zero to probe Winbond controllers. Use 4789 the standard I/O port (0x130) if 1, otherwise the 4790 value given is the I/O port to use (typically 0x1b0). 4791 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE, 4792 0 otherwise. 4793 4794 pata_platform.pio_mask= [HW,LIBATA] 4795 Format: <int> 4796 Supported PIO mode mask. Set individual bits to allow 4797 the use of the respective PIO modes. Bit 0 is for 4798 mode 0, bit 1 is for mode 1, and so on. Mode 0 only 4799 allowed by default. 4800 4801 pause_on_oops=<int> 4802 Halt all CPUs after the first oops has been printed for 4803 the specified number of seconds. This is to be used if 4804 your oopses keep scrolling off the screen. 4805 4806 pcbit= [HW,ISDN] 4807 4808 pci=option[,option...] [PCI,EARLY] various PCI subsystem options. 4809 4810 Some options herein operate on a specific device 4811 or a set of devices (<pci_dev>). These are 4812 specified in one of the following formats: 4813 4814 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 4815 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 4816 4817 Note: the first format specifies a PCI 4818 bus/device/function address which may change 4819 if new hardware is inserted, if motherboard 4820 firmware changes, or due to changes caused 4821 by other kernel parameters. If the 4822 domain is left unspecified, it is 4823 taken to be zero. Optionally, a path 4824 to a device through multiple device/function 4825 addresses can be specified after the base 4826 address (this is more robust against 4827 renumbering issues). The second format 4828 selects devices using IDs from the 4829 configuration space which may match multiple 4830 devices in the system. 4831 4832 earlydump dump PCI config space before the kernel 4833 changes anything 4834 off [X86] don't probe for the PCI bus 4835 bios [X86-32] force use of PCI BIOS, don't access 4836 the hardware directly. Use this if your machine 4837 has a non-standard PCI host bridge. 4838 nobios [X86-32] disallow use of PCI BIOS, only direct 4839 hardware access methods are allowed. Use this 4840 if you experience crashes upon bootup and you 4841 suspect they are caused by the BIOS. 4842 conf1 [X86] Force use of PCI Configuration Access 4843 Mechanism 1 (config address in IO port 0xCF8, 4844 data in IO port 0xCFC, both 32-bit). 4845 conf2 [X86] Force use of PCI Configuration Access 4846 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 4847 the function, IO port 0xCFA, also 8-bit, sets 4848 bus number. The config space is then accessed 4849 through ports 0xC000-0xCFFF). 4850 See http://wiki.osdev.org/PCI for more info 4851 on the configuration access mechanisms. 4852 noaer [PCIE] If the PCIEAER kernel config parameter is 4853 enabled, this kernel boot option can be used to 4854 disable the use of PCIE advanced error reporting. 4855 nodomains [PCI] Disable support for multiple PCI 4856 root domains (aka PCI segments, in ACPI-speak). 4857 nommconf [X86] Disable use of MMCONFIG for PCI 4858 Configuration 4859 check_enable_amd_mmconf [X86] check for and enable 4860 properly configured MMIO access to PCI 4861 config space on AMD family 10h CPU 4862 nomsi [MSI] If the PCI_MSI kernel config parameter is 4863 enabled, this kernel boot option can be used to 4864 disable the use of MSI interrupts system-wide. 4865 noioapicquirk [APIC] Disable all boot interrupt quirks. 4866 Safety option to keep boot IRQs enabled. This 4867 should never be necessary. 4868 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 4869 primary IO-APIC for bridges that cannot disable 4870 boot IRQs. This fixes a source of spurious IRQs 4871 when the system masks IRQs. 4872 noioapicreroute [APIC] Disable workaround that uses the 4873 boot IRQ equivalent of an IRQ that connects to 4874 a chipset where boot IRQs cannot be disabled. 4875 The opposite of ioapicreroute. 4876 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 4877 routing table. These calls are known to be buggy 4878 on several machines and they hang the machine 4879 when used, but on other computers it's the only 4880 way to get the interrupt routing table. Try 4881 this option if the kernel is unable to allocate 4882 IRQs or discover secondary PCI buses on your 4883 motherboard. 4884 rom [X86] Assign address space to expansion ROMs. 4885 Use with caution as certain devices share 4886 address decoders between ROMs and other 4887 resources. 4888 norom [X86] Do not assign address space to 4889 expansion ROMs that do not already have 4890 BIOS assigned address ranges. 4891 nobar [X86] Do not assign address space to the 4892 BARs that weren't assigned by the BIOS. 4893 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 4894 assigned automatically to PCI devices. You can 4895 make the kernel exclude IRQs of your ISA cards 4896 this way. 4897 pirqaddr=0xAAAAA [X86] Specify the physical address 4898 of the PIRQ table (normally generated 4899 by the BIOS) if it is outside the 4900 F0000h-100000h range. 4901 lastbus=N [X86] Scan all buses thru bus #N. Can be 4902 useful if the kernel is unable to find your 4903 secondary buses and you want to tell it 4904 explicitly which ones they are. 4905 assign-busses [X86] Always assign all PCI bus 4906 numbers ourselves, overriding 4907 whatever the firmware may have done. 4908 usepirqmask [X86] Honor the possible IRQ mask stored 4909 in the BIOS $PIR table. This is needed on 4910 some systems with broken BIOSes, notably 4911 some HP Pavilion N5400 and Omnibook XE3 4912 notebooks. This will have no effect if ACPI 4913 IRQ routing is enabled. 4914 noacpi [X86] Do not use ACPI for IRQ routing 4915 or for PCI scanning. 4916 use_crs [X86] Use PCI host bridge window information 4917 from ACPI. On BIOSes from 2008 or later, this 4918 is enabled by default. If you need to use this, 4919 please report a bug. 4920 nocrs [X86] Ignore PCI host bridge windows from ACPI. 4921 If you need to use this, please report a bug. 4922 use_e820 [X86] Use E820 reservations to exclude parts of 4923 PCI host bridge windows. This is a workaround 4924 for BIOS defects in host bridge _CRS methods. 4925 If you need to use this, please report a bug to 4926 <linux-pci@vger.kernel.org>. 4927 no_e820 [X86] Ignore E820 reservations for PCI host 4928 bridge windows. This is the default on modern 4929 hardware. If you need to use this, please report 4930 a bug to <linux-pci@vger.kernel.org>. 4931 routeirq Do IRQ routing for all PCI devices. 4932 This is normally done in pci_enable_device(), 4933 so this option is a temporary workaround 4934 for broken drivers that don't call it. 4935 skip_isa_align [X86] do not align io start addr, so can 4936 handle more pci cards 4937 noearly [X86] Don't do any early type 1 scanning. 4938 This might help on some broken boards which 4939 machine check when some devices' config space 4940 is read. But various workarounds are disabled 4941 and some IOMMU drivers will not work. 4942 bfsort Sort PCI devices into breadth-first order. 4943 This sorting is done to get a device 4944 order compatible with older (<= 2.4) kernels. 4945 nobfsort Don't sort PCI devices into breadth-first order. 4946 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 4947 tuning and use the BIOS-configured MPS defaults. 4948 pcie_bus_safe Set every device's MPS to the largest value 4949 supported by all devices below the root complex. 4950 pcie_bus_perf Set device MPS to the largest allowable MPS 4951 based on its parent bus. Also set MRRS (Max 4952 Read Request Size) to the largest supported 4953 value (no larger than the MPS that the device 4954 or bus can support) for best performance. 4955 pcie_bus_peer2peer Set every device's MPS to 128B, which 4956 every device is guaranteed to support. This 4957 configuration allows peer-to-peer DMA between 4958 any pair of devices, possibly at the cost of 4959 reduced performance. This also guarantees 4960 that hot-added devices will work. 4961 cbiosize=nn[KMG] The fixed amount of bus space which is 4962 reserved for the CardBus bridge's IO window. 4963 The default value is 256 bytes. 4964 cbmemsize=nn[KMG] The fixed amount of bus space which is 4965 reserved for the CardBus bridge's memory 4966 window. The default value is 64 megabytes. 4967 resource_alignment= 4968 Format: 4969 [<order of align>@]<pci_dev>[; ...] 4970 Specifies alignment and device to reassign 4971 aligned memory resources. How to 4972 specify the device is described above. 4973 If <order of align> is not specified, 4974 PAGE_SIZE is used as alignment. 4975 A PCI-PCI bridge can be specified if resource 4976 windows need to be expanded. 4977 To specify the alignment for several 4978 instances of a device, the PCI vendor, 4979 device, subvendor, and subdevice may be 4980 specified, e.g., 12@pci:8086:9c22:103c:198f 4981 for 4096-byte alignment. 4982 ecrc= Enable/disable PCIe ECRC (transaction layer 4983 end-to-end CRC checking). Only effective if 4984 OS has native AER control (either granted by 4985 ACPI _OSC or forced via "pcie_ports=native") 4986 bios: Use BIOS/firmware settings. This is the 4987 the default. 4988 off: Turn ECRC off 4989 on: Turn ECRC on. 4990 hpiosize=nn[KMG] The fixed amount of bus space which is 4991 reserved for hotplug bridge's IO window. 4992 Default size is 256 bytes. 4993 hpmmiosize=nn[KMG] The fixed amount of bus space which is 4994 reserved for hotplug bridge's MMIO window. 4995 Default size is 2 megabytes. 4996 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 4997 reserved for hotplug bridge's MMIO_PREF window. 4998 Default size is 2 megabytes. 4999 hpmemsize=nn[KMG] The fixed amount of bus space which is 5000 reserved for hotplug bridge's MMIO and 5001 MMIO_PREF window. 5002 Default size is 2 megabytes. 5003 hpbussize=nn The minimum amount of additional bus numbers 5004 reserved for buses below a hotplug bridge. 5005 Default is 1. 5006 realloc= Enable/disable reallocating PCI bridge resources 5007 if allocations done by BIOS are too small to 5008 accommodate resources required by all child 5009 devices. 5010 off: Turn realloc off 5011 on: Turn realloc on 5012 realloc same as realloc=on 5013 noari do not use PCIe ARI. 5014 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 5015 do not use PCIe ATS (and IOMMU device IOTLB). 5016 pcie_scan_all Scan all possible PCIe devices. Otherwise we 5017 only look for one device below a PCIe downstream 5018 port. 5019 big_root_window Try to add a big 64bit memory window to the PCIe 5020 root complex on AMD CPUs. Some GFX hardware 5021 can resize a BAR to allow access to all VRAM. 5022 Adding the window is slightly risky (it may 5023 conflict with unreported devices), so this 5024 taints the kernel. 5025 disable_acs_redir=<pci_dev>[; ...] 5026 Specify one or more PCI devices (in the format 5027 specified above) separated by semicolons. 5028 Each device specified will have the PCI ACS 5029 redirect capabilities forced off which will 5030 allow P2P traffic between devices through 5031 bridges without forcing it upstream. Note: 5032 this removes isolation between devices and 5033 may put more devices in an IOMMU group. 5034 config_acs= 5035 Format: 5036 <ACS flags>@<pci_dev>[; ...] 5037 Specify one or more PCI devices (in the format 5038 specified above) optionally prepended with flags 5039 and separated by semicolons. The respective 5040 capabilities will be enabled, disabled or 5041 unchanged based on what is specified in 5042 flags. 5043 5044 ACS Flags is defined as follows: 5045 bit-0 : ACS Source Validation 5046 bit-1 : ACS Translation Blocking 5047 bit-2 : ACS P2P Request Redirect 5048 bit-3 : ACS P2P Completion Redirect 5049 bit-4 : ACS Upstream Forwarding 5050 bit-5 : ACS P2P Egress Control 5051 bit-6 : ACS Direct Translated P2P 5052 Each bit can be marked as: 5053 '0' – force disabled 5054 '1' – force enabled 5055 'x' – unchanged 5056 For example, 5057 pci=config_acs=10x@pci:0:0 5058 would configure all devices that support 5059 ACS to enable P2P Request Redirect, disable 5060 Translation Blocking, and leave Source 5061 Validation unchanged from whatever power-up 5062 or firmware set it to. 5063 5064 Note: this may remove isolation between devices 5065 and may put more devices in an IOMMU group. 5066 force_floating [S390] Force usage of floating interrupts. 5067 nomio [S390] Do not use MIO instructions. 5068 norid [S390] ignore the RID field and force use of 5069 one PCI domain per PCI function 5070 notph [PCIE] If the PCIE_TPH kernel config parameter 5071 is enabled, this kernel boot option can be used 5072 to disable PCIe TLP Processing Hints support 5073 system-wide. 5074 5075 pcie_aspm= [PCIE] Forcibly enable or ignore PCIe Active State Power 5076 Management. 5077 off Don't touch ASPM configuration at all. Leave any 5078 configuration done by firmware unchanged. 5079 force Enable ASPM even on devices that claim not to support it. 5080 WARNING: Forcing ASPM on may cause system lockups. 5081 5082 pcie_ports= [PCIE] PCIe port services handling: 5083 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 5084 even if the platform doesn't give the OS permission to 5085 use them. This may cause conflicts if the platform 5086 also tries to use these services. 5087 dpc-native Use native PCIe service for DPC only. May 5088 cause conflicts if firmware uses AER or DPC. 5089 compat Disable native PCIe services (PME, AER, DPC, PCIe 5090 hotplug). 5091 5092 pcie_port_pm= [PCIE] PCIe port power management handling: 5093 off Disable power management of all PCIe ports 5094 force Forcibly enable power management of all PCIe ports 5095 5096 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 5097 nomsi Do not use MSI for native PCIe PME signaling (this makes 5098 all PCIe root ports use INTx for all services). 5099 5100 pcmv= [HW,PCMCIA] BadgePAD 4 5101 5102 pd_ignore_unused 5103 [PM] 5104 Keep all power-domains already enabled by bootloader on, 5105 even if no driver has claimed them. This is useful 5106 for debug and development, but should not be 5107 needed on a platform with proper driver support. 5108 5109 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 5110 boot time. 5111 Format: { 0 | 1 } 5112 See arch/parisc/kernel/pdc_chassis.c 5113 5114 percpu_alloc= [MM,EARLY] 5115 Select which percpu first chunk allocator to use. 5116 Currently supported values are "embed" and "page". 5117 Archs may support subset or none of the selections. 5118 See comments in mm/percpu.c for details on each 5119 allocator. This parameter is primarily for debugging 5120 and performance comparison. 5121 5122 pirq= [SMP,APIC] Manual mp-table setup 5123 See Documentation/arch/x86/i386/IO-APIC.rst. 5124 5125 plip= [PPT,NET] Parallel port network link 5126 Format: { parport<nr> | timid | 0 } 5127 See also Documentation/admin-guide/parport.rst. 5128 5129 pmtmr= [X86] Manual setup of pmtmr I/O Port. 5130 Override pmtimer IOPort with a hex value. 5131 e.g. pmtmr=0x508 5132 5133 pmu_override= [PPC] Override the PMU. 5134 This option takes over the PMU facility, so it is no 5135 longer usable by perf. Setting this option starts the 5136 PMU counters by setting MMCR0 to 0 (the FC bit is 5137 cleared). If a number is given, then MMCR1 is set to 5138 that number, otherwise (e.g., 'pmu_override=on'), MMCR1 5139 remains 0. 5140 5141 pm_async= [PM] 5142 Format: off 5143 This parameter sets the initial value of the 5144 /sys/power/pm_async sysfs knob at boot time. 5145 If set to "off", disables asynchronous suspend and 5146 resume of devices during system-wide power transitions. 5147 This can be useful on platforms where device 5148 dependencies are not well-defined, or for debugging 5149 power management issues. Asynchronous operations are 5150 enabled by default. 5151 5152 5153 pm_debug_messages [SUSPEND,KNL] 5154 Enable suspend/resume debug messages during boot up. 5155 5156 pnp.debug=1 [PNP] 5157 Enable PNP debug messages (depends on the 5158 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 5159 via /sys/module/pnp/parameters/debug. We always show 5160 current resource usage; turning this on also shows 5161 possible settings and some assignment information. 5162 5163 pnpacpi= [ACPI] 5164 { off } 5165 5166 pnpbios= [ISAPNP] 5167 { on | off | curr | res | no-curr | no-res } 5168 5169 pnp_reserve_irq= 5170 [ISAPNP] Exclude IRQs for the autoconfiguration 5171 5172 pnp_reserve_dma= 5173 [ISAPNP] Exclude DMAs for the autoconfiguration 5174 5175 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 5176 Ranges are in pairs (I/O port base and size). 5177 5178 pnp_reserve_mem= 5179 [ISAPNP] Exclude memory regions for the 5180 autoconfiguration. 5181 Ranges are in pairs (memory base and size). 5182 5183 ports= [IP_VS_FTP] IPVS ftp helper module 5184 Default is 21. 5185 Up to 8 (IP_VS_APP_MAX_PORTS) ports 5186 may be specified. 5187 Format: <port>,<port>.... 5188 5189 possible_cpus= [SMP,S390,X86] 5190 Format: <unsigned int> 5191 Set the number of possible CPUs, overriding the 5192 regular discovery mechanisms (such as ACPI/FW, etc). 5193 5194 powersave=off [PPC] This option disables power saving features. 5195 It specifically disables cpuidle and sets the 5196 platform machine description specific power_save 5197 function to NULL. On Idle the CPU just reduces 5198 execution priority. 5199 5200 ppc_strict_facility_enable 5201 [PPC,ENABLE] This option catches any kernel floating point, 5202 Altivec, VSX and SPE outside of regions specifically 5203 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 5204 There is some performance impact when enabling this. 5205 5206 ppc_tm= [PPC,EARLY] 5207 Format: {"off"} 5208 Disable Hardware Transactional Memory 5209 5210 preempt= [KNL] 5211 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC 5212 none - Limited to cond_resched() calls 5213 voluntary - Limited to cond_resched() and might_sleep() calls 5214 full - Any section that isn't explicitly preempt disabled 5215 can be preempted anytime. Tasks will also yield 5216 contended spinlocks (if the critical section isn't 5217 explicitly preempt disabled beyond the lock itself). 5218 lazy - Scheduler controlled. Similar to full but instead 5219 of preempting the task immediately, the task gets 5220 one HZ tick time to yield itself before the 5221 preemption will be forced. One preemption is when the 5222 task returns to user space. 5223 5224 print-fatal-signals= 5225 [KNL] debug: print fatal signals 5226 5227 If enabled, warn about various signal handling 5228 related application anomalies: too many signals, 5229 too many POSIX.1 timers, fatal signals causing a 5230 coredump - etc. 5231 5232 If you hit the warning due to signal overflow, 5233 you might want to try "ulimit -i unlimited". 5234 5235 default: off. 5236 5237 printk.always_kmsg_dump= 5238 Trigger kmsg_dump for cases other than kernel oops or 5239 panics 5240 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5241 default: disabled 5242 5243 printk.console_no_auto_verbose= 5244 Disable console loglevel raise on oops, panic 5245 or lockdep-detected issues (only if lock debug is on). 5246 With an exception to setups with low baudrate on 5247 serial console, keeping this 0 is a good choice 5248 in order to provide more debug information. 5249 Format: <bool> 5250 default: 0 (auto_verbose is enabled) 5251 5252 printk.debug_non_panic_cpus= 5253 Allows storing messages from non-panic CPUs into 5254 the printk log buffer during panic(). They are 5255 flushed to consoles by the panic-CPU on 5256 a best-effort basis. 5257 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5258 Default: disabled 5259 5260 printk.devkmsg={on,off,ratelimit} 5261 Control writing to /dev/kmsg. 5262 on - unlimited logging to /dev/kmsg from userspace 5263 off - logging to /dev/kmsg disabled 5264 ratelimit - ratelimit the logging 5265 Default: ratelimit 5266 5267 printk.time= Show timing data prefixed to each printk message line 5268 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5269 5270 proc_mem.force_override= [KNL] 5271 Format: {always | ptrace | never} 5272 Traditionally /proc/pid/mem allows memory permissions to be 5273 overridden without restrictions. This option may be set to 5274 restrict that. Can be one of: 5275 - 'always': traditional behavior always allows mem overrides. 5276 - 'ptrace': only allow mem overrides for active ptracers. 5277 - 'never': never allow mem overrides. 5278 If not specified, default is the CONFIG_PROC_MEM_* choice. 5279 5280 processor.max_cstate= [HW,ACPI] 5281 Limit processor to maximum C-state 5282 max_cstate=9 overrides any DMI blacklist limit. 5283 5284 processor.nocst [HW,ACPI] 5285 Ignore the _CST method to determine C-states, 5286 instead using the legacy FADT method 5287 5288 profile= [KNL] Enable kernel profiling via /proc/profile 5289 Format: [<profiletype>,]<number> 5290 Param: <profiletype>: "schedule" or "kvm" 5291 [defaults to kernel profiling] 5292 Param: "schedule" - profile schedule points. 5293 Param: "kvm" - profile VM exits. 5294 Param: <number> - step/bucket size as a power of 2 for 5295 statistical time based profiling. 5296 5297 prompt_ramdisk= [RAM] [Deprecated] 5298 5299 prot_virt= [S390] enable hosting protected virtual machines 5300 isolated from the hypervisor (if hardware supports 5301 that). If enabled, the default kernel base address 5302 might be overridden even when Kernel Address Space 5303 Layout Randomization is disabled. 5304 Format: <bool> 5305 5306 psi= [KNL] Enable or disable pressure stall information 5307 tracking. 5308 Format: <bool> 5309 5310 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 5311 probe for; one of (bare|imps|exps|lifebook|any). 5312 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 5313 per second. 5314 psmouse.resetafter= [HW,MOUSE] 5315 Try to reset the device after so many bad packets 5316 (0 = never). 5317 psmouse.resolution= 5318 [HW,MOUSE] Set desired mouse resolution, in dpi. 5319 psmouse.smartscroll= 5320 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 5321 0 = disabled, 1 = enabled (default). 5322 5323 pstore.backend= Specify the name of the pstore backend to use 5324 5325 pti= [X86-64] Control Page Table Isolation of user and 5326 kernel address spaces. Disabling this feature 5327 removes hardening, but improves performance of 5328 system calls and interrupts. 5329 5330 on - unconditionally enable 5331 off - unconditionally disable 5332 auto - kernel detects whether your CPU model is 5333 vulnerable to issues that PTI mitigates 5334 5335 Not specifying this option is equivalent to pti=auto. 5336 5337 pty.legacy_count= 5338 [KNL] Number of legacy pty's. Overwrites compiled-in 5339 default number. 5340 5341 quiet [KNL,EARLY] Disable most log messages 5342 5343 r128= [HW,DRM] 5344 5345 radix_hcall_invalidate=on [PPC/PSERIES] 5346 Disable RADIX GTSE feature and use hcall for TLB 5347 invalidate. 5348 5349 raid= [HW,RAID] 5350 See Documentation/admin-guide/md.rst. 5351 5352 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 5353 See Documentation/admin-guide/blockdev/ramdisk.rst. 5354 5355 ramdisk_start= [RAM] RAM disk image start address 5356 5357 random.trust_cpu=off 5358 [KNL,EARLY] Disable trusting the use of the CPU's 5359 random number generator (if available) to 5360 initialize the kernel's RNG. 5361 5362 random.trust_bootloader=off 5363 [KNL,EARLY] Disable trusting the use of the a seed 5364 passed by the bootloader (if available) to 5365 initialize the kernel's RNG. 5366 5367 randomize_kstack_offset= 5368 [KNL,EARLY] Enable or disable kernel stack offset 5369 randomization, which provides roughly 5 bits of 5370 entropy, frustrating memory corruption attacks 5371 that depend on stack address determinism or 5372 cross-syscall address exposures. This is only 5373 available on architectures that have defined 5374 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET. 5375 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5376 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT. 5377 5378 ras=option[,option,...] [KNL] RAS-specific options 5379 5380 cec_disable [X86] 5381 Disable the Correctable Errors Collector, 5382 see CONFIG_RAS_CEC help text. 5383 5384 rcu_nocbs[=cpu-list] 5385 [KNL] The optional argument is a cpu list, 5386 as described above. 5387 5388 In kernels built with CONFIG_RCU_NOCB_CPU=y, 5389 enable the no-callback CPU mode, which prevents 5390 such CPUs' callbacks from being invoked in 5391 softirq context. Invocation of such CPUs' RCU 5392 callbacks will instead be offloaded to "rcuox/N" 5393 kthreads created for that purpose, where "x" is 5394 "p" for RCU-preempt, "s" for RCU-sched, and "g" 5395 for the kthreads that mediate grace periods; and 5396 "N" is the CPU number. This reduces OS jitter on 5397 the offloaded CPUs, which can be useful for HPC 5398 and real-time workloads. It can also improve 5399 energy efficiency for asymmetric multiprocessors. 5400 5401 If a cpulist is passed as an argument, the specified 5402 list of CPUs is set to no-callback mode from boot. 5403 5404 Otherwise, if the '=' sign and the cpulist 5405 arguments are omitted, no CPU will be set to 5406 no-callback mode from boot but the mode may be 5407 toggled at runtime via cpusets. 5408 5409 Note that this argument takes precedence over 5410 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 5411 5412 rcu_nocb_poll [KNL] 5413 Rather than requiring that offloaded CPUs 5414 (specified by rcu_nocbs= above) explicitly 5415 awaken the corresponding "rcuoN" kthreads, 5416 make these kthreads poll for callbacks. 5417 This improves the real-time response for the 5418 offloaded CPUs by relieving them of the need to 5419 wake up the corresponding kthread, but degrades 5420 energy efficiency by requiring that the kthreads 5421 periodically wake up to do the polling. 5422 5423 rcutree.blimit= [KNL] 5424 Set maximum number of finished RCU callbacks to 5425 process in one batch. 5426 5427 rcutree.csd_lock_suppress_rcu_stall= [KNL] 5428 Do only a one-line RCU CPU stall warning when 5429 there is an ongoing too-long CSD-lock wait. 5430 5431 rcutree.do_rcu_barrier= [KNL] 5432 Request a call to rcu_barrier(). This is 5433 throttled so that userspace tests can safely 5434 hammer on the sysfs variable if they so choose. 5435 If triggered before the RCU grace-period machinery 5436 is fully active, this will error out with EAGAIN. 5437 5438 rcutree.dump_tree= [KNL] 5439 Dump the structure of the rcu_node combining tree 5440 out at early boot. This is used for diagnostic 5441 purposes, to verify correct tree setup. 5442 5443 rcutree.gp_cleanup_delay= [KNL] 5444 Set the number of jiffies to delay each step of 5445 RCU grace-period cleanup. 5446 5447 rcutree.gp_init_delay= [KNL] 5448 Set the number of jiffies to delay each step of 5449 RCU grace-period initialization. 5450 5451 rcutree.gp_preinit_delay= [KNL] 5452 Set the number of jiffies to delay each step of 5453 RCU grace-period pre-initialization, that is, 5454 the propagation of recent CPU-hotplug changes up 5455 the rcu_node combining tree. 5456 5457 rcutree.jiffies_till_first_fqs= [KNL] 5458 Set delay from grace-period initialization to 5459 first attempt to force quiescent states. 5460 Units are jiffies, minimum value is zero, 5461 and maximum value is HZ. 5462 5463 rcutree.jiffies_till_next_fqs= [KNL] 5464 Set delay between subsequent attempts to force 5465 quiescent states. Units are jiffies, minimum 5466 value is one, and maximum value is HZ. 5467 5468 rcutree.jiffies_till_sched_qs= [KNL] 5469 Set required age in jiffies for a 5470 given grace period before RCU starts 5471 soliciting quiescent-state help from 5472 rcu_note_context_switch() and cond_resched(). 5473 If not specified, the kernel will calculate 5474 a value based on the most recent settings 5475 of rcutree.jiffies_till_first_fqs 5476 and rcutree.jiffies_till_next_fqs. 5477 This calculated value may be viewed in 5478 rcutree.jiffies_to_sched_qs. Any attempt to set 5479 rcutree.jiffies_to_sched_qs will be cheerfully 5480 overwritten. 5481 5482 rcutree.kthread_prio= [KNL,BOOT] 5483 Set the SCHED_FIFO priority of the RCU per-CPU 5484 kthreads (rcuc/N). This value is also used for 5485 the priority of the RCU boost threads (rcub/N) 5486 and for the RCU grace-period kthreads (rcu_bh, 5487 rcu_preempt, and rcu_sched). If RCU_BOOST is 5488 set, valid values are 1-99 and the default is 1 5489 (the least-favored priority). Otherwise, when 5490 RCU_BOOST is not set, valid values are 0-99 and 5491 the default is zero (non-realtime operation). 5492 When RCU_NOCB_CPU is set, also adjust the 5493 priority of NOCB callback kthreads. 5494 5495 rcutree.nocb_nobypass_lim_per_jiffy= [KNL] 5496 On callback-offloaded (rcu_nocbs) CPUs, 5497 RCU reduces the lock contention that would 5498 otherwise be caused by callback floods through 5499 use of the ->nocb_bypass list. However, in the 5500 common non-flooded case, RCU queues directly to 5501 the main ->cblist in order to avoid the extra 5502 overhead of the ->nocb_bypass list and its lock. 5503 But if there are too many callbacks queued during 5504 a single jiffy, RCU pre-queues the callbacks into 5505 the ->nocb_bypass queue. The definition of "too 5506 many" is supplied by this kernel boot parameter. 5507 5508 rcutree.nohz_full_patience_delay= [KNL] 5509 On callback-offloaded (rcu_nocbs) CPUs, avoid 5510 disturbing RCU unless the grace period has 5511 reached the specified age in milliseconds. 5512 Defaults to zero. Large values will be capped 5513 at five seconds. All values will be rounded down 5514 to the nearest value representable by jiffies. 5515 5516 rcutree.qhimark= [KNL] 5517 Set threshold of queued RCU callbacks beyond which 5518 batch limiting is disabled. 5519 5520 rcutree.qlowmark= [KNL] 5521 Set threshold of queued RCU callbacks below which 5522 batch limiting is re-enabled. 5523 5524 rcutree.qovld= [KNL] 5525 Set threshold of queued RCU callbacks beyond which 5526 RCU's force-quiescent-state scan will aggressively 5527 enlist help from cond_resched() and sched IPIs to 5528 help CPUs more quickly reach quiescent states. 5529 Set to less than zero to make this be set based 5530 on rcutree.qhimark at boot time and to zero to 5531 disable more aggressive help enlistment. 5532 5533 rcutree.rcu_delay_page_cache_fill_msec= [KNL] 5534 Set the page-cache refill delay (in milliseconds) 5535 in response to low-memory conditions. The range 5536 of permitted values is in the range 0:100000. 5537 5538 rcutree.rcu_divisor= [KNL] 5539 Set the shift-right count to use to compute 5540 the callback-invocation batch limit bl from 5541 the number of callbacks queued on this CPU. 5542 The result will be bounded below by the value of 5543 the rcutree.blimit kernel parameter. Every bl 5544 callbacks, the softirq handler will exit in 5545 order to allow the CPU to do other work. 5546 5547 Please note that this callback-invocation batch 5548 limit applies only to non-offloaded callback 5549 invocation. Offloaded callbacks are instead 5550 invoked in the context of an rcuoc kthread, which 5551 scheduler will preempt as it does any other task. 5552 5553 rcutree.rcu_fanout_exact= [KNL] 5554 Disable autobalancing of the rcu_node combining 5555 tree. This is used by rcutorture, and might 5556 possibly be useful for architectures having high 5557 cache-to-cache transfer latencies. 5558 5559 rcutree.rcu_fanout_leaf= [KNL] 5560 Change the number of CPUs assigned to each 5561 leaf rcu_node structure. Useful for very 5562 large systems, which will choose the value 64, 5563 and for NUMA systems with large remote-access 5564 latencies, which will choose a value aligned 5565 with the appropriate hardware boundaries. 5566 5567 rcutree.rcu_min_cached_objs= [KNL] 5568 Minimum number of objects which are cached and 5569 maintained per one CPU. Object size is equal 5570 to PAGE_SIZE. The cache allows to reduce the 5571 pressure to page allocator, also it makes the 5572 whole algorithm to behave better in low memory 5573 condition. 5574 5575 rcutree.rcu_nocb_gp_stride= [KNL] 5576 Set the number of NOCB callback kthreads in 5577 each group, which defaults to the square root 5578 of the number of CPUs. Larger numbers reduce 5579 the wakeup overhead on the global grace-period 5580 kthread, but increases that same overhead on 5581 each group's NOCB grace-period kthread. 5582 5583 rcutree.rcu_kick_kthreads= [KNL] 5584 Cause the grace-period kthread to get an extra 5585 wake_up() if it sleeps three times longer than 5586 it should at force-quiescent-state time. 5587 This wake_up() will be accompanied by a 5588 WARN_ONCE() splat and an ftrace_dump(). 5589 5590 rcutree.rcu_resched_ns= [KNL] 5591 Limit the time spend invoking a batch of RCU 5592 callbacks to the specified number of nanoseconds. 5593 By default, this limit is checked only once 5594 every 32 callbacks in order to limit the pain 5595 inflicted by local_clock() overhead. 5596 5597 rcutree.rcu_unlock_delay= [KNL] 5598 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 5599 this specifies an rcu_read_unlock()-time delay 5600 in microseconds. This defaults to zero. 5601 Larger delays increase the probability of 5602 catching RCU pointer leaks, that is, buggy use 5603 of RCU-protected pointers after the relevant 5604 rcu_read_unlock() has completed. 5605 5606 rcutree.sysrq_rcu= [KNL] 5607 Commandeer a sysrq key to dump out Tree RCU's 5608 rcu_node tree with an eye towards determining 5609 why a new grace period has not yet started. 5610 5611 rcutree.use_softirq= [KNL] 5612 If set to zero, move all RCU_SOFTIRQ processing to 5613 per-CPU rcuc kthreads. Defaults to a non-zero 5614 value, meaning that RCU_SOFTIRQ is used by default. 5615 Specify rcutree.use_softirq=0 to use rcuc kthreads. 5616 5617 But note that CONFIG_PREEMPT_RT=y kernels disable 5618 this kernel boot parameter, forcibly setting it 5619 to zero. 5620 5621 rcutree.enable_rcu_lazy= [KNL] 5622 To save power, batch RCU callbacks and flush after 5623 delay, memory pressure or callback list growing too 5624 big. 5625 5626 rcutree.rcu_normal_wake_from_gp= [KNL] 5627 Reduces a latency of synchronize_rcu() call. This approach 5628 maintains its own track of synchronize_rcu() callers, so it 5629 does not interact with regular callbacks because it does not 5630 use a call_rcu[_hurry]() path. Please note, this is for a 5631 normal grace period. 5632 5633 How to enable it: 5634 5635 echo 1 > /sys/module/rcutree/parameters/rcu_normal_wake_from_gp 5636 or pass a boot parameter "rcutree.rcu_normal_wake_from_gp=1" 5637 5638 Default is 1 if num_possible_cpus() <= 16 and it is not explicitly 5639 disabled by the boot parameter passing 0. 5640 5641 rcuscale.gp_async= [KNL] 5642 Measure performance of asynchronous 5643 grace-period primitives such as call_rcu(). 5644 5645 rcuscale.gp_async_max= [KNL] 5646 Specify the maximum number of outstanding 5647 callbacks per writer thread. When a writer 5648 thread exceeds this limit, it invokes the 5649 corresponding flavor of rcu_barrier() to allow 5650 previously posted callbacks to drain. 5651 5652 rcuscale.gp_exp= [KNL] 5653 Measure performance of expedited synchronous 5654 grace-period primitives. 5655 5656 rcuscale.holdoff= [KNL] 5657 Set test-start holdoff period. The purpose of 5658 this parameter is to delay the start of the 5659 test until boot completes in order to avoid 5660 interference. 5661 5662 rcuscale.kfree_by_call_rcu= [KNL] 5663 In kernels built with CONFIG_RCU_LAZY=y, test 5664 call_rcu() instead of kfree_rcu(). 5665 5666 rcuscale.kfree_mult= [KNL] 5667 Instead of allocating an object of size kfree_obj, 5668 allocate one of kfree_mult * sizeof(kfree_obj). 5669 Defaults to 1. 5670 5671 rcuscale.kfree_rcu_test= [KNL] 5672 Set to measure performance of kfree_rcu() flooding. 5673 5674 rcuscale.kfree_rcu_test_double= [KNL] 5675 Test the double-argument variant of kfree_rcu(). 5676 If this parameter has the same value as 5677 rcuscale.kfree_rcu_test_single, both the single- 5678 and double-argument variants are tested. 5679 5680 rcuscale.kfree_rcu_test_single= [KNL] 5681 Test the single-argument variant of kfree_rcu(). 5682 If this parameter has the same value as 5683 rcuscale.kfree_rcu_test_double, both the single- 5684 and double-argument variants are tested. 5685 5686 rcuscale.kfree_nthreads= [KNL] 5687 The number of threads running loops of kfree_rcu(). 5688 5689 rcuscale.kfree_alloc_num= [KNL] 5690 Number of allocations and frees done in an iteration. 5691 5692 rcuscale.kfree_loops= [KNL] 5693 Number of loops doing rcuscale.kfree_alloc_num number 5694 of allocations and frees. 5695 5696 rcuscale.minruntime= [KNL] 5697 Set the minimum test run time in seconds. This 5698 does not affect the data-collection interval, 5699 but instead allows better measurement of things 5700 like CPU consumption. 5701 5702 rcuscale.nreaders= [KNL] 5703 Set number of RCU readers. The value -1 selects 5704 N, where N is the number of CPUs. A value 5705 "n" less than -1 selects N-n+1, where N is again 5706 the number of CPUs. For example, -2 selects N 5707 (the number of CPUs), -3 selects N+1, and so on. 5708 A value of "n" less than or equal to -N selects 5709 a single reader. 5710 5711 rcuscale.nwriters= [KNL] 5712 Set number of RCU writers. The values operate 5713 the same as for rcuscale.nreaders. 5714 N, where N is the number of CPUs 5715 5716 rcuscale.scale_type= [KNL] 5717 Specify the RCU implementation to test. 5718 5719 rcuscale.shutdown= [KNL] 5720 Shut the system down after performance tests 5721 complete. This is useful for hands-off automated 5722 testing. 5723 5724 rcuscale.verbose= [KNL] 5725 Enable additional printk() statements. 5726 5727 rcuscale.writer_holdoff= [KNL] 5728 Write-side holdoff between grace periods, 5729 in microseconds. The default of zero says 5730 no holdoff. 5731 5732 rcuscale.writer_holdoff_jiffies= [KNL] 5733 Additional write-side holdoff between grace 5734 periods, but in jiffies. The default of zero 5735 says no holdoff. 5736 5737 rcutorture.fqs_duration= [KNL] 5738 Set duration of force_quiescent_state bursts 5739 in microseconds. 5740 5741 rcutorture.fqs_holdoff= [KNL] 5742 Set holdoff time within force_quiescent_state bursts 5743 in microseconds. 5744 5745 rcutorture.fqs_stutter= [KNL] 5746 Set wait time between force_quiescent_state bursts 5747 in seconds. 5748 5749 rcutorture.fwd_progress= [KNL] 5750 Specifies the number of kthreads to be used 5751 for RCU grace-period forward-progress testing 5752 for the types of RCU supporting this notion. 5753 Defaults to 1 kthread, values less than zero or 5754 greater than the number of CPUs cause the number 5755 of CPUs to be used. 5756 5757 rcutorture.fwd_progress_div= [KNL] 5758 Specify the fraction of a CPU-stall-warning 5759 period to do tight-loop forward-progress testing. 5760 5761 rcutorture.fwd_progress_holdoff= [KNL] 5762 Number of seconds to wait between successive 5763 forward-progress tests. 5764 5765 rcutorture.fwd_progress_need_resched= [KNL] 5766 Enclose cond_resched() calls within checks for 5767 need_resched() during tight-loop forward-progress 5768 testing. 5769 5770 rcutorture.gp_cond= [KNL] 5771 Use conditional/asynchronous update-side 5772 normal-grace-period primitives, if available. 5773 5774 rcutorture.gp_cond_exp= [KNL] 5775 Use conditional/asynchronous update-side 5776 expedited-grace-period primitives, if available. 5777 5778 rcutorture.gp_cond_full= [KNL] 5779 Use conditional/asynchronous update-side 5780 normal-grace-period primitives that also take 5781 concurrent expedited grace periods into account, 5782 if available. 5783 5784 rcutorture.gp_cond_exp_full= [KNL] 5785 Use conditional/asynchronous update-side 5786 expedited-grace-period primitives that also take 5787 concurrent normal grace periods into account, 5788 if available. 5789 5790 rcutorture.gp_cond_wi= [KNL] 5791 Nominal wait interval for normal conditional 5792 grace periods (specified by rcutorture's 5793 gp_cond and gp_cond_full module parameters), 5794 in microseconds. The actual wait interval will 5795 be randomly selected to nanosecond granularity up 5796 to this wait interval. Defaults to 16 jiffies, 5797 for example, 16,000 microseconds on a system 5798 with HZ=1000. 5799 5800 rcutorture.gp_cond_wi_exp= [KNL] 5801 Nominal wait interval for expedited conditional 5802 grace periods (specified by rcutorture's 5803 gp_cond_exp and gp_cond_exp_full module 5804 parameters), in microseconds. The actual wait 5805 interval will be randomly selected to nanosecond 5806 granularity up to this wait interval. Defaults to 5807 128 microseconds. 5808 5809 rcutorture.gp_exp= [KNL] 5810 Use expedited update-side primitives, if available. 5811 5812 rcutorture.gp_normal= [KNL] 5813 Use normal (non-expedited) asynchronous 5814 update-side primitives, if available. 5815 5816 rcutorture.gp_poll= [KNL] 5817 Use polled update-side normal-grace-period 5818 primitives, if available. 5819 5820 rcutorture.gp_poll_exp= [KNL] 5821 Use polled update-side expedited-grace-period 5822 primitives, if available. 5823 5824 rcutorture.gp_poll_full= [KNL] 5825 Use polled update-side normal-grace-period 5826 primitives that also take concurrent expedited 5827 grace periods into account, if available. 5828 5829 rcutorture.gp_poll_exp_full= [KNL] 5830 Use polled update-side expedited-grace-period 5831 primitives that also take concurrent normal 5832 grace periods into account, if available. 5833 5834 rcutorture.gp_poll_wi= [KNL] 5835 Nominal wait interval for normal conditional 5836 grace periods (specified by rcutorture's 5837 gp_poll and gp_poll_full module parameters), 5838 in microseconds. The actual wait interval will 5839 be randomly selected to nanosecond granularity up 5840 to this wait interval. Defaults to 16 jiffies, 5841 for example, 16,000 microseconds on a system 5842 with HZ=1000. 5843 5844 rcutorture.gp_poll_wi_exp= [KNL] 5845 Nominal wait interval for expedited conditional 5846 grace periods (specified by rcutorture's 5847 gp_poll_exp and gp_poll_exp_full module 5848 parameters), in microseconds. The actual wait 5849 interval will be randomly selected to nanosecond 5850 granularity up to this wait interval. Defaults to 5851 128 microseconds. 5852 5853 rcutorture.gp_sync= [KNL] 5854 Use normal (non-expedited) synchronous 5855 update-side primitives, if available. If all 5856 of rcutorture.gp_cond=, rcutorture.gp_exp=, 5857 rcutorture.gp_normal=, and rcutorture.gp_sync= 5858 are zero, rcutorture acts as if is interpreted 5859 they are all non-zero. 5860 5861 rcutorture.gpwrap_lag= [KNL] 5862 Enable grace-period wrap lag testing. Setting 5863 to false prevents the gpwrap lag test from 5864 running. Default is true. 5865 5866 rcutorture.gpwrap_lag_gps= [KNL] 5867 Set the value for grace-period wrap lag during 5868 active lag testing periods. This controls how many 5869 grace periods differences we tolerate between 5870 rdp and rnp's gp_seq before setting overflow flag. 5871 The default is always set to 8. 5872 5873 rcutorture.gpwrap_lag_cycle_mins= [KNL] 5874 Set the total cycle duration for gpwrap lag 5875 testing in minutes. This is the total time for 5876 one complete cycle of active and inactive 5877 testing periods. Default is 30 minutes. 5878 5879 rcutorture.gpwrap_lag_active_mins= [KNL] 5880 Set the duration for which gpwrap lag is active 5881 within each cycle, in minutes. During this time, 5882 the grace-period wrap lag will be set to the 5883 value specified by gpwrap_lag_gps. Default is 5884 5 minutes. 5885 5886 rcutorture.irqreader= [KNL] 5887 Run RCU readers from irq handlers, or, more 5888 accurately, from a timer handler. Not all RCU 5889 flavors take kindly to this sort of thing. 5890 5891 rcutorture.leakpointer= [KNL] 5892 Leak an RCU-protected pointer out of the reader. 5893 This can of course result in splats, and is 5894 intended to test the ability of things like 5895 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 5896 such leaks. 5897 5898 rcutorture.n_barrier_cbs= [KNL] 5899 Set callbacks/threads for rcu_barrier() testing. 5900 5901 rcutorture.nfakewriters= [KNL] 5902 Set number of concurrent RCU writers. These just 5903 stress RCU, they don't participate in the actual 5904 test, hence the "fake". 5905 5906 rcutorture.nocbs_nthreads= [KNL] 5907 Set number of RCU callback-offload togglers. 5908 Zero (the default) disables toggling. 5909 5910 rcutorture.nocbs_toggle= [KNL] 5911 Set the delay in milliseconds between successive 5912 callback-offload toggling attempts. 5913 5914 rcutorture.nreaders= [KNL] 5915 Set number of RCU readers. The value -1 selects 5916 N-1, where N is the number of CPUs. A value 5917 "n" less than -1 selects N-n-2, where N is again 5918 the number of CPUs. For example, -2 selects N 5919 (the number of CPUs), -3 selects N+1, and so on. 5920 5921 rcutorture.object_debug= [KNL] 5922 Enable debug-object double-call_rcu() testing. 5923 5924 rcutorture.onoff_holdoff= [KNL] 5925 Set time (s) after boot for CPU-hotplug testing. 5926 5927 rcutorture.onoff_interval= [KNL] 5928 Set time (jiffies) between CPU-hotplug operations, 5929 or zero to disable CPU-hotplug testing. 5930 5931 rcutorture.preempt_duration= [KNL] 5932 Set duration (in milliseconds) of preemptions 5933 by a high-priority FIFO real-time task. Set to 5934 zero (the default) to disable. The CPUs to 5935 preempt are selected randomly from the set that 5936 are online at a given point in time. Races with 5937 CPUs going offline are ignored, with that attempt 5938 at preemption skipped. 5939 5940 rcutorture.preempt_interval= [KNL] 5941 Set interval (in milliseconds, defaulting to one 5942 second) between preemptions by a high-priority 5943 FIFO real-time task. This delay is mediated 5944 by an hrtimer and is further fuzzed to avoid 5945 inadvertent synchronizations. 5946 5947 rcutorture.read_exit_burst= [KNL] 5948 The number of times in a given read-then-exit 5949 episode that a set of read-then-exit kthreads 5950 is spawned. 5951 5952 rcutorture.read_exit_delay= [KNL] 5953 The delay, in seconds, between successive 5954 read-then-exit testing episodes. 5955 5956 rcutorture.reader_flavor= [KNL] 5957 A bit mask indicating which readers to use. 5958 If there is more than one bit set, the readers 5959 are entered from low-order bit up, and are 5960 exited in the opposite order. For SRCU, the 5961 0x1 bit is normal readers, 0x2 NMI-safe readers, 5962 and 0x4 light-weight readers. 5963 5964 rcutorture.shuffle_interval= [KNL] 5965 Set task-shuffle interval (s). Shuffling tasks 5966 allows some CPUs to go into dyntick-idle mode 5967 during the rcutorture test. 5968 5969 rcutorture.shutdown_secs= [KNL] 5970 Set time (s) after boot system shutdown. This 5971 is useful for hands-off automated testing. 5972 5973 rcutorture.stall_cpu= [KNL] 5974 Duration of CPU stall (s) to test RCU CPU stall 5975 warnings, zero to disable. 5976 5977 rcutorture.stall_cpu_block= [KNL] 5978 Sleep while stalling if set. This will result 5979 in warnings from preemptible RCU in addition to 5980 any other stall-related activity. Note that 5981 in kernels built with CONFIG_PREEMPTION=n and 5982 CONFIG_PREEMPT_COUNT=y, this parameter will 5983 cause the CPU to pass through a quiescent state. 5984 Given CONFIG_PREEMPTION=n, this will suppress 5985 RCU CPU stall warnings, but will instead result 5986 in scheduling-while-atomic splats. 5987 5988 Use of this module parameter results in splats. 5989 5990 5991 rcutorture.stall_cpu_holdoff= [KNL] 5992 Time to wait (s) after boot before inducing stall. 5993 5994 rcutorture.stall_cpu_irqsoff= [KNL] 5995 Disable interrupts while stalling if set, but only 5996 on the first stall in the set. 5997 5998 rcutorture.stall_cpu_repeat= [KNL] 5999 Number of times to repeat the stall sequence, 6000 so that rcutorture.stall_cpu_repeat=3 will result 6001 in four stall sequences. 6002 6003 rcutorture.stall_gp_kthread= [KNL] 6004 Duration (s) of forced sleep within RCU 6005 grace-period kthread to test RCU CPU stall 6006 warnings, zero to disable. If both stall_cpu 6007 and stall_gp_kthread are specified, the 6008 kthread is starved first, then the CPU. 6009 6010 rcutorture.stat_interval= [KNL] 6011 Time (s) between statistics printk()s. 6012 6013 rcutorture.stutter= [KNL] 6014 Time (s) to stutter testing, for example, specifying 6015 five seconds causes the test to run for five seconds, 6016 wait for five seconds, and so on. This tests RCU's 6017 ability to transition abruptly to and from idle. 6018 6019 rcutorture.test_boost= [KNL] 6020 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 6021 "Maybe" means test if the RCU implementation 6022 under test support RCU priority boosting. 6023 6024 rcutorture.test_boost_duration= [KNL] 6025 Duration (s) of each individual boost test. 6026 6027 rcutorture.test_boost_holdoff= [KNL] 6028 Holdoff time (s) from start of test to the start 6029 of RCU priority-boost testing. Defaults to zero, 6030 that is, no holdoff. 6031 6032 rcutorture.test_boost_interval= [KNL] 6033 Interval (s) between each boost test. 6034 6035 rcutorture.test_no_idle_hz= [KNL] 6036 Test RCU's dyntick-idle handling. See also the 6037 rcutorture.shuffle_interval parameter. 6038 6039 rcutorture.torture_type= [KNL] 6040 Specify the RCU implementation to test. 6041 6042 rcutorture.verbose= [KNL] 6043 Enable additional printk() statements. 6044 6045 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 6046 Dump ftrace buffer after reporting RCU CPU 6047 stall warning. 6048 6049 rcupdate.rcu_cpu_stall_notifiers= [KNL] 6050 Provide RCU CPU stall notifiers, but see the 6051 warnings in the RCU_CPU_STALL_NOTIFIER Kconfig 6052 option's help text. TL;DR: You almost certainly 6053 do not want rcupdate.rcu_cpu_stall_notifiers. 6054 6055 rcupdate.rcu_cpu_stall_suppress= [KNL] 6056 Suppress RCU CPU stall warning messages. 6057 6058 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 6059 Suppress RCU CPU stall warning messages and 6060 rcutorture writer stall warnings that occur 6061 during early boot, that is, during the time 6062 before the init task is spawned. 6063 6064 rcupdate.rcu_cpu_stall_timeout= [KNL] 6065 Set timeout for RCU CPU stall warning messages. 6066 The value is in seconds and the maximum allowed 6067 value is 300 seconds. 6068 6069 rcupdate.rcu_exp_cpu_stall_timeout= [KNL] 6070 Set timeout for expedited RCU CPU stall warning 6071 messages. The value is in milliseconds 6072 and the maximum allowed value is 21000 6073 milliseconds. Please note that this value is 6074 adjusted to an arch timer tick resolution. 6075 Setting this to zero causes the value from 6076 rcupdate.rcu_cpu_stall_timeout to be used (after 6077 conversion from seconds to milliseconds). 6078 6079 rcupdate.rcu_cpu_stall_cputime= [KNL] 6080 Provide statistics on the cputime and count of 6081 interrupts and tasks during the sampling period. For 6082 multiple continuous RCU stalls, all sampling periods 6083 begin at half of the first RCU stall timeout. 6084 6085 rcupdate.rcu_exp_stall_task_details= [KNL] 6086 Print stack dumps of any tasks blocking the 6087 current expedited RCU grace period during an 6088 expedited RCU CPU stall warning. 6089 6090 rcupdate.rcu_expedited= [KNL] 6091 Use expedited grace-period primitives, for 6092 example, synchronize_rcu_expedited() instead 6093 of synchronize_rcu(). This reduces latency, 6094 but can increase CPU utilization, degrade 6095 real-time latency, and degrade energy efficiency. 6096 No effect on CONFIG_TINY_RCU kernels. 6097 6098 rcupdate.rcu_normal= [KNL] 6099 Use only normal grace-period primitives, 6100 for example, synchronize_rcu() instead of 6101 synchronize_rcu_expedited(). This improves 6102 real-time latency, CPU utilization, and 6103 energy efficiency, but can expose users to 6104 increased grace-period latency. This parameter 6105 overrides rcupdate.rcu_expedited. No effect on 6106 CONFIG_TINY_RCU kernels. 6107 6108 rcupdate.rcu_normal_after_boot= [KNL] 6109 Once boot has completed (that is, after 6110 rcu_end_inkernel_boot() has been invoked), use 6111 only normal grace-period primitives. No effect 6112 on CONFIG_TINY_RCU kernels. 6113 6114 But note that CONFIG_PREEMPT_RT=y kernels enables 6115 this kernel boot parameter, forcibly setting 6116 it to the value one, that is, converting any 6117 post-boot attempt at an expedited RCU grace 6118 period to instead use normal non-expedited 6119 grace-period processing. 6120 6121 rcupdate.rcu_task_collapse_lim= [KNL] 6122 Set the maximum number of callbacks present 6123 at the beginning of a grace period that allows 6124 the RCU Tasks flavors to collapse back to using 6125 a single callback queue. This switching only 6126 occurs when rcupdate.rcu_task_enqueue_lim is 6127 set to the default value of -1. 6128 6129 rcupdate.rcu_task_contend_lim= [KNL] 6130 Set the minimum number of callback-queuing-time 6131 lock-contention events per jiffy required to 6132 cause the RCU Tasks flavors to switch to per-CPU 6133 callback queuing. This switching only occurs 6134 when rcupdate.rcu_task_enqueue_lim is set to 6135 the default value of -1. 6136 6137 rcupdate.rcu_task_enqueue_lim= [KNL] 6138 Set the number of callback queues to use for the 6139 RCU Tasks family of RCU flavors. The default 6140 of -1 allows this to be automatically (and 6141 dynamically) adjusted. This parameter is intended 6142 for use in testing. 6143 6144 rcupdate.rcu_task_ipi_delay= [KNL] 6145 Set time in jiffies during which RCU tasks will 6146 avoid sending IPIs, starting with the beginning 6147 of a given grace period. Setting a large 6148 number avoids disturbing real-time workloads, 6149 but lengthens grace periods. 6150 6151 rcupdate.rcu_task_lazy_lim= [KNL] 6152 Number of callbacks on a given CPU that will 6153 cancel laziness on that CPU. Use -1 to disable 6154 cancellation of laziness, but be advised that 6155 doing so increases the danger of OOM due to 6156 callback flooding. 6157 6158 rcupdate.rcu_task_stall_info= [KNL] 6159 Set initial timeout in jiffies for RCU task stall 6160 informational messages, which give some indication 6161 of the problem for those not patient enough to 6162 wait for ten minutes. Informational messages are 6163 only printed prior to the stall-warning message 6164 for a given grace period. Disable with a value 6165 less than or equal to zero. Defaults to ten 6166 seconds. A change in value does not take effect 6167 until the beginning of the next grace period. 6168 6169 rcupdate.rcu_task_stall_info_mult= [KNL] 6170 Multiplier for time interval between successive 6171 RCU task stall informational messages for a given 6172 RCU tasks grace period. This value is clamped 6173 to one through ten, inclusive. It defaults to 6174 the value three, so that the first informational 6175 message is printed 10 seconds into the grace 6176 period, the second at 40 seconds, the third at 6177 160 seconds, and then the stall warning at 600 6178 seconds would prevent a fourth at 640 seconds. 6179 6180 rcupdate.rcu_task_stall_timeout= [KNL] 6181 Set timeout in jiffies for RCU task stall 6182 warning messages. Disable with a value less 6183 than or equal to zero. Defaults to ten minutes. 6184 A change in value does not take effect until 6185 the beginning of the next grace period. 6186 6187 rcupdate.rcu_tasks_lazy_ms= [KNL] 6188 Set timeout in milliseconds RCU Tasks asynchronous 6189 callback batching for call_rcu_tasks(). 6190 A negative value will take the default. A value 6191 of zero will disable batching. Batching is 6192 always disabled for synchronize_rcu_tasks(). 6193 6194 rcupdate.rcu_tasks_trace_lazy_ms= [KNL] 6195 Set timeout in milliseconds RCU Tasks 6196 Trace asynchronous callback batching for 6197 call_rcu_tasks_trace(). A negative value 6198 will take the default. A value of zero will 6199 disable batching. Batching is always disabled 6200 for synchronize_rcu_tasks_trace(). 6201 6202 rcupdate.rcu_self_test= [KNL] 6203 Run the RCU early boot self tests 6204 6205 rdinit= [KNL] 6206 Format: <full_path> 6207 Run specified binary instead of /init from the ramdisk, 6208 used for early userspace startup. See initrd. 6209 6210 rdrand= [X86,EARLY] 6211 force - Override the decision by the kernel to hide the 6212 advertisement of RDRAND support (this affects 6213 certain AMD processors because of buggy BIOS 6214 support, specifically around the suspend/resume 6215 path). 6216 6217 rdt= [HW,X86,RDT] 6218 Turn on/off individual RDT features. List is: 6219 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 6220 mba, smba, bmec, abmc, sdciae. 6221 E.g. to turn on cmt and turn off mba use: 6222 rdt=cmt,!mba 6223 6224 reboot= [KNL] 6225 Format (x86 or x86_64): 6226 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \ 6227 [[,]s[mp]#### \ 6228 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 6229 [[,]f[orce] 6230 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 6231 (prefix with 'panic_' to set mode for panic 6232 reboot only), 6233 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 6234 reboot_force is either force or not specified, 6235 reboot_cpu is s[mp]#### with #### being the processor 6236 to be used for rebooting. 6237 6238 acpi 6239 Use the ACPI RESET_REG in the FADT. If ACPI is not 6240 configured or the ACPI reset does not work, the reboot 6241 path attempts the reset using the keyboard controller. 6242 6243 bios 6244 Use the CPU reboot vector for warm reset 6245 6246 cold 6247 Set the cold reboot flag 6248 6249 default 6250 There are some built-in platform specific "quirks" 6251 - you may see: "reboot: <name> series board detected. 6252 Selecting <type> for reboots." In the case where you 6253 think the quirk is in error (e.g. you have newer BIOS, 6254 or newer board) using this option will ignore the 6255 built-in quirk table, and use the generic default 6256 reboot actions. 6257 6258 efi 6259 Use efi reset_system runtime service. If EFI is not 6260 configured or the EFI reset does not work, the reboot 6261 path attempts the reset using the keyboard controller. 6262 6263 force 6264 Don't stop other CPUs on reboot. This can make reboot 6265 more reliable in some cases. 6266 6267 kbd 6268 Use the keyboard controller. cold reset (default) 6269 6270 pci 6271 Use a write to the PCI config space register 0xcf9 to 6272 trigger reboot. 6273 6274 triple 6275 Force a triple fault (init) 6276 6277 warm 6278 Don't set the cold reboot flag 6279 6280 Using warm reset will be much faster especially on big 6281 memory systems because the BIOS will not go through 6282 the memory check. Disadvantage is that not all 6283 hardware will be completely reinitialized on reboot so 6284 there may be boot problems on some systems. 6285 6286 6287 refscale.holdoff= [KNL] 6288 Set test-start holdoff period. The purpose of 6289 this parameter is to delay the start of the 6290 test until boot completes in order to avoid 6291 interference. 6292 6293 refscale.lookup_instances= [KNL] 6294 Number of data elements to use for the forms of 6295 SLAB_TYPESAFE_BY_RCU testing. A negative number 6296 is negated and multiplied by nr_cpu_ids, while 6297 zero specifies nr_cpu_ids. 6298 6299 refscale.loops= [KNL] 6300 Set the number of loops over the synchronization 6301 primitive under test. Increasing this number 6302 reduces noise due to loop start/end overhead, 6303 but the default has already reduced the per-pass 6304 noise to a handful of picoseconds on ca. 2020 6305 x86 laptops. 6306 6307 refscale.nreaders= [KNL] 6308 Set number of readers. The default value of -1 6309 selects N, where N is roughly 75% of the number 6310 of CPUs. A value of zero is an interesting choice. 6311 6312 refscale.nruns= [KNL] 6313 Set number of runs, each of which is dumped onto 6314 the console log. 6315 6316 refscale.readdelay= [KNL] 6317 Set the read-side critical-section duration, 6318 measured in microseconds. 6319 6320 refscale.scale_type= [KNL] 6321 Specify the read-protection implementation to test. 6322 6323 refscale.shutdown= [KNL] 6324 Shut down the system at the end of the performance 6325 test. This defaults to 1 (shut it down) when 6326 refscale is built into the kernel and to 0 (leave 6327 it running) when refscale is built as a module. 6328 6329 refscale.verbose= [KNL] 6330 Enable additional printk() statements. 6331 6332 refscale.verbose_batched= [KNL] 6333 Batch the additional printk() statements. If zero 6334 (the default) or negative, print everything. Otherwise, 6335 print every Nth verbose statement, where N is the value 6336 specified. 6337 6338 regulator_ignore_unused 6339 [REGULATOR] 6340 Prevents regulator framework from disabling regulators 6341 that are unused, due no driver claiming them. This may 6342 be useful for debug and development, but should not be 6343 needed on a platform with proper driver support. 6344 6345 relax_domain_level= 6346 [KNL, SMP] Set scheduler's default relax_domain_level. 6347 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 6348 6349 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 6350 Format: <base1>,<size1>[,<base2>,<size2>,...] 6351 Reserve I/O ports or memory so the kernel won't use 6352 them. If <base> is less than 0x10000, the region 6353 is assumed to be I/O ports; otherwise it is memory. 6354 6355 reserve_mem= [RAM] 6356 Format: nn[KMG]:<align>:<label> 6357 Reserve physical memory and label it with a name that 6358 other subsystems can use to access it. This is typically 6359 used for systems that do not wipe the RAM, and this command 6360 line will try to reserve the same physical memory on 6361 soft reboots. Note, it is not guaranteed to be the same 6362 location. For example, if anything about the system changes 6363 or if booting a different kernel. It can also fail if KASLR 6364 places the kernel at the location of where the RAM reservation 6365 was from a previous boot, the new reservation will be at a 6366 different location. 6367 Any subsystem using this feature must add a way to verify 6368 that the contents of the physical memory is from a previous 6369 boot, as there may be cases where the memory will not be 6370 located at the same location. 6371 6372 The format is size:align:label for example, to request 6373 12 megabytes of 4096 alignment for ramoops: 6374 6375 reserve_mem=12M:4096:oops ramoops.mem_name=oops 6376 6377 reservetop= [X86-32,EARLY] 6378 Format: nn[KMG] 6379 Reserves a hole at the top of the kernel virtual 6380 address space. 6381 6382 reset_devices [KNL] Force drivers to reset the underlying device 6383 during initialization. 6384 6385 resume= [SWSUSP] 6386 Specify the partition device for software suspend 6387 Format: 6388 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 6389 6390 resume_offset= [SWSUSP] 6391 Specify the offset from the beginning of the partition 6392 given by "resume=" at which the swap header is located, 6393 in <PAGE_SIZE> units (needed only for swap files). 6394 See Documentation/power/swsusp-and-swap-files.rst 6395 6396 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 6397 read the resume files 6398 6399 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 6400 Useful for devices that are detected asynchronously 6401 (e.g. USB and MMC devices). 6402 6403 retain_initrd [RAM] Keep initrd memory after extraction. After boot, it will 6404 be accessible via /sys/firmware/initrd. 6405 6406 retbleed= [X86] Control mitigation of RETBleed (Arbitrary 6407 Speculative Code Execution with Return Instructions) 6408 vulnerability. 6409 6410 AMD-based UNRET and IBPB mitigations alone do not stop 6411 sibling threads from influencing the predictions of other 6412 sibling threads. For that reason, STIBP is used on pro- 6413 cessors that support it, and mitigate SMT on processors 6414 that don't. 6415 6416 off - no mitigation 6417 auto - automatically select a migitation 6418 auto,nosmt - automatically select a mitigation, 6419 disabling SMT if necessary for 6420 the full mitigation (only on Zen1 6421 and older without STIBP). 6422 ibpb - On AMD, mitigate short speculation 6423 windows on basic block boundaries too. 6424 Safe, highest perf impact. It also 6425 enables STIBP if present. Not suitable 6426 on Intel. 6427 ibpb,nosmt - Like "ibpb" above but will disable SMT 6428 when STIBP is not available. This is 6429 the alternative for systems which do not 6430 have STIBP. 6431 unret - Force enable untrained return thunks, 6432 only effective on AMD f15h-f17h based 6433 systems. 6434 unret,nosmt - Like unret, but will disable SMT when STIBP 6435 is not available. This is the alternative for 6436 systems which do not have STIBP. 6437 6438 Selecting 'auto' will choose a mitigation method at run 6439 time according to the CPU. 6440 6441 Not specifying this option is equivalent to retbleed=auto. 6442 6443 rfkill.default_state= 6444 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 6445 etc. communication is blocked by default. 6446 1 Unblocked. 6447 6448 rfkill.master_switch_mode= 6449 0 The "airplane mode" button does nothing. 6450 1 The "airplane mode" button toggles between everything 6451 blocked and the previous configuration. 6452 2 The "airplane mode" button toggles between everything 6453 blocked and everything unblocked. 6454 6455 ring3mwait=disable 6456 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 6457 CPUs. 6458 6459 riscv_isa_fallback [RISCV,EARLY] 6460 When CONFIG_RISCV_ISA_FALLBACK is not enabled, permit 6461 falling back to detecting extension support by parsing 6462 "riscv,isa" property on devicetree systems when the 6463 replacement properties are not found. See the Kconfig 6464 entry for RISCV_ISA_FALLBACK. 6465 6466 ro [KNL] Mount root device read-only on boot 6467 6468 rodata= [KNL,EARLY] 6469 on Mark read-only kernel memory as read-only (default). 6470 off Leave read-only kernel memory writable for debugging. 6471 noalias Mark read-only kernel memory as read-only but retain 6472 writable aliases in the direct map for regions outside 6473 of the kernel image. [arm64] 6474 6475 rockchip.usb_uart 6476 [EARLY] 6477 Enable the uart passthrough on the designated usb port 6478 on Rockchip SoCs. When active, the signals of the 6479 debug-uart get routed to the D+ and D- pins of the usb 6480 port and the regular usb controller gets disabled. 6481 6482 root= [KNL] Root filesystem 6483 Usually this is a block device specifier of some kind, 6484 see the early_lookup_bdev comment in 6485 block/early-lookup.c for details. 6486 Alternatively this can be "ram" for the legacy initial 6487 ramdisk, "nfs" and "cifs" for root on a network file 6488 system, or "mtd" and "ubi" for mounting from raw flash. 6489 6490 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 6491 mount the root filesystem 6492 6493 rootflags= [KNL] Set root filesystem mount option string 6494 6495 initramfs_options= [KNL] 6496 Specify mount options for for the initramfs mount. 6497 6498 rootfstype= [KNL] Set root filesystem type 6499 6500 rootwait [KNL] Wait (indefinitely) for root device to show up. 6501 Useful for devices that are detected asynchronously 6502 (e.g. USB and MMC devices). 6503 6504 rootwait= [KNL] Maximum time (in seconds) to wait for root device 6505 to show up before attempting to mount the root 6506 filesystem. 6507 6508 rproc_mem=nn[KMG][@address] 6509 [KNL,ARM,CMA] Remoteproc physical memory block. 6510 Memory area to be used by remote processor image, 6511 managed by CMA. 6512 6513 rseq_debug= [KNL] Enable or disable restartable sequence 6514 debug mode. Defaults to CONFIG_RSEQ_DEBUG_DEFAULT_ENABLE. 6515 Format: <bool> 6516 6517 rt_group_sched= [KNL] Enable or disable SCHED_RR/FIFO group scheduling 6518 when CONFIG_RT_GROUP_SCHED=y. Defaults to 6519 !CONFIG_RT_GROUP_SCHED_DEFAULT_DISABLED. 6520 Format: <bool> 6521 6522 rw [KNL] Mount root device read-write on boot 6523 6524 S [KNL] Run init in single mode 6525 6526 s390_iommu= [HW,S390] 6527 Set s390 IOTLB flushing mode 6528 strict 6529 With strict flushing every unmap operation will result 6530 in an IOTLB flush. Default is lazy flushing before 6531 reuse, which is faster. Deprecated, equivalent to 6532 iommu.strict=1. 6533 6534 s390_iommu_aperture= [KNL,S390] 6535 Specifies the size of the per device DMA address space 6536 accessible through the DMA and IOMMU APIs as a decimal 6537 factor of the size of main memory. 6538 The default is 1 meaning that one can concurrently use 6539 as many DMA addresses as physical memory is installed, 6540 if supported by hardware, and thus map all of memory 6541 once. With a value of 2 one can map all of memory twice 6542 and so on. As a special case a factor of 0 imposes no 6543 restrictions other than those given by hardware at the 6544 cost of significant additional memory use for tables. 6545 6546 sa1100ir [NET] 6547 See drivers/net/irda/sa1100_ir.c. 6548 6549 sched_proxy_exec= [KNL] 6550 Enables or disables "proxy execution" style 6551 solution to mutex-based priority inversion. 6552 Format: <bool> 6553 6554 sched_verbose [KNL,EARLY] Enables verbose scheduler debug messages. 6555 6556 schedstats= [KNL,X86] Enable or disable scheduled statistics. 6557 Allowed values are enable and disable. This feature 6558 incurs a small amount of overhead in the scheduler 6559 but is useful for debugging and performance tuning. 6560 6561 sched_thermal_decay_shift= 6562 [Deprecated] 6563 [KNL, SMP] Set a decay shift for scheduler thermal 6564 pressure signal. Thermal pressure signal follows the 6565 default decay period of other scheduler pelt 6566 signals(usually 32 ms but configurable). Setting 6567 sched_thermal_decay_shift will left shift the decay 6568 period for the thermal pressure signal by the shift 6569 value. 6570 i.e. with the default pelt decay period of 32 ms 6571 sched_thermal_decay_shift thermal pressure decay pr 6572 1 64 ms 6573 2 128 ms 6574 and so on. 6575 Format: integer between 0 and 10 6576 Default is 0. 6577 6578 scftorture.holdoff= [KNL] 6579 Number of seconds to hold off before starting 6580 test. Defaults to zero for module insertion and 6581 to 10 seconds for built-in smp_call_function() 6582 tests. 6583 6584 scftorture.longwait= [KNL] 6585 Request ridiculously long waits randomly selected 6586 up to the chosen limit in seconds. Zero (the 6587 default) disables this feature. Please note 6588 that requesting even small non-zero numbers of 6589 seconds can result in RCU CPU stall warnings, 6590 softlockup complaints, and so on. 6591 6592 scftorture.nthreads= [KNL] 6593 Number of kthreads to spawn to invoke the 6594 smp_call_function() family of functions. 6595 The default of -1 specifies a number of kthreads 6596 equal to the number of CPUs. 6597 6598 scftorture.onoff_holdoff= [KNL] 6599 Number seconds to wait after the start of the 6600 test before initiating CPU-hotplug operations. 6601 6602 scftorture.onoff_interval= [KNL] 6603 Number seconds to wait between successive 6604 CPU-hotplug operations. Specifying zero (which 6605 is the default) disables CPU-hotplug operations. 6606 6607 scftorture.shutdown_secs= [KNL] 6608 The number of seconds following the start of the 6609 test after which to shut down the system. The 6610 default of zero avoids shutting down the system. 6611 Non-zero values are useful for automated tests. 6612 6613 scftorture.stat_interval= [KNL] 6614 The number of seconds between outputting the 6615 current test statistics to the console. A value 6616 of zero disables statistics output. 6617 6618 scftorture.stutter_cpus= [KNL] 6619 The number of jiffies to wait between each change 6620 to the set of CPUs under test. 6621 6622 scftorture.use_cpus_read_lock= [KNL] 6623 Use use_cpus_read_lock() instead of the default 6624 preempt_disable() to disable CPU hotplug 6625 while invoking one of the smp_call_function*() 6626 functions. 6627 6628 scftorture.verbose= [KNL] 6629 Enable additional printk() statements. 6630 6631 scftorture.weight_single= [KNL] 6632 The probability weighting to use for the 6633 smp_call_function_single() function with a zero 6634 "wait" parameter. A value of -1 selects the 6635 default if all other weights are -1. However, 6636 if at least one weight has some other value, a 6637 value of -1 will instead select a weight of zero. 6638 6639 scftorture.weight_single_wait= [KNL] 6640 The probability weighting to use for the 6641 smp_call_function_single() function with a 6642 non-zero "wait" parameter. See weight_single. 6643 6644 scftorture.weight_many= [KNL] 6645 The probability weighting to use for the 6646 smp_call_function_many() function with a zero 6647 "wait" parameter. See weight_single. 6648 Note well that setting a high probability for 6649 this weighting can place serious IPI load 6650 on the system. 6651 6652 scftorture.weight_many_wait= [KNL] 6653 The probability weighting to use for the 6654 smp_call_function_many() function with a 6655 non-zero "wait" parameter. See weight_single 6656 and weight_many. 6657 6658 scftorture.weight_all= [KNL] 6659 The probability weighting to use for the 6660 smp_call_function_all() function with a zero 6661 "wait" parameter. See weight_single and 6662 weight_many. 6663 6664 scftorture.weight_all_wait= [KNL] 6665 The probability weighting to use for the 6666 smp_call_function_all() function with a 6667 non-zero "wait" parameter. See weight_single 6668 and weight_many. 6669 6670 sdw_mclk_divider=[SDW] 6671 Specify the MCLK divider for Intel SoundWire buses in 6672 case the BIOS does not provide the clock rate properly. 6673 6674 skew_tick= [KNL,EARLY] Offset the periodic timer tick per cpu to mitigate 6675 xtime_lock contention on larger systems, and/or RCU lock 6676 contention on all systems with CONFIG_MAXSMP set. 6677 Format: { "0" | "1" } 6678 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 6679 1 -- enable. 6680 Note: increases power consumption, thus should only be 6681 enabled if running jitter sensitive (HPC/RT) workloads. 6682 6683 security= [SECURITY] Choose a legacy "major" security module to 6684 enable at boot. This has been deprecated by the 6685 "lsm=" parameter. 6686 6687 selinux= [SELINUX] Disable or enable SELinux at boot time. 6688 Format: { "0" | "1" } 6689 See security/selinux/Kconfig help text. 6690 0 -- disable. 6691 1 -- enable. 6692 Default value is 1. 6693 6694 serialnumber [BUGS=X86-32] 6695 6696 sev=option[,option...] [X86-64] 6697 6698 debug 6699 Enable debug messages. 6700 6701 nosnp 6702 Do not enable SEV-SNP (applies to host/hypervisor 6703 only). Setting 'nosnp' avoids the RMP check overhead 6704 in memory accesses when users do not want to run 6705 SEV-SNP guests. 6706 6707 shapers= [NET] 6708 Maximal number of shapers. 6709 6710 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 6711 Limit apic dumping. The parameter defines the maximal 6712 number of local apics being dumped. Also it is possible 6713 to set it to "all" by meaning -- no limit here. 6714 Format: { 1 (default) | 2 | ... | all }. 6715 The parameter valid if only apic=debug or 6716 apic=verbose is specified. 6717 Example: apic=debug show_lapic=all 6718 6719 slab_debug[=options[,slabs][;[options[,slabs]]...] [MM] 6720 Enabling slab_debug allows one to determine the 6721 culprit if slab objects become corrupted. Enabling 6722 slab_debug can create guard zones around objects and 6723 may poison objects when not in use. Also tracks the 6724 last alloc / free. For more information see 6725 Documentation/admin-guide/mm/slab.rst. 6726 (slub_debug legacy name also accepted for now) 6727 6728 Using this option implies the "no_hash_pointers" 6729 option which can be undone by adding the 6730 "hash_pointers=always" option. 6731 6732 slab_max_order= [MM] 6733 Determines the maximum allowed order for slabs. 6734 A high setting may cause OOMs due to memory 6735 fragmentation. For more information see 6736 Documentation/admin-guide/mm/slab.rst. 6737 (slub_max_order legacy name also accepted for now) 6738 6739 slab_merge [MM] 6740 Enable merging of slabs with similar size when the 6741 kernel is built without CONFIG_SLAB_MERGE_DEFAULT. 6742 (slub_merge legacy name also accepted for now) 6743 6744 slab_min_objects= [MM] 6745 The minimum number of objects per slab. SLUB will 6746 increase the slab order up to slab_max_order to 6747 generate a sufficiently large slab able to contain 6748 the number of objects indicated. The higher the number 6749 of objects the smaller the overhead of tracking slabs 6750 and the less frequently locks need to be acquired. 6751 For more information see 6752 Documentation/admin-guide/mm/slab.rst. 6753 (slub_min_objects legacy name also accepted for now) 6754 6755 slab_min_order= [MM] 6756 Determines the minimum page order for slabs. Must be 6757 lower or equal to slab_max_order. For more information see 6758 Documentation/admin-guide/mm/slab.rst. 6759 (slub_min_order legacy name also accepted for now) 6760 6761 slab_nomerge [MM] 6762 Disable merging of slabs with similar size. May be 6763 necessary if there is some reason to distinguish 6764 allocs to different slabs, especially in hardened 6765 environments where the risk of heap overflows and 6766 layout control by attackers can usually be 6767 frustrated by disabling merging. This will reduce 6768 most of the exposure of a heap attack to a single 6769 cache (risks via metadata attacks are mostly 6770 unchanged). Debug options disable merging on their 6771 own. 6772 For more information see 6773 Documentation/admin-guide/mm/slab.rst. 6774 (slub_nomerge legacy name also accepted for now) 6775 6776 slab_strict_numa [MM] 6777 Support memory policies on a per object level 6778 in the slab allocator. The default is for memory 6779 policies to be applied at the folio level when 6780 a new folio is needed or a partial folio is 6781 retrieved from the lists. Increases overhead 6782 in the slab fastpaths but gains more accurate 6783 NUMA kernel object placement which helps with slow 6784 interconnects in NUMA systems. 6785 6786 slram= [HW,MTD] 6787 6788 smart2= [HW] 6789 Format: <io1>[,<io2>[,...,<io8>]] 6790 6791 smp.csd_lock_timeout= [KNL] 6792 Specify the period of time in milliseconds 6793 that smp_call_function() and friends will wait 6794 for a CPU to release the CSD lock. This is 6795 useful when diagnosing bugs involving CPUs 6796 disabling interrupts for extended periods 6797 of time. Defaults to 5,000 milliseconds, and 6798 setting a value of zero disables this feature. 6799 This feature may be more efficiently disabled 6800 using the csdlock_debug- kernel parameter. 6801 6802 smp.panic_on_ipistall= [KNL] 6803 If a csd_lock_timeout extends for more than 6804 the specified number of milliseconds, panic the 6805 system. By default, let CSD-lock acquisition 6806 take as long as they take. Specifying 300,000 6807 for this value provides a 5-minute timeout. 6808 6809 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 6810 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 6811 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 6812 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 6813 smsc-ircc2.ircc_irq= [HW] IRQ line 6814 smsc-ircc2.ircc_dma= [HW] DMA channel 6815 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 6816 0: Toshiba Satellite 1800 (GP data pin select) 6817 1: Fast pin select (default) 6818 2: ATC IRMode 6819 6820 smt= [KNL,MIPS,S390,EARLY] Set the maximum number of threads 6821 (logical CPUs) to use per physical CPU on systems 6822 capable of symmetric multithreading (SMT). Will 6823 be capped to the actual hardware limit. 6824 Format: <integer> 6825 Default: -1 (no limit) 6826 6827 softlockup_panic= 6828 [KNL] Should the soft-lockup detector generate panics. 6829 Format: 0 | 1 6830 6831 A value of 1 instructs the soft-lockup detector 6832 to panic the machine when a soft-lockup occurs. It is 6833 also controlled by the kernel.softlockup_panic sysctl 6834 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 6835 respective build-time switch to that functionality. 6836 6837 softlockup_all_cpu_backtrace= 6838 [KNL] Should the soft-lockup detector generate 6839 backtraces on all cpus. 6840 Format: 0 | 1 6841 6842 sonypi.*= [HW] Sony Programmable I/O Control Device driver 6843 See Documentation/admin-guide/laptops/sonypi.rst 6844 6845 spectre_bhi= [X86] Control mitigation of Branch History Injection 6846 (BHI) vulnerability. This setting affects the 6847 deployment of the HW BHI control and the SW BHB 6848 clearing sequence. 6849 6850 on - (default) Enable the HW or SW mitigation as 6851 needed. This protects the kernel from 6852 both syscalls and VMs. 6853 vmexit - On systems which don't have the HW mitigation 6854 available, enable the SW mitigation on vmexit 6855 ONLY. On such systems, the host kernel is 6856 protected from VM-originated BHI attacks, but 6857 may still be vulnerable to syscall attacks. 6858 off - Disable the mitigation. 6859 6860 spectre_v2= [X86,EARLY] Control mitigation of Spectre variant 2 6861 (indirect branch speculation) vulnerability. 6862 The default operation protects the kernel from 6863 user space attacks. 6864 6865 on - unconditionally enable, implies 6866 spectre_v2_user=on 6867 off - unconditionally disable, implies 6868 spectre_v2_user=off 6869 auto - kernel detects whether your CPU model is 6870 vulnerable 6871 6872 Selecting 'on' will, and 'auto' may, choose a 6873 mitigation method at run time according to the 6874 CPU, the available microcode, the setting of the 6875 CONFIG_MITIGATION_RETPOLINE configuration option, 6876 and the compiler with which the kernel was built. 6877 6878 Selecting 'on' will also enable the mitigation 6879 against user space to user space task attacks. 6880 Selecting specific mitigation does not force enable 6881 user mitigations. 6882 6883 Selecting 'off' will disable both the kernel and 6884 the user space protections. 6885 6886 Specific mitigations can also be selected manually: 6887 6888 retpoline - replace indirect branches 6889 retpoline,generic - Retpolines 6890 retpoline,lfence - LFENCE; indirect branch 6891 retpoline,amd - alias for retpoline,lfence 6892 eibrs - Enhanced/Auto IBRS 6893 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines 6894 eibrs,lfence - Enhanced/Auto IBRS + LFENCE 6895 ibrs - use IBRS to protect kernel 6896 6897 Not specifying this option is equivalent to 6898 spectre_v2=auto. 6899 6900 spectre_v2_user= 6901 [X86] Control mitigation of Spectre variant 2 6902 (indirect branch speculation) vulnerability between 6903 user space tasks 6904 6905 on - Unconditionally enable mitigations. Is 6906 enforced by spectre_v2=on 6907 6908 off - Unconditionally disable mitigations. Is 6909 enforced by spectre_v2=off 6910 6911 prctl - Indirect branch speculation is enabled, 6912 but mitigation can be enabled via prctl 6913 per thread. The mitigation control state 6914 is inherited on fork. 6915 6916 prctl,ibpb 6917 - Like "prctl" above, but only STIBP is 6918 controlled per thread. IBPB is issued 6919 always when switching between different user 6920 space processes. 6921 6922 seccomp 6923 - Same as "prctl" above, but all seccomp 6924 threads will enable the mitigation unless 6925 they explicitly opt out. 6926 6927 seccomp,ibpb 6928 - Like "seccomp" above, but only STIBP is 6929 controlled per thread. IBPB is issued 6930 always when switching between different 6931 user space processes. 6932 6933 auto - Kernel selects the mitigation depending on 6934 the available CPU features and vulnerability. 6935 6936 Default mitigation: "prctl" 6937 6938 Not specifying this option is equivalent to 6939 spectre_v2_user=auto. 6940 6941 spec_rstack_overflow= 6942 [X86,EARLY] Control RAS overflow mitigation on AMD Zen CPUs 6943 6944 off - Disable mitigation 6945 microcode - Enable microcode mitigation only 6946 safe-ret - Enable sw-only safe RET mitigation (default) 6947 ibpb - Enable mitigation by issuing IBPB on 6948 kernel entry 6949 ibpb-vmexit - Issue IBPB only on VMEXIT 6950 (cloud-specific mitigation) 6951 6952 spec_store_bypass_disable= 6953 [HW,EARLY] Control Speculative Store Bypass (SSB) Disable mitigation 6954 (Speculative Store Bypass vulnerability) 6955 6956 Certain CPUs are vulnerable to an exploit against a 6957 a common industry wide performance optimization known 6958 as "Speculative Store Bypass" in which recent stores 6959 to the same memory location may not be observed by 6960 later loads during speculative execution. The idea 6961 is that such stores are unlikely and that they can 6962 be detected prior to instruction retirement at the 6963 end of a particular speculation execution window. 6964 6965 In vulnerable processors, the speculatively forwarded 6966 store can be used in a cache side channel attack, for 6967 example to read memory to which the attacker does not 6968 directly have access (e.g. inside sandboxed code). 6969 6970 This parameter controls whether the Speculative Store 6971 Bypass optimization is used. 6972 6973 On x86 the options are: 6974 6975 on - Unconditionally disable Speculative Store Bypass 6976 off - Unconditionally enable Speculative Store Bypass 6977 auto - Kernel detects whether the CPU model contains an 6978 implementation of Speculative Store Bypass and 6979 picks the most appropriate mitigation. If the 6980 CPU is not vulnerable, "off" is selected. If the 6981 CPU is vulnerable the default mitigation is 6982 architecture and Kconfig dependent. See below. 6983 prctl - Control Speculative Store Bypass per thread 6984 via prctl. Speculative Store Bypass is enabled 6985 for a process by default. The state of the control 6986 is inherited on fork. 6987 seccomp - Same as "prctl" above, but all seccomp threads 6988 will disable SSB unless they explicitly opt out. 6989 6990 Default mitigations: 6991 X86: "prctl" 6992 6993 On powerpc the options are: 6994 6995 on,auto - On Power8 and Power9 insert a store-forwarding 6996 barrier on kernel entry and exit. On Power7 6997 perform a software flush on kernel entry and 6998 exit. 6999 off - No action. 7000 7001 Not specifying this option is equivalent to 7002 spec_store_bypass_disable=auto. 7003 7004 split_lock_detect= 7005 [X86] Enable split lock detection or bus lock detection 7006 7007 When enabled (and if hardware support is present), atomic 7008 instructions that access data across cache line 7009 boundaries will result in an alignment check exception 7010 for split lock detection or a debug exception for 7011 bus lock detection. 7012 7013 off - not enabled 7014 7015 warn - the kernel will emit rate-limited warnings 7016 about applications triggering the #AC 7017 exception or the #DB exception. This mode is 7018 the default on CPUs that support split lock 7019 detection or bus lock detection. Default 7020 behavior is by #AC if both features are 7021 enabled in hardware. 7022 7023 fatal - the kernel will send SIGBUS to applications 7024 that trigger the #AC exception or the #DB 7025 exception. Default behavior is by #AC if 7026 both features are enabled in hardware. 7027 7028 ratelimit:N - 7029 Set system wide rate limit to N bus locks 7030 per second for bus lock detection. 7031 0 < N <= 1000. 7032 7033 N/A for split lock detection. 7034 7035 7036 If an #AC exception is hit in the kernel or in 7037 firmware (i.e. not while executing in user mode) 7038 the kernel will oops in either "warn" or "fatal" 7039 mode. 7040 7041 #DB exception for bus lock is triggered only when 7042 CPL > 0. 7043 7044 srbds= [X86,INTEL,EARLY] 7045 Control the Special Register Buffer Data Sampling 7046 (SRBDS) mitigation. 7047 7048 Certain CPUs are vulnerable to an MDS-like 7049 exploit which can leak bits from the random 7050 number generator. 7051 7052 By default, this issue is mitigated by 7053 microcode. However, the microcode fix can cause 7054 the RDRAND and RDSEED instructions to become 7055 much slower. Among other effects, this will 7056 result in reduced throughput from /dev/urandom. 7057 7058 The microcode mitigation can be disabled with 7059 the following option: 7060 7061 off: Disable mitigation and remove 7062 performance impact to RDRAND and RDSEED 7063 7064 srcutree.big_cpu_lim [KNL] 7065 Specifies the number of CPUs constituting a 7066 large system, such that srcu_struct structures 7067 should immediately allocate an srcu_node array. 7068 This kernel-boot parameter defaults to 128, 7069 but takes effect only when the low-order four 7070 bits of srcutree.convert_to_big is equal to 3 7071 (decide at boot). 7072 7073 srcutree.convert_to_big [KNL] 7074 Specifies under what conditions an SRCU tree 7075 srcu_struct structure will be converted to big 7076 form, that is, with an rcu_node tree: 7077 7078 0: Never. 7079 1: At init_srcu_struct() time. 7080 2: When rcutorture decides to. 7081 3: Decide at boot time (default). 7082 0x1X: Above plus if high contention. 7083 7084 Either way, the srcu_node tree will be sized based 7085 on the actual runtime number of CPUs (nr_cpu_ids) 7086 instead of the compile-time CONFIG_NR_CPUS. 7087 7088 srcutree.counter_wrap_check [KNL] 7089 Specifies how frequently to check for 7090 grace-period sequence counter wrap for the 7091 srcu_data structure's ->srcu_gp_seq_needed field. 7092 The greater the number of bits set in this kernel 7093 parameter, the less frequently counter wrap will 7094 be checked for. Note that the bottom two bits 7095 are ignored. 7096 7097 srcutree.exp_holdoff [KNL] 7098 Specifies how many nanoseconds must elapse 7099 since the end of the last SRCU grace period for 7100 a given srcu_struct until the next normal SRCU 7101 grace period will be considered for automatic 7102 expediting. Set to zero to disable automatic 7103 expediting. 7104 7105 srcutree.srcu_max_nodelay [KNL] 7106 Specifies the number of no-delay instances 7107 per jiffy for which the SRCU grace period 7108 worker thread will be rescheduled with zero 7109 delay. Beyond this limit, worker thread will 7110 be rescheduled with a sleep delay of one jiffy. 7111 7112 srcutree.srcu_max_nodelay_phase [KNL] 7113 Specifies the per-grace-period phase, number of 7114 non-sleeping polls of readers. Beyond this limit, 7115 grace period worker thread will be rescheduled 7116 with a sleep delay of one jiffy, between each 7117 rescan of the readers, for a grace period phase. 7118 7119 srcutree.srcu_retry_check_delay [KNL] 7120 Specifies number of microseconds of non-sleeping 7121 delay between each non-sleeping poll of readers. 7122 7123 srcutree.small_contention_lim [KNL] 7124 Specifies the number of update-side contention 7125 events per jiffy will be tolerated before 7126 initiating a conversion of an srcu_struct 7127 structure to big form. Note that the value of 7128 srcutree.convert_to_big must have the 0x10 bit 7129 set for contention-based conversions to occur. 7130 7131 ssbd= [ARM64,HW,EARLY] 7132 Speculative Store Bypass Disable control 7133 7134 On CPUs that are vulnerable to the Speculative 7135 Store Bypass vulnerability and offer a 7136 firmware based mitigation, this parameter 7137 indicates how the mitigation should be used: 7138 7139 force-on: Unconditionally enable mitigation for 7140 for both kernel and userspace 7141 force-off: Unconditionally disable mitigation for 7142 for both kernel and userspace 7143 kernel: Always enable mitigation in the 7144 kernel, and offer a prctl interface 7145 to allow userspace to register its 7146 interest in being mitigated too. 7147 7148 stack_guard_gap= [MM] 7149 override the default stack gap protection. The value 7150 is in page units and it defines how many pages prior 7151 to (for stacks growing down) resp. after (for stacks 7152 growing up) the main stack are reserved for no other 7153 mapping. Default value is 256 pages. 7154 7155 stack_depot_disable= [KNL,EARLY] 7156 Setting this to true through kernel command line will 7157 disable the stack depot thereby saving the static memory 7158 consumed by the stack hash table. By default this is set 7159 to false. 7160 7161 stack_depot_max_pools= [KNL,EARLY] 7162 Specify the maximum number of pools to use for storing 7163 stack traces. Pools are allocated on-demand up to this 7164 limit. Default value is 8191 pools. 7165 7166 stacktrace [FTRACE] 7167 Enabled the stack tracer on boot up. 7168 7169 stacktrace_filter=[function-list] 7170 [FTRACE] Limit the functions that the stack tracer 7171 will trace at boot up. function-list is a comma-separated 7172 list of functions. This list can be changed at run 7173 time by the stack_trace_filter file in the debugfs 7174 tracing directory. Note, this enables stack tracing 7175 and the stacktrace above is not needed. 7176 7177 sti= [PARISC,HW] 7178 Format: <num> 7179 Set the STI (builtin display/keyboard on the HP-PARISC 7180 machines) console (graphic card) which should be used 7181 as the initial boot-console. 7182 See also comment in drivers/video/console/sticore.c. 7183 7184 sti_font= [HW] 7185 See comment in drivers/video/console/sticore.c. 7186 7187 stifb= [HW] 7188 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 7189 7190 strict_sas_size= 7191 [X86] 7192 Format: <bool> 7193 Enable or disable strict sigaltstack size checks 7194 against the required signal frame size which 7195 depends on the supported FPU features. This can 7196 be used to filter out binaries which have 7197 not yet been made aware of AT_MINSIGSTKSZ. 7198 7199 stress_hpt [PPC,EARLY] 7200 Limits the number of kernel HPT entries in the hash 7201 page table to increase the rate of hash page table 7202 faults on kernel addresses. 7203 7204 stress_slb [PPC,EARLY] 7205 Limits the number of kernel SLB entries, and flushes 7206 them frequently to increase the rate of SLB faults 7207 on kernel addresses. 7208 7209 sunrpc.min_resvport= 7210 sunrpc.max_resvport= 7211 [NFS,SUNRPC] 7212 SunRPC servers often require that client requests 7213 originate from a privileged port (i.e. a port in the 7214 range 0 < portnr < 1024). 7215 An administrator who wishes to reserve some of these 7216 ports for other uses may adjust the range that the 7217 kernel's sunrpc client considers to be privileged 7218 using these two parameters to set the minimum and 7219 maximum port values. 7220 7221 sunrpc.svc_rpc_per_connection_limit= 7222 [NFS,SUNRPC] 7223 Limit the number of requests that the server will 7224 process in parallel from a single connection. 7225 The default value is 0 (no limit). 7226 7227 sunrpc.pool_mode= 7228 [NFS] 7229 Control how the NFS server code allocates CPUs to 7230 service thread pools. Depending on how many NICs 7231 you have and where their interrupts are bound, this 7232 option will affect which CPUs will do NFS serving. 7233 Note: this parameter cannot be changed while the 7234 NFS server is running. 7235 7236 auto the server chooses an appropriate mode 7237 automatically using heuristics 7238 global a single global pool contains all CPUs 7239 percpu one pool for each CPU 7240 pernode one pool for each NUMA node (equivalent 7241 to global on non-NUMA machines) 7242 7243 sunrpc.tcp_slot_table_entries= 7244 sunrpc.udp_slot_table_entries= 7245 [NFS,SUNRPC] 7246 Sets the upper limit on the number of simultaneous 7247 RPC calls that can be sent from the client to a 7248 server. Increasing these values may allow you to 7249 improve throughput, but will also increase the 7250 amount of memory reserved for use by the client. 7251 7252 suspend.pm_test_delay= 7253 [SUSPEND] 7254 Sets the number of seconds to remain in a suspend test 7255 mode before resuming the system (see 7256 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 7257 is set. Default value is 5. 7258 7259 svm= [PPC] 7260 Format: { on | off | y | n | 1 | 0 } 7261 This parameter controls use of the Protected 7262 Execution Facility on pSeries. 7263 7264 swiotlb= [ARM,PPC,MIPS,X86,S390,EARLY] 7265 Format: { <int> [,<int>] | force | noforce } 7266 <int> -- Number of I/O TLB slabs 7267 <int> -- Second integer after comma. Number of swiotlb 7268 areas with their own lock. Will be rounded up 7269 to a power of 2. 7270 force -- force using of bounce buffers even if they 7271 wouldn't be automatically used by the kernel 7272 noforce -- Never use bounce buffers (for debugging) 7273 7274 switches= [HW,M68k,EARLY] 7275 7276 sysctl.*= [KNL] 7277 Set a sysctl parameter, right before loading the init 7278 process, as if the value was written to the respective 7279 /proc/sys/... file. Both '.' and '/' are recognized as 7280 separators. Unrecognized parameters and invalid values 7281 are reported in the kernel log. Sysctls registered 7282 later by a loaded module cannot be set this way. 7283 Example: sysctl.vm.swappiness=40 7284 7285 sysrq_always_enabled 7286 [KNL] 7287 Ignore sysrq setting - this boot parameter will 7288 neutralize any effect of /proc/sys/kernel/sysrq. 7289 Useful for debugging. 7290 7291 tcpmhash_entries= [KNL,NET] 7292 Set the number of tcp_metrics_hash slots. 7293 Default value is 8192 or 16384 depending on total 7294 ram pages. This is used to specify the TCP metrics 7295 cache size. See Documentation/networking/ip-sysctl.rst 7296 "tcp_no_metrics_save" section for more details. 7297 7298 tdfx= [HW,DRM] 7299 7300 test_suspend= [SUSPEND] 7301 Format: { "mem" | "standby" | "freeze" }[,N] 7302 Specify "mem" (for Suspend-to-RAM) or "standby" (for 7303 standby suspend) or "freeze" (for suspend type freeze) 7304 as the system sleep state during system startup with 7305 the optional capability to repeat N number of times. 7306 The system is woken from this state using a 7307 wakeup-capable RTC alarm. 7308 7309 thash_entries= [KNL,NET] 7310 Set number of hash buckets for TCP connection 7311 7312 thermal.act= [HW,ACPI] 7313 -1: disable all active trip points in all thermal zones 7314 <degrees C>: override all lowest active trip points 7315 7316 thermal.crt= [HW,ACPI] 7317 -1: disable all critical trip points in all thermal zones 7318 <degrees C>: override all critical trip points 7319 7320 thermal.off= [HW,ACPI] 7321 1: disable ACPI thermal control 7322 7323 thermal.psv= [HW,ACPI] 7324 -1: disable all passive trip points 7325 <degrees C>: override all passive trip points to this 7326 value 7327 7328 thermal.tzp= [HW,ACPI] 7329 Specify global default ACPI thermal zone polling rate 7330 <deci-seconds>: poll all this frequency 7331 0: no polling (default) 7332 7333 thp_anon= [KNL] 7334 Format: <size>[KMG],<size>[KMG]:<state>;<size>[KMG]-<size>[KMG]:<state> 7335 state is one of "always", "madvise", "never" or "inherit". 7336 Control the default behavior of the system with respect 7337 to anonymous transparent hugepages. 7338 Can be used multiple times for multiple anon THP sizes. 7339 See Documentation/admin-guide/mm/transhuge.rst for more 7340 details. 7341 7342 threadirqs [KNL,EARLY] 7343 Force threading of all interrupt handlers except those 7344 marked explicitly IRQF_NO_THREAD. 7345 7346 thp_shmem= [KNL] 7347 Format: <size>[KMG],<size>[KMG]:<policy>;<size>[KMG]-<size>[KMG]:<policy> 7348 Control the default policy of each hugepage size for the 7349 internal shmem mount. <policy> is one of policies available 7350 for the shmem mount ("always", "inherit", "never", "within_size", 7351 and "advise"). 7352 It can be used multiple times for multiple shmem THP sizes. 7353 See Documentation/admin-guide/mm/transhuge.rst for more 7354 details. 7355 7356 topology= [S390,EARLY] 7357 Format: {off | on} 7358 Specify if the kernel should make use of the cpu 7359 topology information if the hardware supports this. 7360 The scheduler will make use of this information and 7361 e.g. base its process migration decisions on it. 7362 Default is on. 7363 7364 torture.disable_onoff_at_boot= [KNL] 7365 Prevent the CPU-hotplug component of torturing 7366 until after init has spawned. 7367 7368 torture.ftrace_dump_at_shutdown= [KNL] 7369 Dump the ftrace buffer at torture-test shutdown, 7370 even if there were no errors. This can be a 7371 very costly operation when many torture tests 7372 are running concurrently, especially on systems 7373 with rotating-rust storage. 7374 7375 torture.verbose_sleep_frequency= [KNL] 7376 Specifies how many verbose printk()s should be 7377 emitted between each sleep. The default of zero 7378 disables verbose-printk() sleeping. 7379 7380 torture.verbose_sleep_duration= [KNL] 7381 Duration of each verbose-printk() sleep in jiffies. 7382 7383 tpm.disable_pcr_integrity= [HW,TPM] 7384 Do not protect PCR registers from unintended physical 7385 access, or interposers in the bus by the means of 7386 having an integrity protected session wrapped around 7387 TPM2_PCR_Extend command. Consider this in a situation 7388 where TPM is heavily utilized by IMA, thus protection 7389 causing a major performance hit, and the space where 7390 machines are deployed is by other means guarded. 7391 7392 tpm_crb_ffa.busy_timeout_ms= [ARM64,TPM] 7393 Maximum time in milliseconds to retry sending a message 7394 to the TPM service before giving up. This parameter controls 7395 how long the system will continue retrying when the TPM 7396 service is busy. 7397 Format: <unsigned int> 7398 Default: 2000 (2 seconds) 7399 7400 tpm_suspend_pcr=[HW,TPM] 7401 Format: integer pcr id 7402 Specify that at suspend time, the tpm driver 7403 should extend the specified pcr with zeros, 7404 as a workaround for some chips which fail to 7405 flush the last written pcr on TPM_SaveState. 7406 This will guarantee that all the other pcrs 7407 are saved. 7408 7409 tpm_tis.interrupts= [HW,TPM] 7410 Enable interrupts for the MMIO based physical layer 7411 for the FIFO interface. By default it is set to false 7412 (0). For more information about TPM hardware interfaces 7413 defined by Trusted Computing Group (TCG) see 7414 https://trustedcomputinggroup.org/resource/pc-client-platform-tpm-profile-ptp-specification/ 7415 7416 tp_printk [FTRACE] 7417 Have the tracepoints sent to printk as well as the 7418 tracing ring buffer. This is useful for early boot up 7419 where the system hangs or reboots and does not give the 7420 option for reading the tracing buffer or performing a 7421 ftrace_dump_on_oops. 7422 7423 To turn off having tracepoints sent to printk, 7424 echo 0 > /proc/sys/kernel/tracepoint_printk 7425 Note, echoing 1 into this file without the 7426 tp_printk kernel cmdline option has no effect. 7427 7428 The tp_printk_stop_on_boot (see below) can also be used 7429 to stop the printing of events to console at 7430 late_initcall_sync. 7431 7432 ** CAUTION ** 7433 7434 Having tracepoints sent to printk() and activating high 7435 frequency tracepoints such as irq or sched, can cause 7436 the system to live lock. 7437 7438 tp_printk_stop_on_boot [FTRACE] 7439 When tp_printk (above) is set, it can cause a lot of noise 7440 on the console. It may be useful to only include the 7441 printing of events during boot up, as user space may 7442 make the system inoperable. 7443 7444 This command line option will stop the printing of events 7445 to console at the late_initcall_sync() time frame. 7446 7447 trace_buf_size=nn[KMG] 7448 [FTRACE] will set tracing buffer size on each cpu. 7449 7450 trace_clock= [FTRACE] Set the clock used for tracing events 7451 at boot up. 7452 local - Use the per CPU time stamp counter 7453 (converted into nanoseconds). Fast, but 7454 depending on the architecture, may not be 7455 in sync between CPUs. 7456 global - Event time stamps are synchronized across 7457 CPUs. May be slower than the local clock, 7458 but better for some race conditions. 7459 counter - Simple counting of events (1, 2, ..) 7460 note, some counts may be skipped due to the 7461 infrastructure grabbing the clock more than 7462 once per event. 7463 uptime - Use jiffies as the time stamp. 7464 perf - Use the same clock that perf uses. 7465 mono - Use ktime_get_mono_fast_ns() for time stamps. 7466 mono_raw - Use ktime_get_raw_fast_ns() for time 7467 stamps. 7468 boot - Use ktime_get_boot_fast_ns() for time stamps. 7469 Architectures may add more clocks. See 7470 Documentation/trace/ftrace.rst for more details. 7471 7472 trace_event=[event-list] 7473 [FTRACE] Set and start specified trace events in order 7474 to facilitate early boot debugging. The event-list is a 7475 comma-separated list of trace events to enable. See 7476 also Documentation/trace/events.rst 7477 7478 To enable modules, use :mod: keyword: 7479 7480 trace_event=:mod:<module> 7481 7482 The value before :mod: will only enable specific events 7483 that are part of the module. See the above mentioned 7484 document for more information. 7485 7486 trace_instance=[instance-info] 7487 [FTRACE] Create a ring buffer instance early in boot up. 7488 This will be listed in: 7489 7490 /sys/kernel/tracing/instances 7491 7492 Events can be enabled at the time the instance is created 7493 via: 7494 7495 trace_instance=<name>,<system1>:<event1>,<system2>:<event2> 7496 7497 Note, the "<system*>:" portion is optional if the event is 7498 unique. 7499 7500 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall 7501 7502 will enable the "sched_switch" event (note, the "sched:" is optional, and 7503 the same thing would happen if it was left off). The irq_handler_entry 7504 event, and all events under the "initcall" system. 7505 7506 Flags can be added to the instance to modify its behavior when it is 7507 created. The flags are separated by '^'. 7508 7509 The available flags are: 7510 7511 traceoff - Have the tracing instance tracing disabled after it is created. 7512 traceprintk - Have trace_printk() write into this trace instance 7513 (note, "printk" and "trace_printk" can also be used) 7514 7515 trace_instance=foo^traceoff^traceprintk,sched,irq 7516 7517 The flags must come before the defined events. 7518 7519 If memory has been reserved (see memmap for x86), the instance 7520 can use that memory: 7521 7522 memmap=12M$0x284500000 trace_instance=boot_map@0x284500000:12M 7523 7524 The above will create a "boot_map" instance that uses the physical 7525 memory at 0x284500000 that is 12Megs. The per CPU buffers of that 7526 instance will be split up accordingly. 7527 7528 Alternatively, the memory can be reserved by the reserve_mem option: 7529 7530 reserve_mem=12M:4096:trace trace_instance=boot_map@trace 7531 7532 This will reserve 12 megabytes at boot up with a 4096 byte alignment 7533 and place the ring buffer in this memory. Note that due to KASLR, the 7534 memory may not be the same location each time, which will not preserve 7535 the buffer content. 7536 7537 Also note that the layout of the ring buffer data may change between 7538 kernel versions where the validator will fail and reset the ring buffer 7539 if the layout is not the same as the previous kernel. 7540 7541 If the ring buffer is used for persistent bootups and has events enabled, 7542 it is recommend to disable tracing so that events from a previous boot do not 7543 mix with events of the current boot (unless you are debugging a random crash 7544 at boot up). 7545 7546 reserve_mem=12M:4096:trace trace_instance=boot_map^traceoff^traceprintk@trace,sched,irq 7547 7548 Note, saving the trace buffer across reboots does require that the system 7549 is set up to not wipe memory. For instance, CONFIG_RESET_ATTACK_MITIGATION 7550 can force a memory reset on boot which will clear any trace that was stored. 7551 This is just one of many ways that can clear memory. Make sure your system 7552 keeps the content of memory across reboots before relying on this option. 7553 7554 NB: Both the mapped address and size must be page aligned for the architecture. 7555 7556 See also Documentation/trace/debugging.rst 7557 7558 7559 trace_options=[option-list] 7560 [FTRACE] Enable or disable tracer options at boot. 7561 The option-list is a comma delimited list of options 7562 that can be enabled or disabled just as if you were 7563 to echo the option name into 7564 7565 /sys/kernel/tracing/trace_options 7566 7567 For example, to enable stacktrace option (to dump the 7568 stack trace of each event), add to the command line: 7569 7570 trace_options=stacktrace 7571 7572 See also Documentation/trace/ftrace.rst "trace options" 7573 section. 7574 7575 trace_trigger=[trigger-list] 7576 [FTRACE] Add an event trigger on specific events. 7577 Set a trigger on top of a specific event, with an optional 7578 filter. 7579 7580 The format is "trace_trigger=<event>.<trigger>[ if <filter>],..." 7581 Where more than one trigger may be specified that are comma delimited. 7582 7583 For example: 7584 7585 trace_trigger="sched_switch.stacktrace if prev_state == 2" 7586 7587 The above will enable the "stacktrace" trigger on the "sched_switch" 7588 event but only trigger it if the "prev_state" of the "sched_switch" 7589 event is "2" (TASK_UNINTERRUPTIBLE). 7590 7591 See also "Event triggers" in Documentation/trace/events.rst 7592 7593 7594 traceoff_after_boot 7595 [FTRACE] Sometimes tracing is used to debug issues 7596 during the boot process. Since the trace buffer has a 7597 limited amount of storage, it may be prudent to 7598 disable tracing after the boot is finished, otherwise 7599 the critical information may be overwritten. With this 7600 option, the main tracing buffer will be turned off at 7601 the end of the boot process. 7602 7603 traceoff_on_warning 7604 [FTRACE] enable this option to disable tracing when a 7605 warning is hit. This turns off "tracing_on". Tracing can 7606 be enabled again by echoing '1' into the "tracing_on" 7607 file located in /sys/kernel/tracing/ 7608 7609 This option is useful, as it disables the trace before 7610 the WARNING dump is called, which prevents the trace to 7611 be filled with content caused by the warning output. 7612 7613 This option can also be set at run time via the sysctl 7614 option: kernel/traceoff_on_warning 7615 7616 transparent_hugepage= 7617 [KNL] 7618 Format: [always|madvise|never] 7619 Can be used to control the default behavior of the system 7620 with respect to transparent hugepages. 7621 See Documentation/admin-guide/mm/transhuge.rst 7622 for more details. 7623 7624 transparent_hugepage_shmem= [KNL] 7625 Format: [always|within_size|advise|never|deny|force] 7626 Can be used to control the hugepage allocation policy for 7627 the internal shmem mount. 7628 See Documentation/admin-guide/mm/transhuge.rst 7629 for more details. 7630 7631 transparent_hugepage_tmpfs= [KNL] 7632 Format: [always|within_size|advise|never] 7633 Can be used to control the default hugepage allocation policy 7634 for the tmpfs mount. 7635 See Documentation/admin-guide/mm/transhuge.rst 7636 for more details. 7637 7638 trusted.source= [KEYS] 7639 Format: <string> 7640 This parameter identifies the trust source as a backend 7641 for trusted keys implementation. Supported trust 7642 sources: 7643 - "tpm" 7644 - "tee" 7645 - "caam" 7646 - "dcp" 7647 If not specified then it defaults to iterating through 7648 the trust source list starting with TPM and assigns the 7649 first trust source as a backend which is initialized 7650 successfully during iteration. 7651 7652 trusted.rng= [KEYS] 7653 Format: <string> 7654 The RNG used to generate key material for trusted keys. 7655 Can be one of: 7656 - "kernel" 7657 - the same value as trusted.source: "tpm" or "tee" 7658 - "default" 7659 If not specified, "default" is used. In this case, 7660 the RNG's choice is left to each individual trust source. 7661 7662 trusted.dcp_use_otp_key 7663 This is intended to be used in combination with 7664 trusted.source=dcp and will select the DCP OTP key 7665 instead of the DCP UNIQUE key blob encryption. 7666 7667 trusted.dcp_skip_zk_test 7668 This is intended to be used in combination with 7669 trusted.source=dcp and will disable the check if the 7670 blob key is all zeros. This is helpful for situations where 7671 having this key zero'ed is acceptable. E.g. in testing 7672 scenarios. 7673 7674 tsa= [X86] Control mitigation for Transient Scheduler 7675 Attacks on AMD CPUs. Search the following in your 7676 favourite search engine for more details: 7677 7678 "Technical guidance for mitigating transient scheduler 7679 attacks". 7680 7681 off - disable the mitigation 7682 on - enable the mitigation (default) 7683 user - mitigate only user/kernel transitions 7684 vm - mitigate only guest/host transitions 7685 7686 7687 tsc= Disable clocksource stability checks for TSC. 7688 Format: <string> 7689 [x86] reliable: mark tsc clocksource as reliable, this 7690 disables clocksource verification at runtime, as well 7691 as the stability checks done at bootup. Used to enable 7692 high-resolution timer mode on older hardware, and in 7693 virtualized environment. 7694 [x86] noirqtime: Do not use TSC to do irq accounting. 7695 Used to run time disable IRQ_TIME_ACCOUNTING on any 7696 platforms where RDTSC is slow and this accounting 7697 can add overhead. 7698 [x86] unstable: mark the TSC clocksource as unstable, this 7699 marks the TSC unconditionally unstable at bootup and 7700 avoids any further wobbles once the TSC watchdog notices. 7701 [x86] nowatchdog: disable clocksource watchdog. Used 7702 in situations with strict latency requirements (where 7703 interruptions from clocksource watchdog are not 7704 acceptable). 7705 [x86] recalibrate: force recalibration against a HW timer 7706 (HPET or PM timer) on systems whose TSC frequency was 7707 obtained from HW or FW using either an MSR or CPUID(0x15). 7708 Warn if the difference is more than 500 ppm. 7709 [x86] watchdog: Use TSC as the watchdog clocksource with 7710 which to check other HW timers (HPET or PM timer), but 7711 only on systems where TSC has been deemed trustworthy. 7712 This will be suppressed by an earlier tsc=nowatchdog and 7713 can be overridden by a later tsc=nowatchdog. A console 7714 message will flag any such suppression or overriding. 7715 7716 tsc_early_khz= [X86,EARLY] Skip early TSC calibration and use the given 7717 value instead. Useful when the early TSC frequency discovery 7718 procedure is not reliable, such as on overclocked systems 7719 with CPUID.16h support and partial CPUID.15h support. 7720 Format: <unsigned int> 7721 7722 tsx= [X86] Control Transactional Synchronization 7723 Extensions (TSX) feature in Intel processors that 7724 support TSX control. 7725 7726 This parameter controls the TSX feature. The options are: 7727 7728 on - Enable TSX on the system. Although there are 7729 mitigations for all known security vulnerabilities, 7730 TSX has been known to be an accelerator for 7731 several previous speculation-related CVEs, and 7732 so there may be unknown security risks associated 7733 with leaving it enabled. 7734 7735 off - Disable TSX on the system. (Note that this 7736 option takes effect only on newer CPUs which are 7737 not vulnerable to MDS, i.e., have 7738 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 7739 the new IA32_TSX_CTRL MSR through a microcode 7740 update. This new MSR allows for the reliable 7741 deactivation of the TSX functionality.) 7742 7743 auto - Disable TSX if X86_BUG_TAA is present, 7744 otherwise enable TSX on the system. 7745 7746 Not specifying this option is equivalent to tsx=off. 7747 7748 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 7749 for more details. 7750 7751 tsx_async_abort= [X86,INTEL,EARLY] Control mitigation for the TSX Async 7752 Abort (TAA) vulnerability. 7753 7754 Similar to Micro-architectural Data Sampling (MDS) 7755 certain CPUs that support Transactional 7756 Synchronization Extensions (TSX) are vulnerable to an 7757 exploit against CPU internal buffers which can forward 7758 information to a disclosure gadget under certain 7759 conditions. 7760 7761 In vulnerable processors, the speculatively forwarded 7762 data can be used in a cache side channel attack, to 7763 access data to which the attacker does not have direct 7764 access. 7765 7766 This parameter controls the TAA mitigation. The 7767 options are: 7768 7769 full - Enable TAA mitigation on vulnerable CPUs 7770 if TSX is enabled. 7771 7772 full,nosmt - Enable TAA mitigation and disable SMT on 7773 vulnerable CPUs. If TSX is disabled, SMT 7774 is not disabled because CPU is not 7775 vulnerable to cross-thread TAA attacks. 7776 off - Unconditionally disable TAA mitigation 7777 7778 On MDS-affected machines, tsx_async_abort=off can be 7779 prevented by an active MDS mitigation as both vulnerabilities 7780 are mitigated with the same mechanism so in order to disable 7781 this mitigation, you need to specify mds=off too. 7782 7783 Not specifying this option is equivalent to 7784 tsx_async_abort=full. On CPUs which are MDS affected 7785 and deploy MDS mitigation, TAA mitigation is not 7786 required and doesn't provide any additional 7787 mitigation. 7788 7789 For details see: 7790 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 7791 7792 turbografx.map[2|3]= [HW,JOY] 7793 TurboGraFX parallel port interface 7794 Format: 7795 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 7796 See also Documentation/input/devices/joystick-parport.rst 7797 7798 udbg-immortal [PPC] When debugging early kernel crashes that 7799 happen after console_init() and before a proper 7800 console driver takes over, this boot options might 7801 help "seeing" what's going on. 7802 7803 uhash_entries= [KNL,NET] 7804 Set number of hash buckets for UDP/UDP-Lite connections 7805 7806 uhci-hcd.ignore_oc= 7807 [USB] Ignore overcurrent events (default N). 7808 Some badly-designed motherboards generate lots of 7809 bogus events, for ports that aren't wired to 7810 anything. Set this parameter to avoid log spamming. 7811 Note that genuine overcurrent events won't be 7812 reported either. 7813 7814 unaligned_scalar_speed= 7815 [RISCV] 7816 Format: {slow | fast | unsupported} 7817 Allow skipping scalar unaligned access speed tests. This 7818 is useful for testing alternative code paths and to skip 7819 the tests in environments where they run too slowly. All 7820 CPUs must have the same scalar unaligned access speed. 7821 7822 unaligned_vector_speed= 7823 [RISCV] 7824 Format: {slow | fast | unsupported} 7825 Allow skipping vector unaligned access speed tests. This 7826 is useful for testing alternative code paths and to skip 7827 the tests in environments where they run too slowly. All 7828 CPUs must have the same vector unaligned access speed. 7829 7830 unknown_nmi_panic 7831 [X86] Cause panic on unknown NMI. 7832 7833 unwind_debug [X86-64,EARLY] 7834 Enable unwinder debug output. This can be 7835 useful for debugging certain unwinder error 7836 conditions, including corrupt stacks and 7837 bad/missing unwinder metadata. 7838 7839 usbcore.authorized_default= 7840 [USB] Default USB device authorization: 7841 (default -1 = authorized (same as 1), 7842 0 = not authorized, 1 = authorized, 2 = authorized 7843 if device connected to internal port) 7844 7845 usbcore.autosuspend= 7846 [USB] The autosuspend time delay (in seconds) used 7847 for newly-detected USB devices (default 2). This 7848 is the time required before an idle device will be 7849 autosuspended. Devices for which the delay is set 7850 to a negative value won't be autosuspended at all. 7851 7852 usbcore.usbfs_snoop= 7853 [USB] Set to log all usbfs traffic (default 0 = off). 7854 7855 usbcore.usbfs_snoop_max= 7856 [USB] Maximum number of bytes to snoop in each URB 7857 (default = 65536). 7858 7859 usbcore.blinkenlights= 7860 [USB] Set to cycle leds on hubs (default 0 = off). 7861 7862 usbcore.old_scheme_first= 7863 [USB] Start with the old device initialization 7864 scheme (default 0 = off). 7865 7866 usbcore.usbfs_memory_mb= 7867 [USB] Memory limit (in MB) for buffers allocated by 7868 usbfs (default = 16, 0 = max = 2047). 7869 7870 usbcore.use_both_schemes= 7871 [USB] Try the other device initialization scheme 7872 if the first one fails (default 1 = enabled). 7873 7874 usbcore.initial_descriptor_timeout= 7875 [USB] Specifies timeout for the initial 64-byte 7876 USB_REQ_GET_DESCRIPTOR request in milliseconds 7877 (default 5000 = 5.0 seconds). 7878 7879 usbcore.nousb [USB] Disable the USB subsystem 7880 7881 usbcore.quirks= 7882 [USB] A list of quirk entries to augment the built-in 7883 usb core quirk list. List entries are separated by 7884 commas. Each entry has the form 7885 VendorID:ProductID:Flags. The IDs are 4-digit hex 7886 numbers and Flags is a set of letters. Each letter 7887 will change the built-in quirk; setting it if it is 7888 clear and clearing it if it is set. The letters have 7889 the following meanings: 7890 a = USB_QUIRK_STRING_FETCH_255 (string 7891 descriptors must not be fetched using 7892 a 255-byte read); 7893 b = USB_QUIRK_RESET_RESUME (device can't resume 7894 correctly so reset it instead); 7895 c = USB_QUIRK_NO_SET_INTF (device can't handle 7896 Set-Interface requests); 7897 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 7898 handle its Configuration or Interface 7899 strings); 7900 e = USB_QUIRK_RESET (device can't be reset 7901 (e.g morph devices), don't use reset); 7902 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 7903 more interface descriptions than the 7904 bNumInterfaces count, and can't handle 7905 talking to these interfaces); 7906 g = USB_QUIRK_DELAY_INIT (device needs a pause 7907 during initialization, after we read 7908 the device descriptor); 7909 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 7910 high speed and super speed interrupt 7911 endpoints, the USB 2.0 and USB 3.0 spec 7912 require the interval in microframes (1 7913 microframe = 125 microseconds) to be 7914 calculated as interval = 2 ^ 7915 (bInterval-1). 7916 Devices with this quirk report their 7917 bInterval as the result of this 7918 calculation instead of the exponent 7919 variable used in the calculation); 7920 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 7921 handle device_qualifier descriptor 7922 requests); 7923 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 7924 generates spurious wakeup, ignore 7925 remote wakeup capability); 7926 k = USB_QUIRK_NO_LPM (device can't handle Link 7927 Power Management); 7928 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 7929 (Device reports its bInterval as linear 7930 frames instead of the USB 2.0 7931 calculation); 7932 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 7933 to be disconnected before suspend to 7934 prevent spurious wakeup); 7935 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 7936 pause after every control message); 7937 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 7938 delay after resetting its port); 7939 p = USB_QUIRK_SHORT_SET_ADDRESS_REQ_TIMEOUT 7940 (Reduce timeout of the SET_ADDRESS 7941 request from 5000 ms to 500 ms); 7942 Example: quirks=0781:5580:bk,0a5c:5834:gij 7943 7944 usbhid.mousepoll= 7945 [USBHID] The interval which mice are to be polled at. 7946 7947 usbhid.jspoll= 7948 [USBHID] The interval which joysticks are to be polled at. 7949 7950 usbhid.kbpoll= 7951 [USBHID] The interval which keyboards are to be polled at. 7952 7953 usb-storage.delay_use= 7954 [UMS] The delay in seconds before a new device is 7955 scanned for Logical Units (default 1). 7956 Optionally the delay in milliseconds if the value has 7957 suffix with "ms". 7958 Example: delay_use=2567ms 7959 7960 usb-storage.quirks= 7961 [UMS] A list of quirks entries to supplement or 7962 override the built-in unusual_devs list. List 7963 entries are separated by commas. Each entry has 7964 the form VID:PID:Flags where VID and PID are Vendor 7965 and Product ID values (4-digit hex numbers) and 7966 Flags is a set of characters, each corresponding 7967 to a common usb-storage quirk flag as follows: 7968 a = SANE_SENSE (collect more than 18 bytes 7969 of sense data, not on uas); 7970 b = BAD_SENSE (don't collect more than 18 7971 bytes of sense data, not on uas); 7972 c = FIX_CAPACITY (decrease the reported 7973 device capacity by one sector); 7974 d = NO_READ_DISC_INFO (don't use 7975 READ_DISC_INFO command, not on uas); 7976 e = NO_READ_CAPACITY_16 (don't use 7977 READ_CAPACITY_16 command); 7978 f = NO_REPORT_OPCODES (don't use report opcodes 7979 command, uas only); 7980 g = MAX_SECTORS_240 (don't transfer more than 7981 240 sectors at a time, uas only); 7982 h = CAPACITY_HEURISTICS (decrease the 7983 reported device capacity by one 7984 sector if the number is odd); 7985 i = IGNORE_DEVICE (don't bind to this 7986 device); 7987 j = NO_REPORT_LUNS (don't use report luns 7988 command, uas only); 7989 k = NO_SAME (do not use WRITE_SAME, uas only) 7990 l = NOT_LOCKABLE (don't try to lock and 7991 unlock ejectable media, not on uas); 7992 m = MAX_SECTORS_64 (don't transfer more 7993 than 64 sectors = 32 KB at a time, 7994 not on uas); 7995 n = INITIAL_READ10 (force a retry of the 7996 initial READ(10) command, not on uas); 7997 o = CAPACITY_OK (accept the capacity 7998 reported by the device, not on uas); 7999 p = WRITE_CACHE (the device cache is ON 8000 by default, not on uas); 8001 r = IGNORE_RESIDUE (the device reports 8002 bogus residue values, not on uas); 8003 s = SINGLE_LUN (the device has only one 8004 Logical Unit); 8005 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 8006 commands, uas only); 8007 u = IGNORE_UAS (don't bind to the uas driver); 8008 w = NO_WP_DETECT (don't test whether the 8009 medium is write-protected). 8010 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 8011 even if the device claims no cache, 8012 not on uas) 8013 Example: quirks=0419:aaf5:rl,0421:0433:rc 8014 8015 user_debug= [KNL,ARM] 8016 Format: <int> 8017 See arch/arm/Kconfig.debug help text. 8018 1 - undefined instruction events 8019 2 - system calls 8020 4 - invalid data aborts 8021 8 - SIGSEGV faults 8022 16 - SIGBUS faults 8023 Example: user_debug=31 8024 8025 vdso= [X86,SH,SPARC] 8026 On X86_32, this is an alias for vdso32=. Otherwise: 8027 8028 vdso=1: enable VDSO (the default) 8029 vdso=0: disable VDSO mapping 8030 8031 vdso32= [X86] Control the 32-bit vDSO 8032 vdso32=1: enable 32-bit VDSO 8033 vdso32=0 or vdso32=2: disable 32-bit VDSO 8034 8035 See the help text for CONFIG_COMPAT_VDSO for more 8036 details. If CONFIG_COMPAT_VDSO is set, the default is 8037 vdso32=0; otherwise, the default is vdso32=1. 8038 8039 For compatibility with older kernels, vdso32=2 is an 8040 alias for vdso32=0. 8041 8042 Try vdso32=0 if you encounter an error that says: 8043 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 8044 8045 video= [FB,EARLY] Frame buffer configuration 8046 See Documentation/fb/modedb.rst. 8047 8048 video.brightness_switch_enabled= [ACPI] 8049 Format: [0|1] 8050 If set to 1, on receiving an ACPI notify event 8051 generated by hotkey, video driver will adjust brightness 8052 level and then send out the event to user space through 8053 the allocated input device. If set to 0, video driver 8054 will only send out the event without touching backlight 8055 brightness level. 8056 default: 1 8057 8058 virtio_mmio.device= 8059 [VMMIO] Memory mapped virtio (platform) device. 8060 8061 <size>@<baseaddr>:<irq>[:<id>] 8062 where: 8063 <size> := size (can use standard suffixes 8064 like K, M and G) 8065 <baseaddr> := physical base address 8066 <irq> := interrupt number (as passed to 8067 request_irq()) 8068 <id> := (optional) platform device id 8069 example: 8070 virtio_mmio.device=1K@0x100b0000:48:7 8071 8072 Can be used multiple times for multiple devices. 8073 8074 vga= [BOOT,X86-32] Select a particular video mode 8075 See Documentation/arch/x86/boot.rst and 8076 Documentation/admin-guide/svga.rst. 8077 Use vga=ask for menu. 8078 This is actually a boot loader parameter; the value is 8079 passed to the kernel using a special protocol. 8080 8081 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 8082 May slow down system boot speed, especially when 8083 enabled on systems with a large amount of memory. 8084 All options are enabled by default, and this 8085 interface is meant to allow for selectively 8086 enabling or disabling specific virtual memory 8087 debugging features. 8088 8089 Available options are: 8090 P Enable page structure init time poisoning 8091 - Disable all of the above options 8092 8093 vmalloc=nn[KMG] [KNL,BOOT,EARLY] Forces the vmalloc area to have an 8094 exact size of <nn>. This can be used to increase 8095 the minimum size (128MB on x86, arm32 platforms). 8096 It can also be used to decrease the size and leave more room 8097 for directly mapped kernel RAM. Note that this parameter does 8098 not exist on many other platforms (including arm64, alpha, 8099 loongarch, arc, csky, hexagon, microblaze, mips, nios2, openrisc, 8100 parisc, m64k, powerpc, riscv, sh, um, xtensa, s390, sparc). 8101 8102 vmcp_cma=nn[MG] [KNL,S390,EARLY] 8103 Sets the memory size reserved for contiguous memory 8104 allocations for the vmcp device driver. 8105 8106 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 8107 Format: <command> 8108 8109 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 8110 Format: <command> 8111 8112 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 8113 Format: <command> 8114 8115 vmscape= [X86] Controls mitigation for VMscape attacks. 8116 VMscape attacks can leak information from a userspace 8117 hypervisor to a guest via speculative side-channels. 8118 8119 off - disable the mitigation 8120 ibpb - use Indirect Branch Prediction Barrier 8121 (IBPB) mitigation (default) 8122 force - force vulnerability detection even on 8123 unaffected processors 8124 8125 vsyscall= [X86-64,EARLY] 8126 Controls the behavior of vsyscalls (i.e. calls to 8127 fixed addresses of 0xffffffffff600x00 from legacy 8128 code). Most statically-linked binaries and older 8129 versions of glibc use these calls. Because these 8130 functions are at fixed addresses, they make nice 8131 targets for exploits that can control RIP. 8132 8133 emulate Vsyscalls turn into traps and are emulated 8134 reasonably safely. The vsyscall page is 8135 readable. 8136 8137 xonly [default] Vsyscalls turn into traps and are 8138 emulated reasonably safely. The vsyscall 8139 page is not readable. 8140 8141 none Vsyscalls don't work at all. This makes 8142 them quite hard to use for exploits but 8143 might break your system. 8144 8145 vt.color= [VT] Default text color. 8146 Format: 0xYX, X = foreground, Y = background. 8147 Default: 0x07 = light gray on black. 8148 8149 vt.cur_default= [VT] Default cursor shape. 8150 Format: 0xCCBBAA, where AA, BB, and CC are the same as 8151 the parameters of the <Esc>[?A;B;Cc escape sequence; 8152 see vga-softcursor.rst. Default: 2 = underline. 8153 8154 vt.default_blu= [VT] 8155 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 8156 Change the default blue palette of the console. 8157 This is a 16-member array composed of values 8158 ranging from 0-255. 8159 8160 vt.default_grn= [VT] 8161 Format: <green0>,<green1>,<green2>,...,<green15> 8162 Change the default green palette of the console. 8163 This is a 16-member array composed of values 8164 ranging from 0-255. 8165 8166 vt.default_red= [VT] 8167 Format: <red0>,<red1>,<red2>,...,<red15> 8168 Change the default red palette of the console. 8169 This is a 16-member array composed of values 8170 ranging from 0-255. 8171 8172 vt.default_utf8= 8173 [VT] 8174 Format=<0|1> 8175 Set system-wide default UTF-8 mode for all tty's. 8176 Default is 1, i.e. UTF-8 mode is enabled for all 8177 newly opened terminals. 8178 8179 vt.global_cursor_default= 8180 [VT] 8181 Format=<-1|0|1> 8182 Set system-wide default for whether a cursor 8183 is shown on new VTs. Default is -1, 8184 i.e. cursors will be created by default unless 8185 overridden by individual drivers. 0 will hide 8186 cursors, 1 will display them. 8187 8188 vt.italic= [VT] Default color for italic text; 0-15. 8189 Default: 2 = green. 8190 8191 vt.underline= [VT] Default color for underlined text; 0-15. 8192 Default: 3 = cyan. 8193 8194 watchdog timers [HW,WDT] For information on watchdog timers, 8195 see Documentation/watchdog/watchdog-parameters.rst 8196 or other driver-specific files in the 8197 Documentation/watchdog/ directory. 8198 8199 watchdog_thresh= 8200 [KNL] 8201 Set the hard lockup detector stall duration 8202 threshold in seconds. The soft lockup detector 8203 threshold is set to twice the value. A value of 0 8204 disables both lockup detectors. Default is 10 8205 seconds. 8206 8207 workqueue.unbound_cpus= 8208 [KNL,SMP] Specify to constrain one or some CPUs 8209 to use in unbound workqueues. 8210 Format: <cpu-list> 8211 By default, all online CPUs are available for 8212 unbound workqueues. 8213 8214 workqueue.watchdog_thresh= 8215 If CONFIG_WQ_WATCHDOG is configured, workqueue can 8216 warn stall conditions and dump internal state to 8217 help debugging. 0 disables workqueue stall 8218 detection; otherwise, it's the stall threshold 8219 duration in seconds. The default value is 30 and 8220 it can be updated at runtime by writing to the 8221 corresponding sysfs file. 8222 8223 workqueue.panic_on_stall=<uint> 8224 Panic when workqueue stall is detected by 8225 CONFIG_WQ_WATCHDOG. It sets the number times of the 8226 stall to trigger panic. 8227 8228 The default is 0, which disables the panic on stall. 8229 8230 workqueue.cpu_intensive_thresh_us= 8231 Per-cpu work items which run for longer than this 8232 threshold are automatically considered CPU intensive 8233 and excluded from concurrency management to prevent 8234 them from noticeably delaying other per-cpu work 8235 items. Default is 10000 (10ms). 8236 8237 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8238 will report the work functions which violate this 8239 threshold repeatedly. They are likely good 8240 candidates for using WQ_UNBOUND workqueues instead. 8241 8242 workqueue.cpu_intensive_warning_thresh=<uint> 8243 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8244 will report the work functions which violate the 8245 intensive_threshold_us repeatedly. In order to prevent 8246 spurious warnings, start printing only after a work 8247 function has violated this threshold number of times. 8248 8249 The default is 4 times. 0 disables the warning. 8250 8251 workqueue.power_efficient 8252 Per-cpu workqueues are generally preferred because 8253 they show better performance thanks to cache 8254 locality; unfortunately, per-cpu workqueues tend to 8255 be more power hungry than unbound workqueues. 8256 8257 Enabling this makes the per-cpu workqueues which 8258 were observed to contribute significantly to power 8259 consumption unbound, leading to measurably lower 8260 power usage at the cost of small performance 8261 overhead. 8262 8263 The default value of this parameter is determined by 8264 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 8265 8266 workqueue.default_affinity_scope= 8267 Select the default affinity scope to use for unbound 8268 workqueues. Can be one of "cpu", "smt", "cache", 8269 "numa" and "system". Default is "cache". For more 8270 information, see the Affinity Scopes section in 8271 Documentation/core-api/workqueue.rst. 8272 8273 This can be changed after boot by writing to the 8274 matching /sys/module/workqueue/parameters file. All 8275 workqueues with the "default" affinity scope will be 8276 updated accordingly. 8277 8278 workqueue.debug_force_rr_cpu 8279 Workqueue used to implicitly guarantee that work 8280 items queued without explicit CPU specified are put 8281 on the local CPU. This guarantee is no longer true 8282 and while local CPU is still preferred work items 8283 may be put on foreign CPUs. This debug option 8284 forces round-robin CPU selection to flush out 8285 usages which depend on the now broken guarantee. 8286 When enabled, memory and cache locality will be 8287 impacted. 8288 8289 writecombine= [LOONGARCH,EARLY] Control the MAT (Memory Access 8290 Type) of ioremap_wc(). 8291 8292 on - Enable writecombine, use WUC for ioremap_wc() 8293 off - Disable writecombine, use SUC for ioremap_wc() 8294 8295 x2apic_phys [X86-64,APIC,EARLY] Use x2apic physical mode instead of 8296 default x2apic cluster mode on platforms 8297 supporting x2apic. 8298 8299 xen_512gb_limit [KNL,X86-64,XEN] 8300 Restricts the kernel running paravirtualized under Xen 8301 to use only up to 512 GB of RAM. The reason to do so is 8302 crash analysis tools and Xen tools for doing domain 8303 save/restore/migration must be enabled to handle larger 8304 domains. 8305 8306 xen_emul_unplug= [HW,X86,XEN,EARLY] 8307 Unplug Xen emulated devices 8308 Format: [unplug0,][unplug1] 8309 ide-disks -- unplug primary master IDE devices 8310 aux-ide-disks -- unplug non-primary-master IDE devices 8311 nics -- unplug network devices 8312 all -- unplug all emulated devices (NICs and IDE disks) 8313 unnecessary -- unplugging emulated devices is 8314 unnecessary even if the host did not respond to 8315 the unplug protocol 8316 never -- do not unplug even if version check succeeds 8317 8318 xen_legacy_crash [X86,XEN,EARLY] 8319 Crash from Xen panic notifier, without executing late 8320 panic() code such as dumping handler. 8321 8322 xen_mc_debug [X86,XEN,EARLY] 8323 Enable multicall debugging when running as a Xen PV guest. 8324 Enabling this feature will reduce performance a little 8325 bit, so it should only be enabled for obtaining extended 8326 debug data in case of multicall errors. 8327 8328 xen_msr_safe= [X86,XEN,EARLY] 8329 Format: <bool> 8330 Select whether to always use non-faulting (safe) MSR 8331 access functions when running as Xen PV guest. The 8332 default value is controlled by CONFIG_XEN_PV_MSR_SAFE. 8333 8334 xen_nopv [X86] 8335 Disables the PV optimizations forcing the HVM guest to 8336 run as generic HVM guest with no PV drivers. 8337 This option is obsoleted by the "nopv" option, which 8338 has equivalent effect for XEN platform. 8339 8340 xen_no_vector_callback 8341 [KNL,X86,XEN,EARLY] Disable the vector callback for Xen 8342 event channel interrupts. 8343 8344 xen_scrub_pages= [XEN] 8345 Boolean option to control scrubbing pages before giving them back 8346 to Xen, for use by other domains. Can be also changed at runtime 8347 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 8348 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 8349 8350 xen_timer_slop= [X86-64,XEN,EARLY] 8351 Set the timer slop (in nanoseconds) for the virtual Xen 8352 timers (default is 100000). This adjusts the minimum 8353 delta of virtualized Xen timers, where lower values 8354 improve timer resolution at the expense of processing 8355 more timer interrupts. 8356 8357 xen.balloon_boot_timeout= [XEN] 8358 The time (in seconds) to wait before giving up to boot 8359 in case initial ballooning fails to free enough memory. 8360 Applies only when running as HVM or PVH guest and 8361 started with less memory configured than allowed at 8362 max. Default is 180. 8363 8364 xen.event_eoi_delay= [XEN] 8365 How long to delay EOI handling in case of event 8366 storms (jiffies). Default is 10. 8367 8368 xen.event_loop_timeout= [XEN] 8369 After which time (jiffies) the event handling loop 8370 should start to delay EOI handling. Default is 2. 8371 8372 xen.fifo_events= [XEN] 8373 Boolean parameter to disable using fifo event handling 8374 even if available. Normally fifo event handling is 8375 preferred over the 2-level event handling, as it is 8376 fairer and the number of possible event channels is 8377 much higher. Default is on (use fifo events). 8378 8379 xirc2ps_cs= [NET,PCMCIA] 8380 Format: 8381 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 8382 8383 xive= [PPC] 8384 By default on POWER9 and above, the kernel will 8385 natively use the XIVE interrupt controller. This option 8386 allows the fallback firmware mode to be used: 8387 8388 off Fallback to firmware control of XIVE interrupt 8389 controller on both pseries and powernv 8390 platforms. Only useful on POWER9 and above. 8391 8392 xive.store-eoi=off [PPC] 8393 By default on POWER10 and above, the kernel will use 8394 stores for EOI handling when the XIVE interrupt mode 8395 is active. This option allows the XIVE driver to use 8396 loads instead, as on POWER9. 8397 8398 xhci-hcd.quirks [USB,KNL] 8399 A hex value specifying bitmask with supplemental xhci 8400 host controller quirks. Meaning of each bit can be 8401 consulted in header drivers/usb/host/xhci.h. 8402 8403 xmon [PPC,EARLY] 8404 Format: { early | on | rw | ro | off } 8405 Controls if xmon debugger is enabled. Default is off. 8406 Passing only "xmon" is equivalent to "xmon=early". 8407 early Call xmon as early as possible on boot; xmon 8408 debugger is called from setup_arch(). 8409 on xmon debugger hooks will be installed so xmon 8410 is only called on a kernel crash. Default mode, 8411 i.e. either "ro" or "rw" mode, is controlled 8412 with CONFIG_XMON_DEFAULT_RO_MODE. 8413 rw xmon debugger hooks will be installed so xmon 8414 is called only on a kernel crash, mode is write, 8415 meaning SPR registers, memory and, other data 8416 can be written using xmon commands. 8417 ro same as "rw" option above but SPR registers, 8418 memory, and other data can't be written using 8419 xmon commands. 8420 off xmon is disabled. 8421